Transcribing English in Verdurian

Posted by Philip Newton on 5:32 10/1/01

In reply to: (none)

Is there a semi-standard for transliterating English with Verdurian letters?

Here's a guess at one. The basis for the set of phonemes is a combination of the Shavian alphabet and the description of JBR RP (Justin B Rye's pronunciation) as described on his page on a phonemic transcription key, from where I've also stolen the examples for each sound.

The phonemes are written in ASCII IPA (letter case is significant here). In the examples, the capital letter shows the sound that is being described.

Note that JBR RP is non-rhotic while Shavian is rhotic (that is, 'r' is marked in spelling in such things as 'far', which RP pronounces like 'fah'). I've taken the ASCII IPA from JBR's page, so it'll represent non-rhotic pronunciation, but this doesn't make a difference except for /A:/, which can be either <a> (e.g. <father>) or <ar> (e.g. <far>.), /O:/, which can be either <aw>/<au> (e.g. <saw>) or <or> (e.g. <sore>), and /@/, which can be either schwa/wedge (e.g. <AbUndance>) or schwa + -r (e.g. <fathER>). These sounds have split rows for the two representations.

I've included <r> in the Verdurian representation of sounds where speakers of rhotic accents make a difference; a non-rhotic effect can usually be attained by replacing the <r> with an <a>, e.g. "beard" <bird> --> <biad>, or deleting it (for <ar> and <or>) while possibly adding a long mark.

Also, /@/ in JBR RP stands for both unstressed shwa and a stressed vowel, sometimes represented with IPA wedge (ASCII IPA /V/); Shavian distinguishes between them, but since Verdurian can't, I've left them together. In general, vowels are more problematic to transcribe since English's phonemic vowel inventory is rather broad. (JBR RP also uses /@/ for the last syllable of words such as <abundance>, <button>, <bottle>; many accents speak those without a vowel, the <n> or <l> functions as a vowel (like in Ismaîn!).)

Footnotes:

1. This phoneme often turns into /t/ or /s/ when speakers of languages lacking /T/ learn English. But I think that /D/ is closer.
2. There's a letter representing /N/ in Verdurian script, apparently (Verdurian , Maraille ), but since it's not described on the page on Verdurian phonology, it's probably better to transcribe this sound as <ng> (just as the letter eng is not usually used in e.g. newspaper transcription of other languages in English).
3. English <r> and Verdurian <r> do not sound the same, but it's probably still the best correspondence.
4. Using <u> as a semi-vowel here. Maybe <w> (Verdurian ) or <ŭ> (u-breve; Verdurian ) could also be used here.
5. This was the hardest for me to find a good letter for, since Verdurian doesn't have this sound at all. I don't think using <h> (Verdurian ) is a good idea, since that letter is completely silent in modern Verdurian. <h'> (Verdurian ), which represented /h/ in Cadhinor, dropped out on the way from Cadhinor to Verdurian, so it's not that good, either. Using a representation of /x/, as in (modern) Greek or Russian, doesn't work, since Verdurian doesn't have such a sound (though Cadhinor did, and wrote it ). So <rh> seems the best approximation to English /h/.
6. I think this is the closest approximation to the sound of English <ir>/<ur>/<er>.
7. I considered using a lenge cuzea on those letters to show their phonetic (though not really phonemic) length in English -- that is, instead of . However, I guess plain transcription would probably not include it, though specialised scientific/linguistic material might.

Cheers,
Philip