Module talk:User:Erutuon/en-pronunciation

Ping

@Mr KEBAB: I'm finally coming back to the idea of converting between transcription systems. So far, this module can convert Wells's RP transcription system (from Lexical set on Wikipedia) to Lindsey's SSB system in a strictly literal way. But I don't exactly have a plan at this point. — Eru·tuon 05:39, 14 December 2017 (UTC)

@Erutuon That's good news. Thanks for informing me. Mr KEBAB (talk) 14:34, 14 December 2017 (UTC)

@Erutuon But I don't quite understand why you included smoothing. Surely, the phonemic structure is still /aɪə/ etc. regardless of the pronunciation? Mr KEBAB (talk) 14:55, 14 December 2017 (UTC)

@Mr KEBAB: I don't know, but I suppose it must be in the case of flower [flaː] (SSB) or [flɑə] (RP) because there's no */aː/ or */ɑə/ phoneme listed. If it is a merely phonetic process, the hypothetical future template could show a phonetically smoothed transcription (in the same way that phonetic transcriptions with aspiration, flapping, and other allophonic features are given in some entries). So, for instance, the entry for layer could show /lɛjə/, (smoothing) [lɛː]. But it's confusing because in this case /ɛː/ is also a phoneme. — Eru·tuon 20:29, 14 December 2017 (UTC)

@Erutuon: Ok. I think I'm ambivalent in this case. If I had to choose though, I'd stick to phonemic transcription and wouldn't transcribe smoothing except in cases in which it is pretty much universal (mayor /mɛə/ [but note the spelling pronunciation /ˈmeɪə/ which is becoming increasingly common] or theory /ˈθɪəri/). But those are phonemic changes caused by the phonetic process of smoothing.

There's also the issue of frequency, that is which words are more prone to smoothing and which ones aren't. If the phenomenon is unpredictable, there may be no point in considering it in the context of this module. Mr KEBAB (talk) 23:26, 14 December 2017 (UTC)

I should probably get more information on the conditions under which smoothing occurs (phonological, morphological, sociological, or whatever). My impression is that it can happen anywhere inside a word, because I haven't read about any restrictions, but that may be wrong, and I don't speak this dialect anyway. I think it would be helpful to show smoothing. It's not totally obvious to an American like me, and perhaps to others who do not speak the dialects that have it (or don't speak English at all). But at this point I'm just figuring out how to program it with Lua. — Eru·tuon 04:44, 15 December 2017 (UTC)

@Erutuon See [1], for example.

In the case of our, /ɑr/ is a form directly influenced by smoothing, and it could be argued that the underlying phonemic structure is /aʊr/. If you use /ɑr/, you do have smoothing in your accent, it's just rare. Mr KEBAB (talk) 09:31, 15 December 2017 (UTC)

@Mr KEBAB: Smoothing in our happened ages ago. It's no longer productive. I think that word has to have two phonemic representations – /ɑɹ/ and /awɹ/ (in an ad-hoc transcription). I can't smooth any arbitrary word with that phoneme, like flower to (a rough equivalent of) far, the way a RP or SSB speaker can. It might be possible with other vowels plus /ɹ/ – lower /loəɹ/ → /lor/, seer /siəɹ/ → /siɹ/ – but that process is different in that the first vowel sound does not change its phonemic identity. And rawer /ɹɑəɹ/, if it smoothed, might smooth to rawr /ɹɑɹ/. 😁

I might've seen that post before, but not really read it. It does list situations in which smoothing does not apply, which is helpful: the schwa is a separate morpheme, the word is felt to be foreign or is uncommon. (I thought the example of lowered being smoothed (by John C. Wells) in Geoff Lindsey's post might be a counterexample, but the morpheme -er is probably not particularly obvious in the verb to lower as opposed to the adjective low-er.) This suggests that it might be best to have some way to disable smoothing in the hypothetical future template, or to only manually enable it. — Eru·tuon 01:46, 16 December 2017 (UTC)

@Erutuon I'm aware of that (hence the phonemic slashes in my transcription).

If an RP speaker was listening to you, he'd understand the structure of your [lor, sir] as his [lɔː, sɪə] (I'd use brackets for optional, rather than phonemically established smoothing). The second one is an RP-like smoothing, the first one is not.

Yeah. Sometimes it may even confuse native speakers. I remember reading somewhere that a participant of a BBC program pronounced slowered film as [ˈslɜːd ˈfɪlm]. Another participant asked "what is a slurred film?" I don't know whether he was a Brit though. 06:25, 16 December 2017 (UTC)