A few confusing alifs
Neeraj Mathur
apologise for the repetition; I also regret that, in this intellectual
wilderness of suburban Toronto, I have little access to any real
libraries or any decent books and so I would greatly appreciate your
kind satisfaction of my curiosity.
I had a few questions about some odd alifs in Arabic:
1) Dagger alif - as I understand it, this is due to some differences in
dialect? Is clarification possible as to why it exists and what
pronunciation variants are to be inferred from its history?
2) Alif Maqsura - this completely confuses me. What is its history? Did
it reflect a different phoneme at some point, and if so, why is it
present only at the end of words? Why has its spelling been fixed in
certain words only? What is the history of the dots of ya and of this
symbol?
3) Final alifs - were they at some point pronounced long, and a later
cross-the-board sound change shortened them?
4) Quality of alifs - is the distribution of front vs. back formants of
textbook or broadcast MSA of ancient pedigree? Why do languages that
have borrowed Arabic words (particularly Persian and Indo-Aryan tongues)
pronounce all long alifs with a back vowel - is this a spelling
pronunciation of sorts?
5) la:kin 'but' - why does this word have an alif in Arabic, but a ya in
Persian (representing an older /e:/ in Farsi, as preserved in Urdu)? Is
it in origin something to do with maqsura, were there two separate words
in Arabic with some regional distribution or something, or have I got
the direction of borrowing messed up?
6) Not an alif, but ta marbuta: what is its history? What phonemic and
phonetic values did it once represent, or was it created for the sole
purpose of giving graphic unity to the feminine morpheme? Why does it
have such widely differing pronunciations in Arabic [a], Farsi [e], and
Urdu [A:]?
In general, what era's historical spoken reality is represented by
written Arabic and fossilized in the most formal registers?
Thanks again for your answers!
Neeraj Mathur
Yusuf B Gursey
Neeraj Mathur wrote:
> I'm sure that these questions have been discussed here before, and I
> apologise for the repetition; I also regret that, in this intellectual
> wilderness of suburban Toronto, I have little access to any real
> libraries or any decent books and so I would greatly appreciate your
> kind satisfaction of my curiosity.
>
> I had a few questions about some odd alifs in Arabic:
>
> 1) Dagger alif - as I understand it, this is due to some differences in
> dialect? Is clarification possible as to why it exists and what
> pronunciation variants are to be inferred from its history?
mainly because use of non-final alifs for /a:/ came late. a reform
attributed to the caliph `abdulmalik and his governor in `iraq Hajja:j
introduced among other things regular use of alif for medial /a:/.
he also rewrote the Qur'an in the new orthography. but some
relgious words there were scrupples, for superstitious reasons,
to tinker with like alla:h and al-raHma:n . sometimes the rewriting
was irregular, so sometimes a word appears with a dagger alif in the
Qur'an sometimes regular. sometimes differnet scribal developed, so
a Qur'an written say, in Turkey has a word with alif and others don't.
sometiems different traditions of recitation had developed, like in'
the opening chaptewr most read ma:lik "Master", but most in
N. Africa (excl.Egypt) read malik "King" (both are considered
canonical),
so most writing traditions of the Qur'an represent this word without
a proper alif, in deference to the alternative reading.
finally, alif was ommitted in common certain particles, so it is now
represented by a dagger alif
>
> 2) Alif Maqsura - this completely confuses me. What is its history? Did
> it reflect a different phoneme at some point, and if so, why is it
> present only at the end of words? Why has its spelling been fixed in
> certain words only? What is the history of the dots of ya and of this
> symbol?
most archaic arabic dialects either had */ay/ > e: (later > a:)
or /a:/ > /ai/ > e: (later > a:) i.e. the /ai/ is either etymological
or secondary. for words of more than two syllables with final a: the
development /a:/ > /ai/ was usual. in later classical arabic the
distinction was dropped and it came to be pronounced -a: ,
but in reciting the Qur'an carefully one uses [e:] and [a:] for the
rest.
the use of ya' with dots to distinguish it from alif maqsura is modern.
there were some who used the dotted form to represent alif maqsura,
but his reverse convention did not gain popularity.
>
> 3) Final alifs - were they at some point pronounced long, and a later
> cross-the-board sound change shortened them?
in reciting arabic in the classical style, one preserves all long
vowels
and all non-pausal short vowels. in the more relaxed recitation for
secular purposes pauses are more frequent and final long vowels tend to
get shortened. this is influenced by the colloquial where all short
vowels (sporadic exceptions in some conservative colloquials) are
elided and final long vowels are shortened.
>
> 4) Quality of alifs - is the distribution of front vs. back formants of
> textbook or broadcast MSA of ancient pedigree? Why do languages that
postvelars and emphatics produce back formants, and this is ancient.
so is the frequent, but not found in all dialects, back formants
associated with /r/. also archaic is the back /a:/ in alla:h (along
with the
emphatic l). but broadcast arabic also may introduce back or front
formants from foreign languages (french, english, turkish etc.) and let
colloquial habits slip in (of course, not all colloquial habits are
inovations, some may be archaicisms
> have borrowed Arabic words (particularly Persian and Indo-Aryan tongues)
> pronounce all long alifs with a back vowel - is this a spelling
> pronunciation of sorts?
because these are different languages and /a:/ was pronounced
consistently furhter back in these. in ottoman turkish (but not central
asian turkic) this was adopted as a spelling convention for native /a/
(in central asian turkic alif may represent either /a/ or /a"/ since
the uighur script represented both with sogdian script (of aramaic
script origin) alif.
>
> 5) la:kin 'but' - why does this word have an alif in Arabic, but a ya in
> Persian (representing an older /e:/ in Farsi, as preserved in Urdu)? Is
> it in origin something to do with maqsura, were there two separate words
> in Arabic with some regional distribution or something, or have I got
> the direction of borrowing messed up?
in classical and modern arabic la:kin is usually written with the
dagger alif. the orthogrpahy /lykn/ preserved in persian represents a
strongly imalizing old arabic dialect. before the orthography was
standardized
medial /a:/ represented by ya' was not unusual. this represented a
highly fronted /a:/. a few examples remain in the Qur'an. in the
non-canonical yemeni codicies this is quite common. later it was
decided that this pronounciation was inelegant and cased confusion with
/i:/ and medial /a:/ ceased to be represented with y .
>
> 6) Not an alif, but ta marbuta: what is its history? What phonemic and
> phonetic values did it once represent, or was it created for the sole
> purpose of giving graphic unity to the feminine morpheme? Why does it
early in arabic (except for a few old dialects) feminine -t came to be
pronounced in pause as [h] (probably t > th > h) and later dropped
altogether in pause. some old manuscripts use t in context. but on the
whole pausal forms determine the spelling. so with the reform of the
orthography a hybrid was invented. classical, solemn, religious
pronounciation has [h] in pause while the more usual relaxed broadcast
pronounciation is zero in pause.
> have such widely differing pronunciations in Arabic [a], Farsi [e], and
> Urdu [A:]?
because they are different language. Farsi use of [e] is recent,
particularly if close [e] . one leading iranist I had an e-mail
exchange with met close pronounciation, as exemplified by the use of
kasra in Aryanpoor's dictionary, with disaproval as low register. at
least the use of open [e] is better.
in ottoman turkish it is read as /a/ (back vowel) with emphatics and
postvelars, otherwise /e/ (open, the front companion of /a/) and now
romanized as a or e depending on the previous circumstances.
>
> In general, what era's historical spoken reality is represented by
> written Arabic and fossilized in the most formal registers?
>
probably a somewhat formal and conservative register of the 2nd or 3rd
cent. AH, that tried ot mimick earlier pronounciations.
> Thanks again for your answers!
>
I'll probably follow up to myself, this is a vast subject.
> Neeraj Mathur
klei...@astound.net
Yusuf B Gursey wrote:
> Neeraj Mathur wrote:
>> 6) Not an alif, but ta marbuta: what is its history? What phonemic and
>> phonetic values did it once represent, or was it created for the sole
>> purpose of giving graphic unity to the feminine morpheme?
An interesting example which I have not seen cited anywhere occurs in
the 98th surat of the Qur'an. The first seven verses end in ta marbuta
and the eighth in a 3rd person masculine singular suffix -hu. Making
some reasonable assumptions about the "rhyme" used in this short surat
I conclude that ta marbuta must have been pronounced the same as, or
very nearly the same as, ta marbuta.
This might constitute proof about the pronunciation of ta marbuta as
opposed to following grammatical tradition and making deduction from
the script.
Yusuf B Gursey
it has been mentioned in some books that such a conclusion can be drawn
from ancient poetry. the articulation survives in solemn Qur'an
recitation.
Yusuf B Gursey
Yusuf B Gursey wrote:
> Neeraj Mathur wrote:
> >
> > 3) Final alifs - were they at some point pronounced long, and a later
> > cross-the-board sound change shortened them?
>
>
> in reciting arabic in the classical style, one preserves all long
> vowels
> and all non-pausal short vowels. in the more relaxed recitation for
> secular purposes pauses are more frequent and final long vowels tend to
> get shortened. this is influenced by the colloquial where all short
> vowels (sporadic exceptions in some conservative colloquials) are
> elided and final long vowels are shortened.
as for the alif of the accusative, this represents an old rule of pause
where -an (under conditions like if not attached to ta' marbuTa)
became -a: in pause. apparently, this option is still observed in the
standard arabic of Oman.
in older poetry the pausal rule was lengthening final vowels, and this
reflected an archaic state of affairs. by the time of the Qur'an, the
rule was to elide short vowels, except this accusative.
Peter T. Daniels
> I conclude that ta marbuta must have been pronounced the same as, or
> very nearly the same as, ta marbuta.
I should hope so.
--
Peter T. Daniels gram...@att.net
Peter T. Daniels
>
> I'm sure that these questions have been discussed here before, and I
> apologise for the repetition; I also regret that, in this intellectual
> wilderness of suburban Toronto, I have little access to any real
> libraries or any decent books and so I would greatly appreciate your
> kind satisfaction of my curiosity.
The University of Toronto has one of the great libraries of North
America. Your Oxford credentials will surely get you a visitor's pass.
York's is no slouch, either.
Though you should probably not mention Missisauga while you're there ...
look up my friend Doug Frayne (Assyriologist) -- you can trade
Missisauga stories.
I take it you haven't acquired either Wright's grammar (Cambridge) or
(Rodgers's translation of) Fischer's grammar (Harrassowitz, then Yale)?
Both are really essential, and will give you the details that Yusuf
frameworked for you.
klei...@astound.net
Peter T. Daniels wrote:
> klei...@astound.net wrote:
>
> > I conclude that ta marbuta must have been pronounced the same as, or
> > very nearly the same as, ta marbuta.
>
> I should hope so.
> --
There was a ghost of a chance that the ta marbuta was a graphic variant
of a final ta. It is good to have that chance eliminated by evidence
rather than by theory. I gather that the example I mentioned can be
matched by other similar examples. The more evidence the better.
Peter T. Daniels
I think you continue to overlook the thinko that makes both your
statement tautologous and the original intent unrecoverable.