Why Kurdish Has Two Alphabets
The split between Latin and Arabic script does not follow dialect. It follows the borders drawn across Kurdistan after the First World War.
Before the twentieth century, Kurdish — where it was written at all — was written in a version of the Arabic script, the literary script of the Ottoman and Persian worlds the Kurds lived within. Classical Kurdish poetry, including Ahmad Khani's Mem u Zin of 1692, comes down to us in that script.
The division came with the borders. In 1928 the new Turkish republic replaced the Arabic script with a Latin alphabet. Kurdish intellectuals in Turkey and Syria followed: in 1932 Celadet Ali Bedirxan launched a Latin alphabet for Kurdish in the journal Hawar, built so that Kurds already schooled in Turkish Latin letters could read it. In Iraq and Iran, where Sorani speakers lived under states that kept the Arabic script, the older tradition continued and was refined instead.
The result is that script tracks geography rather than speech. A Kurmanji speaker in Turkey and a Sorani speaker in Iraq can share a great deal of vocabulary and still be unable to read one another's writing.
At a glance
- Kurmanji (northern) — Latin Hawar script, used in Turkey and Syria
- Sorani (central) — Arabic-based script, used in Iraq and Iran
- Sorani is a written standard and an official language of Iraq alongside Arabic
- Script follows political borders, not dialect boundaries
The Hawar Alphabet
Thirty-one letters, introduced by Celadet Ali Bedirxan in 1932 and named after the journal that carried it.
31 letters: the 26 of the basic Latin alphabet plus C-cedilla, E-circumflex, I-circumflex, S-cedilla and U-circumflex. Kurdish uses Q, W and X, which the Turkish alphabet does not.
Celadet Ali Bedirxan set out to build an alphabet in which each sound had exactly one letter. He first sought to work with Tewfiq Wehbi in 1931; when no reply came, he and his brother Kamuran Ali Bedirxan launched their own system in 1932 through the magazine Hawar, published in Damascus.
The design was deliberately practical. Kurds in Turkey were already learning the Latin letters of the new Turkish alphabet, so a system built on the same base could be picked up quickly. Hawar remains the standard for Kurmanji today, in Turkey, Syria and across the European diaspora.
The Central Kurdish Alphabet
An Arabic-derived script reshaped for Kurdish sounds — and, unusually for a Perso-Arabic script, one that writes almost all of its vowels.
34 letters as standardised by the Kurdish Academy, counting the double waw separately. Letters such as pe, che, zhe and gaf came from Persian; others were added specifically for Kurdish sounds.
The Arabic script was not so much adopted for Kurdish as inherited: it was the script of religion, administration and literature across the region. What changed in the twentieth century was that it was systematically adapted. Letters were added for sounds Arabic does not have, and — the crucial break with Arabic and Persian practice — vowels were given letters of their own rather than being left to the reader.
That makes Sorani far easier to read aloud than Arabic or Persian, where short vowels normally go unwritten. One short vowel remains an exception: the sound written i in Latin has no letter of its own and must be inferred from context.
Conversion Table
How the letters of the two main scripts correspond. Where a row has no equivalent, the sound exists in one script and not the other.
Vowels
| Latin (Hawar) | Sorani | Sound (IPA) | Example |
|---|---|---|---|
| a | ا | [aː] | با — wind |
| e | ە | [ɛ] | مەزن — great |
| ê | ێ | [eː] | دێ — village |
| i | — | [ɪ] | not written in Sorani — inferred from context |
| î | ی | [iː] | شین — blue |
| o | ۆ | [oː] | تۆ — you |
| u | و | [u], [ʊ] | کورد — Kurd |
| û | وو | [uː] | گەردوون — cosmos |
Central Kurdish has eight vowels; all but the short i are written with a letter of their own.
Consonants
| Latin (Hawar) | Sorani | Sound (IPA) | Note |
|---|---|---|---|
| b | ب | [b] | |
| c | ج | [dʒ] | |
| ç | چ | [tʃ] | |
| d | د | [d] | |
| f | ف | [f] | |
| g | گ | [ɡ] | |
| h | ھ | [h] | |
| j | ژ | [ʒ] | |
| k | ک | [k] | |
| l | ل | [l] | |
| l (ll) | ڵ | [ɫ] | A velarised, dark L. Written ll in Latin when it must be distinguished from plain l. |
| m | م | [m] | |
| n | ن | [n] | |
| p | پ | [p] | |
| q | ق | [q] | |
| r | ر | [ɾ] | |
| r (rr) | ڕ | [r] | A rolled trill, against the single tap of the plain r. Written rr in Latin when it must be distinguished. |
| s | س | [s] | |
| ş | ش | [ʃ] | |
| t | ت | [t] | |
| v | ڤ | [v] | |
| w | و | [w] | This letter is w when it acts as a consonant and u when it acts as a vowel. |
| x | خ | [x] | |
| y | ی | [j] | This letter is y when it acts as a consonant and long i when it acts as a vowel. |
| z | ز | [z] | |
| — | ع | [ʕ] | A pharyngeal sound found in loanwords. The Hawar alphabet has no letter for it. |
| ḧ | ح | [ħ] | Bedirxan proposed a diacritic form for this Arabic sound, but it is not counted as a letter of the alphabet. |
| ẍ | غ | [ɣ] | Likewise proposed as a diacritic form rather than an official letter. |
Latin forms shown in brackets are conventions rather than official letters of the Hawar alphabet.
Other Scripts Kurdish Has Used
Two more alphabets carried Kurdish through the twentieth century, and a third has been proposed to unify the rest.
Cyrillic
From 1946 · Soviet ArmeniaA 40-letter Cyrillic alphabet designed by Heciye Cindi in 1946 for the Kurdish community of Soviet Armenia. It carried a substantial body of Kurdish publishing and broadcasting, and is still met with among older Kurds of the former Soviet Union.
Armenian script
1921 to 1929 · Armenian SSRFor most of the 1920s, Kurdish in Soviet Armenia was written with a modified Armenian alphabet. It gave way to a Latin system at the end of that decade, which was in turn replaced by Cyrillic.
Yekgirtu
Proposed · Kurdish Academy of LanguageA unified Latin alphabet — yekgirtu means unified — proposed by the Kurdish Academy of Language to serve every Kurdish variety with a single system designed for straightforward digital use. It has seen limited take-up and no official adoption.
What Trips Up New Readers
A handful of rules account for most of the difficulty in moving between the two scripts.
Reading Sorani
- The waw does double duty: u as a vowel, w as a consonant. Doubled, it is the long u.
- The ye behaves the same way: long i as a vowel, y as a consonant.
- The short i is never written. The word for great is read mezin, though no letter marks the i.
Reading Hawar
- The dotted and dotless i are different letters: i is short, i-circumflex is long.
- E and e-circumflex are separate vowels, as are u and u-circumflex.
- C is the sound of English j, and c-cedilla is English ch. J is the s in measure.