Page 1 of 1

A paper I wrote on Farsi Ezafe

Posted: 17 Dec 2018 00:24
by Shemtov
This essay presents the morphological and morphosyntactical aspects of the Ezafe Construction in Farsi. It will consider Ezafe as a morphological and morphosyntactical phenomenon, instead of as a purely syntactic one, as some linguists has contested. It will also pay special attention to the use of Ezafe with prepositions.

The Farsi language, also known as Persian, belongs to the Iranian sub-brach of Indo-Iranian branch of the Indo-European language family and has been the dominant language in what is now Iran and the entire Central Asian region since five centuries before the Common Era. However, Modern Farsi has undergone typological changes in the past two and half millennia, especially since the conversion of the region to Islam, but the religious landscape has had little impact on certain major typological changes that were happening before the introduction of Islam; Islam’s main impact is that it created a political landscape that allowed certain divergent dialects to become standard over a large area. During the century after the Islamacization of the region, the area was ruled by Arabs, who made Arabic the sole language of education. However, about a hundred years after, a polity from the Fars region, speaking a divergent dialect, established an independent Empire, though they kept Islam as the official religion. The older forms of the language of this period had great influence on the Turkic languages, including Turkish, as it was the Persian Empire that introduced the Turkic tribes to Islam. They also conquered east to India, allowing the cousin-languages in the Indo-Aryan branch to become Persianized, and borrow many lexemes.

Modern Persian, for political reasons, has three standerdized versions: Farsi, spoken by 70 million total speakers in Iran; Dari, with 33 Million speakers in Afghanistan; and Tajik with 7 million speakers in Tajikistan. There are also a few non-standard versions, such as Tat in the Caucasus, and Bukharian, spoken by Jews originally from Uzbekistan, though most of its speakers now live in the borough of Queens in New York City. The main differences between the standard versions are in the orthography: Farsi and Dari, while both using the Arabic alphabet, represent vowels differently, while Tajik is written in the Cyrillic script. In addition, the Realization of the vowels are different in Farsi and Dari, and Tajik has many morphosyntactic elements borrowed from the Turkic language Uzbek. This essay will focus on Farsi.

Now, we must consider Ezafe. Ezafe is /e/ or, after a vowel /je/ (as Farsi regularly uses /j/ to break hiatus), that is suffixed to a noun to show that it is owned by the following noun, or to an attributive adjective. The name “Ezafe” comes from a similar Arabic morpheme “Idhafa”, but it developed across the Iranian subfamily independently. It has evolved from a Proto-Iranian morpheme /hja/ or /tja/. An Example of Ezafe will now be presented:

In keta:b-e kohne-ye bi arzeʃ-e Maryam

This book-EZ old-EZ without value-EZ Maryam

“This old valueless book of Maryam”

There is some dispute among linguists to the exact nature of Ezafe, especially in relationship to morphosyntax. This paper will present the arguments of P. Samvellian of the Sorbonne. Previous scholars, Samiaan (1993) and Ghomeshi (1997) argue that Ezafe is on a theoretical level, a syntactic function word, that is expressed as a morpheme. Samvellian presents the argument of Ghomeshi that Ezafe never proceeds a phrase. As an example:

*Keta:b-e sorx-i Maryam

Book-EZ red-INDEF Maryam

“Maryam’s red book”

This sentence is ungrammatical.

However, as a counterexample, Samvellian, presents a sentence in which Ezafe proceeds a phrase, from M. Dolatabi’s novel Ruzegare Sapari Shodeye Mardome Salkhorde :

t͡ʃasm-an-e negaren-e forurixtan-e divar-e xane-jaʃ

eye-PL-EZ worried-EZ crumbling-EZ wall-EZ house-3P.SING

“Eyes worried by the crumbling wall of his house

In this sentence, the word ͡ʃasm-an-e have Ezafe before the phrase negaren-e forurixtan-e divar-e xane-jaʃ. Therefore Samvellian suggests that the issue with *Keta:b-e sorx-i Maryam is that Ezafe cannot proceed an indefinite noun. Thus, they compare two sentences, one grammatical, one ungrammatical:

*mard-e nega:ran-e bat͡ʃt͡ʃe-ha-jaʃ-i

man-EZ worrying-EZ child-PLR-3P.SING-INDEF

"A man who worries about his children"

with the grammatical sentence

mard-e nega:ran-e bat͡ʃt͡ʃe-ha-jaʃ

man-EZ worrying-EZ child-PLR-3P.SING

"The man who worries about his children"

Thus, we may conclude that Keta:b-e sorx Maryam “The red book of Maryam” would be grammatical.

Samvellian argues that the evidence for Ezafe to be purely syntactical may be explained by the fact that Farsi has two kinds of nominal affixes: Inflectional and Phrasal, to use their terminology, though a “Pure Morphological affix” vs. “Morphosyntactic affix” label may be clearer. The former is always attached directly to the noun, cannot occur for adjectives, and multiple such affixes cannot be on the same noun. There are two main Inflectional affixes: /ha/, a plural marker, and /he/, a colloquial definite article. The restrictions of these affixes can be shown by the following ungrammatical examples:


*”The books”

*/in pesar-e ahmaq-ha/

*This boy-EZ silly-PL

*”These silly boys”

Ezafe is, on the other hand, a Phrasal affix. These affixes also include the personal possessional affixes and the indefinite article suffix /i/. They can occur after an Inflectional affix, and can occur multiple times in a phrase, and a phrase, though not a lexeme may have more then one. Concentrating on Ezafe, we will use the following examples:

/in pesar-ha-je ahmaq/

This boy-PLR-EZ silly

“This silly boy”

/leba:s-e qermez-e bi a:stin-i/

Dress-EZ red-EZ without sleeve-INDEF

“A Sleeveless red dress”

Now we will consider the interaction of Ezafe with prepositions, according to the analysis of Pantcheva. Pantcheva states that regarding Ezafe, Farsi preposition may be divided into three classes: Those that do not accept Ezafe, those that optionally take Ezafe, and those that must take Ezafe. She calls these classes 1, 2a and 2b, respectively, based on previous works that did not divide between 2a and 2b. We will look at data using each class with and without Ezafe.

Class 1 preposition: /dar/ “at; inside”

/ketab dar keʃu bud/

Book in drawer was

“The book was in the drawer”

*/ketab dar-e keʃu bud/

*Book in-EZ drawer was

*“The book was in the drawer”

Class 2a: tu “inside”

/ketab tu keʃu bud/

Book inside drawer was

“The book was inside the drawer”

/ketab tu-je keʃu bud/

Book inside-EZ drawer was

“The book was inside the drawer”

Class 2b /daxel/ “inside”

/ketab daxel-e keʃu bud/

Book inside-EZ drawer was

“The book was inside the drawer”

*/ketab daxel keʃu bud/

*Book inside drawer was

*“The book was inside the drawer”

Thus, we have considered the function of Ezafe, and its status as a morpheme, and what morphemic class it belongs to, and what its interactions with prepositions are.

Re: A paper I wrote on Farsi Ezafe

Posted: 17 Dec 2018 21:01
by kanejam
Your transcription is totally inconsistent. You variously give book as keta:b, ketab and kitab. You also have at least one <y> for /j/. Why not just use the normal transcription? I.e. ketâb. Also it's not clear what most of your ungrammatical sentences are supposed to demonstrate, you just give an ungrammatical sentence and say 'this is ungrammatical'. It's unclear to me what the example sentence *Keta:b-e sorx-i Maryam has to do with ezafe before phrases. You also say ezafe can't be used with an indefinite noun, but give an example of exactly that later on: leba:s-e qermez-e bi a:stin-i.

Re: A paper I wrote on Farsi Ezafe

Posted: 18 Dec 2018 04:04
by Shemtov
kanejam wrote: 17 Dec 2018 21:01 Why not just use the normal transcription? I.e. ketâb.
The paper is not meant for someone who has any knowledge of Farsi, therefore, I decided to transcribe everything in IPA. You're right that I should clear up the IPA.
kanejam wrote: 17 Dec 2018 21:01 . It's unclear to me what the example sentence *Keta:b-e sorx-i Maryam has to do with ezafe before phrases.
I think it has to do with taking / sorx-i Maryam/ "The red one of Maryam as a phrase, with /keta:be/ being a modifier of red "What kind of red one?" "a Book". I will clear that up.
kanejam wrote: 17 Dec 2018 21:01 You also say ezafe can't be used with an indefinite noun, but give an example of exactly that later on: leba:s-e qermez-e bi a:stin-i.
I think the intention was directly before an indefinite noun, but the article didn't specify that, perhaps because it is written for specialists and the fact that /bi/ would block the Ezafe from being on /a:stin-i/ is taken for granted (the conclusion "not with an indefinite" and the sentence with an indefinite are from the same peer-reviewed article).I will note that.