Jonathan Rosenne wrote:
> in Hebrew there are many exceptions to the rule
> and it is quite complicated or even not possible
> to decide algorithmically which letter to use.
This is wrong.
There are almost no exceptions to the rule in Hebrew.
Although, in this day and age, there are huge computer data
bases -- Bar Ilan University's Hebrew department's data base
has 26 million entries --, Jonny never came up with a list,
or an estimated percentage of exceptions.
The rule is clear: final shape in final position,
canonical shape in initial and middle position.
The few exceptions are:
1.) When peh is pronounced /p/ (as opposed to /f/), it is written
as a non-final.
Usually these are loan words and foreign proper nouns: 'pop', 'tip
top', 'stop' 'filip', 'xrop'.
2.) The name of the Egyptian president: Mubarak
<Both of these cases follow the logic of dagesh as a modifier of
pronunciation: one letter is pronounced /b/ or /v/, an other
/k/ or /kh/, a third /p/ or /f/ -- when it is written with
dagesh or without. This dagesh normally occurs at the start
of a syllable, not at the end. So peh with dagesh is normal for
the initial position, it is pronounced as /p/. In the final
one normally has final peh without dagesh, pronounced as /f/.
So, in Modern (!!) Hebrew -- following Yiddish -- one occasionally
uses the initial form at the end or the pronunciation that in
Hebrew words NEVER occur at the end of a syllable let alone word.>
3.) In advertisements of the MishkaN Bank "giving mortages" is
written "mashkaNtaut" with final-nun so that the word looks like
the banks name.
4.) some abbreviations and Israeli voting symbols.
All these case are clearly exceptions to the rule and very rare.
Since Unicode is about scripts (not languages) and in Yiddish
there are more cases with non-final peh (with dagesh) that ONE
letter-shape might well deserve a code point of its own (??);
the other four made it into Unicode only for compatability
with earlier standards.
So let me repeat Otfried Cheong's question:
Is there a minimal pair in Hebrew word differentiated
only by KAF/FINAL KAF?
This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:54 EDT