Subject: Re: [b-hebrew] Translating - some elementary questions
Date: Sun, 01 Oct 2006 23:52:38 +0100
On 01/10/2006 19:24, Bob MacDonald wrote:
... [tricky to tell the computer where the accent is!]
The accented syllable is (at least nearly always) marked by some kind of
accent mark, in an e-text with accents like the one available from WTS.
But not all accent marks actually mark an accent, so a complex algorithm
is needed. I have worked some of this out, and may be able to send you a
program to do this if you contact me off list.
Even so, there are cases where even in an accented text it is impossible
to distinguish between the two varieties of qamats. An example is
XFK:MFH which can be chokhma "wisdom" but can also be chakhema "she was
wise" (Zechariah 9:2). These two forms are in fact distinguished (in BHS
and the WTS text) by the meteg in the latter, but meteg is not used
consistently in such cases. (I had to make this word, 'okhla and `orma
special cases in my algorithm, but there may have been others I missed.)
Also any consistent algorithm will parse B.FT.IYM "houses" as bottim
(qamats in an unstressed closed syllable being "o"), whereas in fact
this is supposed to be pronounced battim.