Wikipedia:Reference desk/Archives/Language/2022 July 24

= July 24 =

Obstruent and /h/
Are there any languages where combinations of obstruent and Voiceless glottal fricative, such as /kh/, /ph/, /sh/ and /th/ are common and occur inside morphemes? Neither English nor my native Finnish has them. In English they occur only in boundaries of morphemes, such as in word lighthouse. 40bus (talk) 20:16, 24 July 2022 (UTC)


 * Ancient Greek had aspirated p,t,k. In some archaic versions of the Greek alphabet, they were treated as consonant clusters, but most would treat them as single aspirated consonants... AnonMoos (talk) 22:42, 24 July 2022 (UTC)
 * In some Semitic languages, a triliteral consonantal root with the first consonant a voiceless stop and the second consonant [h] could theoretically result in such consonant clusters in certain inflected verbal formations. I only looked at Biblical Hebrew, where the only root of that type seems to be k-h-n, referring to concepts of priesthood, but it's not used in verb conjugations of the type which would produce a [kh] consonant cluster.  Such a consonant sequence would always exist across a syllable boundary... AnonMoos (talk) 22:55, 24 July 2022 (UTC)


 * One Biblical Hebrew example would be tiḏ·hār --2A00:A040:19F:D233:11C2:3A52:D8D9:BE3F (talk) 08:36, 27 July 2022 (UTC)
 * Another would be yiṭ·har --82.166.199.42 (talk) 12:43, 27 July 2022 (UTC)
 * Those could be valid for some purposes, but fell outside my search parameters: Basic p,t,k as first root letter, h as second root letter. AnonMoos (talk) 22:36, 27 July 2022 (UTC)


 * It's possible in Arabic, not only in verbal formations – as in أظهر ʾaẓhara 'show, exhibit' – but in general given how triconsonantal roots work: فقه 'fiqh', أزهار ʾazhār 'flowers'. But it's rare, given that the letter ه [h] is an uncommon radical: although it is one of the most frequent letters in Arabic texts, it mainly occurs in pronouns and pronominal suffixes. --Theurgist (talk) 23:52, 24 July 2022 (UTC)
 * Hindi has it, and I believe so do a bunch of other Indic languages. The word for "India" in Hindi starts with one. Mathglot (talk) 20:41, 25 July 2022 (UTC)
 * I really have only very secondary knowledge of these things, but that has never stopped me from waxing. :) So here goes: All South Asian languages (with a few exceptions) have them. You can hear it in the pronunciation when you travel from Iran or Afghanistan (where it is absent) into Pakistan or India. In fact, historically, when the Muslims conquering armies eventually settled in India, and Urdu arose, they had to invent new forms to accommodate the perso-arabic k (kaaf, ک) with "h" or the same with b (bey, ب)
 * ک + ھ = کھ     or with b (bey)   ب +ھ = بھ
 * I would defer to or  here.  Fowler&amp;fowler  «Talk»  21:14, 25 July 2022 (UTC)
 * What South Asian languages commonly have is aspirated and breathy voiced consonants (where e.g. "ph" is a single phoneme /pʰ/ rather than a sequence of the two phonemes /p/ and /h/). Aspirated (though not breathy voiced) consonants are pretty common cross-linguistically and are also (unexpectedly for many) found in English, see English phonology. However, I don't think that's what the original question was about. I'd be interested to know if there is a language that both has obstruent + h clusters that aren't across syllable boundaries and distinguishes those from aspirated consonants. – Uanfala (talk) 21:47, 25 July 2022 (UTC)


 * Hindi, in addition to having aspirated consonants, also has obstruent + h clusters, although they are either secondary, having arisen after the elision of a short /ə/ (which is still present in the orthography and is enunciated in verse), or occur in borrowings from Arabic. Consider, for example, अपहरण ap'haraṇ /əphərəɳ/, "kidnapping", सतहत्तर sat'hattar /səthətːər/, "seventy-seven", or सुबहान sub'hān /sʊbhaːn/, "praise". Moreover, there seems to be a tendency for obstruent + h clusters to merge with aspirated consonants: e.g. for सुबहान sub'hān /sʊbhaːn/ there is also the alternative form सुभान subhān /sʊbʱaːn/. --2A02:8071:886:2420:51F:EA57:9083:7A85 (talk) 08:33, 26 July 2022 (UTC)


 * Apparently Tagalog tended to borrow Sanskrit words by reanalyzing [Cʰ] as [C'h]. Maybe there could be something there. 惑乱 Wakuran (talk) 13:16, 27 July 2022 (UTC)