Vreleksá

Kiri · Joined: 13 Jun 2009 Posts: 471 Location: Latvia/Italy

One of my conlanging projects is what I call "Future Latvian", which is a hipothesis as to what Latvian could look like in a couple of hundred years. Anyway, it made me think about vowel combinations and their classification.

Contemporary Latvian has a dipthong /ua/ (In Latvian it's alway's the first vowel which is clearly pronounced and second vowel is "shorter" or non-syllabic). If two syllables /ua.jis/ get shortened into one syllable /uais/, the /uai/ part is a definite tripthong, right?

But what about when I have a long vowel instead of a dipthong in the source? If two syllables /a:.jis/ get shortened into one syllable /a:is/, what is the /a:i/ part then?

Just a fun thought:
If I have a word /de.jua.jis/ which gets shortened to /de.juais/ one could arguably say that the /juai/ part is a quadripthong, even though in Latvian /j/ is not considered as a part of dipthongs. The word /jua/ is considered to be composed of a consonant /j/ and dipthong /ua/. Then /jā.jis/ gets shortened to /ja:is/ and I'm a confused little cookie.

eldin raigmore · Posted: Sun Mar 24, 2013 10:19 pm Post subject:

A diphthong happens whenever two pure-vowel-sounds appear together in the same syllable's nucleus. It doesn't matter whether or not the combination is perceived as a single phoneme; what matters is that it be perceived as a single syllable. It also doesn't matter if one of the sounds is perceived as a semivowel instead of as a vowel; vowel-vowel diphthongs, semivowel-vowel diphthongs, and vowel-semivowel diphthongs, are all equally diphthongs.

(Remember that a semivowel is a sound made physiologically just like some vowel, but used in speech as a consonant. Sometimes it may be hard to decide whether it's the last consonant in the onset or the first sound of the nucleus; or whether it's the last sound of the nucleus or the first sound of the coda.)

Likewise a triphthong happens whenever three pure-vowel-sounds appear together in the same syllable's nucleus. It doesn't matter whether or not the combination is perceived as a single phoneme; what matters is that it be perceived as a single syllable. It could be thought of by the language's speakers as three phonemes, or as a single phoneme, or as a diphthong phoneme followed by a vowel phoneme, or s a vowel phoneme followed by a diphthong phoneme. As long as speakers and hearers don't perceive a syllable-boundary between the vowel-sounds, it's still a triphthong. And again, it's alright if one or two of the sounds are seen as semivowels instead of as vowels. I personally would argue that it never happens that a semivowel occurs between two vowels without a syllable boundary happening between the semivowel and one of the vowels; so I don't think there are any VSV triphthongs (V=Vowel, S=Semivowel). But I think VVV, VVS, SVV, VSS, SVS, and SSV triphthongs could all happen, for all I know.

Tetraphthongs have been reported in some natlangs. I am sorry I don't remember a reference for that.

Pentaphthongs and longer polyphthongs have never been reported in any natural language. This doesn't mean that you can't have, for instance, eight vowel-sounds in a row in some grammatical utterance of some natural language. It only means that if that happens there's at least one syllable-boundary perceived as between some two of the vowels.
Again, I'm sorry I don't remember the reference for that.

If you look at UPSID, the UCLA Phonological Segment Inventory Database, you'll see they have no triphthongs or longer polyphthongs included, and also that none of their diphthongs are coded for length. Monophthongs, OTOH, are coded for short and for long and for overshort. UPSID also doesn't distinguish between "rising" diphthongs -- the second sound has the most energy -- "falling" diphthongs -- the first sound has the most energy -- and "even" diphthongs -- both sounds have equal energy.

I think the fact that IPA might allow one to encode sounds not found in UPSID means that they do in fact occur in an occasional natlang (or at least IPA's committee once thought so), but not in as many as 10% of the world's languages (the UPSID database covers about 700 natlangs, and there are probably about 7000.)

eldin raigmore · Posted: Fri May 31, 2013 8:38 pm Post subject:

@Kiri: More to say yet?
_________________
"We're the healthiest horse in the glue factory" - Erskine Bowles, Co-Chairman of the deficit reduction commission