Shavian-info
commented
2 years ago
- arowana — ‹𐑪› or ‹𐑭›?
This should be 𐑭. I missed this in reviewing the suggested spelling.
- Audley — IPA field says it should end with the happy vowel but the only source I know that gives the pronunciation of it (CUBE) equates happy with FLEECE;
This should be the HAPPY vowel. CUBE is not a reliable source on this point as it does not include the HAPPY vowel in any spellings. The authors (with good reason) argue that the HAPPY vowel is a fiction from a time when the pronunciation was transitioning from /ɪ/ to /iː/. Their argument is a good one, but I am not yet ready to take this leap with the ReadLex, on the basis that the final -𐑦/-𐑰 distinction still helps with determining stress. But who knows, maybe one day...
- casuist casuistic casuistical casuistries casuistry casuists — I can't figure which variant is right for RP, but IPA is certainly wrong;
I think the IPA might be correct and the Shavian should be updated. Both /ˈkæz.ju.ɪst/ and /ˈkæ.ʒu.ɪst/ are given for RP in the Cambridge Pronouncing Dictionary, in that order. However, as the US pronunciation is only given as /ˈkæ.ʒu.ɪst/, following the spelling principles used for the ReadLex that pronunciation should be favoured. So the Shavian should be 𐑒𐑨𐑠𐑫𐑦𐑕𐑑.
- countess countesses deaconess deaconesses — Shavian with ‹𐑧› /e/ seems right, but apparently pronunciation with /ə/ or even /ɪ/ is possible so maybe it should be ‹𐑩› /I/ after all?
I have gone with the stronger pronunciation of '-ess' where it is an option in natural speech. Read suggested all -ess should be -𐑩𐑕, but that would give us the less usual 𐑐𐑮𐑦𐑯𐑕𐑩𐑕 for 'princess', and create more homophones like 'adulteress' and 'adulterous'. The -𐑩𐑕 version is used where -𐑧𐑕 never appears in natural speech, e.g. 'actress'.
- employé — I think this should be the same as ‹employee› in English, but if it tries to approximate French then it has wrong vowels and is missing ‹𐑘› (cf. employés);
I followed the Cambridge Pronouncing Dictionary for this one which gives /ɑ̃ːm.plɔɪˈeɪ/ or /ɔ̃ːm.plɔɪˈeɪ/. Frenchified pronunciations are notoriously difficult. I could go with 𐑭𐑥𐑐𐑤𐑶𐑱, maybe.
- jukka — ‹𐑭› or ‹𐑩›?
It's a Finnish first name and not naturalised into English, so I went with a more Finnish pronunciation. I should really move it to the Addendum.dict.
- spiral spiralled spiralling spirally spirals spirals — IPA has one ‹ə› too much in comparison with Shavian, but it's a possible pronunciation too;
I think we should settle on /ˈspaɪ.rəl/, even if Cambridge gives /ˈspaɪə.rəl/. This is one of the very few -aɪ(ə)r- words where it doesn't mark the first schwa as optional in RP which I think might be an error.
- Tremens — if it's ‹delirium tremens› then it's the same as with ‘countess’, but it's marked as a proper name and I couldn't find what it refers to;
Cambridge gives /dɪˌlɪr.i.əmˈtriː.menz/, /-ˈtrem.enz/, /-ənz/. I went with the first listed which possibly reflects that this is a Latin term and not fully naturalised. But yes, it should be NN1.
- Verma — Shavian is probably missing the vowel at the end, but I couldn't find how it should be pronounced.
I can't remember where this came from, but yes it should end in a schwa.
mwgamera
I did some systematic check trying to verify if Shavian spellings can be generated from the IPA field. I skipped entries where IPA is missing, abbreviations, and acronyms. I ignored entries which differed only by having ‹𐑦𐑩› (or ‹𐑰𐑩›) where I expected ‹𐑾›. I also ignored these where I weren't sure how to fix the discrepancy, i.e.:
All the others either looked correct, or I fixed them below. Changes include some obvious errors that should be uncontroversial to fix, but also less obvious changes like those in punctuation to match the spelling (spaces vs hyphens), or a class of words with ‹eəRˈr› which is an impossible sequence that would suggest ‹𐑺𐑮› spelling. I also corrected some ‹r› vs ‹R› which don't influence Shavian spelling at all. But a careful review is needed in any case. Thanks.