Wednesday, 26 June 2013

A comparison of Kurmanji and Sorani Kurdish

Kurmanji and Sorani, sometimes known as Northern and Southern Kurdish respectively, are the two main Kurdish language groups - although they should not be considered fixed languages, but rather two distinct dialect continuums. This map shows that Kurmanji is the larger group (about 80% of all Kurdish speaker) spoken mainly in Turkey and the border regions of Syria, Iraq and Iran and Sorani is spoken in southern Iraq and Iran. This post will outlay the main differences between the two groups.
The first thing one notices upon seeing Kurmanji and Sorani side-by-side are the different scripts employed by the two groups. Sorani is written in an adapted Perso-Arabic script which originates from reforms under the Ottoman Empire which intended to improve Ottoman Turkish spelling and which is why Sorani Kurdish shows vowels more than Persian or Arabic. Kurmanji most often, i.e. in the Turkish region, is written in the Latin script, like Turkish, but is also written in a Perso-Arabic Script in areas where that script is used by other languages.

An example of the two scripts: 
Kurdistan welatekî Rojhilata Navîn e û welatê Kurdan e

 کوردستان ناوچەیەکی لە رۆژهەڵاتی ناوینه و ناوچەیەکەی کوردانە

Kurdistan is a country in the Middle East and is the country of the Kurds

In terms of phonology, there aren't many differences, although from the above example you can see the words 'navîn' or 'ناوین' which both mean 'middle', but the 'v/و' are pronounced /v/ and /w/ respectively. There are a few other sounds in Sorani which don't exist in Kurmanji, such as the dark l /ɫ/. There are other differences, but I won't go into them since phonology isn't something that particularly interests me.

Beyond superficial differences, the main divergence can be found in the fact that Kurmanji has both cases and gender whereas Sorani does not. Like other Iranian languages such as Zazaki, Pashto and Balochi, Kurmanji has a nominative and oblique case as well as a vocative case. It has a masculine and feminine gender -which surprisingly seems to be much greater in number than the former- and one plural form. Sorani, most likely under the influence of languages such as Persian which also has no gender or case anymore, has lost these aspects.

In Kurmanji there is a difference between 'I' and 'me' ('ez' and 'min' respectively) and this goes for all nouns. In Sorani there is only one form, so 'I/me' is 'min' - thus a sentence like 'I know' would, for a Kurmanji speaker, sound like 'me know' in Sorani. The object in Sorani often becomes an enclitic infix in the verb. Take the phrase 'I see you' for example:

K.: Ez dibînim
Here 'tê' is the oblique form of 'tu' - you.

S.: من) دەتبینم) / (min) datbînim
Here, the object infix '-t-' from 'tu' is placed after the verbal prefix and before the root of the verb.

Another example: you see me

K.: Tu min dibînî

S.: تو) دەمبینی) / (tu) dambînî

Both languages have split ergativity which means that in the past tenses the logical subject becomes the grammatical object and vice versa, although only with transitive verbs. This is fine for Kurmanji with its cases because it can easily say something like 'him I saw' to mean 'he saw me' or 'me he saw' to mean 'I saw him'. However, since Sorani has no cases, this becomes more complex and instead the subject becomes object infix or suffix attached wherever it can be, that includes sometimes the end of the verb, giving it the appearance of being non-ergative. Some examples:

I eat the bread > I ate the bread

K.: Ez nan(î) dixwim > min nan xward
Here 'dixwim' in the present agrees with the subject whereas 'xward' in the past agrees with the object. 'Nan(î)' is the oblique of 'nan' -bread, but in most dialects it doesn't carry the masculine oblique marker 'î'.

S.: من) نان دەخۆم > نانم خوارد) / (min) nān dakhom > nānim khwārd
As can be seen, the logical subject becomes an object infix/suffix and, in this case is stuck onto the logical object. If the phrase were simply 'I eat', this would be attached the end of the verb making: khwārdim

The ezafe construction, which is used to connect nouns to other nouns (like 'of' and ''s' in English) and adjectives to nouns, also differs between the two, largely due to the existence, or not, of gender. In Kurmanji the ezafe is simply added to the end of any noun but depends on the gender and number of the noun. Thus the words 'jin' (f.) -woman, 'mirov' (m.) -man, and 'zarrokan' (pl.) -children take different endings when attached to another word such as in the following examples:
jina mezin - the big woman
mirovê mezin - the big man
zarrokên mezin - the big children

In Sorani there are no genders, but there are two ezafe forms. The simple one is the same for all nouns and is made by adding '-i' to the end of the noun, as in:
ئەفرەتی بچووک / afrati biçûk - small woman
منداڵانی بچووک / mindāɫi biçûk - small children

However, there is another slightly different form which is used in when the linked phrase is modified by the demonstrative (this, that) or the definite article (but not the indefinite). The definite article, which doesn't exist in Kurmanji, is formed by adding '-aká' to the end of the noun, or the end of the end of the attached adjective and as can be seen this carries the stress. The demonstrative adds a stressed '-á' to the end of the noun/adjective. Thus we get:

هۆتێلەکە / hotelaká   - the hotel
هۆتێلە باشەکە / hotela bāshaká   - the good hotel

ئەم هۆتێلە / am hotelá   - this hotel
ئەم هۆتێله باشە / am hotela bāshaká   - this good hotel

With an indefinite noun, the normal '-i' ezafe is used:
هۆتێلێکی باش / hotelèki bāsh   - a good hotel

Finally there is a difference in how the two languages form the passive tense. In Kurmanji the verb 'hatin' -to come is used with the infinitive of the verb. E.g.:
Mirovan hatin kuştin   - the men were killed (lit. the men came to-kill)

In Sorani the passive is formed by adding '-rā' to the present stem of transitive verbs to make a new passive infinitive. From this you can make past and present passives as well as as passive participles. In the present the '-rā' becomes '-re', as in:
bîn - present stem of 'to see' -> bînran - to be seen, bînra - it was seen -> dábînre - it is seen/can be seen.

پیاوەکان کوژران / pyawakán kuzhrān   - the men were killed ('kuzh-' is the present stem of 'kushtin' -to kill).

As can be seen from the differences that I've outlined above, there are enough grammatical differences to warrant Kurmanji and Sorani being considered distinct language groups rather than continuum. Of course Kurmanji speakers and Sorani speakers, as well as most Zaza and other Kurdish language speakers consider themselves ethnic Kurds and if that's how they see themselves, then that's what they are - after all language and ethnicity are not the same thing.


  1. "Biji e..." it's Sorani or Kurmanji?

  2. This really is an outstanding post. Thanks so much!

  3. I'm pretty sure most indo-european languages have more feminine words, or at least european languages do