THEORY: NATLANGS: Phonology and Phonetics: Tetraphthongs, Triphthongs, Dipht

>What are some examples of diphthongs in which neither vowel is [i] or [u]
>(or [j] or [u])?
I suspect those with [y]/[H] are the next most common? Finnish has, besides
the [y2] Philip mentioned, also [&y] and [2y]. Difthongs with a back
unrounded glide [M]/[M\] also ought to exist, even if none come to my mind
at the moment.

As others have already mentioned, then you have difthongs with "lowered
glides". AusEng has been brought up already; also, the Savonian dialects of
Finnish commonly lower [i]-final, and sometimes also [u]-final difthongs.

Ray brought up Romanian having rising [ea oa] - but it also has [ia ua],
which makes for a contrast between "lowered" and "non-lowered" glides! I
don't know if the description is diachronically accurate, but it's
nevertheless interesting because, judging by AusEng and SavoFin, difthongs
with "lowered" glides seem to be usually dialectal variants of standard high
glides.

Some Khoisan and Polynesian languages also allow various difthongs built
from some two of [a e o].

---

-Briefly returning to difthongs with high elements: as I might have
mentioned before, Finnish [ie y2 uo] are a bit unusual in that they are
always falling and non-centring. That is, to my knowledge, no dialects
render them as [je H2 wo] nor [i@ y@ u@]. The 2ndmost common variants are
[i& iA y& uA] in western 'lects.

---

And now for a new sub-topic. Do any languages have overtly pharyngeal
glides? That is, akin to the vocalization of German final /R/ (which could
be seen as a uvular glide), are there languages where there would be reason
to analyze /?\/ rather than non-syllabic /a/ or /A/??

---

>Supposing there are triphthongs; aren't most of them an already-acceptable
>diphthong with either a [j] (or [i]) or a [w] (or [u]) stuck on ahead as an
>on-glide or behind as an off-glide?
Yes, and usually even so that the two glides are at opposite ends; not the
same one.


>Don't most of the exceptions take one of the forms V[j]V
>or
>V[w]V
>where the Vs denote possibly distinct vowels?
I suspect they'd rather have lowered, rounded/unrounded and/or central
off-glides added. What you suggest is very prone to breaking down into two
syllabes.

Romanian, the most trifthong-rich language I know of, has the following
ones:
stress-central (circumflex?) [eai eau iai iau iei ieu ioi iou oai]
stress-final (rising) [eoa ioa]

Granted, all 2-vowel segments contained within these are allowed
difthongs...


>Are there any which are clearly VVV
>where none of the Vs is a [i] ([j]?) or [u] ([w]?) or [@] (if that's the
>right way to write a center-mid schwa)?
Well, [eoa] abov. It even contrasts with [ioa]!


>ObConLang;
>Whose has the most creative diphthongs?  (Not a contest, really -- just, if
>you have one, or have seen one, you think is weird, or think I'll think is
>weird, let me know.)
uwjge long vowels /a: e: o: u\:/ are phonetically "semi-centring" [a6 eE oO
u\U] which arguably looks weirder than it actually sounds like.


>Does anyone's lang have five or more vowel-qualities and/or vowel-sounds in
>a row -- for instance, a diphthong followed by a triphthong, or a
>triphthong followed by a diphthong, or a vowel in between two diphthongs?
>eldin
That's easy. As far as Finnish goes, the current champion is 13 consecutive
vowels in <riiuuyöaieoioin> [rii.uu.y2.ai.e.oi.oin]. Even if long vowels
only count as one and unstressed [i] is considered a consonant /j/, you
still can get up to 5.
riiuu = dating
yö = night
aie = plan, thought
oioin = straightening tool
...OK, that last part is something of a cheat, but the rest is perfectly
legible, and 9 vowels in a row isn't bad either.

In English, you could similarily construct maybe "bayoueye hour"
[bI.u.aI.aU@] (7 qualities).

John Vertical
PS. Did anyone else keep typoing dif"thing" instead of "-thong" as you
wrote?

Replies