Language desk
< August 13	<< Jul \| August \| Sep >>	Current desk >

Welcome to the Wikipedia Language Reference Desk Archives
The page you are currently viewing is a transcluded archive page. While you can leave answers for any questions shown below, please ask new questions on one of the current reference desk pages.

August 14

A couple of phonotactics questions

A couple of phonotactics questions:

Is there any language in Europe which allows sounds /j/, /ʋ/ and /w/ in coda position?
Is there any language in Europe which allows a nasal to be followed by heterorganic plosive, i.e. allowing other nasal+plosive combinations than /mp mb nt nd ŋk ŋg/
Is there any language in Europe which allows a C.V split in syllable boundary?
Is there any language in Europe which allows two or more geminates to occur consecutively without vowels in between, e.g. /tːsː/?

--40bus (talk) 17:57, 14 August 2022 (UTC)[reply]

I think Swedish allows /j/ in final positions, although I'm not an expert in phonotactics. 惑乱 Wakuran (talk) 18:32, 14 August 2022 (UTC)[reply]

1: Depends. What's the difference between those semivowels and the corresponding vowels in a diphthong? The latter happens, unless it's analysed as the former.

2: I'm not aware of any, but could be missing an example.

3: Depends. Does it count if a non-phonemic glottal stop gets inserted? Or if the consonant is ambisyllabic?

4: I'm not so familiar with languages with geminates.

PiusImpavidus (talk) 19:12, 14 August 2022 (UTC)[reply]

1 I was thinking languages where central approximant which is not part diphthong can occur in coda. --40bus (talk) 19:30, 14 August 2022 (UTC)[reply]

Dutch has the /ɛi̯/ (<ei> or <ij>), /œy̯/ (<ui>) and /ɔu̯/ or /ʌu̯/ (<ou> or <au>; most people pronounce both as either the first or second, but for some people there's a difference) diphthongs, which are indisputably phonemic, but the situation for /iu̯/ (<ieuw>), /yu̯/ (<uw>), /ui̯/ (<oei>), /eːu̯/ (<eeuw>), /ɔi̯/ (<oi>), /oːi̯/ (<ooi>), /ɑi̯/ (<ai>) and /aːi̯/ (<aai>) is not so clear. Those ending in /u̯/ could also be analysed as a monophthong followed by a /w/, those ending in /i̯/ could also be analysed as a monophthong followed by /j/ (I think that makes more sense). Most varieties of Dutch don't use /w/ outside these might-be-diphthongs, but use /ʋ/ or something similar instead, which is widely used in onset. /j/ is widely used in onset in all varieties of Dutch. Note that I find the description of Standard Dutch in our article overly narrow. It's so detailed that you could pinpoint this pronunciation to a particular city, if there were any city where people actually spoke Standard Dutch. PiusImpavidus (talk) 09:47, 15 August 2022 (UTC)[reply]

2. German Amt, Imker. –Austronesier (talk) 19:46, 14 August 2022 (UTC)[reply]

As for your second question, even English is a good example: input, inbred, dreamt, empty - when pronounced /emtɪ/, scammed - pronounced /skæmd/, Addington - pronounced /ædɪŋtən/, Farringdon - pronounced /færɪŋdən/, Wellingborough - pronounced /wɛlɪŋbərə/, fryingpan - pronounced /fraɪ.ɪŋpæn/, and the like. HOTmag (talk) 20:05, 14 August 2022 (UTC)[reply]

Canberra. -- Jack of Oz ^{[pleasantries]} 00:47, 15 August 2022 (UTC)[reply]

And the other capial city, just next door: Wellington - pronounced: /'wɛlɪŋtən/. As for the first proper name you've mentioned - with "nb", there is another proper name - with the same consonant cluster: Allenby. HOTmag (talk) 03:47, 15 August 2022 (UTC)[reply]

1: In Swedish and Hungarian the answer is unambiguously yes, given the /j/ can occur in coda after a consonant (see Swedish phonology and Hungarian phonology). In many other languages (e.g. Slavic) the answer is usually also yes, because normally the sound in question is analyzed as a phonological consonant, rather than one forming a diphthong with the preceding vowel.

2: Yes, as demonstrated, although this is rare and usually is across a morpheme boundary, whether transparent or only historical (check etymologies of German Imker, Amt and fremd).

3: I think that by the definition of a syllable, phonetically the boundary is -V.CV-, although phonologically other views might be possible. The Arrernte language (of Australia) is said to have no syllable onsets at all, so every boundary must be -VC.V- instead.

4: I highly doubt it, that's not how geminates work. --Theurgist (talk) 01:04, 15 August 2022 (UTC)[reply]

normally the sound in question is analyzed as a phonological consonant, rather than one forming a diphthong with the preceding vowel Did you mean the opposite? Nardog (talk) 04:16, 15 August 2022 (UTC)[reply]

No. Russian phonology says: "Phonological descriptions of /j/ may also classify it as a consonant even in the coda. In such descriptions, Russian has no diphthongs." Phonology pages on other Slavic languages mention diphthongs only in different contexts, or not at all. --Theurgist (talk) 08:36, 15 August 2022 (UTC)[reply]

I must have been sleeping last night, as I missed examples of (2) in my native language: imker, vreemd. Dutch ambt kept, in contrast to its German cognate, its (devoiced) b: /ɑmpt/, which may be dropped when another syllable follows: ambtenaar /ˈɑm.tə.naːr/. They are rare and in many of the /mt/ examples, a /p/ gets inserted in speech and in many /np/ examples, the /n/ is assimilated to an /m/, even when those aren't written. Can we find minimal pairs? I mean, English input has /np/, but as the word imput doesn't exist, we can't rule out that input is actually pronounced as imput. PiusImpavidus (talk) 09:47, 15 August 2022 (UTC)[reply]

In English: impatient —inpatient. In Dutch: impassen —inpassen. Both have a stress difference, though, so they are not true minimal pairs. --Lambiam 06:12, 16 August 2022 (UTC)[reply]

Inpassen is a bit special too. As it's a separable verb, the n is rarely adjacent to the p (only if it's the finite verb in a dependent clause, sometimes the infinitive, the present participle and gerund), so it's logical that in spelling the n isn't turned into an m. The many forms that have a clear /n/ separated from /p/ may act as a barrier to prevent assimilation in the rare forms with /np/. I think we must conclude that /np, nb, mt, md/ exist, but are somewhat marginal. I have the impression that /ŋp, ŋb, mk, mg/ are more common. Maybe the gap from /m/ to /ŋ/ etc. is too large to bridge with assimilation. So, in any case, the answer to OP is yes, they exist. PiusImpavidus (talk) 17:58, 16 August 2022 (UTC)[reply]

And also /ŋt/, /ŋd/. HOTmag (talk) 21:00, 16 August 2022 (UTC)[reply]

Is there any language whic allows V.V split word initially (i.e. a syllable boundary between first two phonemes [which are vowels]), e.g. [#i.aC]? --40bus (talk) 20:03, 18 August 2022 (UTC)[reply]

You mean like French "août" [a.'ut] (regional var.) or "Haïti" [a.it'i], or the English interjection "ahOOga!" [a.'ug@]? SamuelRiv (talk) 20:26, 18 August 2022 (UTC)[reply]

Are there any clusters of three or more non-schwa vowels in English? --40bus (talk) 20:40, 18 August 2022 (UTC)[reply]

Your first question regarding two vowels was about "any language", but your following question regarding three vowels was about "English". Thanks to SamuelsRiv's answer to your first question, I can understand why you didn't continue to ask about two vowels in "English", but why didn't you continue to ask the intermediate question about three vowels in "any language"? HOTmag (talk) 21:30, 18 August 2022 (UTC)[reply]

We used to have an article on languages with vowel-only words that included lots of sequences, but it got deleted as trivia. Japanese aoi 'blue', Swahili maua 'flower', etc. But if you mean in Europe, I think there might be some Celtic stuff, Esperanto iea, etc. — kwami (talk) 21:39, 18 August 2022 (UTC)[reply]

American English phonotactics clearly can handle triple vowel conjunctions, since we've gotten used to borrowing from Hawaiian (the latter, some Americans may pronounce (dropping the glottal stop): [hə'waɪ.i.ən] or [ha'wa.i.ɪn]. We also learn how to pronounce all the vowels in school (YouTube). SamuelRiv (talk) 21:50, 18 August 2022 (UTC)[reply]

Please notice that the OP's question was about "three or more non-schwa vowels in English". HOTmag (talk) 22:03, 18 August 2022 (UTC)[reply]

There are also verb conjugates like "buoying" ['bu.i.ɪŋ] and "KOing" [keɪ.oʊ.ɪŋ] (and some speakers may insert a semivowel /j/, but not all), medical or scientific terms like "osteoarthritis' ['ɑːstɪo.ɑɹθr'aɪtəs] or "paleoanthropology", acronyms like "UAE" [ju.eɪ.iː], and some others that I forgot when I took a nap. I'm not saying they're not unusual, but I don't think they're expressly forbidden by English phonotactics in the same way that we forbid plosive+fricative release in the onset. SamuelRiv (talk) 01:49, 19 August 2022 (UTC)[reply]

According to this website, the biblical proper name Geuel (mentioned in Numbers 13 15) is pronounced: /ge.u.el/. HOTmag (talk) 22:26, 18 August 2022 (UTC)[reply]

That may be intended to be the Hebrew pronunciation, rather than an anglicized form. It's often difficult to tell whether the pronunciations of foreign words/names in English dictionaries are actually supposed to be English. And even if it is supposed to be English, those vowels could be analyzed as /gɛj.ʊw.ɛl/, so the question becomes whether diphthongs such as /ɛi/~/ɛj/ count as vowels or as VC sequences. — kwami (talk) 23:11, 18 August 2022 (UTC)[reply]

1. Even if it's intended to be the Hebrew pronunciation, this is not that different from the other examples you've given from Japanese Swahili and Esperanto.

2. This English book, printed in 1845, whose title is: An explanatory and phonographic pronouncing dictionary of the English language, to which is added a vocabulary of Greek Latin Scripture Christian and geographical names with their pronunciation, gives (p. 893) the pronunciation of Geuel as mentioned above - on a table called (p. 887) "A vocabulary of scripture proper names with their pronunciation". Since they are "scripture proper names", while this book - about the English language - does not call their pronunciation a "Hebrew" one, so the pronunciation of Geuel indicated in this book is supposed to be the "correct" (so to speak) standard universal pronunciation - of a biblical proper name which is (according to this website) "popular in Christian religion".

3. As for your suggestion, that "those vowels could be analyzed as /gɛj.ʊw.ɛl/", please notice that this website states: Geuel is pronounced Geh-uw-eh-L. English pronunciation guide: G as in "grin (G.R.IH.N)" ; EH as in "ebb (EH.B)" ; UW as in "two (T.UW)" ; EH as in "ebb (EH.B)" ; L as in "lay (L.EY)". This website does not call this pronunciation - of this biblical proper name - a "Hebrew" pronunciation, so it's supposed to be the "correct" (so to speak) standard universal pronunciation - of a biblical proper name. Reservation: The website also provides some audios - of individuals from all over the world (without any Hebrew audio though), but the five English audios disagree on the pronounciation of Geuel - whereas none of them pronounce it with three consequtive vowels. The only audios that do agree, though, to the pronunciation mentioned above, are the audio in Finnish (being the OP's native language), the one in Hungarian, and the one in Spanish. HOTmag (talk) 08:33, 19 August 2022 (UTC)[reply]

I wonder if yow would count. Or perhaps yowie, depending on how one treats the semi-vowel /w/. ‑‑ Eiríkr Útlendi │^{Tala við mig} 23:17, 18 August 2022 (UTC)[reply]

For #3, "allows a C.V split in syllable boundary", I can think of examples in English: upend and inexpert jump immediately to mind. Is that the kind of thing you were looking for? ‑‑ Eiríkr Útlendi │^{Tala við mig} 22:05, 18 August 2022 (UTC)[reply]