The Esseintial Núta Blog

A forum for all topics related to constructed languages
User avatar
DesEsseintes
mongolian
mongolian
Posts: 4331
Joined: 31 Mar 2013 13:16

The Esseintial Núta Blog

Post by DesEsseintes »

This is Núta

Hear ye! Hear ye! Núta is here!

Núta is my latest incomplete and incoherent idea for a North American-inspired polysynthetic conlang!

It is but a mere sketch!

And yet I bring unto you a thread! Let us sally forth!

Design Objectives

Núta is heavily inspired by the Iroquoian languages. I have wanted to make something Iroquoianesque for a while now but never hit on an aesthetic I could live with before now. Núta is meant to be somewhat less “messy” than Mohawk/Oneida, and this is reflected in the smaller vowel inventory, less baroque clusters, and lack of nasal vowels.

Phonology coming up!
The profusion of exclamation marks are meant to cover up for the lack of substance in what I have ready.
User avatar
DesEsseintes
mongolian
mongolian
Posts: 4331
Joined: 31 Mar 2013 13:16

Re: The Esseintial Núta Blog

Post by DesEsseintes »

Núta Phonology

Phoneme Inventory

The Núta consonant inventory consists of 9 consonant segments:

/n/ n
/t k ʔ/ t k ‘
/s h/ s h
/r/ r
/j w/ y w

- /t k/ are unaspirated and can be voiced in voiced environments; they become aspirated when followed by /s h/
- the rhotic is generally a tap

There are four phonemic vowel segments:

/a e i u/ a e ı u

Vowel length is a feature of stress and prosody and will be dealt with in that section.

Allophony

Uh, nothing much.

/a/ backs and rounds to [ɔ] before: a) /u/; b) /w/ followed by a back vowel /a u/; c) /kwa khwa hwa ku khu hu/

/k/ is labialised before /w/. Labialisation does not occur before /u/.

Consonants /n t k s r/ are generally palatalised to [nʲ tʲ kʲ sʲ rʲ] before /j/. Palatalisation does not occur before /i/.

/n r w j/ generally devoice to [n̥ ɾ̥ ç ʍ] following /h/. Additional devoicing environments will be listed in the section on prosody in a future post.

Phonotactics

The maximal syllable is CC(w,y)VC.

Relatively complex clusters are permitted. Here is a table of two-consonant clusters:

Code: Select all

    n   t   k   ʔ   s   h   r   y   w
n   -   -   -   -   -   -   -   ny  nw
t   tn  -   -   -   ts  th  tr  ty  tw
k   kn  -   -   -   ks  kh  kr  ky  kw
ʔ   ʔn  ʔt  ʔk  -   ʔs  ʔh  ʔr  ʔy  ʔw
s   sn  st  sk  -   -   sh  sr  sy  sw
h   hn  ht  hk  hʔ  hs  -   hr  hy  hw
r   rn  rt  rk  -   -   -   -   ry  rw
Any cluster ending in one of n t k s h can take a further glide to form a triconsonantal cluster. Furthermore, s h r can occur before any cluster starting in t k.

There might be the odd cluster that goes even further. I’m contemplating tsn ksn, especially.

Note that clusters in Núta tend to remain clusters in all but the most rapid registers of speech. /ts/ is [ts] and not [t͡s], /hn/ is [hn̥], /ksj/ is [ksʲj] and not [kʃ], etc.

The next post will be on stress and tone*. Meanwhile, did I miss anything?
*and it will contain sample word forms! So far, everything has been dreadfully dull and dry, I realise.
User avatar
DesEsseintes
mongolian
mongolian
Posts: 4331
Joined: 31 Mar 2013 13:16

Re: The Esseintial Núta Blog

Post by DesEsseintes »

Prosody - Stress and Tone

Núta has lexical stress. A word normally only has one stressed syllable. Stressed syllables trigger a number of processes involving vowel length and tone.

Open and Closed Syllables

Open and closed syllables in Núta are defined as follows.

A syllable is open if there is no coda consonant or if the coda consonant is h, and the following syllable has a simple onset (maximally C(w,y) ).

A syllable is closed if it contains a coda consonant other than h (usually one of ‘ s r), or if the following syllable has a complex or glottal stop onset.

Vowel lengthening

A stressed syllable nucleus lengthens if the syllable is open. Long vowels are indicated with an interpunct: a· e· ı· u·. If there is no h in coda, the stressed syllable receives high tone. High tone is indicated with an acute accent: á é í ú.

/téyu/ té·yu (open syllable with no h in coda → long vowel + high tone)
/ártsıwı/ ártsıwı (closed syllable with no h in coda → short vowel + high tone)
/ekyá‘u/ ekyá‘u (glottal stop makes preceding syllable closed → short vowel + high tone)

h-Effect

A stressed syllable with h in coda receives falling tone; h then elides. Falling tone is indicated with a grave accent: à è ì ù.
Again, length is assigned according to the guidelines above. (Effectively, the rule regarding vowel lengthening applies before this rule changed high tone to low.)

/kwerısráhte/ kwerısrà·te (open syllable with h coda → long vowel + falling tone)
/íhtsere/ ìtsere (closed syllable with h coda → short vowel + falling tone)

Falling Tone Spread

If a syllable with falling tone has one of n r w y (or a cluster composed solely of these segments) in onset, the falling contour spreads over to the preceding syllable. Depending on the length of the originally accented syllable, two different patterns emerge.

If the originally accented syllable is open (and therefore long), the preceding syllable receives high tone and the originally accented syllable loses its tone.

/uyúhse/ úyu·se

If the originally accented syllable is closed (and therefore short), the preceding syllable lengthens and the originally accented syllable retains its falling tone, resulting in a characteristic drag-and-drop intonation.

/kawáhtne/ ka·wàtne

Devoicing of Unstressed Syllables

Unstressed nuclei (and entire syllables) can be devoiced. This is indicated with a dot above on the vowel in nucleus: ȧ ė i u̇.

There are two situations in which an unstressed vowel is devoiced.
1. The nucleus of any unstressed syllable with h in coda devoices, unless the onset is also h. If the onset of the syllable is one of n r w y (or any cluster composed solely of these segments), it devoices as well.
2. The nucleus of an unstressed syllable with k t s or any cluster containing k t in onset and s in coda devoices.

Some examples

/kahkúhtne/ kȧhkùtne
/uhwasnyé’u/ u̇hwasnyé‘u

Exception: if the first two syllables of a word would both devoice according to the above principles, the initial syllable does not devoice. No word can start in two consecutive devoiced syllables.

/kahkuhtnéhse/ kahku̇htnè·se (first syllable does not devoice as second syllable is devoiced)

This is not quite finished, but the accent rules for underlyingly consonant-final words must await the next post.
User avatar
gestaltist
mayan
mayan
Posts: 1617
Joined: 11 Feb 2015 11:23

Re: The Esseintial Núta Blog

Post by gestaltist »

I quite like this sketch. For some reason, I particularly enjoyed your allophony rules. The difference in how /j w/ affect consonants compared to /i u/ is a nice touch.
Davush
greek
greek
Posts: 671
Joined: 10 Jan 2015 14:10

Re: The Esseintial Núta Blog

Post by Davush »

[:D] I have been looking forward to reading about one of your conlangs for a while now, so I thoroughly enjoyed this! Even though the languages you draw inspiration from are a still a mystery to me, I like the consistent aesthetic and /h/ producing a low tone is nice! Is that type of thing common in Iroquoian languages?
User avatar
DesEsseintes
mongolian
mongolian
Posts: 4331
Joined: 31 Mar 2013 13:16

Re: The Esseintial Núta Blog

Post by DesEsseintes »

gestaltist wrote: 03 Dec 2017 22:23 I quite like this sketch. For some reason, I particularly enjoyed your allophony rules. The difference in how /j w/ affect consonants compared to /i u/ is a nice touch.
That’s nice cos I just popped the allophony in there at the last minute. :mrgreen:
Davush wrote: 03 Dec 2017 22:35 [:D] I have been looking forward to reading about one of your conlangs for a while now, so I thoroughly enjoyed this! Even though the languages you draw inspiration from are a still a mystery to me, I like the consistent aesthetic and /h/ producing a low tone is nice! Is that type of thing common in Iroquoian languages?
I’m glad you enjoyed it and thanks for saying so. [:D] /h/ does not cause low tone in Mohawk/Oneida but my ideas for lengthening stressed syllables are based on those languages, especially Oneida. The h → low tone came from my Híí languages.
User avatar
DesEsseintes
mongolian
mongolian
Posts: 4331
Joined: 31 Mar 2013 13:16

Re: The Esseintial Núta Blog

Post by DesEsseintes »

Vowel Sandhi

In this post, we will make a small foray into Núta diachronics to understand the rather unpredictable outcomes of vowel sandhi in Núta.

Vowel sandhi in Núta occurs when a vowel-final morpheme comes before a vowel-initial morpheme. Núta does not permit vowel sequences in hiatus, apart from those caused by glide-elision, a perfectly regular and predictable process which I’ve yet to outline. Therefore, when two vowels meet at a morpheme boundary they are resolved into a single vowel or glide-vowel sequence.

anı + ıryé·ruanyuryé·ru

To understand why ı + ı resulted in yu, let us first take a look at the vowel system of Proto-Núta.

Proto-Núta had a plain five-vowel system /a e i o u/. Vowel sequences were resolved as follows:

Code: Select all

    a   e   i   o   u
a   a   ai  ai  au  au
e   ja  e   ei  jo  ju
i   ja  e   i   jo  ju
o   wa  we  oi  o   ou
u   wa  we  ui  o   u
At the next stage /u/ fronted to /y/, and at the same time, the diphthongs /au ou ui/ fronted to /øi øi yi/. The changes are summed up in the following table:

Code: Select all

    a   e   i   o   y
a   a   ai  ai  øi  øi
e   ja  e   ei  jo  jy
i   ja  e   i   jo  jy
o   wa  we  oi  o   øi
y   wa  we  yi  o   y
Next, /o oi/ raised to /u ui/ and subsequently the diphthongs /ai ei øi ui yi/ monophthongised to /e e ɥe wi ɥi/.

Code: Select all

    a   e   i   u   y
a   a   e   e   ɥe  ɥe
e   ja  e   e   ju  jy
i   ja  e   i   ju  jy
u   wa  we  wi  u   ɥe
y   wa  we  ɥi  u   y
Finally /y/ merges with /i/ and /ɥ/ merges with /j/. /ɥi jy/ become /ju wi/ respectively*.
No, I haven’t got them the wrong way round. That is what happens.

This last table then gives us the rules for vowel sandhi in modern Núta. Therefore, somewhat confusingly, I’ve used Núta orthography instead of IPA in this last table. <ı2> is used to indicate /i/ descended from Proto-Núta *u via /y/.

Code: Select all

    a   e   ı   u   ı2
a   a   e   e   ye  ye
e   ya  e   e   yu  wı
ı   ya  e   ı   yu  wı
u   wa  we  wı  u   ye
ı2  wa  we  yu  u   ı
This needs some further refining, and there will probably be levelling in some paradigms, but this table is meant to serve as a basis. So yay! I can move on.

Ps. I didn’t proofread this, so please excuse and point out possible errors, etc. [:)]
Davush
greek
greek
Posts: 671
Joined: 10 Jan 2015 14:10

Re: The Esseintial Núta Blog

Post by Davush »

I imagine all that vowel sandhi would be a nightmare for learners of the language... [:D] Interesting to see how the vowel shifts messed it up so much. Does it occur across word boundaries as well?
User avatar
DesEsseintes
mongolian
mongolian
Posts: 4331
Joined: 31 Mar 2013 13:16

Re: The Esseintial Núta Blog

Post by DesEsseintes »

Davush wrote: 05 Dec 2017 10:05 I imagine all that vowel sandhi would be a nightmare for learners of the language... [:D] Interesting to see how the vowel shifts messed it up so much. Does it occur across word boundaries as well?
For now at least, I don’t think it occurs across word boundaries. However, I envisage quite a high morpheme-to-word ratio so my sandhi should have plenty of room to operate in.
User avatar
DesEsseintes
mongolian
mongolian
Posts: 4331
Joined: 31 Mar 2013 13:16

Re: The Esseintial Núta Blog

Post by DesEsseintes »

In my original conception, I envisaged all Núta words ending in a vowel, but now I’m considering word-final consonants. I’d probably restrict these to t k s r, that is no glottals and no glides (and n in Núta seems to share some characteristics with glides). So these kinds of words would then be permissible:

té·rus
awúskrayır
ewí·ryak

Hmm, I need to think about this. Thoughts?
User avatar
Thrice Xandvii
runic
runic
Posts: 2698
Joined: 25 Nov 2012 10:13
Location: Carnassus

Re: The Esseintial Núta Blog

Post by Thrice Xandvii »

Hmmm, while I like coda /k/, it seems like a bit of an odd man out being the most back consonant to be in the coda. Almost seems like /x/ should join it as an allophone of /h/? Or maybe dump it for symmetry?

At any rate, I do have one thing to add, *ahem*.

*bangs fists on table in time to my chanting* Make a script! Make a script! Make a script!

Okay, I'm done now.
Last edited by Thrice Xandvii on 13 Dec 2017 14:00, edited 1 time in total.
Image
User avatar
DesEsseintes
mongolian
mongolian
Posts: 4331
Joined: 31 Mar 2013 13:16

Re: The Esseintial Núta Blog

Post by DesEsseintes »

Thrice Xandvii wrote: 11 Dec 2017 09:33 Hmmm, while I like coda /k/, it seems like a bit of an odd man out being the most back consonant to be in the coda. Almost seems like /x/ should join it as an allophony of /h/? Or maybe dump it for symmetry?
I understand all too well what you mean about /k/ and yes, the fact that - shock horror - a lone velar should upset the otherwise entirely alveolar harmony going on there did rather perturb one. But. I’m not going to fall prey to this kind of thinking this time because that’s what I always do and never get anywhere. Nevertheless, I appreciate the sentiment! [xD]
At any rate, I do have one thing to add, *ahem*.

*bangs fists on table in time to my chanting* Make a script! Make a script! Make a script!
Heh. One day.

Now, I need person markers. Sorely. I have one so far:

h- 3rd person masculine subject

Now, I’m pretty sure the markers are going to be tripartite. There will be a subject series used with statives, intransitive verbs, and transitive verbs with inanimate objects. Then there will be agent and patient series for transitive verbs with two animate arguments. So I need to get to work here.

But first, I’m going to post on Glide Elision.
I love capitalising my morfofo processes.
User avatar
Thrice Xandvii
runic
runic
Posts: 2698
Joined: 25 Nov 2012 10:13
Location: Carnassus

Re: The Esseintial Núta Blog

Post by Thrice Xandvii »

(Sounds like a band name.)
Image
Davush
greek
greek
Posts: 671
Joined: 10 Jan 2015 14:10

Re: The Esseintial Núta Blog

Post by Davush »

I would say nay to coda /k/ - those example words seem quite different from previous examples for some reason
User avatar
DesEsseintes
mongolian
mongolian
Posts: 4331
Joined: 31 Mar 2013 13:16

Re: The Esseintial Núta Blog

Post by DesEsseintes »

Glide Elision

Glide elision applies to the glides w y when these segments follow vowels lengthened due to accent placement.

The following simple rules always apply:
1. w y elide after a homorganic long vowel
2. w y elide after any long vowel before a homorganic vowel

Here are some examples:
/te‘éyırıs/ te‘é·ırıs
/nıtnáwu/ nıtná·u - sky
/kíyat/ kí·at - arrow
/úwehtarı/ ú·ėhtarı (note the devoiced vowel due to coda h)

Furthermore the following two elisions apply sporadically in word-final position (the details remain to be worked out):
3. w elides after long a before another a
4. y elides after long e before another e

/ráwa/ rá·a
/utsnéye/ utsné·e

I’ve spoilered all the possible outcomes of VGV sequences below just for fun:
Spoiler:

áwa → á·wa
or á·a
áwe → á·we
áwı → á·wı
áwu → á·u

éwa → é·wa
éwe → é·we
éwı → é·wı
éwu → é·u

íwa → í·wa
íwe → í·we
íwı → í·wı
íwu → í·u

úwa → ú·a
úwe → ú·e
úwı → ú·ı
úwu → ú·u

***

áya → á·ya
áye → á·ye
áyı → á·ı
áyu → á·yu

éya → é·ya
éye → é·ye
or é·e
éyı → é·ı
éyu → é·yu

íya → í·a
íye → í·e
íyı → í·ı
íyu → í·u

úya → ú·ya
úye → ú·ye
úyı → ú·ı
úyu → ú·yu
The sequences á·a é·e í·ı ú·u are generally pronounced as monosyllabic overlong vowels except in careful speech where they may retain a disyllabic quality. (Note also that á·wa é·ye generally would not contract under such conditions.) Therefore, Núta can be said to have a three-way vowel length distinction. This distinction is even phonemic in monosyllabic words, as some particles in Núta have no accent and therefore consist of a single syllable with a short nucleus. As such, a minimal trio such as the following is possible:

tsı tsí· tsí·ı

So yay! I always wanted to do overlong vowels. [:)]
User avatar
Frislander
mayan
mayan
Posts: 2088
Joined: 14 May 2016 18:47
Location: The North

Re: The Esseintial Núta Blog

Post by Frislander »

This is a bit late but I have kind of mixed feelings about the addition of the new coda consonants. On the one hand I like the new aesthetic they bring, and I definitely think you should keep them if that's what you want. On the other hand, I don't think there are enough conlangs which have consonant clusters but are strictly vowel-final. It's not really a gripe because I'm fine with it either way, but then at the same time that is just my kind of gut feeling on this.

Also that glide elision rule is neat!
User avatar
Thrice Xandvii
runic
runic
Posts: 2698
Joined: 25 Nov 2012 10:13
Location: Carnassus

Re: The Esseintial Núta Blog

Post by Thrice Xandvii »

That's an interesting point, actually. It feels right (though I haven't a clue of it is) that a language with a lot of clustering in the onset should/would have consonants in the coda as well. To have Núta disregard this, would be a cool and somewhat unexpected fact. As always, don't let us stand in your way as I like the shape of your newest words... It just is something interesting to consider.
Image
User avatar
DesEsseintes
mongolian
mongolian
Posts: 4331
Joined: 31 Mar 2013 13:16

Re: The Esseintial Núta Blog

Post by DesEsseintes »

You two will like my Áánene clicklang then. It has consonant clusters and clicks* in word-initial onset, but only allows a very limited subset of sounds anywhere else in a word. Strictly no codas, either. And there may not be velars at all. :mrgreen:

*and psst, perhaps even click clusters, but don’t tell anyone...
shimobaatar
korean
korean
Posts: 10373
Joined: 12 Jul 2013 23:09
Location: UTC-04:00

Re: The Esseintial Núta Blog

Post by shimobaatar »

DesEsseintes wrote: 03 Dec 2017 16:18 - /t k/ are unaspirated and can be voiced in voiced environments; they become aspirated when followed by /s h/
What does this language consider a "voiced environment"?
DesEsseintes wrote: 03 Dec 2017 16:18 The maximal syllable is CC(w,y)VC.

Relatively complex clusters are permitted. Here is a table of two-consonant clusters:
Can these clusters all occur word-initially? Can all consonants occur word-finally? (I see this is addressed below) Are there restrictions regarding what coda consonants can appear before what onsets?
DesEsseintes wrote: 03 Dec 2017 17:25 Unstressed nuclei (and entire syllables) can be devoiced. This is indicated with a dot above on the vowel in nucleus: ȧ ė i u̇.
This makes me wonder something. In the case of tone spread, which syllable is considered stressed? Can an originally stressed syllable be devoiced after the tone spreads? Can an originally unstressed syllable be devoiced after receiving tone?
DesEsseintes wrote: 03 Dec 2017 17:25 Exception: if the first two syllables of a word would both devoice according to the above principles, the initial syllable does not devoice. No word can start in two consecutive devoiced syllables.


Can a word otherwise have multiple devoiced syllables? (I don't know if anyone else can see it, but the spacing between this particular line of text and the quote above it comes out very weird on my end.)
DesEsseintes wrote: 05 Dec 2017 08:41 In this post, we will make a small foray into Núta diachronics to understand the rather unpredictable outcomes of vowel sandhi in Núta.
Very interesting! Was this inspired by anything in particular?
DesEsseintes wrote: 11 Dec 2017 08:46 In my original conception, I envisaged all Núta words ending in a vowel, but now I’m considering word-final consonants. I’d probably restrict these to t k s r, that is no glottals and no glides (and n in Núta seems to share some characteristics with glides). So these kinds of words would then be permissible:

té·rus
awúskrayır
ewí·ryak

Hmm, I need to think about this. Thoughts?
I'd vote against word-final coda consonants, but that's just me. It's totally up to you in the end.
DesEsseintes wrote: 12 Dec 2017 08:34 Now, I’m pretty sure the markers are going to be tripartite. There will be a subject series used with statives, intransitive verbs, and transitive verbs with inanimate objects. Then there will be agent and patient series for transitive verbs with two animate arguments. So I need to get to work here.
[:D]
DesEsseintes wrote: 12 Dec 2017 08:34 But first, I’m going to post on Glide Elision.
I love capitalising my morfofo processes.
I like capitalizing stuff like that, too. I feel like it makes descriptions of conlangs feel more official/scientific.
DesEsseintes wrote: 12 Dec 2017 18:09 1. w y elide after a homorganic long vowel
2. w y elide after any long vowel before a homorganic vowel
"Homorganic" as in /j i/ and /w u/?
DesEsseintes wrote: 12 Dec 2017 18:09 /ráwa/ rá·a
/utsnéye/ utsné·e
Do these sequences, for example, end up as [aː.a eː.e] or [aːː eːː]? I assume the former, but I felt like double checking. (This was answered below.)
DesEsseintes wrote: 12 Dec 2017 18:09 áwa → á·wa or á·a

[…]

éye → é·ye or é·e
What determines how these sequences are realized, whether they're word-final or not? If so, does that mean that /e a/ are considered homorganic with* /j w/, respectively?

*Not sure what preposition to use here.
DesEsseintes wrote: 12 Dec 2017 18:09 So yay! I always wanted to do overlong vowels. [:)]
[:D]
DesEsseintes wrote: 13 Dec 2017 16:41 *and psst, perhaps even click clusters, but don’t tell anyone...
[:O]
User avatar
DesEsseintes
mongolian
mongolian
Posts: 4331
Joined: 31 Mar 2013 13:16

Re: The Esseintial Núta Blog

Post by DesEsseintes »

shimobaatar wrote: 15 Dec 2017 02:31What does this language consider a "voiced environment"?
Probably this: when neither adjacent segment is ‘ s h or word-final position. As a result, the stops would only be voiceless in clusters with those sounds or in word-final position.
Can these clusters all occur word-initially? Are there restrictions regarding what coda consonants can appear before what onsets?
No. Clusters in onset strictly observe the sonority hierarchy except that clusters starting in s are permitted.
DesEsseintes wrote: 03 Dec 2017 17:25 Unstressed nuclei (and entire syllables) can be devoiced. This is indicated with a dot above on the vowel in nucleus: ȧ ė i u̇.
This makes me wonder something. In the case of tone spread, which syllable is considered stressed? Can an originally stressed syllable be devoiced after the tone spreads? Can an originally unstressed syllable be devoiced after receiving tone?
The syllable carrying the tone is the ‘stressed’ syllable, though an alternative analysis would be that the two syllables involved together carry one falling contour.

As for the devoicing, neither of the scenarios you mention occurs. h elides after assigning falling tone and before accent shift occurs. Accent shift never occurs over an unvoiced onset, so the devoicing environment never applies.
DesEsseintes wrote: 03 Dec 2017 17:25 Exception: if the first two syllables of a word would both devoice according to the above principles, the initial syllable does not devoice. No word can start in two consecutive devoiced syllables.

Can a word otherwise have multiple devoiced syllables? (I don't know if anyone else can see it, but the spacing between this particular line of text and the quote above it comes out very weird on my end.)
Yes, a word can have multiple and consecutive devoiced syllables. Here’s an example:

urȧhkȧhtú·sa
DesEsseintes wrote: 05 Dec 2017 08:41 In this post, we will make a small foray into Núta diachronics to understand the rather unpredictable outcomes of vowel sandhi in Núta.
Very interesting! Was this inspired by anything in particular?
It was inspired by vowel sandhi in Oneida. The vowel shift itself is a much simplified version of the Greek shift.
DesEsseintes wrote: 11 Dec 2017 08:46 In my original conception, I envisaged all Núta words ending in a vowel, but now I’m considering word-final consonants.
I'd vote against word-final coda consonants, but that's just me. It's totally up to you in the end.
See upcoming post.
DesEsseintes wrote: 12 Dec 2017 18:09 1. w y elide after a homorganic long vowel
2. w y elide after any long vowel before a homorganic vowel
"Homorganic" as in /j i/ and /w u/?
Correct.
DesEsseintes wrote: 12 Dec 2017 18:09 áwa → á·wa or á·a

[…]

éye → é·ye or é·e
What determines how these sequences are realized, whether they're word-final or not? If so, does that mean that /e a/ are considered homorganic with* /j w/, respectively?

*Not sure what preposition to use here.
I think the word-final realisations may be lexically determined. At any rate the forms with elision are in free variation with uncontracted forms in those words where elision may occur. It’s only a marginal feature.

/e a/ are not considered homorganic with /j w/ but they do agree with them in “backness”.
Post Reply