The Lord’s Prayer in Northern Kurdish: Lexical and orthographic journeys since 1787 Текст научной статьи по специальности «Языкознание и литературоведение»

The Lord's Prayer in Northern Kurdish: Lexical and orthographic journeys since 1787

Молитва «Отче наш» на севернокурдском языке: лексические и орфографические путешествия с 1787 года

N. Bailey Н. Бейли

This article presents a survey of sixteen translations of the Lord's Prayer in Northern Kurdish published from 1787 until 2019. Attention is given to the development of different orthographies and literary standards for Northern Kurdish as well as to the evolution of certain vocabulary as illustrated by the use of 'key terms' in the Lord's Prayer. Both lexical evolution and orthographic development are shown to reflect aspects of Kurdish identity. The discussion also touches on parts of the Prayer that tend to be opaque or misunderstood by Kurdish readers even when the vocabulary and orthography are otherwise described as "intelligible" or "standard".

For the linguistic and orthographic details presented in this study, I greatly benefited from discussions with a number of friends and colleagues, although I take full responsibility for any shortcomings. Thanks are due to Nadire Efo, Emma Casim, and Barise Xelil (especially for help with the Armenian-script texts), to Saeed Othman (especially with the Garzoni and Dalton texts), to Marco Libre and Chiara Libre (especially with Garzoni's Italian), and to Ali Nabhani and Rewar Rahimi Negad for help with Arabic and Persian borrowings. Many thanks are also due to those who read and commented on drafts of this paper, including Elke Karan, Denise Bailey, Andy Faust, M. Maletich and T. A. Luoto (Maletich and Luoto also

Key words: The Lord's Prayer, Northern Kurdish, language history and documentation, orthography, phonology, language and literary standardization, lexical evolution, Bible translation, holy, kingdom, father, Biblical metaphor

В данной статье представлен обзор шестнадцати переводов молитвы «Отче наш» на севернокурдский язык, опубликованных с 1787 г. по 2019 г. Внимание уделяется развитию различных орфогра -фических правил и литературных стандартов для севернокурдского языка, а также эволюции определенной лексики, проиллюстрированной использованием ключевых терминов в молитве «Отче наш». Как лексическая эволюция, так и развитие орфографии отражают аспекты курдской идентичности. Статья также затрагивает те части молитвы, которые, как правило, остаются сложными или непонятными для курдских читателей, даже когда лексика и орфография понятная и стандартная.

Ключевые слова: молитва «Отче наш», севернокурдский язык, история языка и документация, орфография, фонология, языковая и литературная стандартизация, лексическая эволюция, перевод Библии, святой, царство, отец, библейская метафора DOI: 10.37892/2313-5816-2021-2-154-218

shared insights about Kurdish orthography as used in Iraqi Kurdistan). Special thanks are due to Tengiz Siyabendi, who kindly provided me with a digital copy of Lazo's 1921 Kurdish reading primer in Armenian script entitled GuMu (§ems 'The Sun'). I am also grateful to L. Merana who supplied me with copies of difficult-to-obtain materials. I acknowledge with much gratitude others who supported me in accessing digital images of texts: Bugra Poyraz, who photographed countless pages from the archive in the Bible Society in Turkey and Tamar Karasu, the Executive Secretary of the Bible Society in Turkey, for her general support; and Neil Rees (British and Foreign Bible Society) and Martin Reynolds (Mission Assist), who digitalized texts from the Cambridge University Library. Last but not least, I wish to express my appreciation to people at the Institute for Bible Translation for encouragement and support in writing this article, including especially Marianne Beerle-Moor, who first encouraged me to undertake this project.

1. Introduction

... they made a special journey [in 1826] into Kurdistan for the purpose of testing [the Kurdish translation of the four Gospels led by Bishop Shevris of Tabriz]. There it was found to be unintelligible; and Kurdish, it was ascertained, was spoken in so many dialects that it was very difficult to say which was most suitable for a translation. [Canton 1904: 12-13]

Over the last 200 years, those who have invested in Kurdish Bible translation have had a difficult road to travel, and this mirrors the challenging journey that Kurdish-speaking people have had in writing Kurdish and in creating and honing effective literary standards. The above quotation stating that the earliest attempt of a Gospel translation was "unintelligible" is a likely exaggeration since intelligibility is often a matter of degree. But there is no doubt that the journey has been cluttered with obstacles inhibiting clear communication, obstacles due to both dialectal variation in Kurdish as well as different ways the language has been written and viewed.

In this article I have the following goals:

(1) to present a survey of the translations of the Lord's Prayer in different varieties of Northern Kurdish published from 1787 to 2019;

(2) to trace some of the development of literary standards for Northern Kurdish through their application in Bible translation;

and

(3) to trace the lexical evolution of a few terms in the versions of the Prayer. During this period of over two centuries, Northern Kurdish has been written in no fewer than four alphabets (something shared with Turkish). The texts I present here can

serve as historical 'snapshots' of Kurdish orthography and specific lexica.

A fourth goal is

(4) to highlight parts of the Prayer that tend to be opaque to or misunderstood by Kurdish readers even when the vocabulary is otherwise 'intelligible' and the writing system is not problematic. Such opaqueness is generally due to mismatches in background knowledge between typical Kurdish readers and the originally intended audience.

In the next part of this introduction (§1.1), I briefly introduce Northern Kurdish, including some of its linguistic complexity, and the orthographic systems used in the different Northern Kurdish versions of the Prayer. These introductory comments touch on not only the orthographic representation and presumed phonetic correspondences, but also on what can be considered to be the three modern Northern Kurdish literary standards, which have been in extensive use during the last 80 to 100 years. Following these introductory matters, the main survey of the texts is presented in sections §2 (texts from 1787 to 1953) and §3 (texts from 1993 to 2019). Besides issues of orthography, attention is given to various translation challenges, including word choices and lexical evolution. It is illustrated that both the orthographies as well as the lexical choices in the modern translations reflect aspects of Kurdish identity and evolving Kurdish nationalism. Section §4 offers observations on certain 'key terms' (important theological phrases), including especially the renderings for 'holy' and 'kingdom'. Many of the observations derive from my personal notes and reflections while working with Kurdish translation teams since the mid-1980s.

Most of the surveyed texts are from the Gospel of Matthew (Matthew 6:9-13). In two cases, the shorter text from the Gospel of Luke (Luke 11:2-4) is discussed. The following versions of the Lord's Prayer are examined:

• 1787 Garzoni text, Latin script, an Amadiya source (Rome)

• 1806 Adelung text, nearly identical to the Garzoni text, with German glosses (Berlin)

• 1872 Rhea text, from the 1850s or early 1860s, Latin script, a Hakkari source (New Haven)

• 1857 Gospel of Matthew and 1872 New Testament, Armenian script (Istanbul)

• 1870 Dalton text, Arabic and Latin scripts (St. Petersburg)

• 1891 New Testament, Armenian script (Istanbul)

• 1922 Gospel of Matthew, Arabic script (American Bible Society, Istanbul)

• 1923 Gospel of Luke, Arabic script (American Bible Society, Istanbul)

• 1953 K. Bedir-Xan's Gospel of Luke (Bible Society of Lebanon)

• 1993 Gospel of Matthew, Cyrillic-Script Kurmanji Standard (Institute for Bible Translation, Stockholm) (Latin-script version also produced)

• 1998 Four Gospels and 2005 New Testament, LatinScript Kurmanji Standard (Kitabi Mukaddes Çirketi, Istanbul)

• 2000 and 2011 New Testament, Cyrillic-Script Kurmanji Standard (Institute for Bible Translation, Moscow) (Latin-script versions also produced)

• 2004 Bible, Latin-Script Kurmanji Standard (GBV-Dillenburg, Eschenburg)

• 2019 New Testament, Arabic-Script Behdini Standard (Biblica, Erbil/Hawler)

1.1 Northern Kurdish, regional variation

in phonology and writing systems

As a linguistic entity, Kurdish is today recognized in traditional genetic (genealogical) classification terms as a North-West Iranian language within the Iranian and Indo-Iranian branches of the Indo-European language family. Kurdish has been treated in various scholarly studies as consisting of three main linguistic groupings — Southern, Central, and Northern — based on linguistic commonalities and distinctions, with these names referring to relative location [MacKenzie 1961; Fattah 2000; Haig, Opengin 2014]. There are also both shared and distinct social and cultural features throughout the area where Kurdish is used. A number of religious traditions are represented, primarily of Sunni Islam but also of Shiism, Alevism, Yezidism, Christianity, and Judaism.

The majority of Northern Kurdish speakers live in a large and mostly contiguous speech zone that, despite modern borders, stretches from northern Iraq into northern Syria, across much of eastern Turkey, and into northwestern Iran. Important enclaves of speakers reside in Armenia and more distant parts of the former Soviet Union, as well as in central Turkey and northeastern Iran. Northern Kurdish is also used, and even thrives, in a global diaspora.

Given the expansive geographical range, there is, nonetheless, a relatively high degree of mutual intelligibility among these Northern Kurdish speakers, even when separated by large distances (compare [Haig, Opengin 2018: 157-158]). For speakers attempting to communicate from different areas, mutual intelligibility can be predictably diminished, such as when conversation involves specialized lexical domains or when speakers first come into contact with each other. Yet many native speakers who live in the diaspora or who communicate with others on the Kurdish language internet insist that the linguistic differences are not overly challenging or insurmountable. Many strong readers in fact move seamlessly from texts in one variety of Northern Kurdish to another, and skilled

writers readily incorporate vocabulary from multiple varieties.2 However, I have also observed that 'average readers' and people who do not have much contact with speakers from other areas do struggle with understanding other varieties.

In this study, I give attention to the significant challenge that different writing systems and specific writing conventions can create for Kurdish readers and writers. Literary standards are normally based on relatively consistent uses of spelling, vocabulary, and grammar, and consistency in these areas increases effectiveness in communication. Despite all the marvels of the written word, language reduced to marks on paper can set up barriers to communication that do not occur in oral communication. Thus, the existence of multiple literary standards presents additional challenges to speakers of Northern Kurdish, as each standard employs a different alphabet and a different set of orthographic conventions.

Over the centuries, attempts to write Northern Kurdish have resulted in orthographies with differing degrees of fidelity to the phonologies of individual linguistic varieties. We can speak of plural "phonologies" as there exists some variation in the sets of phonemes. Such variation has complicated the efforts of those

Literacy in the mother tongue, however, has been and continues to be a major issue for Kurds. Only in Iraq and parts of the former Soviet Union has there been government support for Kurdish-speaking children to acquire literacy in their own language. In other areas, literacy is normally acquired in a second language (e.g., Turkish, Arabic, Persian, etc.) and learning to read in Kurdish has often been discouraged or even punished. Data is difficult to find, but we assume that there is still a significant number of Kurdish speakers who are functionally illiterate in any language. The International Organization for Migration [2018: 34-36] produced a report for the Kurdistan Region of Iraq, a region where I would expect higher than average rates. Accordingly, for people 6 years and older, the literacy rate is about 79 %, and illiteracy is proportionally higher among people over 55, those in rural settings, and women.

2

who have sought to create useful literary standards, and it has challenged readers who must deal with orthographies that do not completely match their own phonologies.

Until the 1920s, the script of choice for most forms of written Kurdish would have been an Arabic-based script (which is more precisely called an "Arabo-Persian" script by Blau [1996: 23] since Persian additions were normally employed), and Arabic or Persian were the languages in which most writers would have acquired literacy. (An important exception is the use of an Armenian-based script, to be discussed below.) The orthographies of Persian and Arabic underrepresent certain sounds, especially vowels, and most Kurdish texts from before the 1920s imitated such orthographic conventions.3 In contrast, orthographies based on the Armenian script used in Kurdish Bible translations in the 1800s, and then the new Kurdish orthographies from the 1920s on, have favored considerably more phonemic representation.

In the present study, in order to facilitate a comparison between the non-Latin-script orthographies in our surveyed texts, I employ a one-to-one transcription in Latin script that preserves the orthographic distinctions in the old texts. My transcription is based on the modern Kurdish alphabet employing Latin script (this alphabet is used in versions of the Prayer printed in 1953, 1998, 2004, and 2005; and in the table below, the alphabet is shown in the column labeled "1953-2005"). But my transcription also includes elements from the International Phonetic Alphabet (IPA) to represent unique features of these historical orthographies.

3 There are significant examples of Northern Kurdish literature since the 1600s. Noteworthy authors from the earliest period include Meleye Ciziri (ca. 1570-1640) and his disciple Feqiye Teyran (ca. 1590-1660). Probably the best known older text in written Northern Kurdish is the romantic story, Mam and Zin (Mem u Zin) written in the late 1600s by the poet Ahmad-e Kani. Such examples testify to the early use of the written language, especially around the city of Cizira Botan (Turkish: Cirze), even though Kurdish has been overshadowed by the use of other languages.

In the discussion, I also use the following conventions: 'forward slashes' /.../ to indicate phonemic representations; and square brackets to indicate phonetic representations with IPA symbols.

The following table allows comparison of the different alphabets used since 1857 in Bible translation publications.4 In order to facilitate comparison between alphabets, the letters are organized phonetically: Vowels and semi-vowels are presented first; then follow stops (plosives) which are listed according to the points of articulation, and then follow affricates, fricatives (including laryngeals), nasals, and liquids. Not all letters of the different alphabets occur in the texts, and for that reason, I have drawn on more data from the indicated publications to fill out the table. The 'IPA' ('International Phonetic Alphabet') column gives approximate phonetic values, but these values do not exhaustively represent what is found in all varieties of Northern Kurdish. Letters in parentheses (...) indicate letters that represent more than one phoneme in the given alphabet. The abbreviation 'NA' ('not applicable') indicates a phoneme that I believe does not exist for the speakers intended to use that alphabet.5

Summary table of orthographic representations of Kurdish in texts of the Lord's Prayer

IPA 1857 1872 1891 1922, 1953 2019 1993- 1953-

(1911) 1923 2011 2005

J J, h J, h J, h 6 6 6 й y

i h h h (6) (6) (6) и î

e t t t (6) 6, S S е ê

i~i в в в - - - ь i

The orthographies from [Garzoni 1787], [Dalton 1870], and [Rhea 1872] are not included in this table, since it is unclear (and doubtful in the case of Garzoni and Rhea) if their orthographies were ever meant to be used by a community of Kurdish speakers. For example, I assume that 'pharyngealized' (or 'velarized') consonants are not relevant to Kurdish of Transcaucasia and some parts of northeastern Turkey.

4

5

IPA 1857 1872 1891 (1911) 1922, 1923 1953 2019 19932011 19532005

э~е (Ш) Til m often 4 often 4 4 Э e

a~D Ш Ш Ш 1 1 1 а а

w (nL) (4?) (nL) n' (A?)6 nL, (L?) 9 9 9 w w

u nL nL nL (9) (9) 99 у й

u, 0, wi (nL?) (nL?) (nL?, hL?) (9) (9) (9) 0 u

o 0 (n?) 0 (n?) 0 (n?, nA) (9) 9 9 о o

y К hL hL? NA NA 9^ NA NA

P P P P О О о п P

P1 Ф Ф Ф (О) (О) (О) п' (P)

b Щ Щ Щ О О О б b

t q q q О О О т t

th p p p (0) (о) (о) т' (t)

t! NA? NA? NA ь ь (о) NA (t)

d rn rn rn J j J д d

tr 2 2 (*)7 (s) (s) (s) ч' 5

trh * * * s s s ч (5)

d3 fi fi fi г г г щ c

k q q q j j j к k

kh f f P (J) (J) (J) к' (k)

g к к к ^ ^ ^ г g

q (q) q q q q

f Ф Ф Ф j j j ф f

v A 4, l A j j в v

s u u u ^ ^ ^ с s

z q q q 3 3 3 з z

s! NA? NA? NA? NA (s)

z! NA? NA? NA? ь ь (3) NA (z)

.r 2 2 2 ш §

6

The digraph m. is used for /u/ and /w/ as well as where one would expect /u/. The combination n' is used in the 1872 text for /we/

'second person plural oblique' (e.g., Mark 8:18: Smh^t n' /Qave we/ 'your eye(s)'). The letter 4 normally stands for /v/, but I include it here for /w/ with a question mark and in parentheses (4?) given the spelling of the name 'David' as Sm4^s, since possibly this represented /Dawud/ rather than /Davud/. The 1911 Luke publication appears to use the same orthography as the 1891 text, except it reinstated q for /5/ [tj].

IPA 1857 1872 1891 (1911) 1922, 1923 1953 2019 19932011 19532005

3 à à à 3 3 3 ж J

X N N N é é é х x

к n n n t (é)? t (х) or х' x

Ç - - ' t t t э' (or')

h h h h л л л h h

h (h) (h) (h) С С С h' (h)

m iï iï iï ? ? ? м m

n Ь Ь Ь 0 0 0 н n

l L L L J J J л l

H (1) NA NA NA (J) (J) Л NA (l)

г P P P J J J Р r

r n n n (J) (J) Р' (r) or rr

As shown in the above table, there are significant differences between the orthographies, as none of the orthographies represent the same set of phonemic distinctions. The phonemic distinctions that will especially feature in our discussion include:

(1) the +/- aspiration distinction, which I suspect is phonemic in most varieties: /p, ph, t, th, ç, çh, k, kh/;

(2) the distinction between the voiceless glottal fricative /h/ [h] and the voiceless pharyngeal fricative /h/ [h], which is well represented across the varieties, and the voiced pharyngeal approximate AV (i.e., 'ayin);

and

(3) the 'pharyngealized' (or 'velarized') consonant distinction, which in some varieties is relevant for one or more consonants (e.g., especially /sV, /zV, and /lY/, as well as unaspirated pharyngealized /tV), found not only in Semitic loans but also in native Kurdish (Iranian language) words.8

8 For phonological descriptions and sketches (of usually individual varieties), see [KypgoeB (=K'urdo) 1957; MacKenzie 1961; XaMoaH

Given that Northern Kurdish is a large language with different regional phonologies, it is no surprise that the written language has been represented in different ways. However, some of the differences in these systems must be attributed to the fact that the creators of the different alphabets were influenced by the orthographies in which they acquired literacy: Arabic, Persian, Turkish, or Armenian. Such influence is of two types: one in which the creators (consciously or subconsciously) imitated the other system, and one in which they purposely tried to be different.

In the next two sections (§2 and §3), we highlight many differences between the orthographies, and we also trace the

(=Xamo) 1965; Blau 1975; Kahn 1976; Jastrow 1977; Opengin, Haig 2014; Haig, Opengin 2018] (on several varieties). Rizgar's dictionary [1993] is also useful because his pronunciation apparatus accounts for all but the 'pharyngealized' sounds in my table. Chyet's dictionary [2003], following his sources, distinguishes nearly the entire spectrum of sounds, including most pharyngealized ones. The pharyngealized sounds are sometimes called "velarized" [e.g., XaMoaH 1965: 13-15 on Behdini], "emphatic" or "ejective" [e.g., Jastrow 1977: 91]. Besides the three-way distinction for stops and affricates (i.e., "voiced" / "voiceless" / "aspirated"), Kahn [1976] describes a fourth distinction for speakers west of Rezaiyeh Iran: "pharyngealized", which we can represent as /p9, t9, q9/ and uvular /q/ (Kahn suggested that /q/ could be "considered systemically as a pharyngelized /k/" [1976: 23] in a four-by-four set of stops and affricates). I have not included /p9/ and /q9/ [t/9] in my table since none of our orthographies have done so, but in an unpublished phonological analysis Bailey & Bailey [1992] used to supplement teaching Kurmanji to NGO workers in the early 1990s, we also assumed Kahn's four-way contrast since they appeared phonemic in the speech of our main language consultants, primarily Yezidi Kurds from Taqa village near Midyat (Turkey) and Tilxatun village near Qami§lo (Qamishli, Syria). Recently, Barry [2019] reanalyzes the pharyngeal sounds /9/ and /h/ as a property of vowels, which "might be represented as /s9/ and /i9/" (i.e., /e9/ and /i9/) p. 46), but it is unclear to me if Barry's proposal would handle /s9, z9, p9, t9, q9/.

process of increasing phonemic representation in both the Armenian and Arabic scripts.

Looking beyond the alphabets and their sound-symbol correspondences, every relatively systematic attempt to write the language has also involved other orthographic conventions, including spelling and word division, as well as more general linguistic conventions involving grammar, idiom, and vocabulary. In their most mature forms, these systems can be called 'orthographic literary standards'. During the last one hundred years, three such standards have come into use with the development of extensive bodies of literature. Within each standard one can trace the evolution of individual features through the various publications. While these three literary standards currently exhibit relative stability, there is still some irregularity in their application. Brief descriptions of the three literary standards are presented here. The descriptive labels are my own:9

(1) The Arabic-Script Behdini Standard employs a modified Arabic script, which includes both Persian graphemes (o £ j and graphemes unique to Kurdish. Prestigious texts generally reflect the Behdini (also spelled Badini) variety of Northern Kurdish around Dohuk, Iraq, where today Behdini enjoys support in local education and some official capacity. The Arabic-based script used in modern Behdini literature was originally developed by writers of Central Kurdish (e.g., Sorani). This development was a process with contributions by several, including especially Tawfiq Wahby, who in 1923 was commissioned by the Iraqi Ministry of Education to write a Kurdish school grammar [Hassanpour 1992: 360; Ha-sanpoor 1999: 50, 71; Leezenberg 2020: 67], although certain

9 On the 'standardization' of written Kurdish, compare [Hassanpour 1992] and the notes in [Haig, Matras 2002: 4].

conventions can be traced to earlier writers.10 A major feature in the evolution of the Arabic-based script and vocabulary in both Sorani and Behdini can be described in terms of 'language purification', in that foreign words and even sounds, especially if derived from Arabic, have to one extent or another been purged from the literary standard [Hassanpour 1992; Hasanpoor 1999]. This 'purification movement' has been widespread not only in Iraqi Kurdistan but in other parts of the Kurdish world (as we shall see below). For the Arabic-based script, this has meant that certain letters, especially those representing some of the pharyngealized sounds, are not normally represented in Kurdish texts, even though those sounds do still occur in speech.

(2) The Cyrillic-Script Kurmanji Standard (also known as the Kafkaz or Transcaucasian Region Standard) uses a modified Cyrillic script. Its development was entrusted to the Yezidi Kurdish scholar Heciye Cindi by the Soviet government in the 1940s, although there were actually two other Kurdish scripts used earlier in the Soviet era. The first script to be used in the Soviet Union (e.g., in Armenia and Georgia) was an Armenian-based script, promoted by the 1921 alphabet primer §ems, written by Lazo (alias Hakob Ghazaryan, 1869-1926) and published in Etchmiadzin.11 This script was used from at least 1921 until about 1928, but from the late 1920s a Latin

10 Hassanpour [1992: 358-360] states that the first use of J in print for /lV was in the 1909 Mukri (Central Kurdish) translation of the Gospel of Mark (Incili Merqus), and the first use of <J for /v/ was in Khalidi's 1892 dictionary. (I am grateful to Rewar Rahimi Negad for discussing some of these details with me.)

11 See discussions in [Re§id 2020; Siyabendi 2019; Hassanpour 1992: 374, 376]. Siyabendi mentions several Kurdish manuscripts in Armenian script dated before the 1800s, including one apparently from the 10th or 11th century, which is held in the Matenadaran (the Mesrop Mashtots Institute of Ancient Manuscripts) in Yerevan. Below, in footnote 35, I contrast the representations of the stops and

script — known as the "§emo-MarogUlov alphabet" — was used instead.12 This Latin script was used until after World War II when, under Stalin, it was replaced by Cyrillic. In contrast to (1) the modern Arabic Standard and to (3) the Latin Standard, (2) the Cyrillic Standard preserves the +/-aspiration distinction. The +/-aspiration distinction was also represented in the "§emo-MarogUlov" Latin-script standard and in Lazo's Armenian-script standard. This orthographic distinction can presumably be traced back to at least the mid-1800s, as the distinction was made in the 1857 Armenian-script Kurdish Gospel texts. The publications in the Cyrillic Standard (from the 1940s until today) are relatively homogenous in grammar and idiom since most writers have come from the small, mostly Yezidi Kurdish communities in Transcaucasia. Since the early 1990s, many publications that belong to this literary standard actually make use of the Latin script in combination with conventions otherwise unique to the 'Cyrillic' Standard (most notably, the apostrophe, to be described below). A prime example is the periodical R'ya T'eze (P'ua T'33s), which in 2000 switched from Cyrillic to Latin script.13

affricates in the 19th century Gospel publications in comparison to Lazo's alphabet.

12 Re§id [2020] and Siyabendi [2019] note that support to change from Armenian to Latin script mounted during a 1925 conference of Kurdish intellectuals in Leninakan (Gyumri), Armenia, apparently because Latin script was assumed to have more international currency. There is a table comparing alphabets in [Курдоев 1957: 11-13]. Kurdish Wikipedia presents a poster image of the L atin-based "§emo-Marogulov" alphabet with forms that do not entirely match those of Курдоев: https://ku.wikipedia.org/wiki/Alfabeya_Semo-Marogulov

13 Several articles in the March 2020 edition of R'ya T'eze summarize much of the history of the periodical. In that edition, see especially [Xelil 2020; Celilov 2020].

(3) The Latin-Script Kurmanji Standard has been employed especially in Turkey, Syria, and Western Europe (compare comments on "Standard Kurmanji" in [Haig, Opengin 2018: 164-165]. Linguistically, it is probably the most eclectic: originally, it was primarily influenced by the Kurdish of Cizira Botan (Turkish: Cizre), but there has been significant influence from other varieties, perhaps especially from around Amed (Turkish: Diyarbakir). Its birth and early growth are described and illustrated in the periodical Hawar (published 1932-1935 and 1941-1943 in Damascus, Syria), in articles by the periodical's editor and proprietor, Celadet Ali Bedir-Xan (also spelled Bedirxan and Bedir Khan). This orthography was introduced shortly after the new Latin-based Turkish script was officially adopted in 1928 in Turkey, replacing the Arabic script. It shares features with the alphabets of both modern Turkish and French. As explained in early issues of Hawar, Bedir-Xan dispensed with certain sounds on the grounds that they were "foreign" (e.g., the sounds represented by /h, and /x, £■/; see Hawar [1932: vol. 3, p. 3]). Bedir-Xan's decision thus reflects the same sort of 'purifying' goal as those had who created the Arabic-script standard.14 Other simplifications introduced by Bedir-Xan have been justified (by himself or by others) on the grounds that not all varieties of Northern Kurdish employ all of the distinctions.

Early issues of Hawar still did occasionally use /h/ and /x/. Another notable question in issues 6 to 8 concerned the phonetic values of k and q. The initial proposal was that q would symbolize the sound of Arabic A and k would symbolize the sound of Arabic such that 'Kurd' was written as Qurd (there was no discussion about the +/- aspiration distinction). Issue 24 (April 1934) begins with a short announcement stating that the values of the two letters k and q had been exchanged.

Having addressed these introductory matters concerning orthographies and literary standards, we will now turn to our survey of the different versions of the Lord's Prayer.

2. Versions of the Lord's Prayer from 1787 until 1953

2.1 Early versions in non-standard scripts:

Garzoni, Adelung, and Rhea

The oldest version of the Lord's Prayer in our survey comes from a book printed in Rome in 1787, entitled Grammatica e vocabolario della lingua Kurda. This book contains the first known grammatical analysis of a variety of Kurdish. It was authored by the Italian scholar, Maurizio Garzoni, a Dominican friar who moved first to Mosul in 1762 [Blau 2009] and then to Amadiya, Kurdistan, in 1764.15 The city of Amadiya (also spelled Amedi, or in Kurdish, Amedi) and its location were significant for a number of reasons: It sits on a small, high plateau that would be naturally defensible in earlier times. It has a complex and colorful history as it has been home to Jews, Christians, and Muslims.16 For a period lasting from some point in the 1300s until 1842, Amadiya was the seat of the Bahdinan principality [Hassanpour 2011]. Today it is a part of the Dohuk Governorate in the semi-autonomous region of Iraqi Kurdistan, and it is dwarfed by the regional capital, Dohuk. Amadiya also sits about eighteen kilometers from the Great Zab River, beyond which Kurdish can be characterized in terms of Central varieties (e.g., Sorani and related varieties). But

15 Before moving to Amadiya, Garzoni writes that he was in Mosul (p. 7), although he does not state when he first came to Mosul. Moreover, he notes that he was preceded by another Dominican, P. Leopoldo Soldini, "the first missionary to settle in Kurdistan", who arrived in Kurdistan in 1760 and died in Zakho in 1779.

16 According to Petrides [1912: 376], the city had a total of "5,000 inhabitants of whom 2,500 are Mussulmans, Kurds for the most part, 1,900 Jews, and 1,600 Chaldeans."

the primary domain of Northern Kurdish lies on the Amadiya side of the river and follows a general northwestern direction. The Northern Kurdish variety spoken today in the environs of Amadiya is clearly recognized as Behdini.17

Garzoni's text of the Lord's Prayer comes at the end of the work [p. 283], following his grammatical description [pp. 11-74] and Italian-Kurdish vocabulary list [pp. 79-282]. Since Garzoni's aim was to share his knowledge of Kurdish with fellow Italians in ministry planning to live in Kurdistan [pp. 8,11], his reason for using a Latin-script transcription is obvious.18

For comparison, Blau [1975: 21-27] presents a succinct phonological description of Kurdish of Amadiya.

On pages 11-16, Garzoni introduces his transcription of the Kurdish sounds, offering comparisons to Italian, Persian and Arabic. Garzoni's work is an important contribution to the study of Kurdish by a European, and he makes many helpful comments about the Kurdish sounds. However, some discrepancies remain: Although Garzoni's k usually represents what would today be phonemic /k/ and /kh/, he also used k for what both then and today would presumably be phonemic /x/ and /q/. Several examples from his work (with his glosses and my English translations) will illustrate: He writes kalas kem 'salvare (save)' for what today would be /xelas kem/; and kabul kem 'accettare (accept)' for /qabul kem/. He also used q for /k/ (e.g., qo /ku/).

Although he defines hh as equivalent to Arabic 'ayin (£), he doesn't seem to define hhk, which occurs in the word ahhkaft 'parlare (speak)', which is today /axaft/ [axaft].

It also seems Garzoni had difficulty distinguishing the phonemes /j/ [3] from /§/ [f] and /z/ from /s/ as these distinctions are not phonemic in Italian (Marco Libre p.c.). He uses sc, as in Italian, to represent the sound of Arabic thus [f] and the Kurdish phoneme /§/, as illustrated by his spelling of scesc 'sei (six)', today spelled / §e§/, as well as sibi 'come (as, like)', today /§ibi/. But he also uses si (which he does not define) in words where today we would expect /j/ (e.g., bavesium = /bavejem/ 'gettare (throw)'). Similarly, for 'dentro (inside)' he writes zior (today /jor/), and for 'riccio (hedgehog)',

17

One wonders, however, in what form of a "Persian script"19 Garzoni (ever) presented the Prayer to Kurdish readers. The Prayer is followed by the Vulgate version, as seen below.20 I provide an approximate English 'back-translation' of the Kurdish prayer. Note that the letter ' f ' represents non-final lower-case ' s '.

suzi (today juji or juji etc.). Also puzzling is that he writes zen for 'donna, moglie (woman, wife)' (pp. 136, 188, 284), which in most varieties is /jin/, but perhaps his informants were using Persian ¿j [zsn] as a refined synonym. Finally, he writes scierma 'vergogna (shame)' for /§erm e/.

• The vowels are also confusing, assuming their actual phonetic values are equivalent to today's values. Thus, a and a seem to represent /a/ (long low vowel), but also a occasionally (e.g., nav 'name'). But a and occasionally e stand for the mid central vowel /e/ [э ~ e]; and e, e, and e correspond to the mid front vowel /е/ [e] . It seems that e also represented the mid high central vowel (usually represented by /i/ in Latin script), e.g., men /min/ listed under 'io (I)' also with az /ez/.

• Evidently both u and vrepresent back rounded vowels (e.g., mvkaddas 'holy'; u 'and' /u/), but v also represents /v/, as illustrated by t,vem 'volere (want)', which today is spelled /divem/.

19 Here I write "Persian script", following Garzoni (p. 11), who wrote: "Kurds use Persian script, and in all their public documents they use the literary Persian language, so that their writings are understood only by their scholars, who make such a profession in order to earn their food honorably. All villages pay one who is not only able to read Persian, but who is capable of interpreting it in the Kurdish language, and they are called Mella [mullah]. It is true that some personal [i.e., non-official] letters, poems, and songs are composed in their own language, but they are written in Persian script."

20 Oddly, the cited Vulgate excludes veniat regnum tuum, which is nonetheless translated into the Kurdish.

»83

ORAZIONE DOMENICALE In lingua, e frjft Kurd*,

BAbe ma 1« deratiit ferafraan: mvkaddas bit nave ta, Buii a ma baehfcre ta. Dibit amrada ta fet afman, u fer srtl. Au,ro u cht rvi teta nan bdei a ma. U ahi beka ghuna ma, fibi am ifu bek«m ehr

cckiria a ma zerer, ia zabhmet. U na avdia ma naftegerib Amma Kilasbeka ma ei Kaiibia . Ami».

PAter noikr qui es in cslis: fan(Si(íccm!• no-mentuum, Fiat voluntas tua, iicur inc»lop & in terra . Pinrm noihuin quotidianum da nobis hodie . Et dimitte nobis dcbita noftta , (¡cut 8i nos dimittimus dtbitonbus noitris. Ernenos in-ducas in temationctn . Sc<J libera aos % maio . Amen -

Back-translation: Our father who sits on heaven: may your name be holy. Give us your paradise. Your will/desire21 will happen on heaven, and on earth. Today and each day give us enough bread22 And forgive us our sin as we forgive everyone who has done to us harm or hardship.2 And do not throw us into trial/testing. But save us from evils2 Amin.

Under Italian volunta is listed Amr, Amrad. Today, "ami" would be spelled (')emr 'command, will', and "amrad" is more typically mirad or miraz. Both terms are well distributed across Northern Kurdish varieties but avoided by some writers as they were originally Arabic loans.

In modern Kurdish, as in the Biblical languages, nan 'bread' is

commonly used to refer to food in general.

"Harm or hardship" is an attempt to represent poetic zerer ia zahhmet.

The use of couplets (hendiadyses) is common today in all varieties

of Kurdish, both in poetry and in everyday language.

The ending -ya(n) on Karabia is plural (modern spelling: xirabiya(n)),

in which case this may refer to evil things or people, and it is not

21

22

23

Discussion of issues:

(1) It is not known to us who produced or assisted Garzoni in this translation, but the translation mostly follows the constituent order ('word order') of the Latin Vulgate (and the Peshitta), which is unnatural in Kurdish.25

(2) We can only speculate about the decision to render 'May your Kingdom come' by 'Give us your paradise'. In Islam, God's presence may be referred to as 'the heavenly kingdom/domain' (expressed in Persian and some Kurdish varieties

meleküte asmän);26 when the virtuous die, they are said to enter into that divine domain, which is also known as 'paradise' (/behi§t/). But the idea that the Messiah is God's chosen king for the divine kingdom is not commonly understood in Islam.

(3) The other oddity is the form debit, which on p. 25 Garzoni lists as future tense quegli sara 'he will be' and which in modern Behdini would be de bit21 So this appears to be a mistranslation, as it turned a petition into a statement of fact

singular as Adelung has (dem Bösen).

25 Most sentence types in Northern Kurdish conform to a S-O-V-G(oal) constituent order, but for pragmatic purposes as well as poetic purposes a non-Goal constituent can occur postverbally. The postverbal position of baehscte ta 'your paradise', amrada ta 'your will', and ghuna ma 'our sin' are probably the most unnatural.

26 Persian translations of the New Testament since at least the 1800s have also used meleküte asmän for 'kingdom of heaven'. Thomas [2015: 364] notes that ¿L^l pädsähi äsmän "was introduced in the twentieth century translations because of an understanding by many that os^La malakut refers only to the heavenly realm." Both the New Millennium Version and Today's Persian Version use pädsähi äsmän.

21 In Garzoni's transcription, debit cannot be modern /dibit/, present tense 'it happens', and in any case, it is not clear how /dibit/ in Northern Kurdish could make this petition equivalent to 'may it be' (so Saeed Othman, p.c.).

(in which case Adelung's German gloss of sey 'may it be' is

incorrect — see below).

We will return to some of Garzoni's vocabulary choices in section §4, which compares certain key terms in the different versions.

We find essentially the same text, with German glosses, in [Adelung 1806: 298], which was the work of the philologist, Johann Christoph Adelung. Adelung's version, credited to Garzoni, includes both apparent improvements to the text (and orthography) as well as errors. One error in particular — the misspelling of 'father' as baber — may betray his belief that Kurdish was more closely

Alii Gar so ni Gramm. Kur da, S. SgJ.

Vatrr unser, der wohnen über HiBiro«l,

Baber ma, ki deritnit ser Asmä%

B'dei a ma Baehschte ta;

S«v Wille dein im Himmel unJ auf Erie;

Debit Amrada ta, ser Asman ü ser Ard;

Heul« und jeden Tag hinlängliche* Brot gib lins;

Auro u ehr Ruz tera Nan bdei a ma;

TJnd vergib Sünden unsere wie wir vergeben jeden

U afubeka Ghuna ma sibiam afubekem ehr

den- eeüian lut im« Schaden oder Verdrille;

ki tschekiria amaZerer ia Zahhmetj,

fTnd Hiebt wirf uns in Versuchung;

U na avesia ma naf Tegerib;

Sondern befreye udi vom Bösen,

Ami na kalasbeka ma ez Karäbia.

related to Persian than is true (Persian jaj [phedœr] is 'father').28 Garzoni's version was correct (/bab/ [bab] in this Northern Kurdish variety means 'father' and the suffix -ê [e] is the masculine singular 'ezafe' morpheme here used to link the possessive pronoun).

Another early version appeared in a linguistic sketch by Samuel A. Rhea, an American Protestant. This work was entitled a Brief Grammar and Vocabulary of the Kurdish Language of the Hakari District and published posthumously in 1872 by the Journal of the American Oriental Society29 Rhea had lived in the region of Hakkari from 1851 to 1859 and then further east in Orûmiah (Urmia) from 1860 until his death in 1865. His ministry was with Nestorians, and besides being skilled in "modern Syriac" and "perfecting himself in the Oriental or Tatar Turkish", he found it necessary to learn Kurdish [p. 118]. The Prayer appears at the end, along with part of the Parable of the Prodigal Son, and is prefaced by a note of the editors who state that there was no "explanation of how or by whom they were prepared" [p. 155].

Lord's Prayer.

Sdb-è-ma ya Vasmdn-e, ndwë-ta -mokadùs bit y pâdishâhxyë-ta Vêt; rizâyê-ta bit, wekn Casmânëya, weto Carde zhi; bida-ma avro ndri-ê-ruzhê ; zA' karêt-ma Vbora, kurang am zhx borin karddrët-ma ; u ma na ba Ftijerib, beli ma khultis Vka eh? sherë ; sabab yé taya pâdishdh'ifl, u kudset, 11 jelûl, ebed il ebed : amln.

Back-translation: Our father who (is) in heaven, may your name be holy; may your kingdom come; may your pleasure/ desire happen, as in heavens so also in earth; give us today our daily bread; forgive (pass over) our debts, in the same way we also forgive (pass over) our debtors, and do not bring us into trial/testing, but save us from evil; because to you belong the kingdom, and power,30 and glory, forever: amin.

A typical 19th century Western view was that Kurdish was simply a corrupt Persian dialect (e.g., [Adelung 1806: 297]). According to the article, Rhea was with the American Board of Commissioners for Foreign Missions. I am assuming qudret 'power' was meant instead of qudset.

28

29

30

As can be seen, the transcription makes use of a modified Latin script, as this script would have been useful for Rhea and others learning Kurdish. It recognizes most but not all of the typical sounds of Northern Kurdish (e.g., /q/ is not uniquely represented: kar 'debt' is today spelled /qer/ [qar]).31 But certain features in Rhea's text and discussion might lead one to think that some of his informants were not native speakers or represented varieties with different grammar.32 Given these unusual features, my comments will be limited to the vocabulary (see section §4).

2.2 Armenian-script versions

The next few versions in our survey, from 1857, 1872, and 1891, employ modified Armenian scripts,33 which used mostly

The pharyngeals /h/ and /9/ are also not clearly represented. Two entries, haram and kharam, both glossed as 'unlawful', illustrate the apparent confusion about /h/.

For example, the use of the preposition b' for 'in' appears to follow Semitic usage. Also, Rhea states that there is no gender, which of course may have been true for the speech of some of his informants. But he notes that "in some dialects [case is marked] by adding e or a" (p. 120), which may be a reference to the contrast between masculine and feminine ezafe suffixes. In any case, all instances of singular ezafes in his two short texts are -e (e.g., padishahiye-ta 'your kingdom', which in most varieties is a feminine noun requiring -a). For the oblique case ending, most of his examples have -e, which in most varieties is a form that patterns as feminine oblique (e.g., zh' shere = ji §ere; Varde = li erde), with the exception of rtijerib 'trial/testing', which appears unmarked. My Yezidi Kurdish colleague, Nadire Efo (educated in Soviet Armenia), made use of one of the Armenian-script printings of New Testament books when he first began translating Matthew's Gospel in the early 1980s. (Since the volume contained both Matthew and Acts, it was either the 1872 text or the 1891 Matthew text together with 1911 texts.) Mr. Efo had received this book from Casime E't'ar, who was the father of Emma Casim. Mr. E't'ar was a decorated WWII officer and member of the Communist Party in Armenia

31

32

33

Western Armenian conventions.34 As mentioned earlier, unlike the other alphabets, the Armenian alphabet was naturally equipped to represent the Northern Kurdish +/- aspiration distinction for the voiceless stops and affricates.35 But, without modification, it

when he became interested in the Gospel. Mr. E't'ar had received the Armenian-script Kurdish translation from new immigrants from Syria, which he later gave to Mr. Efo. Mr. Efo notes that, in the end, this translation was of limited use to him, given certain aspects of its vocabulary and grammar.

34 Lazo's primer, in contrast, used Eastern Armenian conventions.

35 These texts actually involve more than one system. For the purposes of this study, I have consistently interpreted the stops and affricates in these texts according to the column below labeled "Gospels", which basically follows Western Armenian values, in contrast to Lazo's 1921 primer, which used Eastern Armenian values:

Gospels Lazo's primer

b Щ P

p P 4

ph Ф Ф

d m 4

t 4 m

th P P

c £ 2

5 2 £

5h i i

g к q

k q к

kh P p

To be sure, my system of interpretation sometimes results in phonetic values that are at odds with modern varieties as witnessed by the Cyrillic Standard and Rizgar's Kurmanji-English dictionary. For example, aspirated kh is unexpected in kho 'that' (Matthew 6:10), bikhn 'subjunctive third person plural do' (6:9) and khegh 'girl' (1:23); but these values make sense in other words, such as in: ршр /kher/ 'donkey' (21:2) versus qujn /kerr/ 'deaf'; and qrn /tu/ 'you' versus рпЛЬш /thUnna/ 'nothing'. Moreover, this system does

was not equipped to distinguish certain vowels and the uvular and pharyngeal consonants.

The first Armenian-script text we examine was published in 1857 and then again in 1872, both "in Istanbul".36 Thomas [2000] reports that this translation was the work of "an Armenian preacher in Haineh, Turkey, named Stepan."37 Blincoe [2019: 242] notes that "according to Marcellus Bowen [...] this version was intended for Armenians38 in that part of Kurdistan that extends from Marash

not make sense for many proper names (though I have still applied it), especially in the 1891 text since the translators have used the Classical (=Eastern) Armenian forms and spellings (and many of the stops and affricates have different values than in the Western system).

36 The spelling of proper names changed in these publications, as illustrated by the spellings of 'in Istanbul' on the title pages. In 1857, the phrase was Cupmtf^nrism (li Isthambulda), which, with the exception of the th, probably more closely reflects a Kurdish pronunciation; in 1872, we find Cumm^^oimm (li Isdambolda), which I believe reflects a more common old Armenian spelling with Classical stop conventions, so the first s would be pronounced as unaspirated /t/ (Istambol) and not /d/; then in 1891 it is spelled Upmh^nmm (li Sthambolda), which approaches the 1857 spelling again. The modern Eastern spelling is Usmdpnq_ (pronounced Stambul), and the typical spelling in Northern Kurdish is Stembol, although in Hawar it was usually spelled Stenbol.

37 According to Thomas [2000], Stepan is credited with translating all four Gospels, and "Tamo, an Armenian deacon" with the rest of the New Testament. There were actually three printings of Matthew: (i) the 1856 printing, when Matthew was printed by itself, which I have not had access to (Thomas states this was published by the British and Foreign Bible Society); (ii) the 1857 printing with the other three Gospels (the digital image that I had access to does not mention a publisher); (iii) and the 1872 version, which included the entire New Testament. More details about these translators, their printings and the circumstances of their distribution have been recently discussed by Rzepka [2018: 200-205].

38 That these publications were in Armenian script has been explained on the grounds they were produced by and for native Kurdish-

to Kharput and Diyarbakir and beyond towards Urumia."39 The 1872 text, cited below, included some important orthographic improvements over the 1857 text, although the words of the two texts are identical (see below for a discussion on revisions to proper names in other passages of the Gospels). One important improvement was to add a diacritic " over the letter m (thus iu) to indicate the mid central short vowel /e/ [a ~ ej. The 1857 printing had used m to represent that vowel and the low open long vowel /a/ [a] .40 Another improvement was to add a dot over q (thus q) for uvular /q/ [q] (otherwise q represents /k/ [k] in this orthography), though that letter is not illustrated in our passage. But in these orthographies /?/ is not at all represented, and both /h/ and /h/ are represented by one letter, ^ (h). Other likely under-specifications include that 4 (v) represents both /v/ and /w/, and that m. (u) represents /u/ and /u/.41 The digital image of the Armenian script is followed by my one-to-one transcription into Latin script; a raised "57" indicates where the 1857 text differs.

speakers who were 'Armenians'. But such 'ethnic' designations are problematic as they promote unfortunate stereotypes and obscure degrees of membership in different communities. It is in fact easy to imagine that these translations were intended for an audience that was not entirely homogeneous. In respect to proper names, the translations of 1857 and 1872 in particular have more 'Kurdish' characteristics than the 1891 translation (see below for examples).

39 "Kharput" (also Harpoot, or Xarpet) is just a few kilometers from the modern city of Elazig. Blincoe [2019: 241-245] discusses these and other translations, and also notes the fact that many of the missionaries were primarily serving Christian groups in Kurdistan.

40 The spelling of several words in our sample was also changed in the 1872 text.

41 There are many puzzling representations. For example, by modern pronunciations, where we'd expect the short vowel /u/ ([u~e, 0, wi]) we typically find /u/ (e.g., qmp kurr 'son' 1:23; кпф gur 'wolf' 7:15).

9 \}bi{uj ifai^tu

10 \jui lijimjlj t/uj tpo ^p tur^£ii-tfufbtmiij » tiitiijlj; q ш utr^jjrj^

ufpupu • ш inJi^ut 4 j'Pfjj Ь lJ1 и[Ь • рш if I, t£-tu u^pupu t

£tuijuflM ф о £p uitjfiL. ifijjbutuj* tjm uluIi r/-fl у p input I, t

\\ *ljiuit£ if им

п-oju/h jjn-o itiijlt-ш ujpi/riTi *

12 H1 in^jb^- ifuj г/^ii if in 11 111 iiftiipi^ iiui^ti/ii tp fi tiiiP

13 <//' UfUlfll^ inji /ifiL mtrtJiiintnptiihly filarial t Ц§ tVtu цж^

l^kp firiL- uffcjfc i/ш Ujiii i ft; J £ fuui^iuu ujfj^uj > s^pifiu

4po I,if-Uijiij ifitii mji nt- rj nI '¡'¡^'['^ nl- ^ niiitii u^ /, rnjij

¡¿uft^ui • шilji'h t

9 Anva hun vaha nimec bikhin.42

10 Ya bave me(57me) kho(57ki) li azumanday,43 nave ta aziz bibe(57biba); padi§ahithiye44 ta be; merame ta bibe(57bibi), ?havan kho li azumanda vusan ji li ser arde.

11 Nane me(57ma) hemu rroyan irro (57ji) merra bide.

12 U deyne me ji merra bex§ bikhe ?havan kho em ji bex§ dikhin ji deyndarane xorra.

Concerning the punctuation in the back-translation, Armenian ":" has been interpreted as an English period (.), "." as a semi-colon (;), and "," as a comma (,). Other punctuation has been reproduced

'as is': forwards-leaning ' typically indicates a kind of emphasis, and backwards-leaning ' indicates that something integrally related follows.

In Western Armenian, ^l represents a close (high) front rounded vowel, although this vowel is rare in the sample I have transcribed from this Gospel (another example is tf^r^rqmpmU mucuzatthan 'miracles' in Matthew 7:22). The Northern Kurdish word for 'sky/ heaven' is spelled and pronounced in many different ways, including sometimes with a vowel after the sibilant, e.g., esiman, 'asiman, etc. Compare forms in [Opengin, Haig 2014: 171]. In contrast to most of our other texts, padi^ahithi 'kingdom' is treated here as a masculine noun, shown by the masculine ezafe -ye. The feminine ezafe -a, however, does occur on some nouns, e.g., mal-a Pet'ros 'Peter's home' (Matthew 8:14).

42

43

13 U me(57me) li thercubeye me be(57me bi), le ji §erre(57?SrrS) xalas bikhe(57biki), ?hima kho etaye padi§ahithi u kuvveth u hamd ebedil ebed; amin.

English: So you (should) thus pray: O our father who (are/is) in heaven, may your name be precious/beloved/ worthy. May your kingdom come. May your goal/purpose happen, as in heaven so also on earth. Give us today our bread (of/for?) all days. And forgive us our debts as we also forgive our debtors. And do not let us be brought/ taken in trial/testing, but save us from evil/wickedness,45 because to you belong the kingdom and the power and the praise/glory forever. Amin.

The third Armenian-script version is that of 1891. According to Thomas [2000], it was part of a project "under the general direction of James L. Barton, an American missionary in Harput"46 and the work of "several Armenian pastors, including Bedros Amirkhanian, Bedros Effendi, and Kavine Aflakadian, [who] translated the New Testament and Psalms."47 This version makes use of yet more orthographic innovations: Instead of the diacritic " over m to indicate the mid central short vowel /e/ [a ~ e], m has been inverted: m. A raised apostrophe symbolizes /?/ (e.g., 'm$nL ^ppm /'afu bike/ 'forgive!'). In contrast to 4 (v) representing /v/

As in Rhea's text, so here, §err-e is used as a (feminine) oblique noun with the meaning of 'evil, wickedness'. This sense is also known in Arabic (as well as Persian and Turkish) but in my experience it is not widely known in Northern Kurdish. Today, masculine §err is widely used in the restricted sense of 'war, fighting'. Barton (1855-1936) was with the American Board of Commissioners for Foreign Missions (A. B. C. F. M.), according to https://www.bu.edu/ missiology/missionary-biography/a-c/barton-james-levi-1855-1936/ (accessed 11.07.2020).

The 1891 publication included only Matthew. According to Thomas

[2000], it was published by "the American Bible Society in Constantinople" and "the other Gospels and Acts appeared in 1911."

45

46

and /w/, now nL is used for /w/ as well as /u/ (see below on the use of j for /w/ and /u/).

9 ijuihm ift J-jr m./nrt-rrt ^шщиРт » Ъи'[г пир^ш 9

If hi (Ijim/i ifrnfr* ,po [[t ttirjifiifitfi ) i 0 tjj; ц.т фш if pj- ulpnL,n x Jomb/qpuj^

I'l'Plna tm u[k * ^mttinl-

utm * ^tu,po ^ unL,r,iihX п.риш H Щ и nip mpt/if^ г ifrnfc ^nrp n,o*}

4 2 fipo dp iPmiL-m иц/шт x fjL. mnyb£ tint dp iPmn.m 'шфг/L. "[[*,[* m * ¿шиит ffil/1 </// c/^7 штрлшшр^ jnonum

wpfi * J™ Jnrujnt /¿¿¿pnt-Ufnif?, lu[LFtJJtJ{ UJtiLU • ^o

£ q.mb m ^ипЛт^шр^ [i пи ¿jjnLmjd

hl 'РччЬР ЬщЬ^к* Riffit

9 Idi hUn Uiha nimej bikhn,

10 Ya Bave mei kho li azmani, nave te paqij bibe. XUndkharithia te be, hemde ta bi be, ?haUa kho li azman Uisa ji li ser erde.

11 Nane mei her rroj iro ji merre bi de.

12 U deyne me ji merre 'afu bikhe, ?haUa kho em ji ji deyndare xorre 'afU dkhin.

13 U me mebe thecrUbee, le ji xirab aza bi khe; 5hima kho e teye xUndkharithi U qUeth U 'izzeth ebedi. Amin

English: So you (should) pray thus, O our father who (are) in heaven, may your name be clean/pure. May your sovereignty/rule come, may your intention/will happen, as in heaven so also on earth. Give us today our everyday-bread. And forgive us our debt(s) as we also forgive our debtor(s). And do not take us into trial/testing, but rescue us from evil; because to you belong the sovereignty/ rule and the power and the honor forever. Amen.

Besides the orthographic innovations, the 1891 translation of Matthew (and the 1911 translation of Luke) uses many key terms that differ from those in the 1872 translation (the key terms in the 1872 and 1857 translations are usually identical). The impression is that, while the earlier two translations kept in mind linguistic sensitivities of readers who were culturally 'Kurdish', the 1891 translation was more 'Armenian' (compare [Rzepka 2018: 205]). This is clear in the domain of proper names, as some (but not all) names used in the 1872 translation would certainly have appealed to traditionally 'Kurdish' (i.e., primarily 'islamized') audiences, in contrast to the traditional 'Armenian' forms that replaced them in 1891. For example, for Jesus Christ, Isa El Mesth [1872] was replaced by Yisus Khrisdos ([1891] S^umu -Bp^umnu, today pronounced Hisus Khristos); for Jerusalem, Kuthsu§erif (Matthew 2:1 or Kudsu^erif in Luke 21:20) became Yerusaxem; and for the patriarchs 'Abraham', 'David', and 'Solomon', the 1872 forms Ibrahim, Davud, and Suleman became Abraham, Davith, and Soxomon (Matthew 1:1, 6). In the 1857 text, but not the 1872 one, we even find occasional use of Allah for God (©sog) and Lord (Kupiog) alongside (especially frequent) Xode 'God' and (occasional) Rabb 'Lord'.48

Besides changes in proper names, we find changes in other key terms. For the concept of 'holy', aziz 'precious, beloved,

48 In the first five chapters of the 1857 translation of Matthew, we find these forms: Allah: 1:23; 3:9, 16; 4:7, 10; Xode: 1:20, 24; 2:13, 15, 19; 3:3; 4:3-7, 10; 5:8-9, 34; Rabb: 1:22.

worthy' in 1872 became paqij 'clean, pure' in 1891; for 'will', mêram 'goal/purpose' became hemd 'intention/will'49; and for 'kingdom', padî$ahîthî 'kingdom' became xundkarîtî, roughly 'sovereignty, rule'.50 In section §4, I shall discuss some of these terms in detail, including how xundkarîtî (and related forms) are today rare in the language and how nearly all translations since 1993 use padî$ahîthî (or a related form).

2.3 Other versions, in Arabic and Latin scripts

Hermann Dalton's 1870 text is in Arabic and Latin scripts [p. 73]. Dalton attributed the Latin transcription from the Arabic to a certain Mr. Lerch, who is said to have "travelled the land of the Kurds". This must be Peter Lerch (Пётр Иванович Лерх 1827/8-1884), whose linguistic study of a collection of Kurmanji (=Northern Kurdish) and Zaza texts was published in St. Petersburg, first in Russian [1856] and then in German [1857]. There is, however, no indication who the translator was. My back-translation follows the image from Dalton's volume.51

49 As we will see in Bedir-Xan's Luke, the word is spelled hemd rather than hemd in orthographies that distinguish /h/ and /h/ (in Persian and Arabic it is written with ayin: a^). For many, the use of hemd for 'will, purpose' is odd; today it is limited to phrases like hemde xwe xeber da 'he spoke with self control, intentionally' and be hemde xwe 'accidently, not on purpose' (Nadire Efo and Emma Casim, p.c.).

50 The 1891 Matthew text consistently uses xundkar for 'king' (e.g., 1:6; 2:2; 27:11, 29, 37) and xundkarih for 'kingdom' (e.g., 3:2; 5:19). The use of xundkar - forms persisted in the 1922 and 1953 Gospel publications.

51 Dalton (p. 34, section LXVI) refers to the 1826 failed translation attempt by Bishop Schevris, also known as "the Chaldean Catholic Bishop Shevriz" (according to Thomas [2000]). The failure has been attributed to at least three reasons: (i) it was "found to be unintelligible" (at least for some speakers and to some degree; so [Canton 1904: 12-13]); (ii) there was no established writing tradition;

TO. Knrdisch (vergl. Erl. № LXVI).

jy' 3 * ^ ¿A» JU-c jJ J U L

fiir; Ux^ijj »ju jjjI j3jj* ¿1 cfJj0 3 J1*-® S", 3 *» JJl. il j L jujj^ Lk Jj^' oU L L j\j\xaj ¿¿¿¿.J

Ya Babi ma, ki lisir asmani sakin debi, navi te muqaddds bib6; jfdndkariyi te be we ¿mri te infSz u ed6r bekin liser asmani u liser ardi; nan ei her rOI irO bema bj'de; qarzid ma bibeySe euqis em debeySin qarzdarani ma; nefsi ma niviie nav tesvil hema bederfne £i bel4; milk u qudrct u delali teye Hi hingida fidti ebdiyeti. Amin!

O our Father, who lives (is established?) on heaven, may you name be holy; may your sovereignty/rule come and52 may they achieve and execute your will on heaven and on earth; give us our every-day bread today; forgive our debts as much as we forgive our debtors; do not throw our appetites into testing but ?rescue (us) from misfortune/ calamity; possession and power and glory are yours from then until forever. Amin!

A noteworthy feature of Lerch's Latin transcription is that all nominal suffixes are -i (thus /-i/), even though Arabic could also

and (iii) the multiplicity of dialects. Dalton appeared hopeful that the 1856/1857 translation of Matthew in Armenian script would have better success, but the version Dalton presented (with Lerch's transcription) is significantly different from the 1857/1872 text cited above.

52 I am reading j as /u/ rather than /we/.

be interpreted as /-e/. It may be the gender distinction between feminine and masculine nouns (as reflected in all ezafe and oblique endings being -i) had been leveled in this variety (compare Persian, which also uses only a single form).53 This leveling would be supported by the spelling of ¿i ^jjISaj^ (xundkarii te 'your sovereignty'), assuming it was original, since the second ^ (f) cannot be interpreted as a feminine ezafe (-a).

Dalton's version also involves noteworthy translation features: (1) The translators made use of couplets: (a) the first, jlASl infaz u ecer (kirin) expresses the single verbal idea of'perform' (the words are known in Persian and Arabic but today uncommon in Kurdish); (b) the second, 'from then until forever', concludes the Prayer and expresses the concept of 'eternity'. Such couplets are a popular feature of artistic Kurdish speech, both today and in the past.

(2) The Kurdish requests that one's nefs be spared 'testing'. Although nefs can refer to one's person (being) or character, it most likely refers here to one's bodily appetites (Saeed Othman, p.c.; compare also [Rizgar 1993: 132]).

(3) This translation involves two renderings for 'kingdom', first xundkari 'sovereignty/rule' and then milk. Although milk today can be used generically for a person's 'possessions' or 'property', I assume that here it had a religious connotation. Nadire Efo (p.c.) reports that in his variety, when someone is overly ambitious and attempts to do something beyond their ability, people can comment: Tu nikari ti§teki li ser erd-ezman zede ki yan kem ki, gimkl milke Xwede ye 'You cannot add

53 In Lerch's own 1857 texts, for the masculine ezafe we find -e (e.g., babe wan 'their father' p. 3) but, more frequently, -i (e.g., bera'i te 'your brother'); his texts also testify to the feminine ezafe -a (e.g., zina wi 'his wife' p. 2).

or subtract anything from heaven and earth because it (all) belongs to God.' The assumption is that everything existing ultimately belongs to God. Kamiran All Bedir-Xan (a brother of Celadet Bedir-Xan), who revised Luke's Gospel in the 1940s, also translated part of the Qur'an (Suras 1 through about 4:48), where he used milk several times in verses that state that all things belong to God.54

2.3.1 The 1922 Arabic-script Matthew text (Istanbul):

According to Thomas [2000], the 1891 text in Armenian script was revised and transliterated into Arabic script "by A. N. Andrus of Mardin and H. H. Riggs of Constantinople, both employed by the American Board of Commissioners for Foreign Missions."55 Other than the script, I have not found significant differences in other parts of Matthew, with the exception that, for some well-known Biblical individuals, Armenian proper names were generally replaced by Kurdish forms (e.g., Isaye Mesih 'Jesus Christ'; j^b Daud 'David'; p^aijji Ibrahim 'Abraham').56

The digital image below of the Arabic script is followed by a one-to-one representation in Latin script as well as my proposed interpretation in a modern Latin orthography that disambiguates

54 See, for example, Bedir-Xan's [1971] rendering of Sura 2:255: Xwedê yek e ûji wîpê veXuda nî ne ... Her ti§tên erd û ezmanan milkê wî ne. 'God is one and besides him there is no God/Lord ... All things of the earth and the heavens are his property (milk).'

55 Compare the comments on pp. 21 and 224-226 in the American Bible Society's One Hundred and Seventh Annual Report [American Bible Society 1923].

56 For 'Jerusalem', 1891 Yêrûsaxêm became Or§elîm in the 1922 Matthew Gospel, which then became Qudsa §erîf in the 1923 Mark and Luke Gospels, thus coming full circle with Kûthsû§êrîf in the 1872 Matthew Gospel.

the underspecified letters.57 In contrast to the Dalton text, this Arabic orthography distinguishes /a/ and /e/ (though /e/ is only represented in 22 out of 35 times); but, like the Dalton text, many phonemes are not fully distinguished (e.g., $ represents /w, u, u, o/ and ^ represents /y, i, e/). The text also includes inconsistencies (e.g., the spelling of 'heaven' and of 'our'). The revisions in these verses in comparison to the 1891 text are few (highlighted by underlining). Grammar changes include the use of the plural ezafe ending -ed (e.g., deyned 'debts of') and a more consistent use of the masculine oblique ending -i; both features are by no means universal features of the spoken language, but the use of the plural ezafe (-ed, or more commonly -en) is today considered standard by all writers of Northern Kurdish. The pronoun me 'us' has also been repeated in 13b. Finally, ebedi 'eternally, forever' has been moved before the phrase ye/ya te ye 'to you belong'.

57 I am interpreting the word initial combinations of as representing i and J (I) as representing e.

One-to-one Latin to Arabic Bailey interpretation

9 (e)ydy hwn wha nmyj bkn. Idl hun wiha nimej bikin.

ya bavy meyy kw l esmany, Ya Bave meyl58 ku li esmanl,

navy te paqyj bbe. nave te paqij bibe.

10 xwndkarytya te by, Xundkarltlya te be,

hmdy te bbe, hemde te bibe,

jawa kw l asmany wsa59 jy lsr 'rdy. jawa ku li asmani wisa jl li ser 'erde.

11 nany meyy hr rwj (e)yrw j mra bde. Nane meyl her roj lro ji me ra bide.

2 w dynyd me j mra 'fw bke, U deyned me ji me ra 'efu bike,

jawa kw em jy j dyndaryd xwera 'fw dkn. jawa ku em jl ji deyndared xwe ra 'efu dikin.

13 w me mbe tcrby, U me mebe thecrube,

ly me j xrab aza bke, le me ji xirab aza bike,

jma kw ebdy ya teye xwndkaryty w qwt w 'zt amyn. cima ku ebedi va te ve xundkarltl u quwet u 'izet amln.

English: So you (should) pray thus. O our father who (is) in heaven, may your name be clean/pure. May your sovereignty/rule come, may your purpose/will happen, as in heaven so also on earth. Give us today our everyday-bread. And forgive us our debts as we also forgive our debtors. And do not take us into trial/testing, but rescue us from evil. And do not bring us into trial/testing, but rescue us from evil. Because to you belong the sovereignty/rule and the power and the honor forever. Amen.

2.3.2 The 1953 text (Arabic and Latin scripts, Beirut) and the 1923 text (Arabic-script, Istanbul) of Luke 11

Finally, before turning to versions since 1993, I present the

Prayer from Luke's Gospel as published in 1953 by the Bible Society

58

I am interpreting the suffix 0j here and in verse 11 to represent not the 'primary ezafe' -(y)e but the 'secondary ezafe' -(y)i (in harmony with the 1891 translation).

59 The symbol s represents i.e., pharyngealized [s?].

of Lebanon. This 1953 publication included both Arabic and Latin scripts on facing pages. The 1953 text is presented below in three forms. For comparison, I include a one-to-one transcription of the 1923 Arabic-script Luke (from the same team that produced the 1922 Matthew text), together with a digital image of the same. The 1953 text was prepared by Kamiran All Bedir-Xan with assistance from the Dominican priest, Thomas Bois, who enjoyed a long friendship with the Bedir-Xan brothers [Blau 1985: 11-12]. As Kenneth Thomas states [2000], this translation was "a revision of the Gospel of Luke" of 1923.60 From the perspective of 'language development', this 1953 text illustrates a significant stepping stone between the older translations (i.e., the Armenian-script family of texts) and modern Northern Kurdish literature with near-modern orthographic standards.

There are both significant similarities and differences between the 1923 and 1953 Gospels. In comparing a larger sample of Luke, it is clear that Kamiran Bedir-Xan closely followed the 1923 text while also including occasional refinements in vocabulary and idiom, although this Prayer does not illustrate such refinements (the following key terms remain unchanged: paqij 'holy, clean', xundikariti 'sovereignty/rule', deyn 'debt', nimejkirin 'prayer', tecribe 'trial/test', dibex§in 'forgive').61 But as far as the Arabic-script orthography is concerned, this passage illustrates several

60 As Thomas notes [2000], Bedir-Xan (with Bois) also produced the Book of Proverbs (Metheloken Hezrete Sileman) in 1947, but the only version of Luke's Gospel that I am aware of was published in 1953, not 1947. My copy of the 1953 work is a 1984 reprint by Orientdienst, Wiesbaden.

61 Although the 1911 Luke text in Armenian script used 'afu bikhe for 'forgive' (in harmony with the companion 1891 Matthew text), the 1922 Arabic-script version replaced this with non-Semitic bbxqe (the 1922 and 1953 texts also follow a shorter Greek text, in contrast to the 1911 text). Nevertheless, elsewhere Bedir-Xan preserved the use of Semitic efuti 'forgiveness' (e.g., 1:77) and efubun 'being forgiven' (e.g., 5:20) in harmony with his 1922 source.

improvements. In contrast to the 1923 text, the 1953 text employs J for /o/, £ for non-final /ê/ and a more liberal use of a for /e/. Then, consistent with his brother Celadet's (Hawar) conventions, Kamiran's Latin script reflects yet more specification: e.g., the disambiguation of /w, û, u/ and of /y, î/ as well as the use of /ê/ in all word positions.62

1953 printed Latin script 1953 printed Arabic script 1953 one-to-one Arabic to Latin 1923 one-to-one Arabic to Latin

2 Gava ko hon nimêj bikin, bêjin: gava ko hon nmej bkn, bejn: gava kw hwn nmyj bkn, byjn,

Ya Bav, navê te paqij bibe. 4J ^slj Ь .i^j JâLj ya bav, navy te paqj bbe. ya bav, navy te paqyj bbe.

Xundikarîtiya te bê. xwndkarytya te by. xwndkarytya te by.

3 Nanê me yê her roje roj bi roj bide me; JSJ ^ JSJ ÛJSJ lia ôjJ nany me yy her roje roj b roj bde me; nany meyy hr rwj, rwj brwj bde me.

4 û gunehên me bibxeçe [sic. bibexçe], ia ^âij^S 3 w gwnehyn me bbx§e, w gwnahyd me bbx§e.

62 Not illustrated in this 1953 sample of Arabic script is how Bedir-Xan elsewhere disambiguated /h/ (a) and /h/ (using it also in native heft 'seven'; compare Persian oAa), and occasionally used /9/ (i.e., especially in Semitic words (^At 'fw 'forgive' 5:20, ^^t 'ysy 'Jesus', which is the Arabic spelling, but today people normally write the name as it is pronounced: (')isa). He also occasionally represented pharyngealized sounds (nonexistent in some varieties). This is illustrated by (1) ^ [s9] (e.g., jli^i x(i) laskar 'savior' 1:47), which even occurs in native terms (e.g., a^ sed 'hundred' 7:41 and O^>lo masi 'fish' 5:2) as well as by (2) & [z9] (e.g., ¡oik z(i)lm 'oppression' 11:39).

lewra em ji ki deyndare me ye hemiyan dibex§in; H Ьэ^ dj ¿la ¿b'^-^^ lewra em jy ky deyndary me ye hmyan dbx§n; lwra em jy ky dyndary me ye hmyan dbx§n.

й me me be tecribe. dj ¿la <da э w me me be tcrby. w me me be tcrby.

Digital image of the 1923 Arabic-script, Luke 11:2-4.

3. Versions since 1993

We can now turn to recent versions of the Lord's Prayer in Northern Kurdish. During the last thirty-five years, there has been extensive translation into the three modern literary standards introduced at the end of the Introduction (§1). The production and publication of these translations have occurred in an environment where for decades a prominent theme in Kurdish discourse is how Kurds have been politically splintered by surrounding nations. Additionally, there has been an increasing sensitivity felt by many about their dialectal diversity, especially since at times it has been viewed as creating further divisions and undermining nationalistic ideals. Finally, and perhaps least known, is the division felt by some that results from having to use different alphabets and literary standards.

In fact, as Hassanpour notes, central governments, first in Iraq in the 1930s and then later in the Soviet Union, prevented Kurds

from promoting and using Latin-based orthographies, no doubt at least partially because of fears of a trans-national Kurdish unity [Hassanpour 1992: 376-378, 458; Blau 1996: 24]. As a partial response, since the 1990s, Kurds in the former Soviet states have been gradually replacing Cyrillic with Latin,63 but a few people are still producing materials in Cyrillic script. For areas where Arabic script is used (e.g., in northern Iraq), recurring proposals have been expressed by individuals to transition to Latin script,64 but for practical reasons, Northern Kurdish continues to be published prolifically in both Latin and Arabic scripts.

With these sensitivities in mind, since the 1980s some of the Bible translators representing different varieties of Northern Kurdish and Central Kurdish (Sorani) have periodically met together, informally or in workshops, to learn from each other and to facilitate as much harmony in their work as possible. Initially, there was even hope to produce a unified Northern Kurdish translation, but

63 Older readers who learned the Cyrillic Standard in school still generally prefer the Cyrillic script, but probably most younger readers today prefer Latin script (Baris §amoyan, Tengiz Siyabendi, p.c.). This change in preference is partly due to the advancement of technology (e.g., internet, texting, satellite television). It is also the case that today in Kurdish villages in Armenia, Latin-script Kurmanji is generally taught in the schools, but there are important exceptions, especially where the teacher or community leaders identify with the conservative Yezidi community and claim not to be 'Kurdish' or want to distance themselves from Kurdish nationalism.

64 For many years now, many Kurdish internet websites, especially from Iraqi Kurdistan, have had both Arabic-script and Latin-script menus and pages, but friends have emphasized that, in the current political environment, it is doubtful that there could be governmental support to officially transition to Latin-based script. In his dissertation about the language of instruction in schools in Dohuk province, Saeid [2014] summarizes the opinions of interviewed Behdini speakers, who not only think that Behdini should be the primary language of instruction there but that L atin-script is preferred over Arabic-script.

given the different existing and well-established literary standards, it soon became clear that such a goal was unrealistic. However, such workshops allowed the translators to discover translation solutions that worked well in all of the varieties and in this way built more linguistic unity into the different translations. For this reason, there is now relatively more harmony between several of the translations than if the translations had been done independently. Distinctions in grammar, phonology and literary standards may currently be insurmountable, but lexical harmony in key terms has been partially attainable and seems appreciated by the greater Kurdish community.

In what follows in §3.1 to 3.3, I introduce the translations in our survey published since 1993. Certain features of the orthographies and literary standards are also highlighted. The discussion of key terms is delayed until §4.

The Arabic-Script Behdini Standard

The modern Arabic-Script Behdini Standard is used in the 2019 Biblica New Testament. This modern orthographic standard is the most specified among the Arabic-script texts. It makes distinctions that even K. Bedir-Xan's 1953 Arabic-script Luke did not make, including the digraph 33 (ww) that represents /û/ and ¿3 (ui) that represents /u/ [u] (a vowel lacking in many varieties). Additionally, £ is now used for all instances of /ê/ (not just non-final /ê/) and a for all instances of /e/. Only the high central vowel /i/ [i ~ a] remains unwritten (e.g., ¿,d£j bken = biker). Also, although 3 represents both /u/ and /w/ and ^ represents both /î/ and /y/, syllable structure and position often disambiguates 3 and ^ (e.g., aj^ can only be bibure, not bibwre; is^ can only be çewa, and ¿j must be jî). Furthermore, in contrast to the Latin standard, this orthography distinguishes ^ /h/ from ^ /h/ (e.g., j^ hez 'desire, will, love' versus her 'each'), and represents /x/ (e.g., ¿b bax 'garden') and /'/ (or /Ç/, e.g., 'isa 'Jesus').

:¿¡dSj jfe^ óJ-Э^ l^JdJ «jUL^dJ J ¿J d_a ¿jlj

íO-xj jcjj.xJ dj ¿álj

«O-JgJ dj Ljdjl^ilj «O-J^J d> o dj ljd>-

J l^i^jdA «^jla-^dJ J i^d^ U

.¿J ¿-U4 J4-^ .da íaj cjjJdj «da ij-Jj ^jlj

J pdj ^Sa^jdA «¿j^ da ^Jjdá J

3 da ¿J «ta ¿jjSLslj J d^Jjdj ¿la

^aljjda 3 jiA 3 ^j'djl^lj

.djdj ÓJJ jdü3jdA .ó^alj

Behdini, Bíblica 2019

9 evca hün nvejan hosa bken: babe me ye l esmanan,

bla nave te pyroz byt,

10 bla pa§ayetya te bhet, bla heza te b ch bhet,

ka ?ewa l esmany, herwesa l ser erdy jy.

11 nany tera me, evro bde me.

12 l qeren65 me bbwre,66 herwekw em l qerdaren xo dbwryn.

13 me ne'exe d taqykrne da, le me j xrabkary qwrtal bke, gnkw pa§ayety w hez w meznahy herwher yen tene. amyn.

So you must pray thus: Our father, who (are/is) in the heavens, may your name be holy, may your kingdom come, may your desire/will be fulfilled, just as in heaven, so on earth too. Give us today bread sufficient to fill us. Forgive our debts just as we forgive our debtors. Do not place us in trial/testing, but save us from the evil one/person, because the kingdom67 and the power and the greatness are eternally yours. Amen.

65 Wikiferheng ([https://ku.wiktionary.org/wiki/qer] accessed 21.06.2020) assumes qer 'debt' is ultimately derived from Arabic. Both qer and qerdar 'debtor' (and similar forms in Rhea's version) are undoubtedly related to ^>>3 q(e)rz in Dalton's version. The forms ^>>3 and jIj are also known in Persian (-dar 'having' is an Iranian suffix). The other term used in the Kurdish translations for 'debt(or)', deyn(dar), is also ultimately from Arabic.

66 The Behdini idiom for 'forgive' involves an indirect object: 'pass (over) on our debts'.

67 According to Saeed Othman (p.c.), two forms coexist: pa§ayeti (written, high) and pa§att (informal).

3.2 The Latin-Script Kurmanji Standard

The modern Latin-script standard has been used by two different translations in our survey. The orthographic standard is essentially what C. Bedir-Xan introduced in Hawar, but since the 1930s some changes in word-division rules and spelling have been introduced (e.g., the negative particles and auxiliary 'be' are now attached to the verb, the second person pronoun hon is now written hun, etc.). The most recent version using this standard is in the 2005 New Testament published by Kitabi Mukaddes §irketi (the Turkish Bible Society, Istanbul). The same translation team and publisher issued a nearly identical version in 1998 (with the four Gospels); differences from the 2005 version are indicated by a raised "98".68 During about the same period another New Testament was produced by GBV-Dillenburg (Eschenburg, Germany).69

Latin Kurmanji, Kitabi Mukaddes §irketi, 2005 and 1998

9 Hun bi vi awayi dua bikin:

Bave me ye li ezmanan, (98 bila) nave te piroz be.

10 Bila padi§ahiya te be.

Daxwaza te wek li ezmen,(98 ezmanan,) bila li ser ruye erde ji be

cih (98 bi cih be.)

11 Nane me ye rojane roj bi roj bide me.

12 U li deynen me bibihure,70 wek ku em li deyndaren xwe

bihurtine.

I was involved as an exegetical checker with this team until the 1998 publication. Since then, my primary involvement with this team has been in occasional workshops involving other teams too; this team has met many times with the Institute for Bible Translation team and regularly shared manuscripts.

The GBV translation team did not take part in our pan-Kurdish workshops but did have access to publications by workshop participants.

This is the same idiom for 'forgive' ('pass (over) on our debts') as used in the 2019 Behdini translation.

58

59

13 U me nebe ceribandine, le me ji Ye Xerab xilas bike. £imki padi§ahi, karin(98 pekarin) U rumet her u her en te ne! Amin.

You should pray in this way: Our father in the heavens, may your name be holy. May your kingdom come. May your desire/will be fulfilled in heaven as on the face of the earth. Give us our daily bread day by day. And forgive our debts as we have forgiven our debtors. And to not bring us into trial/testing, but save us from the Evil One. Because the kingdom, power and honor/fame belong to you. Amen.

Latin Kurmanji, GBV-Dillenburg, 2004

9 Ji ber ve yeke hun usa dua bikin:

Ya Bave me yi ezmanan! Bila nave te be piroz kirin.

10 Serweriya te be, mina li ezmanan, bila li erde ji viyana* te bibe. (* Daxwaz)

11 Nane me yi rojane, iroj bide me.

12 U deynen me ji me re bibex§ine, ?awa ku em ji ji deyndaren xwe dibex§inin.

13 U me nexe ceribandine, le bele me ji ye xerab xelas bike. Ji ber ku serweri, qudret u bilindahi heta abadinan yen te ne! Amin.

For this reason you should pray like this: O our father in the heavens, may your name be holy. May your sovereignty/rule come, as in the heavens, let your desire happen on earth. Give us today our daily bread. And forgive our debts as we also forgive our debtors. And do not put/throw us into trial/testing, but save us from the evil one. Because the sovereignty/rule, power and highness belong to you until eternity. Amen.

3.3 The Cyrillic-Script Kurmanji Standard

The Cyrillic-script standard has been used in the 1993, 2000, and 2011 publications of the Institute for Bible Translation, Moscow. Since the dissolution of the Soviet Union, demand has steadily increased for Latin-script versions of this material, and so each of these Cyrillic-script publications has been simultaneously produced in a Latin script. The Latin script used in these publications mostly conforms to the Bedir-Xan Latin-script standard, with the notable exception of the special use of the apostrophe. As indicated in the alphabet table presented earlier, the apostrophe has the following functions in these IBT Cyrillic- and Latin-script publications: On the stops /n' (p'), t' (t'), k' (k')/ it indicates the aspirated phonemes in contrast to corresponding unaspirated phonemes; on the affricate /h' (?')/ [tj] it indicates the unaspirated phoneme, as this phoneme is rarer than aspirated /h (?)/; on /p' (r')/ it indicates the trill, in contrast to the flap, /p (r)/; on /h' (h')/ it indicates [h] (i.e., h and in contrast to /h (h)/; and on /a' (e')/ it indicates a pharyngealized [a~3] (comparable but not identical to /?/ and ^).71 The 2011 version of the Prayer is presented below (it is identical to the 2000 text except in capitalization); differences in the 1993 version are indicated by raised "93".

Cyrillic Kurmanji, Institute for Bible Translation, 2011, 2000, and 1993

9 He hyH aha goa 6bKbH: Le hün aha dua bikin:

EaBe M3HH 8'3MaHa! Bave meyi E'zmana!

HaBe Ta nupo3 6a,* Nave Te piroz be,*

71 The Cyrillic Standard prescribes against writing an apostrophe with other vowels since pharyngealization with other vowels is infrequent. But in Курдоев^ (K'urdo's) discussion of /eV [1957: 17], he lists И'са 'Jesus' as one of the relatively rare words in his variety where pharyngealization occurs with other vowels (i.e., /i, i, o/). Thus, for many readers of these Bible texts, the written form Иса (and Isa) is pronounced as [ i?sa ~ 9isa ].

10 П'адшатийа Тэ бе, э'мьре Тэ бэ, 4awa ль э'змен,

оса жи ль сэр э'рде.

11 Нане мэйи р'оже р'ож бь р'ож бьдэ мэ.

12 Дэйнед мэ бьбахшинэ, 4awa ко эм дьбахшиньнэ дэйндаред хwэ.(93 4awa к0 эм жи

дэйндаред xwe дьбахшиньнэ.)

13 У мэ нэбэ щер'ьбандьне, ле мэ жь йе хьраб* хьлаз кэ, чьмки п'адшати, qэwaт(93 п'адшати) у р'умэт йед Тэ нэ, h'этa-h'этaйе. Амин

*Бь готьнэкэ дьн: «Бьра инсан qэдьре наве Тэйи пироз бьгьрэ».

**Aha жи те фэ'мкьрьне: «Жь мирещьн».

P'ad§atiya Te be, e'mire Te be, gawa li e'zmen,(93 ezmana) usa ji li ser e'rde. Nane meyi r'oje r'oj bi r'oj bide me.

Deyned me bibax§ine, gawa ku em dibax§inine deyndared

XWe (93 gawa ku em ji deyndared xwe dibax§inine.)

U me nebe cer'ibandine, le me ji ye xirab* xilaz ke, gimki p'ad§ati, qewat(93 qewat-p'ad§atl) u r'umet yed Te ne, h'eta-h'etaye. Amin. *Bi gotineke din: «Bira insan qedire nave Teyi piroz bigire».

**Aha ji te fe'mkirine: «Ji mirecin».

But you should pray thus: Our Father of the heavens! May Your name be holy,* may Your Kingdom come, may your will happen, as in heaven so on earth. Give us our daily bread day by day. Forgive our debts as we (1993 also) forgive our debtors. And do not bring us into trial/testing but deliver us from the evil (one),** because the kingdom, power and honor belong to you, forever. Amen.

(Footnotes: * In other words, 'May people show respect for your holy name'. **This can also be understood as 'from the demon-prince'.)

4. Translation issues and vocabulary evolution

The following table (p. 202) allows comparison of several of the key terms in our surveyed texts. A long dash "—" indicates a term that the surveyed text lacks (e.g., several of our texts lack the word 'pray!' since they lack the introductory line 'Pray in this way!').

I now offer some longer comments on the renderings for 'holy', 'kingdom' and 'father', as well as shorter comments on some of the remaining terms. As mentioned earlier, at least since the 1920s, there has been a general trend to purify Kurdish of foreign loans, especially Semitic ones, at least in formal writing, and this trend can at least partially be viewed as a reaction to political and social domination by foreigners [Hassanpour 1992; Hasanpoor 1999]. Moreover, this trend is also evident in these modern translations, albeit, to differing degrees.

The concept of 'holy'72 has been translated by four different words in our texts. Several, including the oldest, used the Semitic borrowing muqaddes (Garzoni, Dalton, and Rhea), which is well known across the region in the religious domain. The 1857/1872

72 Translators often struggle with 'holy', especially where there is no established translation tradition to which they can resort, and their job is complicated by the complexity of the Biblical concept of holiness. Regarding the meaning of 'holy' in the line 'May your name be holy', [Wierzbicka 2001: 237-238] reduces the kernel sense to the 'goodness' and 'uniqueness' of God, paraphrasing 'holy' propositionally as (i.e., in what she has famously promoted as "conceptual primes" and "universal human concepts"): "(a) I know: You are someone good; (b) no one else is like You; (c) nothing else is like You." To this kernel sense I would add that the Biblical concept of 'God's holiness' is expected to evoke feelings of awe, i.e., fear and respect (Exo 3:5-6, 15:11, Psa 111:9-10, Pro 9:10, Isa 8:13, Rev 15:4), which in turn compels one to conform to God's holiness and act in ways that distinguish oneself (or 'separate' oneself or things, including cultic articles) from what is sinful or common (Exo 3:5-6, Lev 20:22-26, 2Co 6:14-18).

Summary table of key term renderings

to o to

father holy kingdom pray will debt forgive tempt/ test power glory

1787 Garzoni bab mvkaddas baehscte (behi§t) 'paradise' — meram (ghuna 'sin') afu beka tegerib (tecerib) — —

1857, 1872 bav aziz 'precious, beloved, worthy' padi§ahithi nimec meram deyn bax§ bike thercübe küvveth hamt

1870 Dalton bab muqaddas xundkari / milk — emr qarz bibexse tesvil qüdret delal

1872 Rhea bab mokadas padishahi — kar (qer) b'bora tijerib kudset (kudret) jelal

1891, 1922 bav paqij' clean, pure' xundkarithi nimej hemd (fiemd) deyn 'afu bike thecrübe qüeth 'izzeth

1953 Bedir-Xan (Luke) bav paqij (2x elsewhere miqedes) xundikariti nimej (elsewhere fiemd, daxwaz)73 deyn bibex§e tec rib e (elsewhere qiwet) (elsewhere izet, rumet)

1993, 2000, 2011IBT bav piroz p'ad§ati du a e'mir deyn bibax-§ine cer'i-bandin qewat rumet

1998,2005 KM§ bav piroz padi§ahi du a daxwaz deyn bibihue ceri-bandin pekarin/ kar in rumet

2004 GBV bav piroz serweri du a viyan deyn bibex-§ine ceri-bandin qudret bilindahi

2019 Biblica bab piroz pa§aieti nvej fiez qer bbure taqikrn hez meznahi

öd

et" Vi

texts then used aziz 'precious, beloved, worthy' (implying deserving of 'respect'), and the 1891/1922 texts usedpaqij 'clean/pure'. Bedir-Xan [1953] also used paqij in the Prayer for 'holy' and usually elsewhere in Luke's Gospel, though twice he used miqedes (1:49, 70)..74 Since 1993, however, all published translations in Northern Kurdish, as well as Central Kurdish, have used piroz in the Prayer. In one of the translation workshops in the early 1990s (described earlier), the Kurdish translators (from different regions and working in different literary standards) soon agreed that piroz was the appropriate 'Kurdish term' for 'holy' to be used throughout the Bible.75 This decision can therefore be described as a quasi

73 As Luke's version of the Prayer is shorter, certain key terms do not occur. So the terms included here in parentheses come from other passages in the 1953 text. For 'will', Bedir-Xan used in Luke 22:42 hemd and in Luke 12:47 he used both hemd and (what is an obvious improvement by today's standards) daxwaz (1922 has only hemd). For Suva^n; 'power' (in the doxology), he usually used qiwet (e.g., 21:27 'coming in a cloud with power and glory'); and for 5o^a 'glory' he usually used izet (2:9, 4:6, 9:26; 12:27; 14:10; 17:18; 19:38; 21:27; 24:26), but occasionally other words: rumet (2:14; 1922 'zzt), serbilindahi (2:32; same in 1922), and celal (9:31-32; same in 1922).

74 In his partial translation of the Qur'an, Bedir-Xan [1971] also used miqedes and paqij to render the concepts of 'holy' and 'clean/pure'. For miqedes, see Sura 2:87, 143-144, 194; for paqij, see 2:125, 129, 222, 232; 3:42. In his translation of the book of Proverbs (Metheloken Hezrete Sileman, 1947), he used the word ewli(ya) (Proverbs 9:10, 30:3).

75 As I recall, in early workshops in the late 1980s,paqij 'pure, clean' was initially a close contender to muqedes, given Bedir-Xan used it. But in February 1992 (in Germany), when a Sorani translator pointed out that they were using piroz in their Gospel drafts, the Northern Kurdish translators soon agreed to also use this term, although the decision was not finalized until August 1994 (in Bulgaria). Interestingly, the IBT translation team, originally based in Armenia, had until 1991 been using ziyaret (see the photograph at the end of this article, which shows the Lord's Prayer on page 15 of the handwritten manuscript by Emma Casim). For most

cross-dialectal consensus, where writers — Bible translators — agreed to use a term that is perceived as pure Kurdish because it is obviously not of Semitic or Turkish origin and because it is an established term.

The term piroz [piroz] is, however, not without difficulty as a translation solution, although it is interesting semantically. It is obviously related to Persian jsj^j piruz [phiruz] but, for most Northern Kurdish speakers, it lacks the sense of 'victorious' as in Persian, and so the idea of 'holy' involves a significant semantic shift from Persian. The point of similarity seems to be that 'to be victorious' (Persian) is 'to be blessed (by God)' (Kurdish), and someone or something that is blessed may also be 'holy'.76 In any case, Kurdish piroz is commonly used in blessings and congratulations77 (but Persian piruz is apparently used much less so, if at all, though in one well known phrase for some speakers it would count as a blessing: jsj^j jbjsjsj newruze tan piruz 'happy/ blessed new year to you'). Thus, the concepts of holiness and

Northern Kurdish speakers, ziyaret is a noun meaning 'visiting' but it is also used to refer to the shrines and graves of holy people, where sacrifices are made and which confer blessing on visitors. As Nadire Efo and Emma Casim (p.c.) note, muqades was not well known in their (Transcaucasian) Yezidi community, and piroz was primarily a synonym for bimbarek 'blessed'; in contrast, ziyaret seemed capable, by semantic extension, of expressing the concept of 'holy'. The 1993 IBT version of Matthew used piroz in 6:9 but used ziyaret in the phrase 'Holy Spirit' (1:18, etc.), bimbarek in 'holy city/place' (4:5; 24:15), and bihurti in 'holy things' (7:6). But after that publication, ziyaret was retired.

76 Besides the semantic differences, it may be that the phonological differences (Northern Kurdish [piroz] versus Persian [phiruz]) also suggest that the linguistic connection is relatively distant.

77 Glosses like 'blessing' and 'congratulations' are the most common ones found in Kurdish dictionaries for piroz. Fewer include glosses like 'sacred, holy' [Rizgar 1993: 148] or "1. muqedes 2. bimbarek" [Demirhan 2007: 306].

blessedness are closely associated in Northern Kurdish. What this implies is that piroz can be ambiguous: A phrase like Ruhe Piroz (or Giyane Piroz) which is meant to translate 'The Holy Spirit', can also be understood as 'The Blessed Spirit'. Therefore, in order to make a distinction, some (but not all) of the translation teams have reserved piroz for the concept of 'holy' (e.g., ayiog) but have used distinct terms for 'blessed' (e.g., su^oyn^svog).78

Beyond the issues of piroz, the rendering of the clause 'May your name be holy' in probably all of the Kurdish translations is not transparent in meaning, although for liturgical purposes, it sufficiently mimics the typical structure of translations in other languages. Several of the translations thus include an explanatory footnote; for example, the 2011 IBT has this note: Bira insan qedire nave Teyipiroz bigire 'May people show respect for your holy name,' which shows that this petition is to be understood as a prayer that God, who is uniquely holy and perfect, be praised and honored in the lives of the disciple and others. Moreover, a complementary element is likely implied in the petition, involving not only the role that people have in 'sanctifying' God's name, but also God's role [Bivin, Tilton 2011]. The 2005 TBS text has essentially the same note as the 2011 IBT text but also refers the reader to Ezekiel 36:23, which reflects well the complementary roles (see Ezekiel 36:20-32).

For the term 'kingdom', I have already commented on Garzoni's use of baehscte (behi§t) 'paradise' and its conceptual link to God's 'heavenly kingdom/domain' (e.g., melekute asman); and I have also noted how Dalton's text used both (a) xundkari 'sovereignty, kingdom' as well as (b) milk 'property', which was likely counted

78 For example, in Matthew 21:9, for 'blessed' the IBT translations have bimbarek, the 2019 Biblica Behdini text has bereketdar, but the 2016 Biblica Sorani (Central Kurdish) text and the TBS Kurmanji have piroz. For the verb 'sanctify, consecrate' (âyiâÇœ), when piroz kirin could be construed as 'congratulate', the IBT translations use buhurtî-jibare kirin 'make holy-separate' orpaqij kirin 'purify'.

as a religious term referring to God's universal possession of all things (milk might also be recognized by Kurds knowing Arabic as related to melekut). In contrast, modern translators (except for GBV 2004) have all chosen pad(i)§a(h/t)i, which was also used in the 1857/1872 translations. It is unknown why the 1891-1923 translators chose instead xund(i)kar- forms, a choice that Bedir-Xan [1953] perpetuated.79 It may be that xund(i)kar- seemed more appropriate to describe the sovereign rule of God in contrast to pad(i)§a(h/t)i, which might connote a mere human king(ship). Today, xund(i)kar- is barely known (and confused with terms for 'reader': xwendevan, xwendekar), although I assume it derives ultimately from Persian j^Li- xodavand-gari 'sovereignty, lordship'.80 Pad(i)§a(h/t)i is well known and is transparently derived from (equally well known) pa(di)§a(h) 'king'. The terms are also manifestly Iranian rather than Semitic or Turkic.

Again, beyond the choice of a word for 'kingdom', the assertion of this second petition ('may your kingdom come') along with its intended function in the disciple's spiritual life, can be opaque in an Islamic context, where it is assumed that all people and things are already subject to God's sovereignty [Brown 2000: 42]. Thus,

79 In the Qur'an portions that Kamiran Bedir-Xan translated, first published in the periodical Hawar (edited by Kamiran's brother, Celadet Bedir-Xan) I have not found, xund(i)kar-, but Sura 3.26 uses padi§ahi. There are also many instances of padisa(hi) in the later volumes of Hawar [1941-1943].

80 Hayyim [1934: V1: 742] defines Persian "jlSL^ (khonkar)" as an "[o] ld title of the Sultans of Turkey. An emperor; & monarch" and adds that this is a "[c]orruption of jH^Ia^" (i.e., xodavand-gar). Turkish hunkar is defined as 'sovereign, sultan' in the Redhouse dictionary [1987]. The cognate xunkar is also known in Kurdish but rare, and, according to Saeed Othman (p.c.), is perhaps best known as a description for the very wealthy (compare Hayyim's entry for aS^Ia^ xodavand on p. V1:701). Jaba & Justi [1879: 166] give the French glosses of "souverain, monarque" for "jlSAj^ khoundkar".

both lines in the Prayer employing 'kingdom' are easily lost on a mostly Islamic audience. Rather than concerning "the realm or territory over which [God] rules," the petition is that God establish His reign on earth, which is "an activity" God is gradually bringing about, and the implication is that the disciple should submit to and promote "God's kingly rule" (see [Newman, Stine 1988] on Matthew 3:2). Moreover, the nature of the divine kingdom is wholly different from human kingdoms ruled by selfish despots, as this kingdom is founded on the principles of divine love and good will towards all.81

Concerning the rest of the terms, I can state that, for the translations since 1993, there is more use of what are considered originally Semitic loans in the IBT, TBS and GBV translations than in the Biblica (Behdini) translation. For example, for 'pray', since 1993, only the 2019 Biblica text has used nvej (a form of nimej), which is counted as 'pure Kurdish',82 while the other translations have used the originally Semitic loan dua because the translators felt that nimej best applies to ritual prayer (especially Islamic), while dua describes spontaneous, heart-felt prayer (and is not exclusively Islamic), and because these translators understood that Christ's prayer should be viewed more like dua. For 'power', the IBT and GBV translations have also used terms that are considered Semitic (IBT: qewat and GBV: qudret) on the grounds that they are the normal terms and widely known. In contrast, the Behdini translators [2019 Biblica] have used a term (hez) that in recent decades has gained ground as a literary substitute. These Behdini translators would say that a Semitic loan would be used only as a last resort, that is, when no other appropriate and sufficiently known Kurdish

81 Modern Kurdish conceptions of 'king' and 'kingdom' typically conjure up despotic images of totalitarian authorities.

82 Thomas [2015: 364] also notes that for 'prayer' many Persian translations have favored either "the Pahlavi word jLi namaaz" (i.e., a term traceable back to Middle Persian) or "the Arabic word Itj do'aa. "

term existed. Such Semitic terms in the Behdini Prayer include qer 'debt', as well as hez 'will, desire' and taqi(krn) 'test/trial', even though some of these terms have through a natural course of use been so altered in form or meaning that even Behdini speakers who know Arabic well do not recognize the connection. Of all the Northern Kurdish translation projects, the Behdini translators (from Northern Iraq) live in the closest proximity to the Arabic linguistic world, and of all of the Northern Kurdish communities, writers from this community are typically the most sensitive to all things Arabic.

Finally, several comments can be made about the key term 'Father'. All of the translations, from 1787 until today, have used the same kinship term, bav (or bab), which for Northern Kurdish speakers is the normal (if not only) term for biological father. Nevertheless, the received metaphor, that 'God is Father', and its typical Biblical implications remain opaque for many people. Typical Biblical implications include that God, who is wholly other, may be addressed not only as a person,83 but as a father, because He cares for people as a father cares for his children (Psalm 103:13, Proverbs 3:11-12), and even desires an intimate relationship with people comparable to that between a father and a child (Luke 11:9-13, Jeremiah 31:33-34, Hebrews 12:5-6). This inspired metaphor has been resisted by many for centuries, not only in the Middle East but also in increasing degrees in the West where ever more distant and impersonal images of God have gained ground. Nevertheless, the language has been embraced by many in Kurdistan and surrounding countries. Although the Messiah's use of the title 'Father' and his instruction to use it when addressing God in prayer is distinct, the metaphor has ancient roots in the Torah's language about 'covenant', that is, how God may treat

83 See [Wierzbicka 2001: 234-236] on the implications in this prayer that God is not an abstract power but a personal 'someone', while still wholly other and unique.

people as family, binding Himself to them as their eternal kin, and cementing a relationship based on bidirectional love and loyalty [Cross 1998; Brown 2001; Farrell 2004; Hahn 2009]. Through his deeds and parables, the Messiah taught that God is less like a stern judge and more like a loving father who welcomes home a wayward child who comes to his senses and returns to a father's embrace (Luke 15, Matthew 9:9-13, John 8). He taught his followers to pray to be delivered from evil, although pain and evil are, as the Messiah well knew, inevitable. In the Messiah, in both his life and death, is the true embodiment of the kinship (parent-child) metaphor, as the Messiah is the divine-human link for all who have the grace to recognize God as their spiritual Father, who is perfect in holiness and love.

5. Conclusion

This survey of the Lord's Prayer in Northern Kurdish has provided several glimpses into the ways Northern Kurdish has evolved in both vocabulary and written form during a period of over 230 years. Without a doubt, Kurdish identity and evolving Kurdish nationalism have left their stamp on the development of the written language — on the orthographies and literary standards as well as on the vocabulary used in written publications. The different translation teams have kept in step with these linguistic trends, and their published translations, especially since 1993, also reflect the translators' desire to harmonize with each other, especially in vocabulary, whenever possible, and thereby promote a pan-Kurdish vocabulary that spans several varieties. We can expect that future Bible translations will continue to be influenced by such trends as Kurdish continues to be a living and ever developing language for millions of people.

PA v ль э'нишке к'учл бьсвкьньн да»' бЬкъи,ко жь к'ьдаш бьи

р'дот жь wsp'л дьбежьм:

Э

swahi\ мдше xv^e ид а ст^-НДоН.

Д-^р-ц^а-я-л-ЯШ1

6 V Г«6А Т D КО ДПА'ДЬКН БЬК'в&д ОДА XW6 ЬйнДР'вД

Даре бьчйФьпин ,, у

ДОЛ' еькй *ь &Аве *wap'A, ко щйе ув^лэтийа Еиве Га, КО 6Ь ДЬЗИбА ДЬБИнв,

we Ашкврв бьдэ те.

ч v глва ко hvn дй а1 дь-КЬН , минй п'уг'л'йрьза п'ьр:

,*эбэр _надь.,ч., , чьм&и е^анд 1|'<9лмин дькьн ко жь Б0йа хвьарДАНА . xwa vve. ee Ььп истом .

!?■ /WA hYH (ЧИНЯ WAH НЭ-

бьи 7 чьмки Блае здна чь,.кй wa лазьмэ, ьа баси ьибикрна we-

« Б А В е М Э • ■ • » 9

Иди A\VA ЬУН АОЛ бЬКЬН-

бдве май э'змлн« ! идаеТе Зйдрэтбб; ■10 П'адшагйй Те ее ¿'мре Твой

ЧАИ/А ко в'ЗМАНА, 0СА-ЖИ

сер в'р'де; /•/ н/ше май nsp'd р'ож бо р'о* бьда мэ •

ч XtiV

iZ У бЬбАШинв деине мэ ,

У.1М

ко ем жи дьбАш*инна Мвиь

д/\р<г xwa;

•в V мая нэбэ^цер'ьБлндьне,

ле хрАбйе азлк« бЬкб;

чонки йд Тэйа qaw/дт, П а-

лшати v шькьри, нет та

h'STre . Амин, м/ вънад

ЖЬ VVAHP'A ACPV бькьи , БАВ€

wati дзмАмА жи жь wap> a®v бьке • ■IS lie wd<и пун сущя

ЖЬ VVA HjP'A А<РУ нэкьн, бдве wa Жи жь \л-эр'а суще

W<? ЛФ.У НЛК.Э,

/5

Matthew 6:5-16 from page 15 of the handwritten manuscript produced by Emma Casim in the 1980s in Soviet Armenia as she assisted Nadire Efo.

Published Bible Translations

(Unless noted otherwise, these publications are in Northern Kurdish.)

1857. Fb&fL, fromtt dt fum ti_ Utu^ht Up^BumUmpU mtupt Ummptnu Umpqou Lnrqmu rn. ¿mUUm [INCIL, Xodee me Isa El Mesihe nivisandin bi desthe Madthews Markos Lukas u Hanna. 'Injil: Our Lord Jesus the Messiah wrote by the hand of Matthew, Mark, Luke and John']. Istanbul. Accessed 01.01.2020. URL: https://archive.org/details/KurdishNTHayadar/page/n21

1872. Otjtfmht Jbo t foofrjt duf ^m fci Utufrh [P'eymane No e Xoiye Me Isa El Mesih. 'The New Testament of our Lord Isa AlMessih']. Istanbul. Accessed 01.01.2020. URL: https://archive. org/details/peymanenoekhoiye00blis/page/14/mode/2up

1891.Mififiiui Puiujuit duifi S^unLu ^p^umnu £o Uuirnptnu "b^frufr [incila Rabbe mei Y(H)isus Khrisd(t)ws Kho Madthews (Mattevos) Nvisi. 'The Gospel of our Lord Jesus Christ that Matthew wrote']. Istanbul. The American Bible Society.

1911. fU^^m fm^^t tfuifr 3^unru Pp^umnu po Lnr^mu L^u^ [Incila Rabbe mei Y(H)isus Khrisd(t)ws kho Lug(k)as Nvisi. 'The Gospel of our Lord Jesus Christ that Luke wrote']. Istanbul.

1922 . ^.o U*—ai J o ¿¿^.^il [incyla 'ysy yy msyh lsr nvysyna mty. 'The Injil of the Messiah Isa (Jesus) according to the writing of Matthew']. Istanbul. Accessed 11.07.2020. URL: http://bnk.institutkurde.org/images/pdf/6X21EJCT3N.pdf

1923. l§sJ L*—.¿ai j—J -o .¿¿^.^il [incyla 'ysy yy msyh lsr nvysyna lwqa. 'The Injil of the Messiah Isa (Jesus) according to the writing of Luke']. Istanbul. Accessed 11.07.2020. URL: http:// bnk.institutkurde.org/images/pdf/6X21EJCT3N.pdf

1953. Incila Luqa Is^J ['The Gospel of Luke' (two script publication)]. 1984 reprint by Orientdienst, Wiesbaden. Copyright: Bible Society of Lebanon.

1993. M3rHHHHa Hca Macnh' nb ropa MaTe. Mizginiya Isa Mesih' li gora Mete. ['The Good News according to Matthew' (two script publication)]. Institute for Bible Translation. Stockholm.

1998. Încîl (Mizgînî). ['Injil (The Good News)' (The Four Gos-pels)]. Kitabi Mukaddes Çirketi. Istanbul.

2000. Mb3rHHH: nnÖMam Hy (Hh^hh). Mizgînî: Peymana Nû (Încîl). ['Good News: The New Testament (Injil)' (Cyrillic and Latin script editions published separately)]. Institute for Bible Translation. Moscow.

2004. Kitêba Pîroz: Peymana Kevin û Peymana Nû. ['The Holy Book: The Old Testament and the New Testament']. GBV-Dillenburg. Eschenburg.

2005. Încîl (Mizgînî). 'Injil (The Good News)']. (New Testament.) Kitabi Mukaddes Çirketi. Istanbul.

2011. Mb3rHHu: nnHMarn Hy (Hh^hh). Mizgînî: Peymana Nû (Încîl). ['Good News: The New Testament (Injil)' (Revision of2000. Cyrillic and Latin-script editions published separately)]. Institute for Bible Translation. Moscow.

2014. ajija ¿a>jj - 'The Holy Bible - The New

Millennium Version'. (Persian.) Elam Publications.

2016. jcj^j [ktêbî pîroz. 'The Holy Book']. (Sorani, Central Kurdish Bible.) Biblica.

2019. ^-jaâdj - ¿jLaj v ^a—^-a ô-l—-a^- U-aU-0 _ jsj^r! [incyla pyroz - mzgynya 'ysayê mesyh b zmanê kurdy-behdyny. 'The Holy Injil - The Good News of the Messiah Isa (Jesus) in Behdini-Kurdish']. (New Testament.) Biblica, Erbil, Iraq.

References

Adelung J. C. Mithridates oder Allgemeine Sprachenkunde mit dem Vater Unser als Sprachprobe. Berlin, 1806.

American Bible Society. Annual Report of the American Bible Society, One Hundred and Seventh. New York, 1923. URL: https:// babel.hathitrust.org/cgi/pt?id=wu.89065732869&view=1up&seq=5 (accessed 12.02.2021).

Bailey N. A., Bailey D. L. A brief technical description of the sounds in Kurmanji-Kurdish with practical tips for the foreign language learner. Unpublished paper. Herborn, Germany, 1992.

Barry D. Pharyngeals in Kurmanji Kurdish: a reanalysis of their source and status // S. Gundogdu, E. Opengin, G. Haig, E. Anonby (eds.). Current issues in Kurdish linguistics. Bamberg Studies in Kurdish Linguistics 1. Bamberg, 2019.

Bedir-Xan C. Hawar (periodical), volumes 1 to 57. Damascus, 1932-1935, 1941-1943. URL: https://bnk.institutkurde.org/ (accessed 31. 07.2020). (In Kurdish)

Bedir-Xan K. Ji tefsira Qurane. Paris, 1971. (Hand typed manuscript). URL: https://en.calameo.com/books/005071515a481f4679416 (accessed 20.06.2020 (Essentially the same text was also published bit by bit in Hawar, volumes 27-57, 1941-1943.) (In Kurdish) Bivin D. N., Tilton J. N. Lord's Prayer // Jerusalem Perspective. Jerusalem, 2011. URL: https://www.jerusalemperspective. com/769/ (accessed 15.09.2020).

Blau J. Biranina Thomas Bois [Remembering Thomas Bois] // Hevi: Kovara gandiya gi§ti. Paris, 1985, 4:11-13. URL: https:// bnk.institutkurde.org/images/pdf/BVXUEJG3ZH.pdf (accessed 14.02.2021). (In Kurdish)

Blau J. Jaziri // EncyclopediaIranica, 2008. URL: http://www. iranicaonline.org/articles/jazir (accessed 23.06.2020).

Blau J. Kurdish Language, ii. History of Kurdish Studies // Encyclopedia Iranica, 2009. URL: http://www.iranicaonline.org/ articles/kurdish-language-ii-history-of-kurdish-studies (accessed 07.07.2020).

Blau J. Kurdish written literature // P. Kreyenbroek, C. Allison (eds.). Kurdish culture and identity. London, 1996, 20-28.

Blau J. Le Kurde de 'Amadiya et de Djabal Sindjar: Analyse linguistique, textes folkloriques, glossaires. Paris, 1975.

Blincoe R. Ethnic realities and the Church: Lessons from Kurdistan (2nd edition). Littleton, Colorado, 2019.

Brown R. The "Son of God": Understanding the Messianic titles of Jesus // International Journal of Frontier Mission, 2000, 17:1.41-52.

Canton W. A history of the British and Foreign Bible Society. Vol. 2. London, 1904. URL: https://archive.org/details/britishbible soc02cantuoft/page/n7/mode/2up (accessed 11.02.2021).

Celîlov H. M. Dîroka R'ojnema kUrdî ya here femendirêj "R'ya Teze" [The history of the longest living Kurdish periodical 'R'ya T'eze']. R'ya T'eze, 2020.3.3-4. URL: https://krd.riataza.com/ wp-content/uploads/2020/04/mart-2020.pdf (accessed 22.06.2021). (In Kurdish)

Chyet M. L. Kurdish-English Dictionary. Ferhenga Kurmancî-Inglîzî. New Haven, 2003.

Cross F. M. From Epic to Canon: History and literature in Ancient Israel. Baltimore, 1998.

Dalton H. (ed.). Das Gebet des Herrn in den Sprachen Russlands. St. Petersburg, 1870.

Demîrhan U. Ferhenga Destî, kurdî bi kurdî. Dogubayazit-Agri. 2007. URL: https://www.academia.edu/473687/Ferhenga_ Destî-kurdî_bi_kurdî (accessed 23.02.2021). (In Kurdish)

Farrell T. Metaphor in Matthew's Gospel: Its cognitive structure, background, culture dependence and translatability. Roehampton University dissertation, 2004.

Fattah I. K. Les dialectes kurdes méridionaux: Étude linguistique et dialectologique. Leuven, 2000.

Garzoni M. Grammatica e vocabolario della lingua Kurda. Rome, 1787. URL: https://archive.org/details/b28777086/ (accessed 05.07.2020).

Hahn S. W. Kinship by covenant: A canonical approach to the fulfillment of God's saving promises. New Haven, 2009.

Haig G., Matras Y. Kurdish linguistics: a brief overview // STUF - Language Typology and Universals, 2002, 55:1.3-14.

Haig G., Opengin E. Kurdish: A critical research overview // G. Haig., E. Opengin (eds.). Kurdish linguistics: Focus on variation and change. Kurdish Studies, 2014, 2.2., 99-122.

Haig G., Opengin E. Kurmanji Kurdish in Turkey: Structure, varieties, and status // C. Bulut (ed.). Linguistic Minorities in Tur-

key and Turkic speaking minorities of the peripheries. Wiesbaden, 2018, 157-229.

Hasanpoor J. A study of European, Persian, and Arabic loans in Standard Sorani. Uppsala University dissertation, 1999.

Hassanpour A. Bahdïnân // Encyclopedia Iranica, 2011. URL: http://www.iranicaonline.org/articles/bahdinan-kurdish-region-river-dialect-group-and-amirate (accessed 07.07.2020).

Hassanpour A. Nationalism and Language in Kurdistan 19181985. San Francisco, 1992.

Hawar (journal in Kurdish). Damascus, Syria, 1932-1935 and 1941-1943. URL of volumes 1-26: https://bnk.institutkurde.org/ images/pdf/C6VQGNJ31S.pdf and of volumes 27-57: https://bnk.institutkurde.org/images/pdf/DX8L5PTACN.pdf (accessed 12.10.2020). (In Kurdish)

Hayyim S. New Persian-English. Tehran, 1934. URL: https:// dsal.uchicago.edu/dictionaries/hayyim/ (accessed 11.07.2020).

International Organization for Migration. Demographic Survey: Kurdistan Region of Iraq, 2018. URL: https://reliefweb.int/ report/iraq/demographic-survey-kurdistan-region-iraq (accessed 11.02.2021).

Jaba M. A., Justi M. F. Dictionnaire Kurde-Français. St. Pe-tersbourg, 1879. URL: https://bnk.institutkurde.org/images/pdf/ V4XCUZ5PFY.pdf (accessed 12.10.2020).

Jastrow O. Zur Phonologie des Kurdischen in der Türkei // Studien zur Indologie und Iranistik, 1977, 3:84-106.

Kahn M. Borrowing and variation in a phonological description of Kurdish // Natural language studies 22. Phonetics Laboratory. University of Michigan, Ann Arbor, dissertation, 1976.

Kreyenbroek P. Kurdish written literature // Encyclopedia Iranica, 2005. URL: http://www.iranicaonline.org/articles/kurdish-written-literature (accessed 23.06.2020).

K'urdo Q. Zimanê Kurdî: Rêziman. [The Kurdish language: Grammar] Frankfurt, 1981. (Published by Komkar, with unfortunate mistakes in the phonetic section). (In Kurdish)

Leezenberg M. Vernacularization as governmentalization: The development of Kurdish in Mandate Iraq // H. Murre-van den Berg, K. Sanchez & T. C. Baarda (eds.). Arabic and its alter-natives: Religious minorities and their languages in the emerging nation states of the Middle East (1920-1950). Leiden, 2020. URL: https:// www.academia.edu/39082080/ (accessed 14.08.2020).

Lerch P. Forschung über die Kurden und die iranischen Nord-chaldäer. St. Petersburg, 1857. URL: https://archive.org/details/ bub_gb_WlGVYoEkr7sC/mode/2up (accessed 10.06.2020).

MacKenzie D. N. Kurdish Dialect Studies. Vol. 1. Cambridge U.K., 1961.

New Redhouse Turkish-English Dictionary. Istanbul, 1987. Newman B. M., Stine P. C. A handbook on the Gospel of Matthew. New York, 1988.

Öpengin E., Haig G. Regional variation in Kurmanji: A preliminary classification of dialects // Kurdish Studies 2, 2014: 143-176.

Petrides S. Amadia and Akra. // C. G. Herbermann et al. (eds.). Catholic Encyclopedia. New York, 1912, Vol. 1, 376.

Re§id T. Ji diroka alfabeyen kurdi ['From the history of the Kurdish alphabets']. // Zimannas (online journal), 2020 (January). URL: https://zimannas.wordpress.com/category/alfabe/ (accessed 04.03.2021). (In Kurdish)

Rhea S. A. Brief Grammar and Vocabulary of the Kurdish Language of the Hakari District // Journal of the American Oriental Society, 1872, 10.3, 118-155. URL: https://ia802907.us.archive. org/4/items/jstor-592123/592123.pdf (accessed 05.07.2020).

Rizgar B. Kurdish-English, English-Kurdish Dictionary. London, 1993.

Rzepka M. Translation, power and domination: The postcolonial explorations of the Bible in the Kurdish context // J. Bochenska (ed.). Rediscovering Kurdistan's cultures and identities. Cham, Switzerland, 2018.

Saeid M. A. Two varieties of Kurdish in competition: The problematic of modelling a language of instruction in Iraqi Kurdistan.

Universität Wien dissertation, 2014. URL: http://othes.univie. ac.at/33109/ (accessed 11.02.2021).

Siyabendî T. Afrandin û Pêçketina Alfabeya KurdîKurmancî li Ermenistana Sovyetê de (1921-1991). Class paper, State University of Yerevan. 2019. URL: https://www.academia.edu/42736280/ Afirandin_û_Pê§ketina_Alfabeya_Kurdî Kurmancî_li_Ermenista na_Sovyetê_de_1921_1991 (accessed 04.03.2021). (In Kurdish)

Thomas K. J. A restless search: A history of Persian translations of the Bible. Atlanta, 2015.

Thomas K. J. BIBLE viii. Translations into other Modern Iranian Languages // Encyclopedia Iranica, 2000 (and 1989), IV/2, 213-214. URL: http://www.iranicaonline.org/articles/bible-viii (accessed 23.06.2020).

Wierzbicka A. What did Jesus Mean?: Explaining the Sermon on the Mount and the Parables in Simple and Universal Human Concepts. Oxford, 2001.

Wîkîferheng, qer entry. URL: https://ku.wiktionary.org/wiki/ qer (accessed 21.06.2020). (In Kurdish)

Wikipedia. Alfabeya §emo-Marogûlov [The Alphabet of Shemo-Marogulov] URL: https://ku.wikipedia.org/wiki/Alfa-beya_%C5%9Eemo-Marog%C3%BBlov (accessed 11.07.2020). (In Kurdish)

Xelîl T'. 90-SalyaRojnema ''Rya T'eze"pirozbe! [Congratulations to R'ya T'eze for its ninetieth year]. R'ya T'eze, 2020.3.1-2. URL: https://krd.riataza.com/wp-content/uploads/2020/04/mart-2020.pdf (accessed 22.06.2021). (In Kurdish)

Lmqn (Lazo = Ghazaryan H.) Güüi u. [§ems 'The Sun' (Kurdish reading primer)]. Etchmiadzin, 1921. (In Kurdish)

Köpgo Q. (K'urdo Q., KypgoeB K. K.) rpaMamuKa 3bMane KöpMüh^u üa Kypm [A short grammar of the Kurmanji language]. Yerevan, 1949. URL: https://bnk.institutkurde.org/images/ pdf/8QVKRB73JL.pdf (accessed 02.09.2020). (In Kurdish)

Köpgo Q. (K'urdo Q., KypgoeB K. K.). 3bMane K'öpdu: P'e3bMan [The Kurdish language: Grammar]. Yerevan, 1978. (In Kurdish)

Курдоев К. К. (K'urdo Q.). Грамматика курдского языка (курманджи): фонетика, морфология [Grammar of the Kurdish language (Kurmanji): phonetics and morphology]. Moscow, Leningrad, 1957. URL: https://www.twirpx.com/file/796967/ (accessed 19.09.2020). (In Russ.)

Лерх П. И. Исследования об иранских курдах и их предках, северных халдеях [Lerch P. I. Research on Iranian Kurds and their ancestors, northern Chaldeans]. Санкт-Петербург, 1856. URL: https://en.calameo.com/read/00422818884046625f47a (accessed 10.06.2020). (In Russ.)

Хамоян М. У. [Maxime Xamo]. Бахдинанский диалект курдов Ирака (Бамарни и Атриш) [Bahdinan dialect of Iraqi Kurds (Bamarni and Atrish). Yerevan, 1965. (In Russ.)

Nicholas Bailey SIL International Николас Бейли SIL International nicholas_bailey@sil.org