How can I write in Tatar using the Latin alphabet?

From a post by Alik Gilmullin <gilmy@open.ksu.ras.ru>.


                                /_/ 
                                (0 0) 
========================----oOO--(_)--OOo----===================== 
  Alik Gilmullin                            
  Artificial Intelligence Lab               
  Kazan State University                    
  Kazan, Tatarstan                          
  (gilmy@open.ksu.ras.ru)                   
===================================================================

Dear tatarophiler!

As is well known the Latin alphabet contains 26 letters, but in Tatar Language there are at least 32 phonemes. So, there is problem of precise and unique representation of the Tatar words on our computers. It is necessary for correct understanding, search operations and, finally, for the future of the Tatar Language. We offer this agreement.


 ********************* 
 Tatar e-mail alphabet 
 ********************* 

 ------------------------------ 
 Tatar phonemes       Examples 
 and its average 
 appearance in       (if it's 
 Tatar texts (%)     necessary) 
 ------------------------------- 
  A    0.113 
  E    0.072     *ni (mother) 
  N    0.070 
  E    0.068     *t (dog), b*r (one) 
  I    0.066     bal*k (fish) 
  R    0.064 
  L    0.063 
  K    0.056 
  T    0.050 
  I    0.042     k*tap (book), *ke (two) 
  Y    0.033     ma* (oil) 
  G    0.030 
  M    0.028 
  B    0.028 
  D    0.028 
  S    0.027 
  U    0.025 
  Sh   0.018     ko* (bird), bi* (fife) 
  Z    0.016 
  C    0.015     ki* (evening) 
  P    0.015 
  Ng   0.013     ta* (dawn) 
  U?   0.011     k*l (like), but not "kul"(arm) 
  O?   0.011     k*l (ashes), but not "kol"(slave) 
  O    0.010 
  W    0.007 
  X?   0.006     *at (letter)  May be X=H ? 
  J?   0.005     *ey (summer) 
  F    0.003 
  H    0.003     *em (and) 
  V    0.002     *okzal (railway station) 
  J  ~ 0.001     *urnal (journal)

P.S. The last two letters appear mostly in Russian loan words.
P.P.S. Of course, this is only average texts statistics prepared by our group working on Tatar computer linguistics.

So, we have following problems:

- - Four vowels (E,I,U,O) and at least one consonant (in our case J) used twice.
- - Two phonemes (Sh,Ng) consist of two Latin letters.

There are a few ways to solve vowels problem = = "kol (ashes) - kol (slave) problem":

- - koel - kul (like in Germany)
- - kol' - kol (make use of the sinharmony phenomena of Tatar by sign ')
- - kol - qol (like in Yangalif,1929)
- - kol - kol (no difference, recognition by context)

and of course

- - special drivers with required letters (we have it)
- - UNICODE project (in the near future as we hope)

Next:What are the major(?) publications in Tatarstan? Up:TOC Previous:How do I communicate electronically with people in Tatarstan?