Arabic-L:LING:Buckwalter transliteration query

[Dilworth Parkinson]

------------------------------------------------------------------------
Arabic-L: Thu 19 Nov 2009
Moderator: Dilworth Parkinson <dilworth_parkinson at byu.edu>
[To post messages to the list, send them to arabic-l at byu.edu]
[To unsubscribe, send message from same address you subscribed from to
listserv at byu.edu with first line reading:
            unsubscribe arabic-l                                      ]

-------------------------Directory------------------------------------

1) Subject:Buckwalter transliteration query

-------------------------Messages-----------------------------------
1)
Date: 19 Nov 2009
From:Andrew Freeman <andyf at u.washington.edu>

Subject:Buckwalter transliteration query


Hello all,
MY Question: How do I represent the contrast between the pairs /ay/:/ee/ and /aw/:/oo/ in Arabic and Buckwalter transliteration? 
بَيْت = /bayt/ Sana’a or /be:t/ in Aden.
I cannot finesse this entirely because I am trying to do a close phonetic transcription and label audio sequences for a speech recognizer.
 
                I am working with several transcriptions that are in a style similar to the “Library of Congress” transcription scheme for Semitic languages:
1)      The pharyngealized/emphatics are represented with a dot under the symbol for the phonetic segment.
2)      The ‘cayn’ is written as a super-script ‘c’, except for in  utf8 text contexts where the superscript does not exist, in which case it is simply written as a lower case ‘c’.
3)      These texts represent late phonological processes such as intervocalic voicing of stops and the backing and rounding of high front vowels in pharyngealized contexts  in the transcription.
4)      For all intents and purposes I can transform this transcription into and out of International Phonetic Alphabet without changing any of the information content.
 
I am also working with tools and transcripts that have been trained on the LDC corpora which uses the Buckwalter transcription or even just Arabic script.  For all intents and purposes the Buckwalter transliteration can be transformed into Arabic script (and vice versa) without changing any of the information content.
 
My problem is that I have texts in both transcription styles that contain utterances from more than one dialect.  The dialects in question are the “San’ani” identified Yemeni urban dialect and the “Ta’izzi/Adeni” identified Yemeni dialects.
 
                Why is this a problem?
 
1)      The Northern Yemeni dialects including Sanaani maintain the /ay/ and /aw/ diphthongs in words inherited from Old Arabic.  According to the only all accounts that I have seen (Feghali, 1990, 1991), Adeni produces these sounds as /o:/ and /e:/ which is what happens in most dialects with which I have any familiarity.
 
Maintaining this contrast in IPA is not problem.  Preserving this contrast in Arabic script and Buckwalter transliteration is not so obvious.  I suppose I can introduce the ‘e’ symbol into the Buckwalter transliteration scheme.  However, using ‘o’ is not an option because it is already being used to represent the sukuun.
 
MY Question: How do I represent the contrast between the pairs /ay/:/ee/ and /aw/:/oo/ in Arabic and Buckwalter transliteration?  Any suggestions are ideas?
 
Andy Freeman
(206)225-0386



--------------------------------------------------------------------------
End of Arabic-L:  19 Nov 2009


-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/arabic-l/attachments/20091119/750927a9/attachment.htm>