Could some European languages get phonemic vowel length in future? I don't like that so few languages in Europe have that. Which would cause phonemic long vowels?
8What do you mean “so few languages in Europe”? Lots of European languages distinguish long and short vowels – Faeroese, Norwegian, Swedish, Danish, German, Dutch, English (according to some, see also Tristan’s answer below), Irish, Scottish, Welsh, Latvian, Lithuanian, Finnish, Estonia, Sami languages, Turkish (in limited circumstances), Czech, Slovak, Serbo-Croatian, Bosnian…– Janus Bahs JacquetNov 18, 2022 at 21:11
1Have any European languages got phonemic vowel length in the past? How did that happen? Why might it not happen again?– Robbie GoodwinNov 19, 2022 at 19:05
I suppose you mean short and long vowel length in phonemes. Languages do not "get" things.– LambieNov 21, 2022 at 12:37
It absolutely can happen, and indeed it has done in the past couple of decades in at least one instance!
In his book, English after RP, Geoffrey Lindsey describes the phonetics Standard Southern British English (SSB). This variety occupies a similar sociolinguistic position to Received Pronunciation (RP) but, having diverged substantially from those earlier descriptions he feels warrants its own name.
This variety undoubtedly has vowel phonemes which are distinguished (almost) exclusively by quantity - i.e. it has phonemic vowel length. This is something RP did not have, and the new long vowels do not align with the tradition "long vowels" in English phonology which have all diphthongised.
The SSB long vowels mostly come about as a result of non-rhoticity and largely descend from monophthongisation of the RP triphthongs & centring diphthongs.
We see the following phonemic pairs distinguished (almost) exclusively by quantity (e.g. RP [i:] and [u:] which were the only "long" vowels that hadn't diphthongised at that stage are now [ɪi̯] & [ʊʉ̯]):
- KIT ~ NEAR : [ɪ] ~ [ɪ:]
- DRESS ~ SQUARE, LAYER : [ɛ] ~ [ɛ:]
- comMA ~ NURSE, LOWER : [ə] ~ [ə:]
Note that when a NEAR, LAYER, or LOWER vowel is prosodically stressed (e.g. when said in isolation) they are pronounced bisyllabicly as [ɪjə], [ejə], & [əwə] respectively. The SQUARE, NURSE, and NORTH vowels are still long monosyllables when they receive prosodic stress.
Some minimal pairs are (note that minimal pairs involving comMA are difficult because of its restriction to unstressed position):
- KIT ~ NEAR : bid ~ beard
- DRESS ~ SQUARE : bed ~ bared
- DRESS ~ LAYER : led ~ layered
- SQUARE ~ LAYER : lair ~ layer (pronounced identically except under prosodic stress)
So SSB has acquired vowel length in the past couple of decades since the last overviews of RP.
@jk-ReinstateMonica I've added some minimal pairs. Also realised I was wrong with LOT ~ FORCE/NORTH/THOUGHT & LOWER, as those are [ɔ] & [o:]– TristanNov 18, 2022 at 16:44
1Australian English also has a vowel length contrast Nov 18, 2022 at 22:48
Hm. I never pronounce lair and layer identically. I guess I don't have an SSB accent after all. Nov 20, 2022 at 0:32
In turn of the 21st century 'SSBE'--which is a term used specifically to refer to RP without the same silly, incorrect connotations that 'RP' has--there was already a well-recognised length-based phonemic contrast between schwa/COMMA and NURSE. Nov 21, 2022 at 3:25
1@Araucaria-him sorry, I meant quantity– TristanNov 21, 2022 at 9:46
Most European language with phonemic vowel length have this as an inherited feature, but some languages have lost this feature during their development, e.g. Latin has phonemic vowel length, but Spanish hasn't.
Acquisition of phonemic vowel length is possible due to several mechanisms like
- heavy borrowing from other languages with phonemic vowel length
- compensatory lengthening: A post-vocalic consonant is lost, but the preceding vowel becomes lengthened
however, we cannot predict that some language will acquire phonemic vowel length in the foreseeable future. Evolution is unpredictable, after all.
1A third process is through intervocalic deletion of consonants. Western Lombard for example did it; compare ⟨andà⟩ "to go" (infinitive) with /a/ vs. ⟨andaa⟩ "gone" (feminine participle) with /a:/. The later was clearly created through consonant deletion, as the consonant still exists in other Romance languages (e.g. Spanish andada, Italian andata).– lvxferreNov 22, 2022 at 23:29
An example in American English (west coast dialects) comes from reduction of Vr and simplification of rr sequences. This gives rise to minimal pairs: [tɛɹ] "tear (rip, not cry)" vs [tɛ:ɹ] "terror"; [hɔɹ] "whore" vs. [hɔ:ɹ] "horror", [bɛɹ] "bear (n,v), bare"; [bɛ:ɹ] "bearer (v→n)", [fɪɹ] "fear", [fɪ:ɹ] "fearer" and [skaɹ] "scar (n,v)", [ska:ɹ] "scarrer" (one who scars).
1Hm not sure that's how I'd transcribe my West Coast AmE pronunciations. Is this specific to California, maybe? Nov 20, 2022 at 1:14
Seattle: I don't know how general this is.– user6726Nov 20, 2022 at 1:46
3I would have thought these were all examples of a long consonant: [tɛɹ] vs [tɛɹ:], etc, since they all involve the loss of an unstressed vowel between two "r"s.– chepnerNov 20, 2022 at 1:50
That is probably the historical precursor to vowel lengthening.– user6726Nov 20, 2022 at 2:29
@user6726 That's where I'm from, not sure I've heard this Nov 20, 2022 at 20:09