University of Bucharest
Center for Arab Studies
ROMANO-ARABICA
VI-VII
2006-2007
Peripheral Arabic Dialects
UNIVERSITY OF BUCHAREST
CENTER FOR ARAB STUDIES
ROMANO-ARABICA
New Series
No 6-7
Peripheral Arabic Dialects
EDITURA UNIVERSITđĮIIăDINăBUCUREŞTI
– 2006/2007
Editor:
Nadia Anghelescu
Associate Editor: George Grigore
Advisory Board:
Ramzi Baalbaki (Beirut)
Michael G. Carter (Sidney)
Jean-Patrick Guillaume (Paris)
Hilary Kilpatrick (Lausanne)
Chokri Mabkhout (Tunis)
Yordan Peev (Sofia)
Stephan Procházka (Vienna)
André Roman (Lyon)
Editor in Charge of the Issue:
George Grigore
(e-mail: gmgrigore@yahoo.com)
Published by:
© Center for Arab Studies
PitarăMoşăStreetăno 11, Sector 2, 70012 Bucharest, Romania
Phone/fax: 0040-21-2123446
© Editura Universităţii din Bucureşti
Şos.ăPanduri,ă90-92,ăBucureştiă– 050663; Telefon/Fax: 0040-21-410.23.84
E-mail: editura@unibuc.ro
Internet: www.editura.unibuc.ro
ISSN 1582-6953
Contents
Werner Arnold,ăTheăArabicăDialectăofătheăJewsăofăIskenderun…………………….
Andrei A. Avram, Romanian Pidgin Arabic...............................................................
Guram Chikovani, Some Peculiarities of Central Asian Arabic From the
PerspectiveăofăHistoryăofăArabicăLanguage………………………………………….
Dénes Gazsi, Shi‗iteă Panegyricală Poemsă fromă theă Townshipă ofă Daﻣt-iă ƥzādagānă
(H~ūzistān).....................................................................................................................
George Grigore,ăL‘énoncéănonăverbalădansăl‘arabeăparléăàăMardin………………
Otto Jastrow, Where do we stand in the research on the Anatolian qəltu dialects?...
Jérôme Lentin,ăL‘arabeăparléăenăSicileăétait-il un arabe périphérique ?.....................
Gunvor Mejdell,ăFromăperipheryătowardsăcentreăstage:ăresearchăonă‗mixedăstyles‘–
results and challenges………………………………………………………………..
Jonathan Owens & Fadila Brahimi, Grundvokabular und idiomatische Struktur:
ArabischăinăNigeriaăundăNordafrika…………………………………………………
Stephan Procházka, Does geographical periphery imply linguistic periphery? The
examples ofătheăArabicădialectsăofăCiliciaăandăUrfaăinăSouthernăTurkey……………
Arlette Roth,ă Quellesă nouvellesă perspectivesă s‘ouvrentă avecă l‘explorationă etă laă
description des dialectes arabes dits périphériques ?...................................................
Thomas Stolz & Andreas Ammann, Beda u Qabad: The Maltese Inchoative /
Ingressive.....................................................................................................................
Catherine Taine-Cheikh, Périphérie géographique et perméabilité aux contacts. Le
casăduăMaghreb………………………………………………………………………
Shabo Talay, The influence of Turkish, Kurdish and other neighbouring languages
on Anatolian Arabic.....................................................................................................
Hristina Tchobanova, Phonetic features and changes in Andalusi Arabic: a case
studyăofătheălaryngealăplosiveă/‘/..................................................................................
7
13
29
39
51
63
71
85
97
109
133
149
159
179
189
FOREWORD
This issue of Romano-Arabica contains the proceedings of the
International Colloquium Peripheral Arabic Dialects, which was held in
Bucharest, from the 18th to the 20th of May 2007.
The colloquium, organized by the Center for Arab Studies of the
University of Bucharest, and chaired by Prof. Dr. Otto Jastrow (FriedrichAlexander University, Erlangen-Nürnberg), joined together a panel of
distinguished specialists from as many as nine countries who have been studying
Arabic dialects spoken outside the official boundaries of the Arab world, covering
a vast territory, from Central Asia to the Iberian Peninsula.
Finally, may we, as organisers, thank each and every participant for
making the International Colloquium Peripheral Arabic Dialects – 2007 an
extremely enjoyable and worthwhile event.
The Colloquium Organisers
The Arabic Dialect of the Jews of Iskenderun
Werner Arnold
University of Heidelberg
Germany
Iskenderun is an important harbour town at the northeastern Mediterranean
seashore in the Turkish province of Hatay. The Arabic speaking population calls
theăcityăIskadrōnaăwhichă isătheăoldăAramaică nameăofătheătown.ăIskenderună isătheă
northernmost city in which a city dialect of the syro-palestinian type is spoken.
Nearly all dialects of this type have the following features in common:
a) Shift of the interdentales to the corresponding dentals
b) Shift of q to the glottal stop.
c) The lowering of u and i in simple closed word-final syllable to o and e
d)ăWordăfinalăImālaăofătheăfeminineăending -a
e) Monophthongization of aw and ay.
f) Loss of genus distinction in plural forms of verbs and pronouns
The Arabic dialect of the Jews of Iskenderun and of Antakya, the capital
city of the province of Hatay, has all these features:
1) talğ >talğ
hāda > hāda
d}uhr > d}əhr
2) qāla > ‘āl
3) yiktub > yəktob
yinzil > yənzel
4) ši*a > šə*e
5) bayt > bēt
mawt > mōt
6) hənnen yəktəbu
7
əntu təktəbu
But the dialects of the Sunni Muslims, Christians and Nus}ayris in Antakya
and Iskenderun as well as the dialect of the city of Gaza do not share all these
features. The phoneme q is preserved in the city of Iskenderun and in the
Christian dialect of Antakya. In the dialect of the Sunnis and Nus}ayris in Antakya
q shifted to a velarized k}. In the city of Gaza q alternates with g. The lowering of i
and u in word-final syllables is not attested in the city dialects of the Christians,
Nus}ayris and Sunnis of Iskenderun and Antakya in the north and in Gaza in the
South. Imala of the feminine ending -a is not attested in Gaza.
It seems as if the common features of the city dialects in the eastern
Mediterranean countries developed somewhere in the centre of the area and
spread to the north and to the south without reaching the fringes of the area,
especially not to the city of Iskenderun in the outermost north and the city of Gaza
in the remotest south.
Only the dialect of the Jews in Iskenderun has all the above mentioned
features so that we can draw the conclusion that the Jews of Iskenderun originally
came from the center of the Syro-Palestinian dialect area and settled down at the
periphery of that area. In that way a central dialect has become a peripheral
dialect, as the dialect of the Jews is now different not only from the two other
Arabic dialects of the city of Iskenderun but also from all other dialects in that
peripheral area of the Syro-Palestinian dialect area. This will be demonstrated by
the following examples from the city of Iskenderun:
Jews
7) Consonant q >
qalb >
8) Short Vowels i/u
ism >
umm>
in final closed
syllables
’
’alb
>ə
əsm
əmm
> e/o
>yəktob
yinzil
>yənzel
9) a in pre-stressed nağğār
closed syllables
opened syllables
katabt
10)Diphthongs aw/ay ō/ē
8
Christians
q
Nus}ayris
q
qalb
qalb
>i
ism
imm
>i
ism
imm
>i
> yiktib
> yinzil
niğğār
>i
> yiktib
> yinzil
niğğār
ktabt
ktabt
aw/ay
aw/ay
>
mawt
bayt
11)ăImālaă— >ē —
nağğārīn
12) Pronoun
independent 3 pl
suffixed 3 sg m
13)
Interrogative
particle
where?
when?
14) Weak verbs
IIIy
IIIă‘ă
I‘/IIIy
mōt
bēt
mawt
bayt
mawt
bayt
nağğārīn
niğğērīn
niğğārīn
hənnen
d}a rab-o
hinni
d}a rab-u
hinni
d}a rab-u
wēn
ēmta
aynah}a ll
aymat
fayn
īmat
nisa
‘əri
ağa
nisi
qiri
iğa
nisi
qara
iğa
In the second part of my paper I want to mention some peculiarities of the
Jewish dialect of Iskenderun which are not attested in the other dialects of the
Turkish province of Hatay or even in the Syro-Palestinian dialects in general.
Only in the Jewish dialect an always stressed suffixed demonstrative
pronoun -āăcanăbeăaddedătoăaănoun:
15) bi yōmā bis}ūmu they fast on this day
bi santā in this year
bi lēltā in this night
badālā instead of this
After a preceding demonstrative pronoun followed by a noun with or
without article an additional demonstrative pronoun can be suffixed. In this case
the suffixed demonstrative pronoun has the form -āy:
bi hayy xs}ūs}āy for that reason
bi hal-xs}us}āy for that reason
bi hād s}ūrāy in this form
bi hal-baladāy in this village
hal-luġāy this language
9
Only in the Jewish dialect a second pronoun of the third person which
developed from the independent personal pronoun, can be added to the first
suffixed personal pronoun. This phenomenon is well known from the
Mesopotamian Qəltu-dialectsăasătheăexampleăofăOttoăJastrowăfromăMardīnăshows:
16) ‘at}aytū́ hu-we ich gab ihm ihn (Jastrow 1978: 296 f.)
In the Syro-Palestinian dialects this kind of suffixation has as far as I know
not been observed in other dialects then the Jewish dialect of Iskenderun.
Examples:
17) labbása he dressed her + hūwe > labbásawe he dressed her (with) it
(m)
them
them
them
(m)
them
labbása he dressed her + hīye > labbásaye he dressed her (with) it (f)
labbsak he dressed you + hūwe > labbsakwe he dressed you (with) it (m)
labbsak he dressedd you +hənnen > labbsakənnen he dressed you (with)
labbása he dressed her + hənnen > labbasánnen he dressed her (with)
labbsáto she dressed him + hənnen > labbsatūnnen she dressed him (with)
bilabbsa he dresses her + hūwe > bilabbsawe he dresses her (with) it (m)
bilabbsa he dresses her + hīye > bilabbsaye he dresses her (with) it (f)
bilabbso he dresses him + hūwe > bilabbsūwe he dresses him (with) them
bilabbsū they dress him + hūwe > bilabbsūwe they dress him (with) it (m)
bilabbsūha they dress her + hūwe > bilabbsūwe they dress her (with) it
bilabbso he dresses him + hənnen > bilabbsūnnen he dresses them (with)
bilabbsūhon they dress them + hənnen > bilabbsūnnen they dress them (with)
them
A second pronoun can be suffixed also to a an indirect pronoun which has
been connected with the verb by the former preposition li:
baha} ww əšlon I pick for them + hūwe > bah}aww əšlunwe I pick for them
it (m)
10
A peculiarity of the Jewish dialect of Iskenderun is the emphatic negation
particle lam. This particle occurs only once in my texts to negate the imperfect
form of the verb with the particle am- expressing continued presence:
17) lə-knīse msakkara w-lam ambis}īr HminyánH
The synagogue is closed and a Minyan is no longer at all coming into
being
I ask my informant to give me some other examples for the use of lam, and
I found out that the particle lam negates only the verbal forms in imperfect with or
without the particle am-. The perfect form of the verb and the so called bimperfect can not be negated by the particle lam but only by the negation particle
ma.
18) bəkra lam yəği
tomorrow he will not come at all
mā bikteb
he does not write
mā katab
he didn't write
I therefore expressed in my book on the Arabic dialects of the Turkish
province of Hatay the opinion that the particle lam in the Jewish dialect of
Iskenderun cannot trace back to classical Arabic lam, which is a negation particle
for the past tense (Arnold 1998: 117). I thought it could probably be a connection
of the two negation particles lā and mā to lāmā, which lost by frequent usage the
word final vowel and then the length of the first vowel in closed syllable.
19) lā + mā >lāmā >lām>lam
The connection of two negation particles would also explain the emphatic
character of the negation particle lam in the Jewish dialect of Iskenderun.
In 2002 Gabriel Rosenbaum published in the Festschrift for Otto Jastrow
hisăarticleă―The Particles ma and lam andăemphaticănegationăinăEgyptianăArabic‖
(Rosenbaum 2002: 583-598) with many records for the use of lam in colloquial
Arabic from Egypt, Syria and the Lebanon. In colloquial Egyptian Judeo-Arabic
texts from the 19th and even of the twentieth century the negation particle lam
appeared even many times. In contrary to the Judeo-Arabic dialect of the Jews of
Iskenderun lam is used in Egypt to negate the past tense too. While other scholars
argued that the particle lam is a hypo-correction that disappeared in living speech
11
and was borrowed from standard Arabic by literate or semi-literate people to rise
to stile of the language at a more classical one, Rosenbaum came in his article to
theăconclusionăthată―oneăshouldănotănecessarilyăassumeăthatăinăallăcasesăitsăuseăină
‗āmmiyya is due to pseudo-correction or elevation of style. If the use of lam in the
colloquial language indeed started as a pseudo-correction, this pseudo-correction
must have become productive, with lam becoming a part of the colloquial lexicon;
its usage should than be regarded as belonging to the colloquial, with different
rulesăthanăthoseăofăstandardăArabic‖.
The usage of lam by a speaker in the Turkish province of Hatay, where it
was always strictly forbidden to teach Arabic and who is therefore illiterate in
ArabicăisăaăstrongăsupportăforăRosenbaum‘săopinion,ăthatălamăhasăbecomeăatăleastă
in Judeo-Arabic a part of the spoken dialect.
In 2000 the last Jewish family of Iskenderun left the city. The end of the
millennium was in the same time the end the Jewish community in Iskenderun
and the end of the negation particle lam in this city.
References
Arnold, Werner. 1998. Die arabischen Dialekte Antiochiens. Wiesbaden.
Jastrow, Otto. 1978. Die mesopotamisch-arabischen Qəltu-Dialekte. Band I: Phonologie
und Morphologie. Wiesbaden.
Rosenbaum,ăGabriel.ă2002.ă―TheăParticlesăma and lam and emphatic negation in Egyptian
Arabic‖,ă in:ă Sprich doch mit deinen Knechten aramäisch, wir verstehen es! 60
Beiträge zur Semitistik. Festschrift für Otto Jastrow zum 60. Geburtstag.
Herausgegeben von Werner Arnold und Hartmut Bobzin. Wiesbaden.
12
Romanian Pidgin Arabic
Andrei A. Avram
University of Bucharest
Romania
1. Introduction
The present paper describes an Arabic pidgin used in the 1980s by Romanian
and Arab (Egyptian and Iraqi) oil workers in Iraq 1. An additional aim of this paper
is to assign this pidgin to the various types of pidgin suggested in the literature.
The description is based on fieldwork carried out in 1984 and 1985. All
examples from Romanian Pidgin Arabic are rendered in the system of
transliteration for Arabic. Examples from other pidgins are transcribed as in the
sources mentioned. The following abbreviations are used: 1 = first person; 2 =
second person; Ar. = Arabic; DEM = demonstrative; E. = English; Eg. Ar. =
Egyptian Arabic; Ir. Ar. = Iraqi Arabic; lit. = literally; NEG = negator; NHJE =
New Hebrides Jargon English; Norw. = Norwegian; O = object; PREP =
preposition; Rom. = Romanian; Rus. = Russian; S = subject; SG = singular; SPPE
= Samoan Plantation Pidgin English; V = verb.
The paper is organized as follows. In section 2 I sketch the main
characteristics of the phonology of Romanian Pidgin Arabic. Section 3 focuses on
its morphology. The syntax is outlined in section 4. In section 5 I discuss the
lexicon. The findings are summarized in section 6.
2. Phonology
The phonology of Romanian Pidgin Arabic is characterized by significant
inter-speakerăvariation,ădueătoătheă influenceăofătheăspeakers‘ă firstălanguages 2, i.e.
Romanian, Egyptian Arabic and Iraqi Arabic.
Consider first some characteristics typical of speakers with Romanian as
their first language. Thus, the distinction between short and long vowels is lost,
both in lexical items of Arabic origin and in those from English:
(1) a. lazim ‗must‘ăŢăAr.ălāzim
1
See also Avram (1997).
Inter-speaker variation of this type is attested in other pidgins, e.g. Chinook Jargon (Thomason
1981).
2
13
b. slip ‗toăsleep‘ăŢăE.ăsleep
The marked consonantal phonemes of Arabic are either replaced or lost 3.
In word-initial position the velar voiceless fricative/h̬/ and the pharyngeal
voiceless /h}/ are replaced by fricative /h/:
(2) a. hamsa ‗five‘ăŢăAr.ăh̬a msa
b. habib ‗friend‘ăŢăAr.ăh}a bīb
In word-final position the reflexes of /h}/ are /h/, /a/ or /ø/:
(3) a. embarih ‗yesterday‘ăŢăIr.ăAr.ămbērih}
b. rua ‗toăleave‘ăŢăAr. rūh}
c. mufta ‗key‘ăŢăIr.ăAr. muftāh}
Theă reflexă ofă theă voicelessă glottală stopă /‘/ă andă ofă theă pharyngeală voicedă
fricativeă/‗/ăisăø:
(4) a. arba ‗four‘ăŢăAr.ă‘arba‗
b. dilwati ‗now‘ăŢăEg.ăAr.ădilwa‘ti
c. saa ‗watch;ăhour‘ăŢăAr.ăsā‗a
d. ašara ‗ten‘ăŢăAr.ă‗ašara
The voiced velarăfricativeă/ġ/ăisăreplacedăbyă/g/:
(5) šogol ‗work‘ăŢăIr.ăAr.ăšuġul
The reflex of the voiceless uvular stop /q/ is /k/:
(6) suk ‗market‘ăŢăAr.ăsūq
The so-calledă―emphatics‖ăareăreplacedăbyăplainăconsonants:
(7) halas ‗ready‘ăŢăAr.ăh̬a lāṣ
Finally, the geminate consonants of Arabic undergo degemination:
(8) sita ‗six‘ăŢăAr.ăsitta
3
As in other Arabic pidgins and creoles (see Versteegh 1984, Smart 1990, Avram 1993 and 1995).
14
As for speakers with Egyptian or Iraqi Arabic as their first language, a typical
feature is the replacement of /p/ and /v/ by /b/:
(9) bibul ‗people‘ăŢăE.ăpeople
3. Morphology
Nouns are invariant in form:
(10) Sadik la
ani work la
sonda.
friend PREP 1SG work PREP rig
‗Myăfriendăworksă/ămyăfriendsăworkăonătheărig.‘
As shown in the translation of the example above, the singular vs. plural
interpretation relies on the context, given the absence of overt plural markers.
Plurality may also be indicated by the occurrence of a cardinal numeral:
(11) talata sadik
three friend
‗threeăfriends‘ăăă
Another possibility is the use of tumač ‗much,ămany‘(ŢăE.ătoo much)4:
(12) Aku sadik tumač la
ani.
be friend many PREP 1SG
‗Iăhaveămanyăfriends.‘
Finally, plurality may also be expressed by kulu-kulu ‗all‘:
(13) sayara kulu-kulu
car
all
‗cars‘
As for the dual, it is replaced by structures with the cardinal numeral itnen
‗two‘ăfollowedăbyătheănoun5:
4
5
Cf. plural formation with dewaya ‗much‘ăinăpidginizedăHausaă(Manessyă1995).
As in e.g. pidginized Gulf Arabic (Smart 1990).
15
(14) itnen dinar
two dinar
‗twoădinars‘ă
The definite article is not attested. Speakers with Romanian as their first
language occasionally use the cardinal number wahed ‗one‘ă asă ană indefiniteă
article:
(15) Ani šuf wahed sayara.
1SG see one car
‗Iăsawăaăcar.‘
The number of adjectives is extremely small. All of them are invariant.
Adjectives can also used as adverbs, as illustrated by the following example:
(16) Inte work zen.
2SG work good
‗Youăworkăwell.‘
Only a few pronominal forms are attested. The system of personal
pronouns is poorly developed and consists only of the two forms listed below:
(17) a. ani ‗1SG‘
b. inte ‗2SG‘
For plural referents, an additional form, the noun pipol ‗people‘,ă isă
occasionally used:
(18) Pipol rumani drink mai tumač.
people Romanian drink water much
‗We,ăRomanians,ădrinkămuchăwater.‘
Similar pronominal uses of nouns are reported in the literature on pidgin
languages. Consider the following example from Queensland Pidgin English
(Mühlhäusler 1997), where the noun kanaka ‗people‘ă functionsă asă aă personală
pronoun:
(19) Kanaka work plenty.
‗Weăworkăaălot.‘
16
Like in other Arabic pidgins and creoles 6, the affixed pronouns of Arabic
are replaced by personal pronouns. The examples below illustrate the use of
personal pronouns in structures with a direct object, with a (prepositional) indirect
object and expressing possession respectively:
(20) a. Inǧiner šuf inte.
engineer see 2SG
‗Theăengineerăseesăyou.‘
b. Ani spik la
inte.
1SG speak PREP 2SG
‗Iătoldăyou.‘
c. Inte sadik la
ani.
2SG friend PREP 1SG
‗Youăareămyăfriend.‘
The only demonstrative is hada ‗this;ă that‘,ă whichă functionsă eitheră asă aă
pronoun or as a pronominal adjective:
(21) hada čakuš.
DEM hammer
‗Thisăisăaăhammer.ă/ăthisăhammer.‘
Finally, there are no relative pronouns.
The following question words are recorded:
(22) a. min / minu ‗who‘
b. eš / šinu / šu ‗what‘
c. fen / wen ‗where‘
d. šwakit / ešwakit ‗when‘
e. šlon ‗how‘
f. leš / lieš ‗why‘
g. šged ‗howămany‘
h. kam / čam ‗howămuch‘.
A few remarks are in order here. First, note the existence of variants.
Second, some of these variants are merely alternative pronunciations, e.g. leš and
lieš. Third, other variants can be accounted for in terms of their different
etymology: fen < Eg. Ar. fēn vs. wen < Ir. Ar. wēn; kam < Eg. Ar. kam vs. čam <
6
See Versteegh (1984) and Avram (1994 and 1995).
17
Ir. Ar. Finally, forms such as šwakit lit.ă‗whatătime‘ăoră lieš lit.ă‗forăwhat‘ăareănotă
perceived as bimorphemic by speakers with Romanian as their first language, for
whom the system of question words is therefore opaque7.
The only class of numerals attested is that of cardinal numerals. All of
them are invariant in form. Arabic dual forms of cardinal numbers are replaced by
structures consisting of the cardinal numeral itnen and a numeral:
(23) itnen miya
two hundred
‗twoăhundreds‘
Verbs are invariant in form. There is no copula:
(24) Ani šabab.
1SG young
‗Iăamăyoung.‘
Existence and possession are both expressed by means of aku (< Ir. Ar.
aku):
(25) a. Aku manğariya.
be food
‗Thereăisăfood.‘
b. Aku fulus la
inte?
be money PREP 1SG
‗Haveăyouăgotăanyămoney?‘
Only a small number of adverbs are attested. These include the following:
(26) a. baden / baaden ‗then‘
b. hon ‗here‘
c. šuwaya(-šuwaya) ‗aălittle‘
Recall, however, that most adjectives may be used as adverbs as well.
Function words hardly occur. Thus, no conjunctions or complementizers
are attested. As for prepositions, Romanian Pidgin Arabic has a one-preposition
7
For a survey of the systems of question words in pidgins see Muysken and Smith (1990).
18
system8. The only preposition is la. It occurs e.g. in structures with an indirect
object and in those expressing location, direction and possession, as shown below:
(27) a. Gib sigara la
ani.
bring cigarette PREP 1SG
‗Giveămeăaăcigarette.‘ă
b. Ani slip la
karavan.
1SG sleep PREP caravan
‗Iăsleepăinăaăcaravan.‘
c. Inte rua la
sonda?
2SG go PREP rig
‗Areăyouăgoingătoătheărig?‘
d. La ani maku fulus.
PREP 1SG not be money
‗Iădon‘tăhaveăanyămoney.‘
Strikingly similar situations are mentioned in the literature. Russenorsk,
foră instance,ă hasă theă ―all-purpose‖ă preposition po, which occurs in structures
expressing e.g. location, direction or possession (Broch and Jahr 1981 and 1996):
(28) a. Principal po lan.
‗Theăcaptainăisăashore.‘
b. Moja po vater kastom.
‗I‘llăthrowăyouăinătheăwater.‘
c. klokka po ju
‗yourăwatch‘
This survey of the lexical categories attested in Romanin Pidgin Arabic
would not be complete without a discussion of categorial multifunctionality 9. This
is a direct consequence of the small size of the vocabulary, on the one hand, and
of the total lack of inflections, on the other hand. Thus, many words can be
assigned to more than one category and can be analyzed as lexically
underspecified10:
(29) a. halas ‗ready‘ăandă‗toăfinish‘
b. šogol ‗work‘ăandă‗toăwork‘
8
See the discussion of such systems in Mühlhäusler (1997).
See Voorhoeve (1981) and Mühlhäusler (1997) for a theoretical discussion.
10
In the sense of Silverstein (1972a and 1972b).
9
19
c. zen ‗good‘ăandă‗well‘
Categorial multifunctionality is attested in other pidgins. Here are some
examples from Butler English (Mehrotra 1999):
(30) a. fire ‗fire‘ăandă‗toăburn‘
b. like ‗similar‘ăandă‗like‘
c. wet ‗wet‘ăandă‗toăbathe‘
Finally, the only morphological operation attested is total reduplication.
However, reduplication is not productive. It may express an intensifying meaning:
(31) zen- zen
good good
‗veryăgood‘
In addition, there are a number of pseudo-reduplicated forms11:
(32) a. fikifiki ‗sexualăintercourse‘ă
b. sawasawa ‗together‘
c. semsem ‗similar,ăidentical‘
These are lexicalized forms, for which there are no corresponding simplex
forms, i.e. *fiki, *sawa and *sem. Pseudo-reduplicated forms are found in other
pidgins, as illustrated by the following example from English Japanese Pidgin
(Goodman 1967):
(33) saymo-saymo ‗similar,ăidentical‘ă
4. Syntax
Like most pidgins, Romanian Pidgin Arabic has SVO word order, including in
questions:
(34) Inte šuf hada?
2SG see DEM
‗Didăyouăseeăthis?‘ă
11
Also called quasi-reduplicated (Bakker 2003)
20
The only exceptions are sentences with aku ‗toă be‘ă oră maku ‗notă toă be‘,ă
which may display VSO or OVS word order12:
(35) a. Maku saa la
ani.
be
watch PREP 1SG
‗Iădon‘tăhaveăaăwatch.‘
b. La
ani maku sigara.
PREP 1SG not be cigarette
‗Iădon‘tăhaveăcigarettes.‘ăăăăă
There are no tense and aspect markers, unlike other Arabic pidgins 13. In
the absence of tense and aspect marking, the temporal and aspectual interpretation
of sentences mainly relies on the context. Also recorded is the unsystematic use of
time adverbials and of halas ‗toăfinish‘14. Consider the following examples:
(36) a. Leš rua dilwati?
why go now
‗Whyăareăyouăleaving?‘
b. Baaden spik la
mudir.
then
speak PREP boss
‗[Afterwards]ăI‘llătellătheăboss.‘
c. Inte halas it?
2SG finish eat
‗Haveăyouăeatenă[everything]?‘ăăă
Negation is expressed by means of an invariant negator, occurring in
preverbal position:
(37) Ani no aref.
1SG NEG know
‗Iădon‘tăknow.‘
Interestingly, unlike other Arabic pidgins, the negator is not derived from
an Arabic etymon, but from E. no.
Sentence coordination is achieved mainly by mere juxtaposition of
sentences. Less frequently, the adverb baaden is used:
12
Cf. Russenorsk, which alternates SVO and SOV (Broch and Jahr 1981).
See Versteegh 1984) and Avram (1994, 1995 and 2005).
14
Cf. the use of halas in pidginized Gulf Arabic (Smart 1990).
13
21
(38) Tal hon baaden giv hada sikina la
ani.
come here then
give DEM knife PREP 1SG
‗Heăcameăhereăandăgaveămeăthisăknife.‘
A similar use of adverbs for sentence coordination is reported for other
pidgins, e.g. Russenorsk (Broch and Jahr 1981):
(39) Moja po anner skip nakka vin drikkom, so moja nakka lite pjan.
‗IădrankăsomeăwineăonăanotherăshipăandăIăgotăaălittleădrunk.‘
A consequence of the absence of relative pronouns, conjunctions and
complementizers is the fact that subordinate clauses occur very rarely. The main
strategy is the use of paratactic structures. These structures are iconic and are
therefore instances of the so-calledă―natural‖ăsyntax15:
(40) Inte no work, maku fulus la
inte.
2SG NEG work not be money PREP 2SG
‗Ifăyouădon‘tăwork,ăyouăhaveănoămoney.‘
Iconic paratactic structures occur in other pidgins, as illustrated by the
following example from Samoan Plantation Pidgin English (Mühlhäusler 1997):
(41) No mani, no kam.
‗IfăIădon‘tăhaveămoney,ăIăwon‘tăcome.‘
At first sight, a case could be made for positing a zero-complementizer.
Accordingly, the example below would be ambiguous since it might be
interpreted as an instance of either direct speech or of the zero-complementizer:
(42) Inte spik la
ani hada muzen.
2SG speak PREP 1SG DEM bad
‗Youătoldăme:ăThisăisănotăgood.‘ă/ă‗Youătoldămeăthisăwasănotăgood.‘
This is the interpretation proposed by Romaine (1988) for similar cases in
Russenorsk:
(43) Moja ska si ju grot ligom.
‗I‘llăsay:ăYouălieăaălot.‘ă/ă‗I‘llăsayăthatăyouălieăaălot.‘
15
In the sense of Haiman (1985).
22
However, such an analysis runs into conceptual and empirical problems
since there is no independent evidence for a zero-complementizer. I therefore
conclude that Romanian Pidgin Arabic has no complementizers (overt or zero).
5. The lexicon
The size of the core vocabulary of Romanian Pidgin Arabic amounts to
approximately 150 words. This accords well with the figures reported for some
pidgins: 50-100 for New Hebrides Jargon English (Mühlhäusler 1997); 150-200
for Russenorsk (Romaine 1988); 300 for Samoan Plantation Pidgin English
(Mühlhäusler 1997).
As for the origin of the vocabulary, Egyptian and Iraqi Arabic account for
some 75% of the lexical items, and Romanian and English for the remaining 25%.
The vocabulary includes a number of lexical hybrids16:
(44) a. giv / gib ‗toăgive;ătoăbring‘,ăcf.ăE.ăgive and Eg. Ar. gib ‗bring!‘ăăăă
b. la ‗PREP‘,ăcf.ăRom.ăla ‗at,ăto,ăin‘ăandăAr.ăli ‗for‘ă
c. no ‗NEG‘,ăcf.ăE.ăno and Rom. nu
d. problem [pro’blem]ă ‗problem‘, cf. E. problem and Rom. problemă
[pro’blemə]
The forms in (45a) are a consequence of the replacement of /v/ by /b/,
mentioned in section 2. The form in (45d) ends in [m], as in English, but has the
stress pattern of its Romanian counterpart. Lexical hybrids are found in other
pidgins, e.g. Russenorsk (Broch and Jahr 1981):
(45) po ‗on‘,ăcf.ăRus.ăpo, Norw. på
In spite of the extremely small size of the vocabulary, lexical doublets and
even synonymic series do occur. Their existence is due to their different
etymology:
(46) a. mašina < Rom. maşină and sayara < Ar. sayyāra ‗car‘
b. zen < Ir. Ar. zēn, kuwais < Eg. Ar. kuwayyis and gud < E. good
16
Lexical items identified across languages (Mühlhäusler 1997).
23
Similar cases are recorded in other pidgins such as Russenorsk (Broch and
Jahr 1981):
(47) bra < Norw. bra, god < Norw. god, dobro < Rus. dobro and korosjo <
Rus.
xorošo (Broch and Jahr 1981)
Several instances of reanalysis of morphemic boundaries have also been
identified. Failure to identify morphemic boundaries sometimes depends on the
first language of the speakers of Romanian Pidgin Arabic. Thus, for speakers with
Romanian as their first language the form in (48a) is monomorphemic. On the
other hand, the morphemic boundaries of forms such as the one in (48b) are
reinterpreted by all speakers.
(48) a. muzen ‗bad‘ăŢăIr.ăAr. mū zēn ‗notăgood‘
b. tumač ‗much,ămany‘ăŢăE.ătoo much
Reanalysis of morphemic boundaries is attested in other pidgin languages.
Consider the following example from Pacific Pidgin Englishes (Mühlhäusler
1997):
(49) aidono ‗Iădon‘tăknow‘ăŢăE.ăI don‘t know
6. Conclusions
As is well known, pidgins have been classified, among others, on the basis
of social criteria. One such typology is proposed by Bakker (1995), according to
the social situation in which pidgins are used. Exemplified below are the types of
pidgin distinguished by Bakker (1995): (i) maritime pidgins: Lingua Franca,
Russenorsk; (ii) trade pidgins: Pidgin Eskimo, Chinese Pidgin English; (iii)
interethnic contact languages: Chinook Jargon, Mobilian Jargon; (iv) work force
pidgins: Butler English, Fanagalo, Hawaiian Pidgin English.
Sebba (1997) suggests a typology according to the social context of the
language‘să originsă andă identifiesă theă followingă types:ă (i)ă militaryă andă policeă
pidgins: Lingua Franca, Hiri Motu; (ii) seafaring and trade pidgins: Russenorsk,
Chinook Jargon, Chinese Pidgin English; (iii) plantation pidgins: Tok Pisin; (iv)
mineă andă constructionă pidgins:ă Fanagalo,ă Ewondoă populaire;ă (v)ă immigrants‘ă
pidgins: Gastarbeiterdeutsch; (vi) tourist pidgins: Tarzanca (= Turkish tourist
pidgin); (vii) urban contact vernaculars: Fly Taal.
24
Romanian Pidgin Arabic can be quite straightforwardly assigned to
Bakker‘să(1995)ăworkăforceăpidgins.ăOnătheăotherăhand,ăSebba‘să(1997)ătypologyă
needs to be extended so as to include a class of oil industry pidgins 17.
Another well-known typology (Mühlhäusler 1997) seeks to classify pidgin
languages on purely linguistic criteria and distinguishes three developmental
stages: (i) prepidgin 18; (ii) stable pidgin; (iii) expanded pidgin 19 . Each of these
phases is characterized by a specific set of phonological, morphological, syntactic
and lexical diagnostic features.
Now, at various points in this paper I have noted similarities between
Romanian Pidgin Arabic described and other pidgins. In the table below I
compare Romanian Pidgin Arabic with three pidgin languages, generally
considered to be representatives of the prepidgin stage: Russenorsk and two
English-lexified pidgins, Butler English and Pacific Pidgin Englishes. The latter is
a cover term for non-specified varieties of Pidgin English spoken in the Pacific in
the 19th century. The table is based on data from Broch and Jahr (1981) for
Russenorsk, Hosali and Aitchison (1986), Mehrotra (1999) and Hosali (2005) for
Butler English, Mühlhäusler (1997), and Tryon and Charpentier (2004) for Pacific
Pidgin Englishes. With one exception, the diagnostic features are those suggested
byăMühlhäusleră(1997).ăIăhaveăaddedă―non-productiveăreduplication‖,ăproposedăbyă
Bakker (1995 and 2003), who explicitly states that this feature is typical of
expandedă pidgins.ă Ină theă table,ă theă signă ―+‖ă showsă thată aă featureă isă attestedă andă
―‖ﻭăthatăităisănot;ătheăabsenceăofărelevantăinformationăisăindicatedăbyăblankăspaces.ăăăăă
(29) Comparison of selected diagnostic features
Feature
inter-speaker variation in phonology
minimal personal pronoun system
copula
conjunctions, complementizers
one-preposition system
non-productive reduplication
multifunctionality
tense and aspect markers
coordination with adverb
iconic paratactic structures
small size of core vocabulary
Romanian Pidgin
Arabic
+
+
ﻭ
ﻭ
+
+
+
ﻭ
+
+
150
17
Russenorsk
+
+
ﻭ
Butler
English
+
ﻭ
+
+
ﻭ
+
+
150-200
+
ﻭ
+
Pacific Pidgin
Englishes
+
+
ﻭ
ﻭ
+
+
+
ﻭ
+
NHJE 50-100
SPPE 300
To which the pidginized Gulf Arabic described by Smart (1990) could also be assigned.
Alternativeănamesăincludeă―jargon‖ăandă―minimalăpidgin‖.
19
Alsoăcalledă―extendedăpidgin‖.
18
25
lexical hybrids
lexical doublets
reanalysis of morphemic boundaries
+
+
+
+
+
+
As can be seen, Romanian Pidgin Arabic shares its characteristics with at
least one of the other prepidgins. From the point of view of its developmental
stage, then, Romanian Pidgin Arabic is a prepidgin.
References
Avram,ă A.ă A.ă 1993.ă ―Onă theă phonologyă ofă Arabică pidginsă andă creoles‖,ă in:ă Revue
roumaine de linguistique XXXVIII, 403-412.
Avram,ă A.ă A.ă 1994.ă ―Onă theă morphologyă ofă Arabică pidginsă andă creoles‖,ă in:ă Revue
roumaine de lingistique XXXIX, 121-129.
Avram,ăA.ă A.ă1995.ă―SomeăstructuralăcharacteristicsăofăArabicăpidginsăandăcreoles‖,ăin:ă
N. Anghelescu and A. A. Avram (eds.): Proceedings of the Colloquium on Arabic
Linguistics, Bucharest: Center for Arab Studies, 73-83.
Avram,ăA.ăA.ă1997.ă―Unăjargonăarab‖,ăin:ăStudii şi cercetări de lingvistică XLVIII, 27-38.
Avram,ăAndreiăA.ă2003.ă―Arabicăpidginsăandăcreolesăfromăaăcomparativeăperspective‖,ăin:ă
Romano-Arabica 3, 25-40.
Bakker,ăP.ă1995.ă―Pidgins‖,ăin:ăJ.ăArends,ăP.ăMuysken,ăandăN.ăSmithă(eds.):ă Pidgins and
creoles. An introduction, Amsterdam / Philadelphia, John Benjamins, 25-39.
Bakker,ăP.ă2003.ă―Theăabsenceăofăreduplicationăinăpidgins‖,ăin:ăSilviaăKouwenbergă(ed.):ă
Twice as meaningful. Reduplication in pidgins, creoles and other contact languages,
London: Battlebridge, 37-46.
Broch, I. and Jahr, E. H. 1981. Russenorsk – et pidginspråk i Norge, Oslo: Novus Forlag.
Goodman,ă J.ă S.ă 1967.ă ―Theă developmentă ofă aă dialectă ofă English-Japaneseă pidgin‖,ă in:ă
Anthropological Linguistics 9, 43-55.
Haiman, J. 1985. Natural syntax: iconicity and erosion, Cambridge: Cambridge
University Press.
Hosali,ăP.ă2005.ă―ButlerăEnglish‖,ăin:ăEnglish Today 81, 34-39.
Hosali,ă P.ă andă Aitchison,ă J.ă 1986.ă ―Butleră English:ă aă minimală pidgin?‖,ă in:ă Journal of
Pidgin and Creole Languages 1, 51-79.
Manessy, G. 1995. Créoles, pidgins, variétés véhiculaires, Paris: CNRS Éditions.
Mehrotra,ăR.ăR.ă1999.ă―LexicalăpolysemyăinăIndianăPidginăEnglish‖,ăin:ăJournal of Pidgin
and Creole Languages 14, 131-135.
Muysken,ăP.ăandăSmith,ăN.ă1990.ă―Questionă wordsă inăpidginăandăcreoleă languages‖,ă in:ă
Linguistics 28, 883-903.
Mühlhäusler, P. 1997. Pidgin and creole linguistics, expanded and revised edition,
London: University of Westminster Press.
Romaine, S. 1988. Pidgin and creole languages, London: Longman.
Sebba, M. 1997. Contact languages. Pidgins and creoles, London: Macmillan.
26
Silverstein,ăM.ă1972a.ă―ChinookăJargon:ălanguageăcontactăandătheăproblemăofămultilevelă
generativeăsystemsăI‖,ăin:ăLanguage 48, 378-406.
Silverstein,ăM.ă1972b.ă―ChinookăJargon:ălanguageăcontact and the problem of multilevel
generativeăsystemsăII‖,ăin:ăLanguage 48, 596-625.
Smart,ă J.ă R.ă 1990.ă ―Pidginizationă ină Gulfă Arabic:ă aă firstă report‖,ă ină Anthropological
Linguistics 32, 83-119.
Thomason,ăS.ăG.ă1981.ă―ChinookăJargonăinăarealăandăhistoricalăcontext‖,ăin:ăUniversity of
Montana Occasional Papers in Linguistics 2, 295-396.
Tryon, D. T. and Charpentier, J. M. 2004. Pacific pidgins and creoles. Origins, growth
and development, Berlin · New York: Mouton de Gruyter.
Versteegh, K. 1984. Pidginization and creolization: the case of Arabic, Amsterdam: John
Benjamins.
Voorhoeve,ă V.ă J.ă 1981.ă ―Multifunctionalityă asă aă derivatională problem‖,ă in:ă P.ă Muyskenă
(ed.): Generative studies on creole languages, Dordrecht: Foris, 25-34.
27
28
Some Peculiarities of Central Asian Arabic From the
Perspective of History of Arabic Language
Guram Chikovani
Tbilisi Institute of Asia and Africa, Tbilisi
Georgia
1.0. Central Asian Arabic, like Maltese, became isolated from the Arab
World and over the course of the time developed as a peripheral Arabic dialect.
This, with newly emerged elements due to independent development carries signs
of the oldest characteristics of phonological, morphological and syntactic signs of
Semitic languages and in particular Arabic language.
In the following article I present examples recorded in the Central Asian
regions of Bukhara and Qashqa-daryā, which clearly shows general Semitic
language peculiarities preserved in Central Asian Arabic up to nowadays. I will
distinguish important morphological and lexical material which, in our opinion, is
important from the prospect of the history of Arabic language.
Formation of plural.
Plural nouns in both gender categories use the -āt suffix element.
} āt bōyatin kibīra walatumāt insōr.
sabiyāt gālaw: – nahn
―Theăboysăsaid,ă―WeăareătheăsonsăofăaăGreatăBey‖ă(QAD) *.
zōkāt isnēyta i‗darb dah}a l mad. ziklōnāt salās iōm dard sāraw.
―Theyă(theătwoăofăthem)ăsetăoutăinăaăjourney.ăTheyăwalkedăforăthreeădays‖ă
(QAD).
In the above sentence the sabiyāt ―theă boys‖, nah}nāt ―we‖ă and walatumāt
―hisăsons‖ăwereăformedăbyă -āt suffix. From here, nah}nāt – is the 1st person plural
pronoun, which expresses a dual pronoun with suffix -āt. Dual is not uncommon
in Central Asian Arabic verb formations as well.
– ey, adəmiyāt, kul-kumāt ta‗āluwāt
*
QAD - Qashqa-daryā dialect.
29
―Hey,ăpeople,ăallăofăyouăgatherăhereă(come)‖ă(QAD).
nāb gāil: –i‗dilmurād fat faras intuāt!
―Theăoldămanăsays,ă―GiveăDilmuradăaăhourse!‖ă(QAD).
salās walad salās bint šāfawāt
‖Theăthreeăboysăfellăinăloveăwithătheăthreeăgirls‖ă(QAD).
In northwestern Semitic languages, we find masculine suffix *-āt in
addition to suffixes -ū/-ī (m)... We find such examples frequently in Hebrew and
Syrian languages. Parallelism of suffixes is also typical. For example, in Hebrew
the common gender is ‗āb ―cloud‖:ă pl. ‗ābīm/‗abōt; m.g. dōr: dōrim/dōrōt
―generation‖,ă ―relation‖ ―family‖,ă ―tribe‖;ă in Ugaritic r īš: ršm/ršt ―head‖; in
Phoenician im: imōt/imū ―day‖. In some cases it is possible for these forms to
appear analogous to one another. For example, the influence of the words ‘umm
(―mother‖)ăandă lil (―night‖)ăexplainsătheăexistence of the suffix -āt in plural forms
of the nouns ‘b (―father‖)ăand yum (―day‖).ă
At times, forms with -*ū and -*āt suffixes are in contradiction to each other
as is evidenced in the case of collective plural, for e.g. Meshn. yāmīm ―day/days‖ă
(pl.): yāmōt ―season‖ă(collect.),ăšānīm ―years‖ă(pl.):ăšanōt ―age‖,ă―time‖ă(collect.).ă
Central Asian Arabic formation of plural utilizing -āt suffix follows the
pattern of northwest Semitic languages where, in masculine gender along with
suffixes -ū/-ī(m)..., we have *-āt. Consequently, the existing condition of Central
Asiană Arabic,ă despiteă it‘să relationă toă Ugaritică orăHebrew,ă isă significantă fromă theă
perspective of general Semitic language. It is possible that the formation of plural
with the suffix *-āt may not reflect the oldest stage of language development, but
could instead be the result of its internal development. In this case first suffix *-āt
should have developed as a marker for masculine plural, and afterwards spread
analogous to the verbal forms. In any case we deal with element of general
Semitic formation, whose area of functioning is large in the peripheral Arabic
dialect of Central Asia.
2.0. It is significant to mention the forms of broken plural, which carry
singular meaning in Qashqa-daryā and Bukhara dialects.
i‗kulhum beitum hast. binti bonka šuġl misū. suhūr zēn hast.
―Everyoneăhasăaăhouse.ăMyădaughterăworksăinătheăbank.
My son-in-lawăisăaăgoodă[man]‖ă(BAD) *.
*
BAD – Bukhara dialect.
30
dük ‗iyāl-ke kēyin ‗andun.
―Heăhadăoneăson‖ă(BAD).
ilay qubūr salās yumāt qaraúlya sú!
―Guardămyăgraveăforăthreeădays!‖ă(BAD).
In the Classical Arabic for the broken plural forms fu‗ūl, fu‗ul and fi‗āl we
have the following formation of the singular ones : fu‗ūl: a) fa‗l b) fi‗l c) fu‗l: bahr} buhū} r ―sea‖,ă bah}t-buh}ūt ―study‖,ă ―research‖,ă ―discussion‖;ă ğild-ğulūd ―skin‖, h}ilmhu} mūl ―load‖, ğund-ğunūd ―army‖, burd-burūd ―(woman‘s)ăjacket‖,ă―blouse‖,ă―top‖,ă
―cloak‖,ă―cape‖.ă
d) fa‗al, e) fa‗il, f) fā‗il: ‘asad-‘usūd ―lion‖,ă malik-mulūk ―king‖,ă ğālis-ğulūs
―sitting‖.ă
fu‗ul: a) fa‗l, b) fa‗al, c) fa‗il, d) fa‗ul: saqf-suquf ―roof‖,ă ‘asad-‘usud ―lion‖,ă
namir-numur ―leopard‖,ăda} bu‗-du
} bu‗ ―hyena‖.
e) fa‗al, f) fi‗al, g) fu‗al: sā} na‗-su} nu‗ ―doer‖,ă dilāt-dulūt ―quick-footed‖,ă kurā‗kuru‗ ―bracelet‖.
h) fa‗īl, i) fa‗ūl: nadīr-nudur , sa} būr-su} bur ―patient‖.ă
j) fa‗īlat: madīnat-mudun ―town‖
k) fā‗īl: tāğir-tuğur ―merchant‖.
fi‗āl: a) fa‗l, b) fu‗l, c) fi‗l: bahr} -bihā} r ―sea‖,ă rumh-} rimāh} ―spear‖,ă di‗b-di‗āb
―wolf‖;
d) fa‗al, e) fa‗ul, f) fa‗il, g) fa‗īl, h) fu‗al, i) fā‗il, j) fa‗lānu//fa‗lānun//
fu‗lānun, k) fu‗la: ğamāl-ğimāl ―camel‖,ă rağul-riğāl ―man‖,ă ğarib-ğirāb ―scabby‖,ă
―mangy‖;ă kabīr-kibār ―big‖,ă ruta} b-ritā} b ―newlyă pluckedă date‖,ă sā} hi} b-si} hā} b ―friend‖ă
} b ―angry‖,ă hu~ msā} n-hi~ mās} ―hungry‖,ă ‘untā-‘ināt ―woman‖,ă ―female‖ă
ġadb} ān-ġidā
(Lekiashvilli 1963: 34-35; Wright 1967: 25, 26).
A number of broken plural forms in the Classical Arabic are used in
singular as well. Among them are fu‗ūl//fu‗ul and fi‗āl constructions, which are
present in Central Asian as well. The abovementioned examples from Central
Asian Arabic clearly show that in literary Arabic the forms of broken plural with
plural connotation, such as ‗iyāl ―sons‖,ă qubūr ―graves‖ă andă su} hūr ―sons-in-law‖ă
carry singular meanings in dialect. We can suppose that Central Asian Arabic
reflects the old case with regards to development of the language. According to
the presenting material we can assume that forms C luC2ūC3//C1uC2uC3 and
C1iC2aC3 initially expressed the singular form and gained the Plural meaning in
the following stage of language development progress.
3.0. In Qashqa-daryā Arabic we find the formation of plural with mīm.
31
adrūn h}ōyit} kisīr hamrayām kānat
―Thereăwasăaălotăofăgoldă(goldenăcoins)ăinătheăhouse‖ă(QAD).ă
ziklonāt hamrayām-tangayām ha~ zuwa, ġaduwa
―Theyătookăgoldăandăsilveră[and]ăwentăaway‖ă(QAD).ă
kokōyata kulla hamrāt h~a zuwa, kisatumāt-kisayām hašuwa
―Theă brothersă tookă theă wholeă goldă [thată wasă there];ă filledă theă purses‖ă
(QAD).
The forms h~a zuwa, ġaduwa and hašuwa (Comp. with Arabic h}a ša (u))
represent plural forms of the 3rd person (h~a zaw, ġadaw, hašaw). With the addition
of the pronominal suffix we get h~a zuwa, ġaduwa, hašuwa.
~ na kisīr mōšīnām ademiyāt kānaw.
i‗čoyhō
―Ină theă tea-house there was a lot of people, who came for entertainment
(walkers)‖.
mōšīnām contains the following elements mōš-īn-ām. From the two forms
of presented plural one is used with mīm.
nah}nāt madīna naġade, pōšō binta inšūfa, salāsatnam naġade,
fahadna hama bint noh~usa.
―Weăshallăgoătoătheătown;ăweăshallăseeătheădaughterăof the Pasha; the three
ofăus‘llăăcomeă[and]ăoneăofăus‘llă marryăthisă beautyă(daughter) (we shall take this
girl for oneăofăus)‖.ă
In the word salāsatnam m is added to the pronominal suffix of the plural
form of the 1st person. From one side, this might express dual plural form or,
from the other, 2nd and 3rd person plural pronominal suffixes, developed in
analogy with the forms of the construction. In both cases the expression of plural
should be taken into consideration with mīm, the reflection of which in the
Classical Arabic is the construction of plural for 2nd and 3rd persons with m:
comp. -kum//-hum: baytukum ―yourăhouse‖,ăbaytuhum ―theirăhouse‖.
zōka i‗ibsōt i‗wača ‘abu ‘umma isimumāt nasağa
―Heă madeăanăembroideryăofăthe names of his mother and father [parents]
on the carpet‖.
32
The plural form of ism in dialect differs from literary Arabic isimumāt and
is constructed with dual plural form. From the forming elements (-um-āt), -um
containing mīm is notable.
} yit} daha~ lt, il-morti rōsa gaseyta, il-bōyām qatalta, il-bētī maddeyt.
ugub il-hō
―ThenăIăwentăintoătheăpalace,ăcutăoffămyăwife‘săhead,ăkilledătheăBeyătooă
and returnedăhome‖.ă
In this sentence the form il-bōyām is plural. Even though it has a singular
meaning, it is important because of its construction. This example has been
evidenced in only one circumstance and is actually a mistake made by an
Informant during his speech.
ziklōnāt moyōtumāt ufraġat.
―Theyăhaveătheirăwaterăfinished‖.ă
mūy ―water‖ăisăa feminine form. The plural of muyōt< muyāt is constructed
with plural mīm suffix -um//ūm. The dual plurals of the same pattern words are
constructed with -āt, i.e. we get a duplicate of the forming element. In the end we
get a tripled form of plural.
The subject matter regarding the mīm plural in QAD appears naturally.
Plural m×m construction is typical for northwest Semitic languages – Ugaritic and
Hebrew. In Ugaritic, masculine nouns in the status absolutus plural have the
ending -m [-ūma] in Nominative and -m [īma] in Genitive-Accusative form. We
can compare the presented elements to suffixes of regular plural in literary Arabic
-ūna/īna. It is worth mentioning that in both dialects of Central Asian Arabic the
numerals are fundamentally different form the Classical Arabic in the second
decade (10+1) and correspond fully with Ethiopian. We often find the same
pattern of construction in Phoenician and Nabatean (Akhvlediani 1985: 47;
Brockelmann 1908: 489).
In Biblical Hebrew the nouns of masculine gender in plural have -īm
suffixes: sūs > pl. sūsīm ―hourses‖.ăWeăalsoăfindăsomeă instancesăwhenăthisăruleă isă
of not followed, i. e. when -īm forms the plural from the feminine nouns as well:
šānā > pl. šānīm. From the masculine nouns from time to time we find other forms
for the formation of plural: -īn, -ī, -ay, and -ām (Tsereteli, K. 2001: 75). Some of
the masculine nouns add suffix -ōt, ‘āb-‘ābōt ―father‖ă , šēm -šemōt ―name‖ă (lookă
above, 1.0.).
The element -īn is analogous to plural forms in Chaldaic and Syriac. It is
often found in Old Testament books of later age and in poems from the oldest
33
periods. For e.g. meleh~īn ―kings‖, Proverbs 31, yam× n ―days‖,ăDanielă12,ă13,ăetc.ă
(Gezenius 1874: 332).
In the element -ī - mīm is dropped. As for -ām it is a rarely used archaic
form sullam < salal ―stairs‖ă(Gezeniusă1874:ă332).
The feminine plural is constructed with the suffix -ōt. If in singular the
feminine noun ends with -ōt//īt, in the feminine form īm is added: almenūtīm
―widowhood‖,ă―widows‖.
Some words have double plural forms. This lexeme ending with ōt in
plural additionally receives -im//īm: hō} ma ―wall‖ă > hō} mōt > hō} mōtīm (Gezenius
1874: 335).
Some words we find only in plural forms, for e.g.: metīm ―people‖,ă
―humanăbeings‖,ă―mortal‖,ăelōhīm ―God‖ăareăpresentedăonlyăin the plural form.
In addition to Ugaritic and Hebrew, the element -m/-īm is characteristic in
Phoenician (īm). With regard to Aramaic and Ethiopian, in these languages
consequently is found īn and ān forms. In Classical Ethiopian the element of
feminine plural -āt is found in masculine nouns as well: kāhem-kāhenāt ―priest‖,
falāsefāt ―philosophers‖.ăInăArabică ină indirectăcases,ăasăweăknowăweăhaveăending
-īn and in Nominative case we have ending -ūn.
In Akkadian, plural masculine forms are constructed with -ē, -ani, -ānu, ūti,
and less often with ū. For e.g.: sisē ―horses‖,ă ilāni ―Gods‖,ă duppānu ―earthenwareă
plates‖,ă rabūti ―big‖,ă tabbanūtu ―buildings‖,ă paršū ―decisions‖.ăFeminineăplurală ină
Akkadian is constructed with suffix -ātī.
Mimation is typical for Akkadian. The names in status absolutus receive m and sometimes it is replaced with -ma. We also find parallel forms without
mimation in the language: bītum//bītu ―house‖,ă rieum//rieū ―shepherd‖.ăPredictablyă
mimation is dropped in the words in status constuctus and with pronominal suffix
forms. Mimation is also characteristic for Epigraphic South Arabic, where an
indefinite noun ends with (-m).
4.0. To explain the plural form of mīm in QAD dialect I believe we must
consider linguistic factors as well as extra-linguistic ones.
In my opinion the existing toponym in the Karshi area of the Qashqa-daryā
side is noteworthy, and is linked to the old Arab tribe in the area. Qakhlai must be
distinguished, as it derives from qah}tān and is an indication of southern Arab tribe.
Arabs from the Arab Peninsula in the pre-Islamic period were divided into
two groups: Arabs of pure blood (al-‗arab al-‗arība ) and non-native Arabs, i. e.
assimilated Arabs (al-‗arab al-musta‗riba ). There existed a third group of Arabs,
which were unified groups of Aramaic descent (al-bā‘ida ţă ―lost‖,ă ―unknown‖).ă
The pure Arabs (al-‗arība ) related themselves to the yoktān, i. e. qaht} ān. Joktan
(Qahtan) was the son of Eber. He was a founder of the southern Arab tribe of
34
Qahtan. It is known that southern Arabs, i.e. Yemenites, were distinguished in one
way with their originality and in the other with the linguistic peculiarities of their
language contacts. This was the main reason for having Hebrew Community
beside the other southern Arab tribes in the area. Influence of Hebrew in southern
Arabic and Ethiopian languages has been documented on numerous occasions by
specialists. Southern Arabic is linguistically closer to Hebrew and Ethiopian than
to northern Arabic. G. Tsereteli in his articleă―NewestăTheoriesăaboutătheăPlaceăofă
OriginăofăSemites‖ăquotesăA.ăLein‘săfollowingăwords,ă―Himyariteădialect,ătheăareaă
of which is surrounded with al-Yemen, is closer to the Ethiopian and Hebrew than
to ma‗dad (i.e. to the pure classical Arabic of northernăArabicătribes)‖ă(Tsereteli,ă
G. 2004: 172).
According to A. Ungnad, southern Arabic dialect is further from northern
Arabic than from Hebrew. The similarity of southern Arabic and Hebrew is
apparent in morphology, grammar and onomastics. The remarkable resemblance
is in nomenclature of Old Hebrew – Israelites and Minaeans and Sabaeans
(Tsereteli, G. 2004: 172). As for Wolf Leslau, he purports that the influence of the
Hebrew Community of southern Arabia on Ethiopian is noticeable. The evidence
of Judaist elements in the Ethiopian church, which then spread later in Coptic
(Leslau 1965: 5) is also noteworthy. It is necessary to mention that mimation is
usual in southern Arabic epigraphic inscriptions as opposed to northern Arabic. In
the aforementioned inscriptions, the name becomes indefinite, gaining [m] in the
end. Thus mīm as a formative element is not unfamiliar for the southwestern
Semitic language (Bauer 1966: 47-48; Grande 1972: 43). The abovementioned
indicatesă linguisticăcontactăofăQakhlai‘săsouthern Arab tribe with Hebrew, which
explains the existence of the Hebrew element of plural formation in QAD. The
southernă Arabică toponym‘să fixationă ină Qashqa-daryā region gives us a basis to
relate Arabs living in Jeinau and Kamashi if not entirely, at least partially to
Joktan (Qahtan) descendents.
4.1. Arab-Hebrew linguistic contacts are not peculiar within the Arabic
speaking world. In addition to the aforementioned southern Arabia, we can
mentionă theă Fertileă Crescentă andă ină particulară Baghdad‘să easternă province Al‘anbār in Iraq. There lives an Arab tribe Dulaym, in the speech of which it is
characteristic to use element mīm in the 3rd person singular perfect verb.
e.g. akalam < ‘akalū ―(they)ăhaveăeaten‖
šarabam < šarabū ―(they)ăhaveădrunk‖ă
katabam < katabū ―(they)ăhaveăwritten‖ă
35
In Imperfect it is ioklūn, išrabūn and iktubūn. For plural in Arabic it is
difficult to argue the influence of the Hebrew language alone in the foreign
originated formations, but in our opinion emergence of mīm in verb constructions
must be indicative of certain language contacts1.
4.2. In the Fertile Crescent and southern Arabia, as well as in internal
regions of the Arabian Peninsula the existence of Jews has been historically
documented. The State of Israel, as we know, was established in the second part
of the last century of the 2nd millennium B.C. In the second part of the 8th
century in Asia Minor, Assyria rose to power, took control of Syria in 738 B.C.
and became a danger to Israel. The Assyrian king Tiglath-pileser III (745-727)
attacked Israel and took over the northern part of Galilee including the eastern
side of the river Jordan, added the territory and began a system of taxation on the
rest of Israeli Kingdom. In 727 B.C., after the death of Tiglath-pileser III,
rebellion in Syria and Palestine broke out against Assyria. In response, the
Assyrians conducted a punishment campaign; they arrived with a big army and
after 3 years of siege in 722 B.C. took the capital of Israel Samaria. The king of
Assyria, Sargon the 2nd (722-705), took a majority of the Israeli population
captive and returned with them to Mesopotamia (Mamulia 1988: 304).
In the end of 7th century B.C. the Fertile Crescent was ruled over by the
Babylonian Kingdom or Chaldean Babylonia. The greater role of its creation and
the further existence played by Semitic tribes of Chaldeans living around the
Persian Gulf had relocated from Arabia in the end of the 2nd millennium B.C. The
Bible is an important source in studying the history of Babylon, in particular the
Books of the Prophets Jeremiah, Ezekiel and Daniel. The value of the inscriptions
found due to archeological searches combined with the written accounts of
antiquity about Babylon is colossal, including Herodotus (5th century B.C.);
Xenophon (5th-4th B.C.) Ktesion (5th -4th B.C. ) and others.
In 605 B.C., Nebuchadnezzar (604-562 B.C.) was able to conquer Assyria
and acquired a vast majority of its territories. At this point he decided to take over
Syria, Phoenicia and Judea, which he robbed and then forcibly relocated to
Babylon thousands of Judeans (according to some recourses 7 000), mostly from
dignified households (Mamulia 1988: 327). Nebuchadnezzar put on the thrown of
Judea Zedekiah and took over all of Palestine.
In 587 B.C., Nebuchadnezzar once more struck Jerusalem with his army,
due to the disloyalty of Zedekiah and other small kingdoms. Judea was not ready
to fight Babylon and turned for help to Egypt. Ezekiel describes this story in the
TheăinformationăaboutătheămentionedăpeculiarityăofătheătribeăDulaym‘săspeechăwasăgivenătoămeă
by an Iraqi Doctorate student of Tbilisi Institute of Asia and Africa, Adnan Jasim
1
36
Bible:ă ―Bută heă rebelledă againstă himă ină sendingă hisă ambassadors into Egypt, that
they might give him horses and much people. Shall he prosper? shall he escape
thatădoethăsuchăthings?ăorăshallăheăbreakătheăcovenant,ăandăbeădelivered?‖ă(Ezekielă
17, 15).
The Egyptian army did not save Judea, they retreated and left it alone
facing the enemy. The large army of Nebuchadnezzar attacked and burned down
the city. Again, thousands of Judeans were taken away to Babylon.
After Judea, Nebuchadnezzar forced Tyros to surrender. In 574 B.C., the
King of Tyros, Ithobal, made a fettering peace agreement with the king of
Babylon. Like the conquered Judeans, Phoenicians were taken to Babylon. Such
Jewish-Phoenician migration waves were habitual in Mesopotamia in the
abovementioned time period. It is known that Nebuchadnezzar during his reign
conducted a military operation to Yemen. The enchanted wealth of this country
has attracted every strong kingdom. According to the story, Nebuchadnezzar
reachedăYemen‘săwesternă boardersăandăthanădecidedătoăstopăandănotăovertireătheă
army and did not continue the operation (Maksutov 1905: 331). The beauty and
wealth of Yemen was very attractive for forcibly relocated Jewish people in
Babylonia. Babylonia during the reign of Nebuchadnezzar had a highly advanced
culture which was only possible with highly developed trade conditions. In
Babylonia local and imported trade property and their barter were protected by
law. The traditional sphere of this business was prospective to the forcibly
displaced Jews.
From the Babylonian captivity, the Jewish Community migrated to Arab
and other, among them Semitic and non Semitic, populated areas. Thus we do not
see it impossible to consider the plural suffix (mīm) found in QAD to be a
reflection of these circumstances in Arab-Hebrew linguistic contacts, which
predates the migration of Arabs of the Qakhlai tribe from southern Arabia to
Central Asia. As for preserving the abovementioned plural forming element in
CentralăAsianăArabic,ăităisămostlyădueătoăQAD‘săcenturyălongăisolation.ă
The plural formation with mīm is not seen today in Yemenite dialect. In
1989 during field work conducted at Aden, located in Abyan Province, and to
Mukalla in Yemen, I recorded the dialectological material where the
abovementioned construction has not been evident. Since we do not have the
written source of the old Arabic dialect of Yemen, it is impossible at this stage to
know if there was ever in this dialect linguistic contacts showing plural formation
with mīm. Only the data from QAD gives us the possibility to relate this
construction to the Yemen Qakhlai tribe. Therefore, the material of Peripheral
Arab dialects is important and must be taken in serious consideration from the
perspective of researching the Arabic dialects and, in general, the history of
Arabic language.
37
References
***. 1989. Biblia, Sakartvelos sapatriarko,ă Tbilisi:ă Publishingă Houseă ―Patriarchateă ofă
Georgia‖.
Axvlediani, V. G. 1985. Buxarskiy arabskiy dialect, Tbilisi: Publishing House
―Metsniereba‖.
Bauer, G. M. 1966. Yazik iuzhno-araviyskoy pis‘mennosti, Moscow: Publishing House
―Nauka‖.
Bertsol‘d,ăK.ă1904. Assiriya i Vaviloniya, S.-Peterburg:ăPublishingăHouseă―Nauka‖.
Brockelmann, K. 1908. Grundriss der Vergbichenden Grammatik der semitischen
Sprache, Bd. I, Berlin: Verlag Von Reuther und Reichard.
Chikovani, G. 2002. Shua aziis arabuli dialektebi. Kashkadariuli dialekti. Phonologia,
gramatika, leksika,ăTbilisi:ăPublishingăHouseă―LanguageăandăCulture‖.
Delich, F. 1909. Kolibel‘ xristianstva. Vavilon i Bibliya, tri lekcii, S.-Peterburg.
Gezenius, V. 1874. Evreyskaya grammatika, S.-Peterburg.
Grande, B. M. 1972. Vvedenie v sravnitel‘noe izuchenie semitskix yazikov, Moscow:
PublishingăHouseă―Nauka‖.
Lane, A. W. 1863. Arabic-English Lexicon, Frederick Ungar Publishing & Co.
Lekiashvili, Al. 1963. Saxelta skesisa da ricxvis carmoeba semitur enebshi, Tbilisi:
Publishing House Tbilisi State University (TSU).
Leslau, W. 1965. The Land of Prestler John, Faculty Research Lecture, University of
California, Los Angeles: Publishing House University of California.
Maksutov, V. P. 1905. Istoriya Drevnego Vostoka, t. II, knigi V-VIII,ăNevskiy,ă№139,ăS.Peterburg: Printing-houseă―Berezhlivost‖.ă
Mamulia, G. 1988. Dzveli Palestina, dzveli agmosavletis istoria, Tbilisi: Publishing
Houseă―Metsniereba‖.
Segert, S. 1965. Ugaritskiy yazik,ăMoscow:ăPublishingăHouseă―Nauka‖.
Tsereteli, G. 2004. Uaxloesi teoriebi semitta pirvandeli samshoblos shesaxeb, rcheuli
txzulebani xut tomad, I, urartologia, semitologia, ebraistika, Tbilisi: Publishing
Houseă―LanguageăandăCulture‖.
Tsereteli, K. 2001. Bibliis ebrauli, Tbilisi: Publishing House Tbilisi State University
(―Nekeri‖).
Vinkler, G. 1913. Vavilon, Published by P. P. Soykin, S.-Peterburg.
Wright, W. 1967. A Grammar of the Arabic Language, Third edition, vol. I-II,
Cambridge: Cambridge University Press.
38
Shi‘ite Panegyrical Poems from the Township
of Dašt-i z dag n (H~ūzist n)
Dénes Gazsi
Eötvös Loránd University of Sciences, Budapest
Hungary
1. Introduction
Poetry has beyond doubt played an immense role in the history of Arabic
culture as a whole, and in the history of Arabic literature in particular. Poetry, on
the other hand, has never been confined to the literary speech; from the very
beginning it had a vast output in dialectal pieces as well. Many of the dialectal
poems were written down, probably in an effort to safeguard the oral lore of its
people from loss for the future generations.
Even in our present century, when travelling through Arabic-speaking
regions, one may meet many people who compose poems in the tradition of their
forefathers. This seems to be the case in the present-day Iranian province of
H~ūzistānă asă well;ă aă factă whichă isă demonstratedă ină aă bookă ofă twoă volumesă titledă
Nigāhī ba mašāhīr-i ‗ilm wa adab-i H~ūzistān1 that I managed to obtain when I
spent two months in the Islamic Republic of Iran at the beginning of 2005.2 These
volumes contain hundreds of poems composed by well-known and lesser known
poets from various regions of H~ūzistān,ă andă ină theă pastă fewă monthsă Iă startedă toă
transcribe them into Latin script with the help of a native speaker. My aim with
this is analysing their most important linguistic features while, at the same time,
incorporating them into a wider scope of the description of the local dialects.
‗A look into the famous scholars and writers of H~ūzistān‘
I thank the Islamic Culture and Relations Organization (ICRO) for granting me the scholarship
thatăenabledămeătoăconsultăwithă‗AlīăAﻣrafăS}ādiqīăonăArabicăandăPersianălinguisticsăandăvisitămanyă
libraries and cultural institutes in Tehran. Sincere thanks are also due to Werner Arnold for kindly
inviting me to Heidelberg on a DAAD scholarship and to Stephan Procházka for letting me use the
library at the University of Vienna where I was able to get acquainted with some western sources
on this topic. I also wish to offer my thanks for the stimulating comments from the participants of
the Colloquium.
1
2
39
The present article is intended to give an inkling of this continued work on
the H~ūzistāniă dialects.ă Foră theă sakeă ofă brevity,ă onlyă someă examplesă fromă theseă
poems will be shown and a set of peculiarities will be discussed that can be
discerned from them. The chapters of the above-mentioned book are divided
according to the townships of H~ūzistān,ă andă underă eachă townshipă (šahristān in
Persian) the poets living there are discussed in detail. Any of the townships could
have been chosen for the above-mentioned purpose, but what struck the eye in
connectionă withă theă townshipă ofă Daﻣt-i ƥzādagānă isă thată manyă ofă itsă poetsă hadă
composedăpanegyricalăpoemsăinăwhichătheyăpraisedăanyăofătheătwelveăŠī‗īăImams.ă
This topic is not to be found so frequently in any other region of the province. The
reason for this phenomenon remains yet unclear, although it is correct to ascertain
thatămanyă‗Alids,ătheădescendantsăofăImamă‗Alī,ăhaveăbeenălivingăforăcenturiesăină
this area.3
2. The locality and its population
Daﻣt-iăƥzādagānămeansăinăPersianătheă‗fieldăofătheăfree-born‘.ăItăisăsituatedă
to the west of theăprovincialăcapitală Ahwāzăandă hasăaăcommonă borderăwithăIraq.ă
ToătheănorthăliesătheătownshipăofăDizfūlăandăŠūﻣ,ătoătheăsouthăH~urramﻣahrăwithătheă
Iranian Gulf region. In Iran, in general, the changing of town names and township
names are never easy to follow, the changing of names in the Arabic-speaking
areas has particularly been complex throughout the history of the country. A fact
behind this may well be the constant efforts of consecutive regimes to Persianize
the indigenous Arabic-speaking population.
The region in question was first known in modern times by the name of
the most numerous Arab tribe inhabitingă it,ă theă Banīă T}uruf. 4 Its name was
subsequently changed to H}uwayza,ăwhileă ină1935ăităwasăagainăchangedătoăDaﻣt-i
MīﻣānăbyătheăthenărulingăRid}āăŠāh.ăThisăreflectsătheăMiddleăPersianăformăMēﻣān,ă
the name of the whole of lower Mesopotamia and parts of H~ūzistān,ăthatăwasăusedă
until the late Middle Ages. 5 Until 1944 when it became an independent
administrativeă unit,ă ită hadă formedă aă partă ofă theă townshipă ofă Ahwāz.ă Afteră theă
Pūr-Kāz}im 1378/1999, 334.
SeeăunderăDaﻣt-iăMīﻣānăinăDihh~udāă1377/1998,ă10905.ăSeeăalsoătheămapăshowingătheănamesăofă
tribes in Ingham 1982, 18.
5
Ingham 1997, 47 (Footnote 22). The area has been known under different forms of this name in
variousălanguagesăsuchăasăGreek,ăSyriac,ăArmenianăandăHebrew.ăEvenăYāqūtămentionedăităunderă
the name Dast-iăMaysān.ăForăaălistăofătheseăforms,ăseeăPūr-Kāz}im 1378/1999, 334., Streck-Morony
1991, 918f and 920f. For Dast-iăMīsān,ăseeăalsoăDihh~udāă1377/1998,ă10878.
3
4
40
Islamică Revolutionă ină 1979,ă theă nameă wasă yetă againă altered,ă thisă timeă toă Daﻣt-i
ƥzādagān.
The centre of the present-dayă townshipă isă Sūsangard,ă formerlyă calledă
H~afāğiyaăinăArabic6,ăandătheămajorătownsăareăBustānă(previouslyăcalledăBisaytīn)ă
asăwellăasăHūzegānă(formerlyăknownăasăH}uwayza).7
The residents of this area have generally been Arab nomads who were to a
large extent forcefully sedentarized in the past decades. 8 This nomadic past is still
evident from their speech, for they speak the ‗arab oră ‗nomadic‘ă typeă ofă theă
H~ūzistāniă dialects,ă ină contrastă withă theă h}a d}a r oră ‗sedentary‘ă typeă ină theă southernă
partsăofătheăprovince,ăalongătheăbanksăofătheăŠat}t} al-‗ArabăandătheălowerăKārūn. 9
In general, the dialect of both groups is distinguishably a Mesopotamian gilitdialect. Thus, H~ūzistānīăArabic,ăinătheălinguisticăsense,ăisănotăperipheral because it
shows all the features of the Arabic dialects on the Iraqi side of the border, it can
only be considered peripheral in a geographical sense. It is true, though, that since
H~ūzistānīsă liveă underă aă majoră Persiană influence,ă theyă tendă toă use more Persian
words than their Iraqi counterparts.
AccordingătoăIngham‘săclassificationăofătheăH~ūzistāniădialects,ătheăspeechă
in the northern region of the province 10 is typologically closer to that of the
‗Amāraăarea.ăThisăareaăcomprisesătheămarshlands called H}awr al-H}uwayza to the
eastă ofă theă townă ofă ‗Amāra,ă nowă situatedă ină Iraq,ă rangingă acrossă theă Iraq-Iran
border as far as H}uwayzaă (Hūzegān)ă andă Bisaytīnă (Bustān)ă onă theă Iraniană side.ă
This territory is still largely populated by the so-calledăMi‗dānăoră‗marshăArabs‘.ă
AsăregardsătheăregionalăfeaturesădiscussedăinăIngham‘săarticle,ătheălocalăH~ūzistāniă
speechăwhichăisăofătheă‗nomadic‘ătypeăhasăsimilaritiesăwithătheădialectsăspokenăină
theăŠat}t} al-‗ArabăandătheălowerăKārūn.ăForătheăcityăofăSūsangardăonly one feature
characteristicăofătheă‗Amāra-variety was attested, every other feature was absent.
Thisăexistingăpeculiarityăisăthatăinătheă‗nomadic‘ătypeăofătheădialectăcertainăverbală
ForăSūsangard,ăseeăDihh~udāă1377/1998,ă13842.ăForăH~afāğiya,ăseeăDihh~udāă1377/1998,ă9886.
Dihh~udāă1377/1998,ă9242.ăWrittenăalsoăasăHuwayza,ăsee Dihh~udāă1377/1998,ă23595.
8
The population in the southern parts of H~ūzistānăunderwentătheăprocessăofăsedentarizationăsomeă
time earlier due to the bad effects they had had on the exploitation of oil in the region. Earlier
writers such as Lorimer testifiedătheănomadicăconditionsăofătheănorthernăareas.ăHeăgaveăAhwāzăandă
Íuwayzaă asă theă maină pastorală areasă andă Banīă Tamīmă alongă withă Banīă T~uruf in H}uwayza as the
tribes with a majority of nomadic sections. See Ingham 1997, 50 (Footnote 66).
9
In Iran the division between the h}a d}a r and ‗arab dialects usually reflects the discrepancy
between the population living from palm cultivation on the one hand, and those occupied in cereal
cultivation and pastoralism away from the rivers on the other hand. See Ingham 1997, 29.
10
The areas to the north and west of theăprovincialăcapital,ăAhwāz.
6
7
41
and nominal forms involving only the three radicals followed by a vowelbeginning suffix drop the vowel between the first and the second radical, e.g.
nšidat ‗sheăasked‘.ăOtherăcharacteristicsăofătheădialectăofătheă‗Amāraăregionăandăitsă
surroundingă marshlandsă includeă theă pronunciationă /ž/ă asă aă reflexă ofă jīm 11 , the
occurrence ofă/ī/ăasăaăreflexăofătheăClassicalăArabicădiphthongă/ay/ăinăcertainănonverbal forms12,ăandătheăuseăofă/‗ēb/ăasăaănegativeăparticleăforăverbs. 13
3. Poems
The poems in the volumes are written in Arabic with some necessary
vowel signs for an easier comprehension14 , but no Persian translation is given
anywhere throughout the books. The explanatory panels between the poems as
well as the biographies of the poets and the general introduction to the townships
are all in Persian. This implies that the book was written for an audience who
basically speak Persian and either belong to the Arabic-speaking minority or have
some knowledge of Arabic and want to get acquainted with the literature of this
minority.
I am completely aware of the fact that a written dialectal material should
always be treated carefully. One should by no means come to final and clear-cut
conclusions regarding the dialectal features based on written texts. Such material,
especially poems, does not by all means mirror the actual spoken language. My
goal with this material is to show to what extent these texts are reliable in setting
up the characteristics of the given dialect and how the well-known linguistic
features peculiar to the region are attested in them. In the following some random
excerpts from these panegyrical poems can be read. Through the first complete
poem I intend to demonstrate some of the linguistic features of the area, whereas
afterwards only some lines from other poems will be shown that have interesting
characteristics. I, of course, do not have in mind to give a full phonological and
morphological description of the dialect now because the features deduced from
the texts all fit into the general characteristics of the Southern Mesopotamian gilitdialects.
In the rest of Southern Mesopotamia and H~ūzistānă theă regulară reflexă ofă jīm is /y/ or /j/, see
Johnstone 1965, 236-237ff. and Johnstone 1967, 9-11ff.
12
Examples such as /h}wīza/ăforăH}uwayzaăandă/daﻣtămīﻣān/ăareăgivenăinăIngham 1997, 33.
13
For a thorough elucidation of these regional contrasts, see Ingham 1997, 31-35.
14
There are also poems originally written in Persian but these are out of the scope of our interest
this time.
11
42
3.1. The composer of the first poem is al-Marh}ūmă Sayyidă Ğa‗fară
Mūsawī. 15 He was born in 1935 into a family of renowned scholars. Both of his
fatherăandăgrandfatherăwereăŠī‗īătheologians.ăHeăwasăaădescendantăofă‗Alī,ăthroughă
theăseventhăImam,ăMūsāăal-Kāz}im. This could have been one of the reasons why
he composed so many qas}īdas ină whichă heă praisedătheă Prophet‘să familyă andă theă
Šī‗īă Imams.ă Furthermore,ă heă belongedă toă theă familyă ofă theă chieftainsă ofă alBūgadīmī,ă andă wasă aă veryă famousă andă respectedă citizenă ină hisă nativeă townă ofă
Bustān.ă Unfortunately,ă heă didă notă getă toă publishă hisă poems,ă soă theseă areă onlyă
known in hand-written versions. The following qas}īda praises the first Imam,
‗Alī.
1. ‗Alī yā nūr il-hidāya wa bah}ər yā-l mā la h}a dd // ‗Alī yā bah}r il- buh}ūr
il-kaləf mā yinsidd.
‗Alī,ăohălightăofă[divine]ăguidanceăandăseaăthatăhasănoăboundary,ă//ă‗Alī,ăohă
seaăofăseas,ătheăfondnessă[forăyou]ăcannotăbeăblocked‘.
2. ‗Alī s}a rs}a r ib-yōm il-kōn mā yinridd // ‗Alī isma l-bārī h~tār wa (m‘īd).16
‗Alīăletăoutăaăpiercingăcryăonătheăday of the battle that cannot be answered,
//ă‗Alī,ăGodăchoseăhisănameăandă….‘
3. Nāda bi s-samā Ğibrīl bi-sma gām yitraddad // ‗Alī bi l-bašar h~a s}s}ā(h),
ib-s}ifāta bi l-bašar mufrad.
‗InătheăskyăGabrielăcalledăoutăhisănameăthatăstartedăechoing,ă//ă‗Alīăisă
distinguished among mankind, with his attributes he is unique among
mankind‘.
4. Rabb il-‗arəš min h~alqā(h) āya li-nabī Ah}mad // Bi l-Ka‗ba in wulid
mowlūd, ġīrak yā ‗Alī mah}h}a d.
‗TheăLordăofătheăThroneăisătheăoneăwhoăcreatedăanăāya (sign, wonder) for
the Prophet Ah}mad,ă//ăIfăthereăwereăanăinfantăbornăbyătheăKa‗ba,ăohă‗Alī,ăită
wouldăbeănoăotherăthanăyou‘.
5. Bi l-mowlūd farh}ānīn farh}a mā ad}unn tūğad // Fağr in-nūr hall wa lāh}
li l-‘Islām bi l-mowlid.
‗Byătheă[birthăofăthe]ăinfantăweăfeelăsuch joy that, I think, does not exist, //
TheădawnăofălightăappearedăandăglitteredăforăIslamăină[his]ăbirthplace‘.
15
16
Pūr-Kāz}im 1378/1999, 449-450.
The correct reading and meaning of the last word remains unclear.
43
6. Ib-mīlādak yā bū l-H}asanīn, čam h}āġī lladī thaddad // ‗Alī yā tamrat
il-‘Islām, šağara t}ayyiba aw warəd.
‗Byăyourăbirth,ăohăFatherăofăÍasan and Íusayn, how many an oppressor was
demolished,ă//ă‗Alī,ăohăfruităofăIslam,ăaăpleasantătreeăorăaărose‘.
7. ‗Alī yā h}ād}ir iš-šiddāt il-yidkura ib-kill wakət yištadd // ‗Alī bi-sma
niraddid dōm biyya š-kitīr nitmağğad.
‗Alī,ăohăwhoălivedăthroughămiseries, the mentioning of which becomes
moreăintenseăeveryătime,ă//ă‗Alī,ăweăconstantlyărepeatăhisăname,ă[and]ăhowă
oftenăareăweăpraisedăthroughăhim‘.
In the field of phonology the following characteristics based on the poem
can be outlined:
– The affrication of the kāf when preceding a front vowel, e.g. čam (Line
6) stands for Classical Arabic kam ‗howămany‘. 17
– The voiced velar plosive /g/ as a reflex of qāf, e.g. gām (Line 3) for Cl.
qāma ‗startedădoingăsomething‘. 18
– The realization of qāf as the voiceless velar plosive /k/, e.g. wakət (Line
7) for waqt ‗time‘.19
– The Classical Arabic preposition bi- has two forms depending on its
phonological environment. It either remains bi- when it is followed by a
consonant cluster, e.g. bi s-samā (Lineă3)ă‗inătheăsky‘, bi-sma (Lineă3,ă7)ă‗(by)ăhisă
name‘,ăbi l-bašar (twiceăinăLineă3)ă‗amongămankind‘,ăbi l-Ka‗ba (Lineă4)ă‗byătheă
Ka‗ba‘,ă bi l-mowlid (Lineă 5)ă ‗ină [his]ă birthplace‘,ă bi l-mowlūd (Lineă 5)ă ‗byă theă
[birthăofăthe]ăinfant‘,ăorăităbecomesă ib- when it is followed by a sequence of one
consonant and a vowel, e.g. ib-yōm (Lineă2)ă‗onătheăday‘,ăib-s}ifāta (Lineă3)ă‗withă
hisă attributes‘,ă ib-mīlādak (Lineă 6)ă ‗byă youră birth‘,ă ib-kill (Lineă 7)ă ‗everyă time‘.ă
Furthermore, the dialectal form of bihi is biyya (Lineă7)ă‗throughăhim‘.
A certainly more interesting peculiarity is that lexemes or even syntagms
from Classical Arabic quite often occur in the poems, by which they become not
exclusively dialectal. This phenomenon has also been attested on many occasions
in the texts collected by Ingham in H~ūzistān.ăInă hisădataăthisăusuallyăoccurredăină
the speech of šuyūh~, to whom much prestige is attached and are on a higher social
This sound change is a very common feature not only in the H~ūzistāniă dialectsă bută ină mostă
Mesopotamian vernaculars as well, see Blanc 1964, 25., Fischer-Jastrow 1980, 143.
18
See Fischer-Jastrow 1980, 142.
19
This phenomenon is well-known in some Iraqi dialects in the case of certain words, see Blanc
1964, 27.
17
44
level within their community. 20 This seemingly makes them entitled to use forms
that would otherwise not be current in everyday conversations. In the previous
poem such classicisms are in wulid (Lineă 4)ă ‗ifă …ă wereă born‘,ă tūğad (Line 5)
‗exist‘ă andă alladī (Lineă 6)ă ‗that‘.ă Theă wordă kitīr ‗much,ă often‘ă ină Lineă 7ă hasă ană
intermediate form between the classical katīrun and the true dialectal čitīr. This
shows that the poet has intended to draw his style nearer to Classical Arabic,
while at the same time keeping the general framework for the dialectal forms. In
Line 5 the cognate accusative (maf‗ūl mut}laq) farh}ānīn farh}a mā ad}unn tūğad
‗[we]ă feelă suchă joyă that,ă Iă think,ă doesă notă exist‘ă isă alsoă aă signă ofă makingă theă
language more literary, since this structure is not as frequent in Arabic dialects as
it is in Classical Arabic. Of course, this and the next poems were not written by
šuyūh~ asăităisătheăcaseăinăIngham‘sătexts,ăbutătheăgenreăitselfămakesăitănecessaryătoă
include classicisms in the otherwise dialectal text.
3.2.ăTheă followingăpoetăisătheăcontemporaryăMullāăNa‗īmă Kurūﻣādī. 21 He
isă aă memberă ofă theă Kurūﻣātă clană thată belongsă toă theă Baytă Sa‗īdă familyă ofă Banīă
T}uruf.ă Heă isă originallyă fromă theă townă ofă Bustān,ă bută rightă nowă aă residentă ofă
Ahwāz.ăTheăfollowingăshortăpoemăwasăwrittenăinăpraiseăofătheătwelfthăImam,ătheă
Mahdī.
1. Inta rah}ma d-dīn rabbna, inta sūra w-inta āya // ya‘ti l-h~īr ib-wus}ūlak
w-inta li- h}āġī nihāya.
‗Youăareătheămercyăofătheăreligionă(Islam),ăourăLord,ăyouăareăaăsūra, and
you are an āya, // Blessing comes upon your arrival, and you put an end to
theăoppressor‘.
2. Inta l-kill ‗ūd nudūwa w-inta li l-‗at}šān māya // Inta l-t}āġīn hadma,
inta li l-mad}lūm tāya.
‗Youăareămoistureăforăeveryăwoodăandăyouăareăwaterăforătheăthirsty.ă//ăYouă
bring destruction to the oppressors, [and] you are shelter for the
tyrannized‘.
Following the logic of poetic language discussed at the previous verse, a
Classical Arabic word can also be found in here, ya‘ti (Lineă1)ă‗heăcomes‘.ăApartă
from that, the rest of the bayt is purely dialectal. Besides, there are two unusual
words in Line 2. The first one is māya ‗water‘,ăwhichăis,ăasăfarăasăIăcould find out,
20
21
Ingham 1982, 173.
Pūr-Kāz}im 1378/1999, 447-448.
45
either mayy or māy in H~ūzistān. 22 And the second word is tāya ‗shelter‘,ă whichă
seems to be very rare in the dialects of the region. The verb tawa originally means
‗toăstayăatăaăplace‘,ăandătāwa and tāya standăforă‗shelterăofăcamels‘. 23 In our poem
itămeansă‗shelter‘ăinăaăbroaderăsense.
3.3. Another example of a Classical Arabic form is to be seen in a verse by
Mullāă H}abībă S}ayyāh}īă Sālimī. 24 Heă isă aă descendantă ofă theă thirdă Shi‗iteă Imam,ă
H}usayn and was born in 1969 in a village not far from Bustān.ă Heă hasă beenă
writing poems from an early age, and a huge number of his poems are in praise of
the third Imam. Here is only one of his mis}rā‗s.
1. Tabqā (i)b-had}a m wa b-dilla wa-ms}āyib //ă…
‗Youă[will]ăstayăinăoppression,ăinăsubmissivenessăandădisasters‘.
The word tabqā is clearly classical, with the retention of the qāf, and in the
Arabic text even an ‘alif maqs}ūra has been put on the final yā‘ to mark length
(ăٰ) تﻘﻰ.
But in another poem composed by Íaydară Ma‗ﻣalīă fromă theă famousă Baytă
Ma‗ﻣalăclan25, the real dialectal version of the indefinite form of the verb baqiya
can be found.
1. (‘Ağrak ‗alā l-Bārī l-wāh}id il-maškūr) // Yabga hnāk tilgā(h) ibma‗ālī(h).
(‗Yourărewardăisă[incumbent]ăuponătheăCreator,ătheăOne,ătoăwhomăoneă
shall be grateful) // If he stays there, you will find him in his excellent
deeds‘.
In this line the Arabic text writes g as a reflex of the qāf, and the final -a is
written with a hā‘ ( )يتگهwhich corresponds to its dialectal pronunciation.
3.4. The following line is from Sa‗dăibnăĞabbārăSuwārī.26 He was born in
the first quarter of the 20th century in a small village in the border region of
Bustān.ă Althoughă heă remainedă illiterateă foră hisă wholeă life,ă heă hasă composedă
22
Ingham 1982, 173 and 174.
Al-Munğid 1997, 76.
24
Pūr-Kāz}im 1378/1999, 442-443.
25
Pūr-Kāz}im 1378/1999, 468.
26
Pūr-Kāz}im 1378/1999, 474-475.
23
46
poems from the age of 20 which were written down by one of his friends. Many
ofăhisăpoemsăareădedicatedătoătheăMahdī,ăwhileăothersătreatătheăfeudsăandăkillingsă
between the segments of the Arab population of H~ūzistānă dueă toă theă dividingă
policies of Rid}āăŠāh.
Līš galli-tlōm h~āyib yā Sa‗d // wa ġīr amr illā(h) fa-lā yanzil wa‗d.
‗Whyădidăheătellăme:ă‗youăblameăabortively‘,ăohăSa‗dă(addressingă
himself)?ă//ăSinceăwithoutătheăorderăofăGodănoăpromiseădescends‘.
In the second mis}rā‗ă theă poetă usedătheă syntagmă fa-lā yanzil, an insertion
from Classical Arabic, with not only the negative marker lā but the vocalization of
theăverbătoo.ăAăpeculiarityăofătheă‗Amāra-variety of H~ūzistāniăArabicăisăattestedăină
thisăline,ătheăClassicădiphthongă/ay/ăisăpronouncedăasă/ī/. 27 It should be noted that
although the verses were read out by a native of the region, the vowel signs in the
writtenă textă conceală theă differenceă betweenă theă /ī/ă usedă ină thisă areaă andă theă /ē/ă
typical of the rest of the province.
3.5.ăTheăsonăofăthisăpreviousăpoet,ă‗AlīăibnăSa‗dăSuwārīăwasăbornăină1947ă
in the same village as his father.28 Due to the poverty his family was living in, he
spent 13 years working in Kuwait. The majority of his poems are panegyrics of
theăŠī‗īăImamsăincludingătheăoneăwhoseămis}rā‗ reads as follows:
Wa kill yōm agūlan tihill gālu yihill bāčir //ă…
‗Everyă timeă Iă sayă ‗youă [will]ă appear‘,ă theyă sayă ‗heă [will]ă appeară
tomorrow‘.
A feature of frequent occurrence in H~ūzistāniă Arabică isătheăpresenceăofăaă
suffix -an, appearing optionally in the 1st person singular of the imperfect of
hollow and doubled verbs. It also appears with other types of verbs when followed
by object pronoun suffixes.29 The exact explanation of this phenomenon is yet to
beăclarified,ăbutăitămayăseemătoăbeătheăreminiscenceăofătheă‗energeticăform‘ă(al-fi‗l
al-mu‘akkad) used widely in Classical Arabic.
3.6. Since we are dealing with a dialect spoken in Iran, we should not be
surprised at the use of Persian words within the poems. This is exactly what
The example līš isăalsoăfoundăinăIngham‘sămaterial,ăseeăInghamă1997,ă32.
Pūr-Kāz}im 1378/1999, 476-477.
29
For further examples, see Ingham 1997, 16.
27
28
47
‗Abbāsă H}urayzāwī,ă aă contemporaryă poetă andă aă nativeă ofă Sūsangard,ă doesă ină hisă
mis}rā‗:30
…ă//ăYā s}āh}bi čarh~ iz-zimān ib-dōr.
‗Ohămyăcompanion,ătheăwheelăofătimeăisăinărotation‘.
The Persian word čarh~ meaningă ‗wheel,ă fortune,ă fate‘ 31 is put in a
construct state where it retains its original meaning. This word also appears in
Iraqi Arabic, even though there is confusion as to what it exactly means there. In
hisă dictionaryă onă theă foreignă loansă ină spokenă Baġdādī,ă Majīdă Muh}ammad gives
čarih~ ină theă meaningă ‗hurry‘,ă andă čarh~a la ină theămeaningă ‗aă walk‘,ă soă ită bearsă aă
slightly different meaning. 32 For both words he nevertheless gives Persian as
origin, but in Persian none of the two meanings can actually be found. The Iraqi
Arabic dictionary of Woodhead-Beene, on the other hand, mentions čarih~ (Pl.
čurūh~)ăasă‗wheel‘ăandăčarh~a la (Pl. -āt)ăasă‗spin,ăturn,ărevolution‘. 33
4. Conclusions
During the analysis of the poems it occurred to me that the motive for the
use of classicisms and other non-dialectal forms may as well have metrical
reasons. This can be the case when dialectal short vowels were substituted for
classical long ones. But poetry as a genre requires to some extent a language close
to literary Arabic given the eloquence of the speech and the sublimity of the
subject matter. Having this in mind, one should be cautious about trying to define
the characteristics of a dialect reflected in the texts. Moreover, the poets
themselves could have introduced literary forms as hypercorrections because
these give a better expression to the purport of their message.
Fromă theă featuresă ofă theă ‗Amāra-variety of H~ūzistāniă Arabică discussed
previously, one is certainly not to be found in the corpus. The letter jīm is in every
case written in its regular form ()ﺝ, though its pronunciation should be a voiced
postalveolară fricativeă /ž/.ă Ină viewă ofă theă factăthată otheră soundsă alienă toă Classical
Arabic are written with the letters borrowed from Persian (g as گand č as )چ, one
may expect the letter ژtoăbeăusedăforă/ž/.ăWhereasăthroughoutăIraq,ătheă/g/ăandă/č/ă
Pūr-Kāz}im 1378/1999, 473-474.
Dihh~udāă1377/1998, 8100-8102.
32
Muh}ammad 1990, 126-127.
33
The dictionary even gives a verbal form čirah~ ‗toăturnăonăaălathe‘ăasătheăbasicăword,ăfromăwhichă
the other ones are derived, see Woodhead-Beene 1967, 84.
30
31
48
sounds in many location names are in fact written with the Persian characters, in
no case can the use of the ژbe observed. Following this logic, it was merely
unthinkable for the poets to incorporate this character into their text, even if by
this they did not render the exact phonetic value of the dialect. However, the
poems in the two volumes are representative of a treasure of popular literature in
the Arabic-speaking regions of Iran and because of this, they should not be left
aside and thrown into oblivion for reason that they are not fully adequate for a
linguistic study.
References
al-Munğid fī l-luġa wa l-‘a‗lām.ă1997.ăBayrūt:ăDārăal-Maﻣriq.
Blanc, Haim. 1964. Communal Dialects of Baghdad, Cambridge, Massachusetts: Harvard
University Press.
Dihh~udā,ă ‗Alīă Akbar,ăMu‗īn,ă Muh}a mmad,ă Šahīdī,ă Ğa‗fară (eds).ă 1377/1998.ă Luġatnāma,
15 vols., T}ahrān:ăMu‘assasa-yiăLuġatnāma-yi Dihh~udā.ă[Lexicon]
Fischer, Wolfdietrich, Jastrow, Otto. 1980. Handbuch der arabischen Dialekte,
Wiesbaden: Harrassowitz.
Ingham, Bruce. 1982. North-East Arabian Dialects, London: Kegan Paul.
Ingham, Bruce. 1997. Arabian Diversions. Studies on the Dialects of Arabia, London:
Ithaca Press.
Johnstone,ă Thomasă M.ă 1965.ă ―The sound change j > y in the Arabic dialects of
peninsular Arabia‖,ăin:ăBSOASăxxviii,ă2ă(1965),ă233-41.
Johnstone, Thomas M. 1967. Eastern Arabian Dialect Studies, London: Oxford
University Press.
Majīd,ăMuh}a mmad. 1990. Mu‗ğam al-mus}t}a lah}āt wa l-‗alfāz} al-‘ağnabiyya fī l-luġa al‗āmmiyya al-‗irāqiyya,ăBaġdād:ăDārăaﻣ-Šu‘ūnăat-Taqāfiyyaăal-‗ƥmma.
Pūr-Kāz}im, H}āğğăKāz}im. 1378/1999. Nigāhī ba mašāhīr-i ‗ilm wa adab-i H~ūzistān, Vol.
2.,ăSūsangard:ă Intiﻣārāt-iă Sarzamīn-i H~ūz.ă [Aă lookă intoă theă famousă scholarsăandă
writers of H~ūzistān].
Streck,ă M,ă Morony,ă M.ă 1991.ă―Maysān‖,ă in:ă C.ă E.ă Bosworth,ă E.ă vană Donzel,ă B.ă Lewis,ă
Ch. Pellat (eds.): The Encyclopaedia of Islam, New Edition, Volume VI, Leiden:
E. J. Brill, 918-923.
Woodhead, D. R., Beene, Wayne. 1967. A Dictionary of Iraqi Arabic, Arabic-English,
Washington D. C.: Georgetown University Press.
49
50
L’énoncé non verbal dans l’arabe parlé à Mardin
George Grigore
Université de Bucarest
Roumanie
Leă travailă queă l‘onă vaă présenteră comporteă uneă précisionă préalableă poură
éliminer une possible confusion due à la superposition des termes qui englobent le
mot « Mardin »ă poură dénommeră diversesă variétésă d‘arabe.ă J‘entends ici par
« l‘arabeăparléăàăMardin » (désormais, le mardini), le parler de la ville de Mardin
(Turquie)ăqui,ăconformémentăàăOttoăJastrow,ăformeăavecăd‘autresăparlersă– Ebene,
Kōsa, Mh}a llami, Qart}mīn, Azəh~ etc. – le groupe dialectal de Mardin qui, à son
tour,ă appartientă àă laă brancheă anatolienneă deă l‘arabeă mésopotamienă deă typeă qəltu
(Jastrow 1994 : 121).
Dans ce travail, je propose d‘illustrerăparăquelquesăexemplesălesădifférentsă
typesăd‘énoncé nonăverbalăexistantădansăcetteăvariétéăd‘arabe.
Je crois nécessaireă d‘ajouteră queă leă tableauă queă jeă dresseraiă utiliseraă ună
corpusă deă donnéesă tiréă deă conversationsă spontanées,ă enă mardini,ă queă j‘aiă
enregistré sur le terrain, à Mardin, durant mes séjours temporaires pendant une
période comprise entre 2002 et 2006.
Pour cetteă étude,ă j‘aiă commeă pointă deă départ,ă laă descriptionă effectuéeă pară
Dittersă (2001)ă poură laă structureă phrastiqueă deă l‘arabeă dansă uneă approcheă
déclarative – c‘est-à-dire non interprétative, qui vise à décrire les faits langagiers
en se basant sur des structuresă syntaxiquesă etă sémantiques.ă Enă partantă d‘uneă
description élaborée par le biais de fonctions spécifiques et en terme de
constituantsăimmédiats,ăDittersădistingueădeuxătypesăd‘énoncés :
– un énoncé verbal avec une fonction prédicative et ayant comme élément
principal un verbe. Les autres constituants sont définis par la structure valencielle
du verbe.
– un énoncé non verbal à fonction énonciative possédant une fonction
topique qui comprend comme catégorie principale un syntagme nominal, et une
fonction commentaire comprenant soit un autre syntagme nominal ou un
syntagme adjectival, adverbial, prépositionnel ou bien une proposition. Le terme
employé par Ditters et qui se retrouve dans la majorité de travaux sur ce sujet est
celuiă d‘énoncé nominal, mais ayantă enă vueă qu‘ilă s‘agită deă propositionsă dansă
lesquelles le prédicat est assuré non seulement par un mot appartenant à la classe
51
nominale,ămaisăaussiăparăunăsyntagmeăadverbial,ăprépositionnel,ăjeăconsidèreăqu‘ilă
faudraităplutôtăparlerăd‘énoncé non verbal (averbal).
L’énoncé non verbal
À partir du modèle de classification des énoncés non verbaux en
h}assāniyya élaboré par Catherine Taine-Cheikh (1994 :186),ă j‘auraiă enă vue,ă poură
l‘analyseă quiă suit,ă lesă typesă suivants d‘énoncésă nonă verbauxă dansă l‘arabeă parlé à
Mardin:
- l‘énoncéăexistentiel ;
- l‘énoncéălocatif ;
- l‘énoncéăattributif ;
- l‘énoncéăéquatif ;
- l‘énoncéădeămonstration.
L’énoncé existentiel formulé à l’aide de fīyu
Leă mardiniă emploieă poură formuleră ună énoncéă existentielă l‘auxiliaireă deă
prédication fīyu « existe », « il y a », une forme figée, qui peut être expliquée par
laă grammaticalisationă d‘ună syntagmeă contractéă fī-hu « en lui » ou fī-wu. Pour la
forme négative, on emploie seulement fi (en): mā-fi « ilăn‘yăaăpas ». Cet auxiliaire
de prédicationăn‘estăpasăspécifiqueăauămardini,ă maisăilăestădesăplusăfréquentsă dansă
les dialectes arabes actuels auxquels il fournit une gamme vaste des moyens
d‘expression.
L‘auxiliaireădeă prédicationă fīyu/mafi a des équivalents de même type dans
toutes les autres langues parlées à Mardin : en turc, var/yok, en t}uroyo, kīt/layt ou
kito/layto (pour le passé : kitwa/laytwa)1 , en kurde, heye/tune. Comme ceux-ci
(pour le turc voir Deny 1920 : 813), cet auxiliaire indique en mardini :
a)ăl‘existenceăouăl‘inexistenceăenăgénéralăd‘unăconcept :
ayš fīyu ?
Qu‘yăa-t-il ?
ayš fīyu ayş mā-fi ?
Qu‘yă a-t-il,ă queă n‘yă a-t-ilă pas?ă (c‘est-à-dire : « qu‘est-ceă qu‘ilă yă aă deă
nouveau ? »).
təmm fīyu, lsēn mā-fi.
Boucheăilăyăa,ălangueăilăn‘yăaăpasă(seădităd‘uneăpersonneăquiăneărépond pas
OttoăJastrowăexemplifieăl‘emploieădeăcetăauxiliaireădeăprédication en t}uroyo par la proposition :
layt talgo ؟al i=ar؟o « es liegt kein Schnee auf der Erde » (ilăn‘yăaăpasădeăneigeăsurălaăterre) (1992 :
64)
1
52
aux questions).
mā-fi ši
Ilăn‘yăaărien.
mā-fi ah}h}ad.
Ilăn‘yăaăpersonne.
b) existence localisée :
hawne fīyu ‗aš byūt mērdīnīye.
Ici il y a dix maisons (familles, n.n.) mardiniennes.
‗ənde məne2 fīyu čaqmaq ?
Qui a un briquet (lit.: « chez qui il y a un briquet ? ») ?
ğamb ōtēl-na fīyu h}ammām.
À coté de notre hôtel il y a un bain (public).
fə-l-bayt mā-fi m}a yy.
Ilăn‘yăaăpasăd‘eauădansălaămaison.
Dansă leă cadreă deă l‘énoncéă existentiel,ă leă syntagmeă nominală quiă exprimeă
l‘entitéădontăonăsignaleăl‘existenceăestătoujoursăindéterminéăetăilăestăprécédéăparăleă
syntagmeă nominală prépositionnelă (quiă indiqueă leă lieuă d‘existence).ă Ceux-ci sont
lesă traitsă fondamentauxă quiă leă différencientă deă l‘énonceă locatif,ă introduită pară laă
copule enclitique, à qui il ressemble beaucoup par ses composantes. La différence
entre un énoncé locatif (exemple a) et un énoncé existentiel (exemple b) est la
suivante :
al-h}ammām fə-wast} əl-wəlāye we.
Le bain (public) est au centre de la ville.
fə-wast} əl-wəlāye fīyu h}ammām.
Dans le centre de la ville il y a un bain (public).
Dans tous les exemples ci-dessus, fīyu a le sens de présent ; pour le placer
dans le passé, on emploieăl‘invariableăkān « il a été » : kān fīyu « il y a eu » ; mākān fi « ilă n‘yăaăpasă eu » ; et pour la placer dans le futur, on emploie ta-ykūn « il
sera : ta-ykūn fīyu « il y aura » ; mō-ta-ykūn fi « ilăn‘yăauraăpas ».
2
Le /e/ final qui apparaît à ‗ənde et məne est euphonique.
53
Les énoncés attributif, locatif et équatif formulés à l’aide d’une copule
enclitique
Le mardini emploie un suffixe prédicatif ou une copule enclitique, comme
elle a été nommée par H. J. Sasse en parlant du mh}allami, « Die enklitische
Kopula » (Sasse 1971 : 247), qui a les mêmes formes que le pronom personnel
autonome duquel il a évolué et avec lequel il peut être confondu : ana ; ənt ; ənti ;
we (forme réduite de hūwe) ; ye (forme réduite de hīye) ; nəh}ne ; əntən ; ənne
(forme réduite de hənne). Ce type de copule, basée sur les pronoms personnels, est
présenteăaussiădansăd‘autresăparlersămésopotamiens : dans le parler de Mossoul et
dans les parlers chrétien et juif de Bagdad elle se forme de la particule yā à
laquelle on ajoute les pronoms suffixés : kalebna kelleš zēn yānu « Notre chien est
très bon » (Blanc 1964 : 124) ;ă dansă leă parleră arabeă deă l‘Asieă Centraleă elleă seă
forme de l‘élémentă inn + les pronoms suffixés : nōxuš-inni « Je suis malade »
(Blanc 1964 : 125).
La grammaticalisation de cette copule a comme point de départe, je
suppose, le pronom proleptique ou emphatique qui figure dans les constructions
de mise en relief et quiă s‘accordeă enă genreă etă enă nombreă avecă sontă referant.ă Ceă
genreă deă pronomă estă employéă dansă toutesă lesă variétésă d‘arabe,ă yă comprisă leă
classique :
Zaydun huwa al-wazīru.
Zayd est, lui, le ministre.
La liaison avec cette structure proleptique est plus évidenteă dansă l‘arabeă
parléăàă Siirtă oùă cetteă copuleă n‘estă pasă enclitiqueă commeă elleă estă enă mardini,ă maisă
elle est placée entre le sujet et le nom prédicatif comme ,,, le pronom proleptique
en classique :
āvi lbənt īye bōš malīha}
Cette fille est très bonne (Jastrow 1980 : 148).
Cetteăsérieă deăsuffixesă prédicatifsă (enăfond,ă desăpronomsă enclitiques)ă s‘estă
développée, je crois, à partir des modèles offerts par les autres langues parlées à
Mardin.
Ainsi,ădansăleăTurcăosmanliăetălesădialectesăturcsădeăl‘Anatolieăorientale – y
compris Mardin – deă tellesă structuresă existent,ă c‘est-à-dire des propositions non
verbales réalisées avec des suffixes prédicatifs – nommés par Deny verbes
substantifs – qui sont les mêmes pronoms personnels (Deny 1920 : 351-352) :
Män yaş män – « Je suis jeune » ; Sen yaş sen – « Tu es jeune » etc.
En t}uroyo, les suffixes prédicatifs se rapprochent davantage des pronoms
personnels. Singulier : 1. ono – no ; 2.m. hat – hət ; 2.f. hat – hat ; 3.m./f.
54
hiye/hiya – yo ; Pluriel : 1. ah}na – na ; 2. hatu – hatu ; 3. hənnək – ne (v. Jastrow
1992 : 23).
La ressemblance entre les suffixes prédicatifs en t}uroyo et les suffixes
prédicatifs en mardini en ce qui concerne leur origine, leur formation et leur
emploi, est évidente (voir les exemples inscrits dans le tableau ci-dessousă oùă l‘onă
a employé comme nom prédicatif en t}uroyo, rabo (f. rabto; pl. rabe) « grand » et
en mardini, gbīr (f. gbīre; pl. gbār}) « grand » :
Personne
I, sg.
II, m., sg.
II, f., sg.
III. m., sg.
III., f. sg
I. pl.
II. pl.
III. pl.
t}uroyo
ono rabo-no
hat rabo-hət
hat rabto-hat
hiye rabo-yo
hiya rabto-yo
ah}na rabe-na
haut rabe-hatu
hənnək rabe-ne
mardini
ana gbīr ana
ənt gbīr ənt
ənti gbīre ənti
hūwe gbīr we
hīye gbīre ye
nəh}ne gbār} nəh}ne
əntən gbār} əntən
hənne gbār} ənne
français
Je suis grand
tu es grand
tu es grande
il est grand
elle est grande
nous sommes grands
vous êtez grands
ils sont grands
Par ailleurs, ce phénomène peut être soutenu par des structures similaires
du kurde qui emploie le verbe bûyîn (être) comme copule qui institue la jonction
entre le sujet et le nom prédicatif. Passant en revue la conjugaison de ce verbe au
présent, indicatif, ez im « je suis » ; tu yî « tu es » ; ew e « il est » ; em în « nous
sommes » ; hûn in « vous êtes » ; ew in « ils sont », on peut observer la similitude
– due au hasard linguistique, sans doute, mais qui aurait pu jouer un rôle dans le
renforcement de ces suffixes en mardini – à la troisième personne du singulier et
du pluriel : kurde e – mardini we; kurde in (prononcé ən) – mardini ənne.
L‘ordreădesămotsădansălesăénoncésănonăverbauxăenămardini,ăkurde,ăturcăetă
t}uroyo est identique : Sujet – Nom prédicatif – Copule :
mardini : Abū-y rā‗i we.
kurde : Bavê min şivan e.
turc : Babam çobandır.
t}uroyo : Babi rə‗yo-yo.
/Père-mon berger est/
Mon père est berger.
Ce suffixe prédicatif est employé pour exprimer un énoncé non verbal de
type :
a)
attributif – composé de deux syntagmes, un syntagme nominal
sujet fortement déterminé et un syntagme nominal très peu référentiel (adjectif,
55
participe ou syntagme indéfini), qui peut être introduit aussi par la préposition
kama (comme):
šəġəl-na qlayyəl zōr we.
Notre travail est un peu difficile.
t}ālə‗ ana əlle nēzəl ana ?
Je monte ou bien je descends (lit. : « Je suis montant ou bien je suis
descendant ? ») ?
h}əss-u kama h}əss əd-dīk we.
Sa voix est comme la voix du coq.
maymət la ar}ayna fə-l-lōqānt}a mən mērdīn ye.
La grand-mère que nous avons vue dans le restaurant est de Mardin.
ta‘ān ! tūtāyāt-i agbar } ənne !
Viens ! Mes mûres sont plus grandes !
b)
locatif – composé de deux syntagmes nominaux ; le syntagme
nominal sujet, qui est le thème, est toujours fortement déterminé et le second
syntagme est prépositionnel ou adverbial :
mērdīn fōq əğ-ğabal ye.
Mardin est sur la montagne.
c) équatif – composé de deux syntagmes nominaux, non prépositionnels,
toujoursădéterminés,ăliésăentreăeuxăparăuneărelationăd‘identification :
arqad}āš zaynab əbn ‗amm-i we.
L‘amiădeăZaynab est mon cousin [paternel].
En mardini, la présence de la copule, dans les énoncés non verbaux, mène
àă l‘omissionă systématiqueă duă pronomă sujet,ă parceă queă lesă informationsă qu‘ilă
renferme sont reprises par la copule (comme dans les énoncés verbaux, elles sont
reprises par le verbe) :
əl-yawme, arba‗t əyyēm, marīd} ana.
Aujourd‘hui,ăilăyăaăquatreăjours,ădepuisăqueăjeăsuisămalade.
ayš fī-k? mō-trīd təqčəm ma‗-i wēya h}əlqān ənt ?
56
Qu‘estăqueătuăas ? Tu ne veux pas parler avec moi ou bien tu es fatigué ?
On peut observer dans les exemples ci-dessus que le suffixe prédicatif
s‘accordeă enă genreă etă enă nombreă avecă leă sujet.ă Enă dépită deă cela,ă notreă baseă deă
données fournit parfois des exemples qui mettent en évidence la tendance de
généralisationă deă l‘emploiă desă suffixesă prédicatifsă deă laă troisième personne du
singulier :
-məne bəl-bāb we?
- nəh}ne we!
- Qui est à la porte?
- C‘estănous!
La négation
Laă négationă desă syntagmesă prédicatifsă formulésă àă l‘aideă d‘uneă copuleă
enclitique se fait au moyen de mo3, qui est très répandu dans les parlers
mésopotamiens. Il est communément admis que mo représente une forme
contractée du syntagme ma+hūwe « ilă n‘estă pas », employé encore tel quel dans
les parlers mésopotamiens méridionaux. Cette étymologie expliquerait la tendance
d‘exclureălaă co-occurrence de la copule de la série we (il est) et le morphème de
négation mo (ilă n‘estă pas)ă dansă leă mêmeă énoncé,ă parceă queă mo est la forme
négative de we. Même si les locuteurs mardiniens préfèrent la commutation des
deuxă élémentsă (lesă exemplesă aă etă b),ă j‘aiă enregistréă desă énoncésă oùă lesă deuxă
formes co-existentă (l‘exempleă c)ă ouă mêmeă l‘emploiă deă we figé,ă quiă neă s‘accordeă
plusăavecăleăsujetă(l‘exempleăd) :
a) bətl} ān ana.
Je suis fatigué.
b) ana mō- bətl} ān.
Je ne suis pas fatigué.
c) ana mō- bətl} ān ana
Je ne suis pas fatigué.
d) ana mō- bətl} ān we.
ăDansă leă parleră arabeă d‘Asəh~, la négation deă laă copuleă n‘estă pasă encoreă réduiteă àă ună seulă
morphème, étant dans une phase antérieure de grammaticalisation par rapport au mardini : 1 sg.
mana « je ne suis pas » ; 1 pl. mənnəh}ne « nous ne sommes pas » ; 2 sg. m. mənt « tuăn‘esăpas » ; 2
sg. f. mənti « tuăn‘esăpas » ; 2 pl. məntən « vousăn‘êtesăpas » ; 3 sg. m. mū/mō « ilăn‘estăpas » ; 3
sg. f. mi « elleăn‘estăpas » ; 3 pl. mən « ils ne sont pas ». Par exemple : ənt mənt masīh}ī ? « du (m.)
kein Christ ? » (tu – m.– n‘esăpasăchrétien ?) (Wittich 2001 : 34-35)
3
57
Je ne suis pas fatigué.
Le morphème de négation mo fait, habituellement, corps commun avec la
séquenceă niéeă (lorsqu‘elleă estă représentée par un nom, un adjectif), par
conséquent,ălaăvoyelleăn‘étantăplusăfinaleădevientălongue,ămō :
əl-bas}ale mō-t}ayyəbe ‗a-t-tūme.
L‘oignonăn‘estăpasăbonăsurăl‘ail.
šap}qat la fə-r }ās mh}ammad mō-ğdīde.
La casquette quiăestăsurălaătêteădeăMuhammadăn‘estăpasănouvelle.ă
h}ār}r} hawnake mo kama h}ār}r} hawne we; hāke nēšəf we.
Laă chaleură deă làă (deă Mardin,ă n.ă n.)ă n‘estă pasă commeă laă chaleură d‘iciă
(d‘Istanbul,
n.n.) est ; celle-là est sèche.
Des exemples précités, on remarque que la copule we (et toute sa serie)
indique le présent ; tout changement au non-présent ou au non-indicatif est
accompagnéăparăl‘apparitionăd‘uneăformeăduăverbeă« être » : kān (pour le parfait),
kān ykūn (pourăl‘imparfait),ă kān kəl-kān (pour le plus-que-parfait), ta-ykūn (pour
le futur), kān ta-ykūn (pour le conditionnel).
Ceci nous amène à considérer que la construction dite non verbale à
copule we est une proposition à verbe kān auă présentă deă l‘indicatif,ă laă formeă deă
présentădeă l‘indicatifădeăceă verbe étant phonologiquement vide, mais présente au
niveauă structurale.ă Pară conséquent,ă l‘énoncéă nonă verbal pour tête syntaxique le
verbe kān, qui est remplacé ici par la copule we (etă saă série).ă Dansă l‘arabeă
classique et dans la plupart de dialectes arabes actuels, le verbe kāna dans une
proposition nominale est réalisé par une forme vide, autrement dit par un zéro qui,
dans le mardini, est remplacé par we. La co-occurrence de we et kān est
impossible parce que les deux éléments ont la même fonction, ceux qui les rend
incompatibles. Si we aurait eu la fonction de pronom proleptique ou emphatique
quiăfigureădansălesăconstructionsădeămiseăenăreliefădansălesăautresăvariétésăd‘arabe,ă
alors son occurrence avec kān serait été admise.
Les formes négatives du verbe kān s‘obtiennentăenăajoutantăl‘adverbeă ma
ou mo avant le verbe pour exprimer successivement la négation dans le passé ou
dans le futur.
La commutation de la copule we avec zéro est possible lorsque le
syntagme prédicatif est introduit par le locatif ‗ənd « chez » :
58
Fawzi ‗ənda-na.
Fawzi est chez nous.
Mais, pour des raisons stylistiques, de tels énoncés ne sont pas très
employés par les locuteurs qui leur préfèrent des phrases nominales à copule :
Fawzi ‗ənda-na we.
Fawzi est chez nous.
Si, copule et préposition locative sont compatibles, la co-occurrence de we
et ‗ənd n‘estăpasăadmiseăquandă lesădeuxăélémentsăontăuneăfonctionă verbaleă(‗ənd
lorsqu‘ilă exprimeă l‘appartenanceă ouă laă possession,ă ceă quiă équivaută auă verbeă
« avoir »).ăDansăl‘exempleăFawzi ‗ənd-ən we « Fawzi est chez eux », la présence
de copule est possible, parce que ‗ənd a ici sa fonction primaire de préposition
locative. La reconnaissance des fonctions de la préposition ‗ənd peut se fonder sur
les indices suivants :
– en fonction verbale,ăelleăestăplacéeăenătêteădeăpropositionă(‗ənd-i əl-čaqmaq
– « chez moi est le briquet », « je possède le briquet », elle est parfois
renforcéeăparăl‘auxiliaireădeăprédicationăfīyu « il est »:ă‗ənd-i fīyu əl-čaqmaq ;
mais en fonction prépositionnelle elle vient selon le thème : əl-čaqmaq ‗ənd-i
we « le briquet est chez moi ».
– la négation est aussi définitoire :ăenătantăqueăprépositionăd‘appartenance,ăelleă
porte la négation spécifique aux verbes « ma » : mā ‗ənd-i əl-čaqmaq « jeăn‘aiă
pas le briquet » ; en tant que préposition locative, il porte la négation
spécifique aux nominaux : mo : əl-čaqmaq mō-‗ənd-i «Leă briquetă n‘estă pasă
chez moi ».
L’énoncé de monstration formulé à l’aide de l’auxiliaire kwā
Le mardini emploie un auxiliaire de prédication pour laă réalisationă d‘ună
énoncé de monstration dont la seconde partie est soit un syntagme nominal
déterminé, soit un syntagme verbal. Cet auxiliaire, kwa « c‘est », « voici » est
formé du kū – une forme tronquée de ykūn « il est » – et du démonstratif universel
hā4 (Jastrow 1981 : 164). Je crois que ce mot est formé du k et du wa. Le k est le
démonstratif k‘ deăl‘araméenăbabylonienă– le même qui est renfermé dans aku « il
Dans un autre arabe anatolien, celui parlé à Qart}mīn,ăkū se combine également avec hīya (elle)
résultant kyā,ăunăauxiliaireăemployéăpourăl‘introductionăd‘unănomăféminin : kya ar}nabe « C‘estăună
lièvre » (Jastrow 1980 : 148).
4
59
y a »,ălaăparticuleăd‘existenceăenăirakienă(Müller-Kessler 2003 : 643) – et wa vient
de huwa « il »,ăc‘est-à-dire « le voilà ».
SN:
kwa l-h~bayz !
Voici le pain !
kwa l-masa} w l-kərsi !
Voici la table et la chaise !
SV:
a) unăparticipeăactif,ăayantăleăsensăd‘unăprésentăcontinu :
kwa r }āyəh} !
Le voici partant !
Observation : kwa peut être suivi directement par un verbe, mais il est clair
que le sujet nominal ou son substitut pronominal a été élidé :
a)ăunăverbeăàăl‘inaccompli,ăayantăleăsensăd‘unăprésentăcontinu :
kwa yākəl !
Leăvoiciăqu‘ilămangeă(c‘est-à-dire, il est en train de manger).
b)ăunăverbeăàăl‘accompli,ăayantăleăsensăd‘unăpasséăproche :
kwa ğa !
Le voici venu !
Références
Blanc, Haim. 1964. Communal Dialects in Baghdad. Massachusetts : Harvard University
Press.
Boucherit, Aziza. 2002. L‘arabe parlé à Alger. Aspects sociolinguistiques et énonciatifs.
Paris-Louvain : Editions Peeters.
Deny, J. 1920. Grammaire de la langue turque (Dialecte osmanli). Paris : Éditions
Ernest Leroux.
Ditters, Everhard. 2001. « The Description of Modern Standard Arabic Syntax in Terms
of Functions and Categories », in Langues et Littératures du Monde Arabe,
2(2001): 115-151.
Grigore, George. 2003. « Quelques traces du contact linguistique dans le parler arabe de
Mardin (Turquie) », Romano-Arabica 3 : 119-134.
Grigore, George. 2007. L‘arabe parlé à Mardin – monographie d‘un parler arabe
périphérique. Bucarest :ăEdituraăUniversitĒįiiădinăBucureşti.
Jastrow, Otto. 1978. Die mesopotamisch-arabischen qəltu-Dialecte. vol. 1. Phonologie
und Morphologie. Wiesbaden : Steiner.
60
Jastrow, Otto. 1980. « Das Mesopotamische Arabisch », Handbuch der arabischen
Dialekte (éditeurs: Wolfdietrich Fischer et Otto Jastrow). Wiesbaden :
Harrassowitz :140-173).
Jastrow, Otto. 1981. Die mesopotamisch-arabischen qəltu-Dialekte. vol. 2. Stuttgart :
Franz Steiner.
Jastrow, Otto. 1992. Lehrbuch der T}uroyo-Sprache. Semitica Viva – Series Didactica.
Wiesbaden : Otto Harrassowitz.
Jastrow, Otto. 1994. « The qəltu dialects of Mesopotamian Arabic », Actas del Congreso
Internacional sobre interferencias linguisticas arobo-romances y paralelos
extra-iberos. Zarogosa : 119-123.
McCarthy, R. J. ; Raffouli, Faraj. 1964. Spoken Arabic of Baghdad. Beyrouth.
Müller-Kessler, Christa. 2003. « Aramaic ‘k‘, lik‘ and Iraqi Arabic ‘aku, maku. The
Mesopotamian Particles of Existence » , Journal of the American Oriental
Society 123.3 : 641-646.
Sasse, Hans-Jürgen. 1971. Linguistische Analyse des Arabischen Dialekts der
Mh}allamīye in der Provinz Mardin (Südossttürkei). Berlin.
Taine-Cheikh, Catherine. 1994. « Le H}assāniyya de Mauritanie, un dialecte nonmarginal de la périphérie », Actas del Congreso Internacional sobre
interferencias linguisticas arobo-romances y paralelos extra-iberos. Zarogosa :
173-199.
Talay, Shabo. 1999/2003. Der arabische Dialekt der Khawetna, I : Grammatik (1999),
II : Texte und Glossar (2003). Wiesbaden : Harrassowitz.
Wittich, Michaela. 2001. Der arabische Dialekt von Azəx. Wiesbaden : Harrassowitz.
61
62
Where do we stand in the research on the Anatolian qəltu
dialects?
Otto Jastrow
Friedrich-Alexander University, Erlangen-Nürnberg
Germany
Anatolian qəltu dialects are conventionally divided into four groups
(Jastrow 2006):
1
Mardin group
Mardin town (Muslims; Christians, mostly emigrated)
Mardin villages (Muslims; Christians, emigrated)
Plain of Mardin (Muslims ; Christians, extinct)
Kōsa and Mh}a llami dialects (Muslims, 1 Christian village)
ƥzəx (Christians, now emigrated)
Nusaybin and Cizre (Jews, emigrated)
2
Siirt group
Siirt town (Muslims; Christians, extinct)
Siirt villages (Muslims)
3
Diyarbakırăgroup
Diyarbakırătownă(Christians,ăextinct;ăJews,ăemigrated)
Diyarbakırăvillagesă(Christians, extinct)
Siverek,ăČermik,ăUrfaă(Jews,ăemigrated)
4
Kozluk-Sason-Muşăgroup
Kozluk (Muslims; Christians extinct?)
Sason (Muslims; Christians extinct?)
Hasköy (Muslims)
Most of these dialects were only discovered in the second half of the 20 th
century but one of them, the dialect spoken in the town of Mardin, has been
known for over 120 years. I was first described in an article by Albert Socin
(Socin 1882-3). By that time, of course, only few Arabic dialects had been
described at all, and Mardin, together with Mossul, formed just one of the few
isolated points on the Arabic dialect map. It was much later, with the appearance
63
ofă Haimă Blanc‘să epochală bookă Communal Dialects in Baghdad (1964), that the
position of Mardin Arabic was adequately defined, namely as one of the so-called
Anatolian Arabic dialects who in turn formed a subgroup of the qəltu dialect
group of Mesopotamian Arabic (Blanc 1964, 5 ff.). But only after Jastrow had
discovered a number of additional Anatolian qəltu dialects which were described
in part by him and in part by his former students (Jastrow 1973, 1978, 1981, 2003,
2005, Talay 2001, 2002, Wittrich 2001) emerged the actual picture of the
Anatolian Arabic dialect group as given above.
The dialect of the town of Mardin as well as most of the other dialects
belonging to the Mardină group,ă toă theă exceptionă ofă ƥzəx, represent the most
conservative dialect type of Anatolian Arabic. This corresponds to the fact that,
geographically, they are closest to north-eastern Syria and northern Iraq. In an
outer circle, the Mardin dialects are surrounded by a number of more progressive
dialects,ă namelyă theă isolatedă dialectă ofă ƥzəx, and the dialect groups of Siirt,
Kozluk-Sason-Muş,ăandăDiyarbakır.ăWhenăIăfirstădiscoveredăandădescribedăthoseă
dialects I was fascinated by their diversity and the various interesting linguistic
developments which had taken place in them, to the point that for a long time I
overlooked the fact that, notwithstanding their many differences, those dialects
have much in common as compared to Mardin. This can be demonstrated by a
number of features:
t
d
d}
a) Different reflexes of the interdentals
ƥzəx Siirt
Daragözü
Mardin
tāte
sāse
fāfe
sāse
axad
axaz axav
aġaz
d}a rab
z}a rab v}a rab
z}a rab
Diyarbakır
tlāte
axad
d}a rab
―three‖
―heătook‖
―heăshot‖
Table a) shows that only the Mardin group dialects have retained the
original pronunciation of the Arabic interdental fricatives t, d and d} (the latter
sound resulting from the merger of the two phonemes of Old Arabic, d}ād and
d}ā‘), whereas in the remaining dialects the interdentals have been shifted to dental
stopsă (Diyarbakıră group),ă toă sibilantsă (ƥzəx, Kozluk-Sason-Muşă groupă – as
represented by Daragözü) or – a rare sound change – to labio-dental fricatives
(Siirt group).
b) Devoicing of ʕ
Mardin
ƥzǝx
bāʕ, ybīʕ bāʕ, ybīʕ
Siirt
bāḥ, ybīḥ
Daragözü
Diyarbakır
bāḥ, ībīḥ
bāḥ, ybīḥ
“he sold”
64
bǝʕtu
bǝʕtu
bǝḥtu
bǝḥtu
bǝḥtu
“I sold”
Ină theă Mardină groupă dialects,ă includingă ƥzǝx, the voiced pharyngeal
fricative, ʕ, is never unvoiced, neither preceding a voiceless consonant nor in
word final position. This, however, occurs regularly in the remaining three dialect
groups, with ʕ yielding voiceless ḥ, as demonstrated by the above examples. In
Siirt as well as in some of the Kozluk-Sason-Muşădialectsăfinalăḥ furthermore can
be reduced to zero, especially in high frequency words, e.g. Siirt ṭallo, yṭallo
(<ṭalloḥ, yṭalloḥ < ṭalloʕ, yṭalloʕ) “to look”, ṛā, yṛō (< ṛāḥ, ṛōḥ)ă―toăgo‖.
c) Loss of initial h in pronouns
ƥzǝx
Siirt
Mardin
hāḏa
hāza
āva
hawn(e)
hawn(e)
awne
Daragözü
āza
ōn(i)
Diyarbakır
āda
awne
―this‖
―here‖
Whereas in the Mardin group dialects initial h of the independent
pronouns and adverbs is retained, it has been mostly elided in the remaining three
dialectă groups,ă withă aă fewă examplesă toă theă contraryă ină theă Diybakıră group,ă e.g.,ă
hāda alongside āda foră―this‖.
d) Relative pronoun
Mardin
ƥzǝx
Siirt
lallē
Daragözü
lē
Diyarbakır
lē
―which‖
The relative pronoun in the Mardin group dialects is a prefixed l with a
short vowel (la- in Mardin, lǝ- inătheăKōsaăandăMḥallami dialects) or vowelless l(inăƥzǝx). In the Siirt, Kozluk-Sason-MuşăandăDiyarbakırăgroups,ăhowever,ăaăfreeă
form lē is used.
II
mix‖
III
X
e) Formation of verbal stems II, III, V, VI, X
Mardin
ƥzǝx
Siirt
Daragözü Diyarbakır
ḥammal
ḥammal ʕallem xallǝṭ
xalleṭ ―toă load, to teach, to
dēxal
staxbar
stanḍar
dēxal
staxbar
ʕayyoṭ
dēxel
*staxber
stanṿor
dēxǝl
staxbǝr
65
dēxel
staxber
―toăscream‖
―toăinsert‖
―toăask‖
―toăwait‖
As shown in table e) only the Mardin group dialects have retained the
traditional formation of the derived verb stems which is characterized by the
vowel a in the last syllable of the perfect stem, whereas in the remaining three
dialect groups the a vowel has been changed to ǝ by analogy to the imperfect. In
the Siirt group dialects ǝ in word-final simply closed syllables has secondarily
been split into e and o, according to the phonetic environment, hence ʕallem ―heă
taught‖ă vs.ă ʕayyoṭ ―heă screamed‖. Ină theă Diyarbakıră groupă dialectsă ǝ in this
position is generally realized as [e].
To sum up, the dialects of the Mardin group, with the partial exception of
ƥzǝx, almost present themselves as conservative mainstream dialects when
compared to the three other groups of Anatolian Arabic. Although they differ
from the neighbouring Syrian Arabic dialects they are rather close to the north
Iraqi qǝltu varieties. This is why Albert Socin called his paper from the year 18823 ―DerăarabischeăDialektăvonăMosulăundăMärdin‖,ăinăotherăwords,ăheăviewedătheă
two as manifestations of a single dialect type, notwithstanding the existing
differences. Compared to Mardin, the remaining three dialect groups of Anatolian
Arabicăconformămuchămoreătoătheăideaăofătypicală―peripheralădialects‖.ăThusătheyă
are separated from the continuum of the Arabic language area
show no influence of Standard Arabic or mainstream Arabic dialects
whatsoever,
reflect a strong influence of regional languages
are very progressive in their internal development.
With increasing geographical isolation the latter tendency becomes more
and more conspicuous. Thus in the Kozluk-Sason-Muşă dialectă groupă one can
notice further phonetic changes, e.g.
Partial loss of pharyngeals *ʕ and *ḥ: : ata < *ʕaṭa ―heă gave‖,ă hanta <
*ḥǝnṭa ―wheat‖
Loss of emphasis: madar ―rain‖,ăzarab < *ẓarab < *ḏ̣a rab ―heăshot‖
Intervocalic voicing of stops and fricatives: mǝšīdu < mǝšītu ―Iă went‖,ă
bǝge < *bǝkǝt ―sheăwept‖,ăbaġa < *baqa ―heăstayed‖,ăaġaz < *axaz ―heă
took‖ă
Present Research Situation
From the point of view of linguistic investigation and description the
Mardin dialects so far have received the best coverage. Besides the concise
comparative description in Jastrow 1978 and the text collections of six dialects of
this group in Jastrow 1981 (p. 1-215) there are three grammatical monographs:
66
Sasse 1971 on Mḥallami, Wittrich 2001 on ƥzǝx, and, most recently, Grigore
2007 on Mardin. Thus 125 years after Mardin Arabic was for the first time the
subject of an article it finally received a comprehensive and up-to-date
monographic treatment. In 2003 Jastrow published a large collection of texts in
the Mḥallami dialect of Kinderib, preceded by a short linguistic introduction.
Based on this text edition, in 2005 the same author published his Kinderib
glossary, so far the largest glossary of any Anatolian dialect. The author plans to
publish a second volume of Kinderib texts of the same size and has enough text
recordings for several additional volumes on other dialects of the Mardin group.
The research situation for the Siirt dialect group is less favourable.
Although this is a small group, comprising only seven dialects (Siirt town and six
neighbouring villages) our knowledge is mainly based on the comparative
description in Jastrow 1978 and the text collections of four dialects in Jastrow
1981 (p. 217-307). A description of the dialect of Tǝllo by Ablahad Lahdo still
awaits publication. Although Jastrow has unpublished recorded material it is
probably not sufficient for an independent text volume. On the other hand, the
Siirt group dialects are still spoken in situ by a Muslim population and thus are
available for further investigation.
SimilarlyăunsatisfactoryăisătheăresearchăsituationăforătheăDiyarbakırădialectă
group.
Here again our knowledge is mainly based on the comparative description
in Jastrow 1978 and the text collection from the village of Kaʕbīyeă neară
Diyarbakıră ină Jastrowă 1981 (p. 309-371). The few Christian speakers of this
dialect group which survived the massacres during WW I have long since died
and their dialects must be considered extinct. The Jewish speakers left for Israel
only around 1950 but competent speakers are now getting rare. Thus new
investigations are not very promising. On the other hand a lot of Jewish material
either collected by or generously supplied to Jastrow has not yet published.
A similar situation obtains, mutatis mutandis, for the Kozluk-Sason-Muşă
group. Here we have the small monograph by Jastrow (1973) on Daragözü and,
publishedă 30ă yearsă later,ă Talay‘să twoă articlesă onă theă dialectă ofă Hasköy.ă
Nevertheless this dialect group must be considered the less investigated area of all
of Anatolian Arabic. There are many small Arabic-speaking villages between
Kozlukă andă Muşă whereă theă mostă progressiveă andă exotică varietiesă ofă theă wholeă
Mesopotamian qǝltu area are still spoken by a Muslim population. It is therefore
still possible, albeit probably difficult and dangerous, to work in this area and
67
enlarge our present knowledge of these dialects. In my eyes this is the most
important desideratum in the research on Anatolian Arabic.
Thus, to sum up, the Mardin group dialects, although being the most
conservativeă andă lessă ―peripheral‖ă ofă allă Anatoliană dialectsă haveă neverthelessă
received more scholarly attention than the three remaining groups which, from the
point of view of peripherality and linguistic innovation, are perhaps even worthier
of dialectological investigation. For some of these dialects new fieldwork may
still be possible but for others it is too late because meanwhile they must be
considered extinct.
References
Blanc, Haim. 1964. Communal Dialects in Baghdad. Cambridge, Mass.: Harvard
University Press
Grigore, George. 2007. L‘arabe parlé à Mardin – monographie d‘un parler arabe
périphérique. Bucharest :ăEdituraăUniversitĒįiiădinăBucureşti.
Jastrow, Otto. 1973. Daragözü: Eine arabische Mundart der Kozluk-Sason-Gruppe
(Südostanatolien). Nürnberg: Hans Carl
Jastrow, Otto. 1978. Die mesopotamisch-arabischen qǝltu-Dialekte. Band I: Phonologie
und Morphologie. Wiesbaden: Steiner
Jastrow, Otto. 1981. Die mesopotamisch-arabischen qǝltu-Dialekte. Band II:
Volkskundliche Texte in elf Dialekten. Wiesbaden: Steiner
Jastrow, Otto. 1989. ―The Judaeo-Arabică Dialectă ofă Nusaybin/Qāməﻣli.‖ă In:ă Studia
linguistica et orientalia memoriae Haim Blanc dedicata. Ed. by Paul Wexler,
Alexander Borg, Sasson Somekh. Wiesbaden: Harrassowitz. 156-169
Jastrow, Otto. 2003. Arabische Texte aus Kinderib. Wiesbaden: Harrassowitz
Jastrow, Otto. 2005. Glossar zu Kinderib (Anatolisches Arabisch). Wiesbaden:
Harrassowitz
Jastrow,ă Otto.ă 2006.ă ―Anatoliană Arabic‖.ă Encyclopedia of Arabic Language and
Linguistics. I, 87-96
Jastrow,ă Otto.ă 2007.ă ―Iraq‖.ă Encyclopedia of Arabic Language and Linguistics. II, 414424
Sasse, Hans-Jürgen. 1971. Linguistische Analyse des arabischen Dialekts der Mḥallamīye
in der Provinz Mardin. München (Dissertation)
Socin, Albert. 1882-3.ă ―Deră arabischeă Dialektă vonă Mosulă undă Märdin‖.ă Zeitschrift der
Deutschen Morgenländischen Gesellschaft 36.1-53, 238-277; 37.293-318
68
Talay, Shabo. 2001. ―Deră arabischeă Dialektă vonă Hasköyă (Dēră Khāṣ), Ostanatolien. I.
GrammatischeăSkizze―. Zeitschrift für arabische Linguistik 40.71-89
Talay,ă Shabo.ă 2002.ă „Deră arabischeă Dialektă vonă Hasköyă (Dēră Khāṣ), Ostanatolien. II.
TexteăundăGlossar―. Zeitschrift für arabische Linguistik 41.46-86
Wittrich, Michaela. 2001. Der arabische Dialekt von ƨzǝx. Wiesbaden: Harrassowitz
69
70
L’arabe parlé en Sicile était-il un arabe périphérique ?
Jérôme Lentin
INALCO, Paris
France
Dansă lesă mêmesătermesăouăpresqueăqu‘ilăavaităutilisésădixăansăauparavantă
(Borg 1994, p. 42-43), Alexander Borg a récemment proposé à nouveau (Borg
2004, p. xıx) trois critères pour définir les dialectes arabes périphériques :
(a) un isolement géographique et culturel par rapport aux pays arabes ;
(b) des contacts linguistiques historiques ayant substantiellement altéré
leur profil ancien, les rendant virtuellement inintelligibles à des locuteurs natifs
contemporains de l‘arabe ;
(c) une acculturation linguistique faite de façon prédominante à une langue
étrangère déterminée, 1 la diglossie dans sa version arabe monolingue étant
remplacée par une situation de bilingualisme (ou de multilingualisme) stable. La
compétence des locuteurs en arabe périphérique implique donc une certaine
connaissance de langues étrangères.
Il rappelle par ailleurs (2004, p. 59) qu‘onădisposeădeătroisătermes,ă« quasisynonymes » : latéral, marginal et périphérique etă préciseă que,ă dansă l‘usageă des
arabisants, le dernier objective métaphoriquement l‘isolementă géographiqueă etă
culturel des dialectes arabes ainsi qualifiés par rapport aux dialectes arabes
« mainstream ».2
Siăl‘on retient ces trois critères comme définitoires, il est clair queăl‘arabeă
dialectal sicilien – poură autantă qu‘onă leă connaisse – ne saurait être qualifié de
« périphérique ». Un certain nombre des traits linguistiques et des données
sociolinguistiquesă qu‘onă penseă pouvoiră reconnaîtreă commeă caractéristiquesă deă
cetteă variétéă d‘arabe disparue porte cependant à ne pas considérer la question
commeărégléeăavantăd‘êtreăexaminéeăplusăavant.
De façon rapide mais suggestive, David Cohen avait, en 1994 également,
proposé de distinguer entre dialectes arabes périphériques, extra-périphériques et
en expansion sur un terrain en voie de conquête : « Les statuts diffèrent pour un
1
Aboutissant, ajoutait-il dans sa formulation de 1994, à une relation de Sprachbund avec cette
langue étrangère.
2
Dansă leă casă deă l‘arabeă chypriote,ă saă continuitéă avecă leă Sprachraum arabe suggère plutôt une
position de latéralité par rapport à la zone dialectale de Grande Syrie (ibid. p. 60).
71
groupe périphérique vivant aux confins, un groupe extra-périphérique inclus dans
un domaine linguistique hétérogène ou un groupe en expansion sur un terrain en
voie de conquête. Et des phénomènes linguistiques sont propres à chacun, en
partieăenăraisonădeă ―l‘idéal‖ă linguistiqueăquiăestăleăsien,ă imposantăuneă normeăquiă
peut différer selon le statut du groupeă ouă sonă modeă d‘expression » (D. Cohen
1994, p. 16).
Dans cette perspective,ă l‘arabeă sicilienă présenteă uneă configurationă
particulière et intéressante : il a certainement été, dans la première phase de sa
courte histoire, « en expansion sur un terrain en voie de conquête » ; à certains
égards, on peut considérer que, vu sa situation géographique, il vivait sur les
confins du monde arabophone ;ă enfin,ă ilă n‘aă jamaisă complètementă cessé,ă malgréă
une période probableădeădomination,ăd‘êtreă« inclus dans un domaine linguistique
hétérogène », et même doublement : son environnement était non sémitique, mais
aussi diversifié (grec et roman principalement). On voit en particulier que le
caractériser comme « périphérique » est discutable : la Sicile a eu, surtout peutêtre pendant la domination « normande », un rôle commercial central dans la
Méditerranée ; les contacts, pacifiques ou belliqueux, entre arabophones siciliens
etămaghrébins,ăd‘Ifrîqiyaănotamment,ăontăétéăconstants ; en même temps, dans de
nombreusesăzonesădeăl‘île,ălesăcommunautésăarabophones,ăpuisăcellesăquiăavaientă
été arabisées,ă étaientă enă contactă constantă avecă desă communautésă parlantă d‘autresă
languesă et,ă deă ceă fait,ă certainesă d‘entreă ellesă étaientătrèsă probablementă bilingues,ă
voire trilingues. Si donc les arabophones de Sicile vivaient à la marge ou à la
périphérie du monde arabophone, le qualificatif de « périphérique » ne peut
s‘appliqueră àă leură langueă queă deă façonă relative,ă auă moinsă pendantă laă périodeă deă
dominationămusulmane.ăEnădernièreăanalyse,ăc‘estălaăquestionăduăstatutădeăl‘arabe,ă
aux différentes époques, qui doit être examinée, comme y invite D. Cohen, pour
déterminer le degré de « périphéralité » deă l‘arabeă sicilien ; on verra que la
questionă n‘estă pasă simple.ă Cetă aspectă desă chosesă estă fondamental,ă etă saă priseă enă
compte certainement nécessaire pour la définition même de la notion de
périphérie linguistique (une variété linguistique minoritaire enclavée dans un
environnement linguistique hétérogène pouvant par exemple, suivant les cas, être
ou non périphérique).
Onă ajouteraă d‘abordă qu‘auă coursă deă l‘ensembleă desă conquêtes arabomusulmanesă l‘arabeăs‘estătrouvéă leăplusăsouventă« en expansion sur un terrain en
voie de conquête »,ă etă qu‘enă celaă laă situationă deă laă Sicileă n‘étaită pasă a priori
différenteă deă beaucoupă d‘autres ;3 maisă l‘histoireă aă faită queă cetteă expansionă aă étéă
Onăobserveraăd‘ailleursăqu‘enăphaseădeăconquête,ălesăzonesăoùălaănouvelleălangueăs‘installeăsontă
pară définitionă situées,ă temporairement,ă auxă margesă duă domaineă enă voieă d‘expansionă etă qu‘ellesă
cessentăpeuăàăpeuădeăl‘êtreăàămesureăque le « front » se déplace.
3
72
plusălimitéeăqu‘ailleurs,ălaăprésenceădeăl‘arabeărelativementăbrèveă(quatreăsiècles),ă
etă sonă refluxă totală (àă l‘exceptionă notableă desă communautésă juivesă sură leă casă
desquelles on reviendra) et rapide.4 C‘estă auă coursă deă ceăreflux,ă etă àă partiră d‘uneă
époque difficileă àă détermineră avecă précision,ă qu‘onă pourraită sansă doute qualifier
l‘arabeă sicilienă deă « périphérique », dans un autre sens encore, car il a été
marginalisé avant de disparaître, restant vivant seulement dans une communauté
elle-même précisément marginale,ăjusqu‘àăl‘expulsionădesăJuifsăenă1492.
Onăajouteraăensuiteăqueăsiăl‘onăpeutăraisonnablementăimaginerăqu‘ilăaităpuă
exister pendant une courte période des pidgins, puis des créoles arabes (comme
cela a dû être le cas souvent lors des conquêtes, même si les témoignages directs
demeurentă rares),ă rienă neă permetă deă caractériseră l‘arabeă deă Sicileă commeă ună
pidgin ;ădeă même,ă l‘hypothèseăduă Siculo-Arabic de D. Agius (1996, p. 120, 122,
359)ă quiă considèreă qu‘ilă yă auraită euă uneă variétéă pidginisée,ă etă parfoisă créolisée,
hybrideă d‘arabeă etă deă romană (et,ă dansă uneă moindreă mesure,ă deă grec)5 reste une
hypothèse.
Nature et état de la documentation
L‘arabeă deă Sicileă estă encoreă relativementă peuă étudié.ă Pendantă longtemps,ă
ceăsontăsurtoutălesăinfluencesăetălesătracesădeăl‘arabe sur les dialectes siciliens qui
ontă retenuă l‘attentionă desă chercheursă (citonsă parmiă lesă plusă illustresă aujourd‘huiă
Girolamo Caracausi et le regretté Giovan Battista Pellegrini). Si des arabisants
italiens ont publié, dès le 19e siècle, un certain nombreădeădocumentsăd‘archives,ă
quiă ontă suscitéă deă nombreusesă études,ă ceă n‘estă queă récemmentă queă desă travauxă
approfondisăsurăl‘arabeăsicilienăontăvuăleăjour,ăenăparticulierăleălivreădéjàăcitéădeăD.ă
Agius,ăenă1996,ăetăceluiădeăl‘historienăA.ăMetcalfeăenă2003.ăMais il reste beaucoup
à faire, tant pour la recherche de documents nouveaux que pour la réédition 6 et
l‘analyseăetăl‘interprétationădeăceuxăquiăsontădéjàăconnus. 7
4
On ne sait pas si des Chrétiens ont continué à parler arabe comme première langue après 1300.
Du point de vue de la langue écrite, la dernière inscription tombale en arabe connue date de 1275,
et la dernière signature arabe (probablementă pară ună chrétien)ă d‘ună acteă publică deă1282ă (Metcalfeă
2003, p. 186).
5
Cf. Metcalfe 2003 p. 144 pour la période normande. De même on rejettera, avec H. Bresc (2001
p. 50)ă l‘affirmation de Wansbrough (1984, p. 21) selon laquelle la langue des documents judéoarabes tardifs (15e siècle) publiés dans Giuffrida-Rocco 1976] « pourrait, avec quelque justice, être
décrite comme un pidgin ».
6
J. Johns et A. Metcalfe préparent une nouvelle édition des textes publiés par Salvatore Cusa dans
ses Diplomi. A. Metcalfe signale (2003, p. 149) 1300 « variations » entre le texte édité et les
manuscrits, non inclus 2000 différences mineures.
7
Telles sont les tâches du Groupe International pour l‘Étude de l‘Arabe de Sicile (GIRAS, créé en
juin 2005 à Raguse).
73
Onă n‘aă pasă retrouvéă deă textesă siciliensă enă arabeă dialectală ouă trèsă
dialectalisant (on ne connaîtăpas,ăparăexemple,ăd‘équivalentăduăzağal andalou pour
laăSicile).ăPourăessayerăd‘appréhenderălaăréalitéădialectaleăancienne,ăonăenăestădoncă
réduit,ăcommeăc‘estăsouventăleăcas,ămaisăplusăencoreăqueăpourălesăautresărégions,ăàă
l‘analyseă deă textesă quiă relèventă peuă ouă prouă duă Moyenă Arabe,ă enă particulier les
célèbres ğarā‘id, 8 ouă leă texteă d‘Al-Idrīsī non encore étudié, sauf erreur, de ce
point de vue. Le traité de lah}n al-‗āmma d‘Ibnă Makkī, seul de son genre connu
pour la Sicile, souvent mis à contribution, reste à exploiter plus complètement.
Pour le judéo-arabe, plusieurs documents ont été publiés, de façon plus ou moins
satisfaisante, et le plus souvent dans une perspective uniquement historique.
Si,ă onă leă voit,ă lesă informationsă sură l‘arabeă dialectală sicilienă queă nousă
fournissent tous ces documentsă sontă surtoută indirectes,ă etă s‘ilă faută naturellementă
lesă exploiteră avecă toutesă lesă précautionsă deă méthodeă quiă s‘imposent,ă laă matière,ă
pourtant relativement réduite par rapport à la considérable documentation dont
nousă disposonsă poură l‘Espagneă musulmaneă pară exemple,ă estă loină d‘êtreă
négligeableă etă onă doită êtreă plusă optimisteă qu‘A.ă Metcalfeă (2003)ă qui,ă toută enă
apportant pourtant une contribution très utile, parle de « most deceptive of
subjects ».ăOnăaăenăeffetăsuffisammentăd‘éléments,ăàălaăfoisădansălesătextes et dans
lesă notations,ă bienă qu‘éparsesă etă auă totală peuă nombreuses,ă d‘ordreă
sociolinguistique,ă poură supposeră raisonnablementă qu‘ilă yă avaită ună arabeă dialectală
sicilien, certainement diversifié, dans lequel on doit inclure le judéo-arabe
sicilien ; que, bienă queă s‘étantă élaboréă auă départ9 àă partiră d‘élémentsă disparatesă
(commeă l‘étaientă lesă populationsă quiă ontă arabiséă – en partie – l‘île),ă ilă s‘estă
constitué une identité propre, même si elle a sans doute, au moins dans certaines
zones, évolué plus profondémentă queă celleă d‘autresă dialectesă compteă tenuă desă
apportsă deă populationsă successifs,ă duă Maghreb,ă deă l‘Andalusă ouă deă l‘orientă
égyptien ;ă queă cetă arabeă fută pendantă quelquesă sièclesă laă langueă maternelleă d‘uneă
partie de la population, qui pouvait par ailleurs être bilingue ou trilingue10 ; que
l‘arabeădeăSicileăconnaissaităuneăsituationădeădiglossieă(quiătouchaitănaturellementă
deă façonă inégaleă lesă locuteursă deă l‘arabe) 11 . Les contacts de langues nés de la
situation de plurilinguisme de la Sicile et de beaucoup de ses habitants ont donné
Dont ilăestăd‘ailleursăloinăd‘êtreăcertainăqueătousăaientăétéărédigéăparădesăSiciliens.
C‘estă malheureusementă poură laă premièreă période,ă leă premieră siècleă auă moins,ă queă laă
documentation est la plus pauvre.
10
En outre, à la période musulmane, « l‘arabe,ă commeă langue des colons les plus nombreux et
victorieux,ăaădûăêtreălaălangueădeăl‘acculturationăet,ăcommeă lingua franca, le moyen principal de
communication » (Metcalfe 2003 p. 23).
11
Le Moyen Arabe de Sicile se présente fondamentalement comme une variété semblable à celles
connues ailleurs (cf. Lentin à paraître a).
8
9
74
lieuă àă deă multiplesă influences,ă deă l‘arabeă etă sură l‘arabe. 12 La documentation,
souventăétudiée,ăqueănousădonnentăl‘anthroponymieăetălaătoponymie,ăenăparticulieră
dans les traductions grecques ou latines de divers documents, dont des listes de
vilains, est (Metcalfe 2003 passim, par ex. p. 150 ou 164) à manier avec plus de
précautionsă encoreă qu‘ilă n‘estă souventă fait ;ă maisă ilă n‘enă resteă pasă moinsă qu‘onă
peut en tirer de nombreuses informations additionnelles. Il faut insister sur
l‘importance de la documentation judéo-arabe en provenance avérée de la Sicile :
ilăn‘yăaăpasăenăeffetădeăraisonădeăpenserăqueălesăvariétésădialectalesăduăjudéo-arabe
sicilien aient été plus éloignées des variétés parlées par les Musulmans ou les
Chrétiensă qu‘ailleurs au Maghreb, même si la mobilité entre les communautés
juivesăsiciliennesăpouvaităêtreăplusăgrandeăqu‘entreăautresăcommunautés,ăouăsiălesă
contacts avec des coreligionnaires établis au Maghreb ou en Égypte, ou venus
commeă enseignants,ă ouă poură s‘établir définitivementă dansă l‘île,ă pouvaientă êtreă
plus intenses et exposer davantage à la diversité linguistique. On observe
d‘ailleursăqueălesăplusăanciensădocumentsăjudéo-arabes de Sicile nous fournissent
des exemples de traits linguistiques caractéristiques de l‘arabeă sicilienă deă façonă
générale. Pour les documents tardifs (comme les soixante documents du 15 e siècle
publiés en 1976 par A. Giuffrida et B. Rocco), ils sont évidemment
particulièrement précieux car, même si le dialecte a dû évoluer pendant ces deux
siècles et demi supplémentaires où il est assurément devenu périphérique dans le
paysageălinguistiqueădeăl‘île,ăilăneăpeutăpasăneăpasăavoirăconservéăunăgrandănombreă
de traits anciens.
Pour une tentative de première réponse, forcément sommaire, à la question
qui nous occupe ici, celle de savoir si on peut parleră deă l‘arabeă sicilienă commeă
« périphérique », on partira donc des informations cumulées que nous apporte la
documentation, au risque, assumé, de pécher par anhistorisme.
Un dialecte maghrébin
Bien que le peuplement arabophone ait sans doute été diversifié, le sicilien
était de toute évidence de type « maghrébin », au sens de ce terme en dialectologie
arabe.ăIlăn‘estăpasăinutileădeăleăredireăcarăleăfait,ăaprèsăavoirăétéăanecdotiquementă
contesté (cf. la mise au point très brève de J. Blau 1968), semble encore parfois
12
Metcalfe 2003 p. 171 étudie des formes qui montrent ces interférences et viennent conforter
l‘idéeă queă « un certain degré, peut être élevé, de mélange linguistique est allé de pair avec les
processusă d‘acculturation »,ă cesă influencesă caractérisantă l‘arabeă sicilienă pară rapportă àă d‘autresă
dialectesămaghrébins.ăOnăseăpermettraădeăsignalerăiciăqueăl‘analyseăproposéeă(p. 169) pour h`anzārī
―pig-farmer‖ăestăàărectifierăsurăunăpoint :ăonăneăpeutăàăproprementăparlerăd‘adjonctionăduăsuffixeăari(us) puisque le r se trouve déjà dans le mot arabe ;ămaisăilăestăprobableăqueăl‘analogieăavecălesă
dérivés romans a joué un rôle non négligeable dans la création de ce mot, vu sa forme.
75
être mis en doute. Il n‘estăpasăquestionădeăfaireăiciăl‘inventaireăexhaustifădesătraitsă
maghrébins du sicilien ; on indiquera rapidement ci-après quelques traits
caractéristiques, dont un certain nombre sont typiquement maghrébins (et
« préhilaliens »),13 et quelques phénomènes remarquables.
Phonétique et phonologie.
Les trois interdentales passent souvent aux occlusives correspondantes.
Les « emphatiques » semblent avoir été articulées sans (grande) emphase. Le ğīm
semble être le plus souvent affriqué. Pour /h/, /h}/, /h~/, /</ et /ġ/ voir le paragraphe
suivant.ă Commeă exempleă d‘assimilationă d‘uneă chuintanteă àă uneă sifflanteă (peutêtre caractéristique du parler juif) on notera zzm « bottes » (=
, Bresc-Goitein
1970 p. 909 l. 5 du verso). À noter un *h apparemment réalisé comme proche de ﻣ
(ich-laut ?) : wl ‘wldš ―etăàăsesăenfants‖ă(Giuffrida-Rocco 1976 p. 97).
L‘‘imāla de ā (souvent rendue par έ en grec, cf. Agius 1996 p. 410), est
bien attestée, par ex. dans les noms de métier de schème CaCCāC ; cf. aussi
mhryz (< *mihrās) « mortier » (Bresc-Goitein 1970 p. 909). Le passage *a i, en
particulier dans la première syllabe des noms de schèmes CaC(C)VVC et
maCCaC est fréquent (Agius 1996 p. 210-213), cf. les formes grecques
correspondant à *niğğār ou à *minzil (Metcalfe 2003 p. 160 ou 165,-8) ; citons
encore, pour les formes verbales, *kānit (Giuffrida-Rocco 1976 document 7,5).
Le passage a u est attesté par exemple dans les transcriptions grecque
(avec μου-) et latine (Machalub) de Mah~lūf (Metcalfe 2003 p. 147) ou avec -μουde *‗Umar (ibid. p. 152), au voisinage de labiales (comme dans 7 des 12 exemples
de Agius 1996 p. 221-222) ; cf. dans Giuffrida-Rocco 1976 mwrtw « son épouse »
(doc. 24,2 et 29 2), ‘wrb‗myya « 400 » (doc. 32,1), yw‗ny « c‘est-à-dire » (doc.
32,4 et 36, col. 2, 2) ; ibid. col 2,1 ywt}lwbw « sont exigibles ».14
Les exemples de *VV V ne sont pas rares, à commencer par les
fréquents mt‗ (*mtā‗) ou dananīr « dinars » ; autres ex. : manadīl « mouchoirs »,
bawaqīl « bocaux » (Bresc-Goitein 1970 p. 908), *‗at}t}a r « épicier » (GiuffridaRocco 1974 l. 7 du document).
Desăexemplesădeăstructuresăsyllabiquesăfamilièresăàăl‘arabeămaghrébinăsontă
attestées indirectement *Boubker (en grec, Agius 1996 p. 410) ; cf. aussi le
traitementădeăl‘articleădansă « )ﻰﻌﻘﻻ <( ﺔﻌﻔvipère » Ibn Makkī 114,11 ou les noms
en Bul (Metcalfe 2003, p. 62). Metcalfe (2003 p. 172) cite une forme grecque où
l‘accentănotéăsuggèreăunăaccentăsurălaăfinaleădansăla forme arabe h~a zár.
13
Pour certains des points abordés ici, on trouvera des référence supplémentaires dans Lentin à
paraître a.
14
Il faut corriger (p. 92) « perf. » en « imperf. » ; de plus la référence au vocalisme du maltais est
sans objet puisqu‘ilăs‘agităd‘uneăformeăpassive.
76
Pronoms. Démonstratifs : attestations, à côté des formes en hā-, de dā
(Golb l. 10 du doc.), dī 15; réfléchi :
( ﺤCusa 495,8) ; interrogatifs : ﺪﻘ
ﺶ
« quelle quantité, combien » (Golb p. 119, ligne 4).
Adverbes : « ﻰﻔ ﻄen (contre)bas » (Cusa 241,11) ; « ﻦoù ? » (Golb l.
5 du doc.).
Quantification : « ﺝdeux » (Cusa 233,11-12 ; 236,10 ; 238,9).
Numéraux : st‗šr « seize » (Giuffrida-Rocco 1974 ligne 9) ; h}tāš « onze »
(Giuffrida-Rocco 1976 n° 7 l. 3) ; *(‘)tnāš « douze » (ibid. documents 35,3 et 48,3)
Subjonctions : ( ﻒ ﻜCusa 503,15 « de la même façon que ») ; ﻰ
« lorsqu‘ilăarriveraăque », « si » (Cusa 129,-3 ; Gálvez 176,4).
Morphologie verbale : 1e pers.ăsing.ădeăl‘inaccompliăavecăn– : ( ﻞ ﻘCusa
495,3 ; Golb 120, ligne 12) ; 1e pers. plur. avec n — u : Cusa 604,12 et -2.
Verbes auxiliaires : ? ( ﻰwallā = ?) : ﺸ
ﻰ
( ﻊيCusa 42,10). Voir
aussi les tournures modales avec syntagmes prépositionnels pour « devoir »
signalées par Giuffrida-Rocco 1976 p. 61, 81, 87 et 89.
Détermination nominale : « ﺸ ﻦ ﺸun, deux mois » (‘Idrīsī 74,20).
*mta‗/mtā‗ : Cusa 217,4 ; 495,4 et 7 etc. ; bt‗ Bresc-Goitein 1970 p. 909 l. 7 du
verso ; 36 ex. de mta/ā‗ et 6 de bta/ā‗ dans Giuffrida-Rocco 1976.
Verbes usuels : žāb « apporter » (par ex. Cusa 495,5).
Lexique maghrébin : h}ūt, plur. h}ītān « poisson » ‘Idrīsī 57,2; 59,1 et 3 etc.;
Ibn Makkī 102,1; 111,10. h}ānūt « boutique » ‘Idrīsī 63,18; cf. Ibn Makkī 214,-3 et
de nombreux documents judéo-arabes ; (*s}a wm‗a >) s}a m‗a (d‘aprèsă l‘éditeur ;
plutôt s}um‗a) « minaret » Ibn Makkī 29,-2; ma/isyid « mosquée » Ibn MakkI
276,3/4. ğawāb « lettre » Giuffrida-Rocco 1976 doc. 45,1.
Une forme rā , seule ou combinée semble-t-il à un pronom personnel
suffixe, préverbale ou dans un emploi prépositionnel (« dans » ?),ă ainsiă qu‘uneă
forme tarā, annoncent peut-être les emploisăconnusăaujourd‘huiăduă« présentatif »
maghrébin rā-16 ; elles sont attestées tardivement en judéo-arabe (Giuffrida-Rocco
1976 p. 82-83 et Rocco 1992 lignes 2, 4 et 9 de l'inscription [de 1450] et p. 350,
351 et 355).
A. Metcalfe a été frappé, à juste titre, par les locutions, fréquentes dans les
Diplomi, du type *at}-t}a rīq at}-t}a rīq « le long de la route », « en suivant la route »
et en traite longuement (2003 p. 119-124 : Noun duplication), en donnant des
relevésă trèsă utiles.ă Onă seă rangeraă àă l‘avisă deă Sgroiă (1986,ă p. 28-29 L‘iterazione
nominale),ăpourăquiă l‘arabeă aăamplifiéă laădiffusionăd‘uneăconstructionăattestée,ăilă
est vrai avec des emplois légèrement différents, dans de nombreuses langues, dont
15
Ce dernier est utilisé aussi comme adjectif démonstratif (masculin et féminin) dans GiuffridaRocco 1976 doc. 8,4 et 32B,2.
16
Pour un exemple andalou de 1637, avec la valeur moderne, v. Lentin, à paraître b.
77
leăgrec.ăOnăajouteraăqueăl‘originalitéădeăl‘arabeăsicilien/maghrébinăestăpeut-être ici
laăsystématisationădeăl‘emploiădeăl‘articleădéfini.
Traits « périphériques »
On conviendra que les traits dialectaux signalés au paragraphe précédent
et,ăparmiăeux,ăceuxăqu‘onăpeutăconsidérerăcommeăspécifiquementămaghrébins,ăneă
contribuentă pasă vraimentă àă donneră deă l‘arabeă sicilienă l‘imageă d‘ună dialecteă
périphérique.ă Maisă ună certaină nombreă d‘autresă traitsă linguistiquesă qu‘onă peută
identifieră peuventă rappeleră ceuxă qu‘onă attribue,ă enă dialectologieă arabe,ă auxă
dialectes catalogués comme périphériques.
On reviendra, dans la conclusion, sur certains traits déjà signalés.
Signalons-en quelques autres :
– de nombreux indices laissent à penser que */d/ était souvent réalisé
commeă uneă sourdeă (mêmeă siă l‘onă tientă compteă queă laă notationă grecqueă pară τ,
fréquente, ne peut, pour des raisons liées à la phonétique du grec sicilien, être
invoquée sans plus de précaution). Agius 1996 cite un ex. p. 177 : tastīğa en
Siculo-lah}n Arabic,ă c‘est-à-dire en arabe dialectal sicilien, deux ex. p. 269 en
Siculo-Arabic (laă variétéă pidginiséeă dontă ilă supposeă l‘existence) : Misita et
màrcatu (et leurs variantes), un exemple p. 413 en Sicilian Middle Arabic (mais
d‘aprèsăuneătranscriptionăgrecque)ăoùăleă d de *al-‘afrad est rendu par un τ. Golb
1973 (II l. 22) a tananīr « dinars », Giuffrida-Rocco 1976 (n° 7 l. 3) *h}tāš
« onze » et (n° 36 l. 3) s}yt}nh < *sayyidnā17 ; Bresc-Goitein 1970 a tswt = tusūt <
*dusūt. Voir surtout Metcalfe 2003 p. 153-156, qui suggère lui aussi le parallèle
avecă l‘arabeă deă Chypreă (etă avecă leă maltais),ă etă donneă desă relevésă précisă desă
transcriptions grecques du dāl.
– Le traitement historique des phonèmes /h/, /h}/, /h`/, /‗/ et /ġ/ en Sicile
présenteădesăsimilaritésăfrappantesăavecăceluiăqu‘ilsăontăsubiăàăChypreăetăàăMalte :
Chypre
*(h)
*h}
*h~
*‗
*ġ
17
Malte
h~
‗
*(h)
*h}
*h~
*‗
*ġ
Sicile
h}
*h
*h} ?
*h~
*‗
*‗
*ġ
h~
ġ
Maisăilăn‘yăaăpasăd‘autreăexempleăauăn°ă9ăl.ă1ăcommeăilăestăindiquéăparăerreurăp. 92.
78
– Un phénomène de nasalisation (Metcalfe 2003 p. 171-172), qui affecte
enă particulieră desă groupesă deă deuxă consonnesă identiquesă etă qu‘onă peută
vraisemblablementă attribueră auă contactă avecă leă grecă n‘estă pasă sansă rappeleră desă
phénomènesăsemblablesăobservésădansăl‘arabeădeăChypre.ăOnăciteraăl‘exemple de
*al-mu‘addib, pour ajouteră lesă transcriptionsă latinesă qu‘on trouve dans Gálvez
1995 : 178,-1 Almanahbed ; 180,6 Almandam ; 180,-4ă Almnaddeb.ă L‘exempleă
(ibid. 181,3) *‗Abd al-mawlā Abdulmanla pourrait faire penser, dans une
traduction latine du texte arabe de 1152 copiée en 1641, à une confusion
graphique entre w (ou v) et n ;ă maisă l‘exempleă (180,-1) de *al-yawšī Algianasci
porteă àă supposeră qu‘ilă s‘agită peut-êtreă d‘autreă chose.ă Ilă faudraă d‘autresă exemplesă
cependantăpourăs‘assurer,ăcommeăleăfaităremarquerăA.ăMetcalfe,ăqu‘ilăneăs‘agităpasă
uniquementă deă laă prononciationă grecqueă sicilienneă deă l‘arabe,ă maisă bienă d‘ună
phénomèneă propreă àă l‘arabeă sicilien.ă Poură desă phénomènesă d‘interférence à un
autre niveau, voir la note 12.
– Sură plusă d‘ună point,ă desă phénomènesă caractéristiquesă duă systèmeă
vocalique du sicilien,ă quiă n‘ontă puă qu‘êtreă àă peineă évoquésă ici,ă rappellentă ceuxă
qu‘onăpeutămettreăenăévidenceădansăleăparlerăarabeădeăKormakitiăàăChypre (v. par
exemple Borg 2004 p. 24-25).
Conclusions provisoires
Lesătravauxăd‘historiensăcommeăA.ăNefăetăsurtoutăA.ăMetcalfeăsouventăcitéă
dansă cetteă brèveă étude,ă ontă commencéă àă jeteră quelqueă lumièreă sură ceă qu‘ontă puă
êtreă leă statută deă l‘arabeă etă desă autres langues de la Sicile médiévale à certaines
périodes de son histoire. Pour la période normande par exemple, il semble que les
souverains, qui professaient une certaine connaissance des langues parlées par
leurs sujets, aient jugé utile de continuer à le faireăpourăl‘arabeăjusqu‘auămilieuăduă
13e siècle.ă L‘arabe,ă quiă étaită alorsă uneă langueă deă minoritéă etă deă basă prestige,ă
semble avoir paradoxalement conservé alors un certain statut prestigieux dans les
plus hautes couches de la société. Mais une fois la transition au latin établie par le
biaisă desă traductions,ă c‘estă cetteă langueă quiă devintă laă langue de prestige. Les
dialectes « francs » avaient eu une brève période de prestige à la cour vers le
milieu des années 1160 ;ăversă1180,ăleă latinăetăl‘arabeăavaientăsemble-t-il émergé
commeă languesă favoritesă deă laă royautéă etă deă l‘administration,ă leă grec,ă aprèsă uneă
brève période de renaissance dans la 1e moitié du 12e siècleăn‘ayantăpuămainteniră
sa position de prestige (Metcalfe 2003 p. 106, 110, ou encore 182 pour une
synthèse). De son côté, A. Nef (2004) avait présenté18 des hypothèses suggestives
18
Dans une partie de sa communication malheureusement non conservée dans la version publiée,
elle suggérait une prééminence de genres non religieux (géographie, poésie, conseils aux
79
surăceăqu‘onăpourraităappelerălaădivisionă linguistiqueădesăgenresă littérairesăauă12 e
siècle. Ce sont là de premiers éléments de réflexion, mais qui ne nous informent
que peu sur leărôleăetăleăstatutădeăl‘arabeăparléăenăSicile,ămêmeăsiăcesăquestionsăneă
peuventă évidemmentă êtreă disjointes.ă Dansă l‘étată actuelă deă laă recherche,ă onă neă seă
risquera donc pas à tirer des conclusions sur ce point ; on se bornera à faire
remarquer que le statutădeăl‘arabe,ăetădeăsesădifférentesăvariétés,ăaăétéăvariableăetă
s‘estămodifiéăprofondémentăauăcoursădeăpériodesădeătempsălimitées,ăceăquiăaăsansă
aucună douteă dûă joueră ună rôleă dansă l‘évolutionă deă saă variétéă dialectale,ă quiă avaită
peut-être entamé la marche quiă l‘auraită amenéeă àă devenir,ă duă pointă deă vueă
linguistique, un arabe périphérique (comme, toutes choses égales par ailleurs, sa
voisine de Malte), si cette marche avait pu se poursuivre.
Cetteădernièreăconsidérationăn‘estăpasăentièrementăsansăfondement ; elle se
baseă sură l‘impressionă qu‘onă aă essayéă deă faireă partageră queă l‘arabeă dialectală deă
Sicile présente, sans avoir eu peut-êtreăleătempsădeăcesserăd‘êtreăcomparableăauxă
autresă variétésă maghrébinesă deă l‘époque,ă desă traitsă déjàă enă partieă développés,ă
qu‘onă reconnaît par ailleurs comme caractéristiques de dialectes périphériques
commeă leăchyprioteăouăleă maltais.ăIlăpourraităs‘agir,ăpourăleădireăautrement,ăd‘ună
dialecteă quiă auraită puă deveniră périphériqueă siă l‘histoireă leă luiă avaită permisă (siă lesă
communautés arabophones avaient continué à vivre plus longtemps après la fin de
laă dominationă musulmane).ă Onă pourraită objecteră àă ceă schémaă hasardéă l‘exempleă
du judéo-arabeăsicilien,ăquiădevaităsurvivreăsurăl‘îleădeuxăsièclesăetădemiăencore,ăetă
ne présente pas vraiment, autantăqu‘onăpuisseă l‘entrevoir,ălesăcaractéristiquesădesă
dialectes périphériques. Sans doute faut-il attribuer ce fait aux contacts maintenus
pară sesă locuteursă avecă leursă coreligionnairesă duă mondeă arabeă d‘uneă part,ă avecă
l‘Afriqueă duă Nordă d‘autreă partă (où,ă comme on le sait, certains servaient comme
interprètes). On risquera pour finir une idée plus aventureuse encore : ne pourraiton considérer que, dans le cas des dialectes périphériquesădeăl‘arabe,ăouăplutôtăduă
« devenir périphérique » des dialectes arabes, ce qui est fondamental, outre le
statut des communautés qui les parlent, et celui, complexe et étroitement lié à
l‘identité,ă deă laă langueă qu‘ellesă parlent,ă autrementă dită laă permanenceă desă liensă
linguistiques, culturels, et pour finir sociaux, que conservent entre eux ceă qu‘onă
pourrait appeler les « colocuteurs », que ce qui est fondamental donc est aussi la
disparitionăd‘autresăliens,ătoujoursăprécairesăetăsansăcesseăàăretisser,ăavecăd‘autresă
communautés de « colocuteurs » ? On a souvent fait remarquer queă l‘histoireă duă
maltais,ă etă d‘autresă dialectesă périphériques,ă enă Turquie,ă enă Asieă centrale,ă seă
caractérisaităparăcetteădisparitionădesăliensă(avecălaăcultureăarabe,ăl‘islam,ălesădeuxă
souverainsă…)ădansălaăproductionăécriteăenălangueăarabeădeăcetteăpériode.ăDeămêmeăMetcalfeă2003ă
p. 106 à propos du Livre de Roger seădemandeăsiăceătypeăd‘ouvrageăn‘avaităpasămoinsădeăprestigeă
que ceux de sciences ou de philosophie.
80
parfois)ă etă desă contactsă avecă d‘autresă locuteursă deă l‘arabe.ă Ceă qu‘onă voudrait
suggéreră ici,ă c‘estă queă s‘ilă s‘agită bienă deă cela,ă onă peută considéreră alorsă queă toută
dialecte arabe est toujours potentiellement en voie de « périphérisation ». Il est
frappant en effet que certains traits toujours présents dans les dialectes
périphériques (comme le passage des dentales aux occlusives correspondantes, ou
la perte, relative ou totale,ă deă l‘articulationă diteă « emphatique ») se retrouvent
aussi dans les plus anciens dialectes, citadins, et les plus « évolués » –
linguistiquement – du mondeă arabophone.ă Cesă vieuxă dialectes,ă lorsqu‘ilsă restentă
au coeur du monde arabophone, et que les communautés qui les parlent gardent un
quelconque prestige, sont tout sauf périphériques du point de vue
sociolinguistique,ă bienă qu‘ilsă présententă cesă traits.ă Il suffirait par contre que ce
statutăchangeăpourăqu‘ilsăs‘acheminent,ăsansădouteăd‘autantăplusărapidementăqu‘ilsă
ont déjà franchi techniquement une partie du chemin, vers la « périphéralité ». La
frontièreă linguistiqueă estă enă réalitéă ténue.ă D‘ailleurs,ă cesă vieux dialectes citadins
ne meurent-ilsăpasăaujourd‘hui ?ăC‘estăsansădouteăqu‘avecăceuxăquiălesăparlentăilsă
sont devenus périphériques au sein de leur propre société. Cela signifie aussi que
le maintien de la cohésion linguistique entre les groupes composant une
communauté de « colocuteurs » se fait par des ajustements incessants, qui ne vont
pasă nécessairementă toujoursă dansă leă sensă deă l‘innovation ; dans le monde
arabophone « non périphérique » en tout cas, il faut certainement voir dans cette
mise au pas de certaines innovations (pouvant mener à leur disparition et donc, de
fait, à un retour en arrière historique), une des raisons principales de ce que
l‘historienă desă dialectesă arabesă estă bienă souventă amenéă àă constater : la
conservation remarquable sur la longueă duréeă deă l‘essentielă deă leursă structuresă
linguistiques fondamentales.
Références
Agius, Dionisius A. 1996. Siculo Arabic, Londres/New York: Kegan Paul International.
Blau, Joshua. 1968. « To which dialect group did Sicilian Arabic belong? », in: Journal
of the American Oriental Society, LXXXVIII/3, 522-523.
Borg, Alexander. 1994. « Some evolutionary parallels and divergencies in Cypriot Arabic
and Maltese », in: Mediterranean Language Review 8, 41-67.
Borg, Alexander. 2004. Comparative Glossary of Cypriot Maronite Arabic (ArabicEnglish) – With an Introductory Essay. Leiden/Boston: Brill.
Bresc, Henri. 2001. Arabes de langue, juifs de religion : l‘évolution du judaïsme sicilien
dans l‘environnement latin, XIIe-XVe siècles, Paris: Bouchène.
Bresc, Henri & Shelomo Dov Goitein. 1970. « Un inventaire dotal de juifs siciliens
(1479) », in: Mélanges d‘archéologie et d‘histoire, 82, 903-917.
Bucaria, Nicolò. 1996. Sicilia judaica, Palermo: Flaccovio.
81
Caracausi, Girolamo, 1987-88. « I documenti medievali siciliani in lingua araba », in:
Journal of Maltese Studies, 17-18 (= Giuseppe Brincat, ed.: Incontri siculomaltesi – Atti del II Convegno su Malta – Sicilia. Contiguità e Continuità
Linguistica e Culturale, Malta 4-6 aprile 1986), 13-26.
Cohen, David. 1994. « Préface », in: Actes des premières journées internationales de
dialectologie arabe de Paris, INALCO (Paris) :ăLangues‘O,ă9-19.
Cusa, Salvatore. 1868 [-1882]. I Diplomi greci ed arabi di Sicilia..., vol. I, Palerme.
De Simone, Adalgisa. 1979. Spoglio antroponomico delle giaride (JarA>id) arabogreche dei Diplomi editi da Salvatore Cusa, Roma.
De Simone, Adalgisa. 1983. « Perăunălessicoădell‘araboădiăSicilia », in: Joseph M. Brincat
(ed.), Languages of the mediterranean. Substrata – The Islands – Malta.
Proceedings of the Conference held in Malta 26-29 sept. 1991. The Institute of
Linguistics, University of Malta 1, 105-118.
De Simone, Adalgisa. 1984. « Alcuneă osservazioniă sulă lessicoă dell‘araboă diă Siciliaă traă
l‘XIă eă ilă XIIIă secolo », in: Renato Traini (ed.): Studi in onore di Francesco
Gabrieli nel suo ottantesimo compleanno, Rome: Università di Roma « La
Sapienza », Dipartimento di Studi Orientali, vol. I, 261-268.
De Simone, Adalgisa. 1988. « I diplomi arabi di Sicilia », in: Giornata di studio Testimonianze degli Arabi in Italia (Roma, 10 dicembre 1987), Rome:
Accademia Nazionale dei Lincei, 57-75.
Gálvez, M. Eugenia. 1995. « Noticia sobre los documentos árabes de Sicilia del Archivo
ducal de Medinaceli », in: Del nuovo sulla Sicilia musulmana, Rome: Accademia
Nazionale dei Lincei, Fondazione L. Caetani, 167-182.
Gil, Moshe. 1995. « Sicily 827-1072, in Light of the Geniza Documents and parallel
sources », in: Italia Judaica, vol. V, 96-171.
Giuffrida, Antonino & Benedetto Rocco. 1974. « Una bilingue arabo-sicula », in: Annali
dell‘Istituto Orientale di Napoli, n.s. 24 (34), 109-122.
Giuffrida, Antonino & Benedetto Rocco. 1976. « Documenti giudeo-arabi nel secolo XV
a Palermo », in: Studi magrebini 8, 53-110.
Golb, Norman, 1973. « A Judaeo-Arabic Court Document of Syracuse, A. D. 1020 », in:
Journal of Near Eastern Studies 32 (= Seele Memorial Volume), 105-123.
Ibn Makkī. 1981 [1e éd. 1966]. Tatqīf al-lisān wa talqīh} al-ğanān, ed. ‗A. Mat}ar, Le
Caire: Dār al-Ma‗ārif.
‘Idrīsī (Al-). 1970, 1971, 1972, 1974. Opus geographicum, éds. A. Bombaci, U.
Rizzitano, R. Rubinacci, L. Veccia-Vaglieri, fasc. 1 [1-100], 2 [103-214], 3 [217346], 4 [347-521], Neapoli-Romae: Istituto Italiano per il Medio ed Estremo
Oriente.
Italia Judaica, vol. V = Gli ebrei in Sicilia sino all‘espulsione del 1492 – Atti del V
convegno internazionale – Palermo, 15-19 giugno 1992, Rome: Pubblicazioni
Degli Archivi Di Stato, (Ministero per Beni Culturali E Ambientali, Ufficio
Centrale Per I Beni Archivistici, Saggi 32), 1995.
Johns, Jeremy. 2002. Arabic Administration in Norman Sicily – The Royal DIwAn,
Cambridge: Cambridge University Press.
82
Johns, Jeremy & Alex Metcalfe. 1999. « The Mystery at Chùrchuro : conspiracy or
incompetence in twelfth-century Sicily ? », in: Bulletin of the School of Oriental
and African Studies 62, 226-259.
Lentin, Jérôme. [à paraître en 2007 a]. « Sur quelques spécificités du Moyen Arabe de
Sicile », in: Atti del XII Incontro italiano di linguistica camito-semitica
(afroasiatica), Ragusa, 6-9 giugno 2005, 29-37.
Lentin, Jérôme. [à paraître en 2007 b]. « Unité et diversité du Moyen Arabe au Machreq
et au Maghreb :ă quelquesă donnéesă d‘aprèsă desă textesă d'époqueă tardiveă (16ème19ème siècles) », in : Jacques Grand'Henry et Jérôme Lentin (eds): Actes du
Premier Colloque international Moyen arabe et variétés moyennes de l‘arabe à
travers l‘histoire : état des connaissances, problèmes de définition et
perspectives de recherche, Louvain-la-Neuve, 18-22 mai 2004, Louvain: Peeters
(Publications de l'Institut Orientaliste de Louvain).
Metcalfe, Alex. 2001. « De Saracenico in Latinum transferri: causes and effects of
translation in the fiscal administration of Norman Sicily », in: Al-MasAq 13, 4386.
Metcalfe, Alex, 2003. Muslims and Christians in Norman Sicily: Arabic-Speakers and the
End of Islam, Londres/New York: Routledge Curzon.
Metcalfe, Alex. [à paraître en 2007]. « Sicilian Arabic », in: K. Versteegh, M. Eid, A.
Elgibali, M. Woidich & A. Zaborski (eds): The Encyclopaedia of Arabic
Language and Linguistics, vol. 4, Leiden: Brill (consulté grâce à la courtoisie de
son auteur).
Nef, Annliese. 2000. « La langue écrite des juifs de Sicile au XVe siècle », in: Henri
Brescă etă Christianeă Veauvy,ă avecă laă participationă d‘Élianeă Dupuyă (eds):
Mutations d‘identités en Méditerranée – Moyen Âge et époque contemporaine,
Paris: Bouchène, 85-97.
Nef, Annliese. 2004. « Peut-onă parleră deă ―politiqueă linguistique‖ă dansă laă Sicileă duă XIIe
siècle ? Quelques réflexions préliminaires », in: Jocelyne Dakhlia (ed.), Trames
de langues – Usages et métissages linguistiques dans l‘histoire du Maghreb,
Paris: Maisonneuve & Larose, 41-57.
Rocco, Benedetto. 1992. « Iscrizione giudeo-araba a Messina », in: Vetera Christianorum
29, 345-357.
Rocco, Benedetto. 1995. « Le tre lingue usate dagli ebrei in Sicilia dal sec. XII al sec.
XV», in: Italia Judaica, vol. V, 355-369.
Rocco, Benedetto. 1998. « Il giudeo-arabo e il siciliano nei secoli XII-XV : influssi
reciproci », in: G. Ruffino (ed.): Atti del XXI congresso internazionale di
linguistica e filologia romanza, Palermo 18-24/9/1995, Tübingen: Max Niemeyer
Verlag, 539-545.
Scandaliato, Angela, 2006. Judaica minora sicula. Indagini sugli ebrei di Sicilia nel
Medioevo e quattro studi in collaborazione con Maria Gerardi. Firenze: Giuntina
(Associazione italiana per lo studio del giudaismo, Testi e studi 18).
83
Sgroi, Salvatore Claudio. 1986. Interferenze fonologiche, morfo-sintattiche e lessicali fra
l‘arabo e il siciliano, Palerme: Centro di studi filologici siciliani (Biblioteca del
Centro di studi filologici e linguistici siciliani, N. S. 7).
Wansbrough, John. 1967. « A Judaeo-Arabic Document from Sicily », in: Bulletin of the
School of Oriental and African Studies 30, 305-313.
Wansbrough, John. 1984. « Diplomatica Siciliana », in: Bulletin of the School of Oriental
and African Studies 47, 10-21.
84
From periphery towards centre stage:
research on ‘mixed styles’– results and challenges
Gunvor Mejdell
University of Oslo
Norway
Fourty years ago, little attention had been given to the linguistic outcome
of the interaction between fus}h}ā and the vernacular in the speech of educated
Arabs. That it is an ancient issue to Arabic, is obvious from the long native
tradition of lah}n al-‗āmma literature, where scholars were concerned about the
‗mistakes‘ă committed by the educated and uneducated when using the language
for formal purposes. Vernacular, or non-standard, features were similarly
observed by Western scholars working on medieval manuscripts, and the study of
theseă‗MiddleăArabic‘ătextsăcameătoăbe considered a worthy and separate field of
research, as early as the end of the 19th century.
Whată Iă callă ‗mixedă styles‘ă ofă contemporaryă spokenă Arabic,ă andă whichă
native speakers and researchers refer to as luġa wust}ā, or luġat al-mutaqqafīn, was
first mentioned rather en passant by Ferguson in his seminal 1959 study on
‗Diglossia‘:ă
―Theăcommunicativeătensionsăwhichăariseăinătheădiglossiaăsituationămayăbeă
resolved by the use of relatively uncodified, intermediate forms of the language
[…]ăaăkindăofăspokenă Arabică muchăusedă inăcertainăsemiformalăorăcross-dialectal
situations has a highly classical vocabulary with few or no inflectional endings,
with certain features of classical syntax, but with a fundamentally colloquial base
ină morphologyă andă syntax,ă andă aă generousă admixtureă ofă colloquială vocabulary‖ă
(1959: 332).
The following year, in a series edited by Ferguson, Haim Blanc published
―Stylistică Variationsă ină Spokenă Arabic:ă Aă sampleă ofă Interdialectală Educated
Conversation‖,ă introducingă theă notionă ofă ‗classicization‘ă andă stylistică variation,ă
and providing a linguistic analysis of a short piece of spoken data demonstrating
fluctuating and intermediate forms (Blanc 1964 [1960]). Stimulated by these
studies (as wellă asă byă theă growthă ofă theă ‗linguistică variability‘ă paradigmă ină
85
general), and greatly assisted by the availability of recording facilities, the 1970´s
saw a number of studies dedicated to the issue of intermediate levels,
sociostylistic variation, code-switchingă andă mixing:ă Palvaă 1969,ă Badawīă 1973,ă
Diem 1974, Schmidt 1974, Meiseles 1977, Mitchell 1975, Salib 1979, and more.
Since then the field has multiplied, reflecting an awareness that linguistic practice
in the modern Arabic speaking world is increasingly characterized by this
constant interaction between, and mutual influence of, the vernacular mother
tongue and the standard form. As the standard is aquired through education (or
passively through the media), I assume that this kind of mixed style is restricted as
aăphenomenonătoăcommunitiesăwhichăhaveăstandardăArabicăasăaă‗Dach‘-language.
I do not reject the possibility of mixed forms being produced outside the Arab
statesăbyă‗peripheralădialect‘ăspeakingăpeopleămaybeăliterateăinăArabicăinătermsăofă
Qur‘anică language,ă bută Iă doubtă ită hasă theă kindă ofă fluencyă weă encounteră withă theă
‘mixed‘ăspeechăofăurbanăeducatedăinăArabicăculturalăcenters.ă
Let me state from the outset, that one should not consider this kind of
speechă productionă simplyă asă ‗missingă theă target‘,ă assumingă thată theă targetă isă
standard Arabic. Non-intentional interference from the vernacular mother tongue
on attempts to speak fus}h}ā, is only one of many psycholinguistic processes which
may produce some kind of mixed style in Arabic. There is much evidence that
speakersăperceiveămixedăstyleăasă‗the target‘ăinăcertainăcommunicativeăsituations,ă
and Arabic speakers, scholars and non-scholars, identify the link between luġa
wust}ā and semi-formal situations.
This paper will address what research during the last decades has produced
ofă claimsă andă evidenceă regardingă theă linguistică propertiesă ofă theseă ‗intermediateă
forms‘,ă‗varieties‘,ăoră‗mixedăstyles‘ă– in terms of principles and rules, constraints,
patterns, and tendencies in the way elements and features from the two basic
codes combine. The inherent variability of this kind of data poses great challenges
to the analyst. What adds to the complexity, is the wide range of styles, modes of
speaking, settings, and situations which fall under these notions – making
comparisons and generalizations difficult. For instance, Boussofara-Omar in a
recent contribution in ZAL (Boussofara-Omar 2006), where she applies MyersScotton‘să code-switching model to a political (mixed-style) speech by former
president Bourguiba – finds no trace of standard Arabic (SA) negative markers,
while in my data on academic monologues (Mejdell 2006), individual speakers
vary from very low to very high usage level of SA variants of this feature.
‘Unstable’ or ‘predictable’?
In the literature we find, on the one hand, characterizations in line with
Ferguson‘să ―unstableă intermediateă formsăofătheă language‖,ăsuchăasăDiemă(1974),
whoseă ―Mischspracheă [ist]ă ohneă Festgelegteă Formenă undă Regeln‖,ă andă Blauă
86
(1999)ă writingă ofă ―infinitelyă variedă mixturesă ofă CAă andă NAă elements‖.ă Onă theă
otherăhand,ăweăfindăconfidentăclaimsătoătheăeffectăthată‗mixedăstyles‘ăorăwhateveră
they are labelled – includingă ‗diglossică code-switching‘ă – are not the product of
random mixing or switching or selection, but regulated by rules and constraints, of
both structural and pragmatic nature, to the extent that switching is, at least to a
considerable extent, predictible (e.g. Bassiouny 2006; Boussofara-Omar 2006).
Sometimes such claims seem closer to articles of faith than to claims supported by
evidence.ă―Partialăsystematicity‖ă(Schmidtă1974:223)ăseemsătoămeăaămoreărealistică
assessment, while I admit that more ambitious claims also may be more
motivating in the search for systematicity and regularity, for establishing more
order in the apparent chaos.
I shall in the following present some of the principles and/or structural
constraints on the combination of SA and vernacular features and items, which
have been supported by evidence across data sets and in various studies. In fact,
there has only in recent years been published real data on this kind of speech –
apart from Blanc 1960 and the pioneering study by Werner Diem: Hochsprache
und Dialekt im Arabischen (1974), which included some 20 pages of transcribed
and commented extracts of radio monologues and dialogues, and the extracts in
Schulză 1981.ă Sinceă thenă manyă contributionsă haveă appearedă onă ‗diglossia‘,ă
‗levels‘,ă andă ‗styles‘,ă ‗variation‘,ă ‗code-switching‘,ă andă ‗luġa wust}ā‘ă – however,
much is repetitive and much is assumed, and most studies only include a few
examples of corroborating evidence to a certain claim made, and not a fuller
picture, which more extensive empirical data can provide.
Asymmetry of codes
The most significant general principle is the effect of the asymmetry of
codes on the linguistic contact between the vernacular and the standard. This
observation is of long standing in language contact literature (going back to such
classics as Uriel Weinreich and Nils Hasselmo). It is captured by Diem in the
followingă way:ă ―Dieă Interferenză zwischenă zweiă Sprachenă gehtă überwiegendă vonă
derjenigen der beiden Sprachen aus, die das Individuum am besten beherrscht‖ă
(Diem 1974: 24). This principle, with apparent universal applicability for mixing
between codes of unequal psycholinguistic status, has been variously formulated
and applied – most often to phonological processes, where – in our case –
standard Arabic items may be adjusted to vernacular phonology, but not the other
way around.
87
The asymmetry in status in reflected in the restrictions on the combination
of morphemes in a single word arrived at by R.W. Schmidt 1974 1. He found, that
acceptable combinations of SA and dialect (Egyptian Arabic = EA) always
involve an SA stem + EA grammatical suffix, 2 not EA stem + SA suffix, as in (*
notesă‗notăfound‘/‘notălikelyătoăoccur‘):
stem + suffix
SA + SA
e.g.: h}a:rabată‖sheăfought‖ă(Schmidt:175)
EA + EA
h}a:ribit
SA + EA
h}a:rabit
*EA + SA
*h}a:ribat
The same principle has been formulated as a constraint operating on
‗lexicală item‘ă +ă ‗grammaticală morpheme‘ă (Petersenă 1988) 3 ,ă oră onă ‗contentă
morpheme‘ă+ă‗systemămorpheme‘ă(Myers-Scotton).4
It is found operating on the feature pronoun suffixation:
SA lexical item + SA PRON
SA lexical item + EA PRON 5
EA lexical item + EA PRON
*EA lexical item + SA PRON
It has been demonstrated across word boundaries, as in:
In an excellentă studyă onă ‗rule-governed‘ă hierarchiesă betweenă variousă phonologicală andă
morphological variants in various styles of Arabic in Egypt based on acceptability judgments on
‗theoreticalăhybridăforms‘ă+ăreadingăaloudădataă+ăinterviewsăfromătwoădifferentăsocial groups.
2
Someă SAă stems/lexicală itemsă ‘resist‘ă combiningă withă EAă suffixes,ă stemsă withă interdentals,ă foră
instance,ămayăseemătooă‗high‘ăflown,ăandătheăvariantăwithăsibilantăz}a nnat/z}a nnit more appropriate.
―Theălexicalăhypothesis‖ăisăaboutăconstraintsăworking on specific lexical items, as sociostylistic
connotation of a lexical item with specific cultural layers, never will occur other than with either H
or L properties, be they phonological or morphological. Cf. discussion in Mejdell 2006:55-57;
Badawi 1973:156-57; Holes 1987:100-105; Owens and Bani-Yasin 1987.
3
―Theă dominant-language hypothesis states that in word-internal code-switching, grammatical
morphemes of the DOMINANT language may occur with lexical morphemes of either the
dominant or the non-dominant language. However, grammatical morphemes of the NONDOMINANTălanguageămayăcooccurăonlyăwithălexicalămorphemesăofătheănondominantălanguage.‖ă
(Petersenă1988:486,ăbasedăonăbilingualăchildren‘săgrammar.)
4
Myers-Scotton‘să code-switching model has been, and still is, very influential. As it has been
constantly elaborated, I do not list a specific reference to her work, which is presented e.g. in both
Bassiouny 2006 and Boussofara-Omar 2006.
5
In cases where EA=SA pronoun suffixes (-ni, -na, -ha, hum, -ka. ki, -kum) speakers and
observersă alikeă mayă consideră suffixationă onă SAă stemsă asă justă aă caseă ofă orală SAă ‘pausal‘ă forms,ă
without considering the effects suffixation has by applying EA phonotactics and syllable structure,
e.g. that in EA taslīm+hum > taslimhum;
d}uyūf + ha > d}uyufha. This is discussed more thoroughly in Mejdell 2006:345-7)
1
88
Demonstratives: 6
Relatives:
Complementizers:
ha:za + SA lexical item
*ha:za + EA lexical item
allazi + SA lexical item
*allazi + EA lexical item
‘an/‘anna + SA lexical item
EA lexical item + da
SA lexical item + da
illi + EA lexical item
illi + SA lexical item
‘inn(u) + EA lexical
*‘an/‘anna + EA
‘inn(u)+ SA lexical
item
item
With the feature Negatives + verbs, restrictions apparently go both ways:
SA neg. + SA verb
*SA neg. + EA verb
EA neg. + EA verb
*EA neg. + SA verb
This was convincingly argued by Eid (1988) to be caused by lack of
system congruence, in that SA marks tense in the negative marker (lā, lam, lan),
while EA marks tense on the verb IPF, PF, h}a -), not on the negative.
All these constraints are corroborated by the data in Mazraani 1997,
Mejdell 2006, and Bassiouny 2006 7 – as well as the media excerpts in Diem 1974,
Schmidt 1976, and Schulz 1981.8 The scant empirical evidence providedă byă Abūă
Tāyehă (1995),ă al-Šāyibă (1976),ă andă Aboussofraă (2006)ă alsoă supportsă theă
constraints above, or at least do not provide counter-examples.
Predication, probabilities, frequencies, and usage levels
These constraints, however, have predictive force only in a negative
manner: they claim that certain combinations will not occur – so in a sense we can
say that mixed styles are not without constraints and are not wholly randomly
constituted. However, there are noă regulationsă oră ‘rules‘ă whichă predictă whichă ofă
alternative constructions will be chosen/preferred, for instance, following the SA
verb istat}a ‗a or the shared noun fursa} :
Badawi 1973:166 operates only with the EA variants of demonstratives at the level of ‗āmmiyyat
al-mutaqqafīn, which is surprising considering the high usage level of SA variants in my data, and
someă otheră dataă sets,ă ată least.ă ‗Misobservation‘ă byă Badawi,ă oră doesă ită reflectă spreadingă ofă SAă
variants in subsequent years?
7
In Bassiouny also explicitly stated: EA [ECA] negative marker + MSA-like verbs; no occurrence
of MSA negative marker + EA verbs (2006:141); ECA demonstrative + MSA-like nouns; no
occurrence of MSA demonstrative + ECA nouns (ibid.).
8
No substitution of negative constructions either way, i.e. in either SA text environment or in
spoken interview (EA) environment (1974:122)
6
89
‘istat}a‗na ‘an nuh}a qqiq ha:za n-nas}r / wa ‘istat}a‗na ‘in nih}na [‘inn-ih}na]
nihzim id-duwal l-kubra (both in a 1957 speech by Nasser, in Mazraani 1997:
227)
ha:zihi l-furs}a / il-furs}a di (produced alternatively by one speaker, Mejdell
2006)
What we may establish, however, are probabilities of occurrence, based on
frequencies of occurrence of variants in specific contexts. Through the process of
genre formation, or conventionalization, variants may be seen as competing
forms, progressive and regressive variants, pointing to the emergence of generic
norms.
In my academic mixed styles data, the individual speakers (AUC 1-4, NA
1-3) varied greatly in their relative usage levels of SA and EA variants of the
features investigated:
The distribution of SA : EA variants across speakers and features: 9
COMPLEMENTIZERS (‘an/‘anna, asyndet., ‘inn(u)):
AUC1
SA:EA 15:15
AUC2
28:29
AUC3
14:34
AUC4
6:27
NA1
27:15
NA2
89:17
DEMONSTRATIVES
AUC1
SA:EA 17:12
AUC2
44:16
AUC3
26:32
AUC4
17:18
NA1
42:3
NA2
132:6
NEGATION
AUC1
SA:EA 6:5
AUC2
33:10
AUC3
17:16
AUC4
2:20
NA1
20:0
NA2
20:2
AUC2
16:15
AUC3
3:28
AUC4
2:24
NA1
17:4
NA2
72:14
AUC3
6:39
AUC4
2:46
NA1
35:27
NA2
61:6
RELATIVES
AUC1
SA:EA 5:12
PRONOUN SUFFIXATION
AUC1
SA:EA 7:26
AUC2
39:33
9
Cf. Mejdell 2006: 375-76, where the tables are more detailed. AUC refers to a seminar-setting on
campus where the topic was higher education; NA refers to a literary seminar-setting (nadwa
‘adabiyya). Speakers are academics, the talks are monologues.
90
However, we could observe a clear tendency in the hierarchy between
features across speakers – with DEM and NEG representing the highest usage
level of SA variants (if only attributive DEM is counted, DEM is the by far the
highest ranking feature of all) and the feature PRON SUFF the lowest:
highest value SA
>
>
>
highest value EA
AUC 1: DEM > NEG > REL > COMP > PRON
AUC 2: NEG > DEM > REL > COMP > PRON
AUC 3: NEG > DEM/REL > COMP > PRON
AUC 4: DEM > REL/COMP > NEG > PRON
NA 1 : NEG/DEM > REL > COMP > PRON
NA 2 : DEM /NEG > COMP > REL > PRON
The high usage level of EA pronoun suffixation, means that it frequently
occurs also in SA-dominant environment, e.g.:
qad}iyyit /kayfa tusa:him ha:zihi l-‘afka:r allati nukawwin-ha ‗an ‘al‗a:lam (GM:6.179)
il-‘ibda:‗ fi gawhar-u ‘anna-ka tastat}i:‗... (GM:2.208)
il-fikra ‘illi ‗imba‗atat min mas}r / tuna:di / bi ‘anna-na ‘ah}r a:r fi bilad-na
/ nuqarrar siyasit-na fi bilad-na (Mazraani 1997:225)
la‗alla-na nas}il ‘ila s}i:ġa tawfi:qiyya / tabda‗ ‗ala ‘asas-ha l-mufa:wad}a :t
(political speech by Mubarak, Bassiouny:313)
- yuma:ris h}uku:k-u ‘as-siya:siyya [h}uqūqahu] (―‗āmmiyyată almutaqqafīn‖ăAbūăTāyeh:ă83)
- h}ud}u:r d}uyūfna (classsifiedăasă―fus}hā} ăal-‗as}r al mant}ūqa al-mutadāwila‖,ă
ibid.: 81)
- al-mu‗t}[ i]yy nafsu wa usr[a/i?]tu wa mujtama‗u al-haqa [alh}a qqa](―fus}h}āăal-turāt‖,ăfromătheăprogrammeă―ma‗aărijālăal-dīn‖,ăibid.)
91
Unequal status of variables as sociostylistic markers
Another feature in which the EA variant occurs even in otherwise SAoriented context, is the fem.stat.constr. ending -it (vs. SA -at 10 ) (following the
same asymmetric constraint:11 SA lexical item + it,12 but *EA lexical item + -at):
ha:zihi l-fikra / bir-raġm min ‘annaha fikra ‗ami:qatu l-guzu:r fi saqafìtna
(GM:2.91)
ha:zihi l-‘afka:r at-ta:bita ‘allati tatagassad fi fikrit al-‘umm (GM:6.81)
wa-yufakkir fi ‗alaqtu bi-zamilt-u wa- kayfa ‘anna ‘iz}a :hit ha:za l-‘ab ‘aw
mawt-u sa-tu‘addi ‘ila... (GM:6:244)
l-qara:ra:t al-luġawiyya ‘allati tantahi minha lagnit il-‘us}u:l (Diem
1974:76,ăclassifiedăasă―HocharabischămităDialektinterferenz‖)
However, the verbal contrast 3fs PF ending -at vs. -it , does not show the
same distribution as the nominal -at vs. -it contrast in my data: verbal stem +EA it
(nearly) always occurs in otherwise EA environments, and mostly suffixed to an
EA stem. A cursory reading of the data in Mazraani 1997 gives the same result,
while Bassiouny´s texts show a more balanced distribution. This suggests that the
probability of -it occurring on SA stems is higher with fem. nouns than with fem.
verbs. It also suggests that speakers´ perception of -at vs. -it may be different in
the two categories of a similar feature.
We are here at a crucial issue – if we want to move from a purely
descriptive analysis and attempt to interpret the linguistic facts in a sociolinguistic
– including stylistic – meaningful way. As we could see from the tables above, the
speakers in my academic monologue data, had different usage levels of the SA
and EA variants resp., but they agreed on the relative ranking of the features in
terms of code. This makes it natural to conclude, that the features do not all play
the same sociostylistic role. The problem, or challenge, was in fact noted by
Schmidt:ă―Colloquialăvariantsăappearăatădifferentălevelsăofăstyleăandăcorrelateăwithă
different sociological factors. These phonological contrasts between the two codes
do not all mean the same thing either stylistically or sociolinguistically‖ă
(1974:107, underlined in orig.). It was recently raised again by J. Lentin, this time
with reference to dialectology research: what appears to the descriptive
analyst/outsider to be pertinent differences/contrasts in a variety may appear
insignificant and of little sociostylistic relevance to the insiders/the actual
‘Hybrid‘ form:ă‖kasusvokalloseăForm‖ă-at (Diem 1974:37) considered SA variant here.
al-Šāyib:ă-it : wust}ā follows al-da:rija (1976:52, 61)
12
But again there may be limits to acceptability, as when informants (Schmidt 1974:179-80)
accepted ‘iza:‗it lubna:n, but rejected *‘ida:‗it lubna:n (withăinterdental,ămarkedăforă‘high-flown‘ă
style)
10
11
92
language users – and vice versa: apparently minor and subtle distinctions may
have a distinctive sociostylistic value to the users (2002:43, paraphrased by me).
Andă evenă ―theă twoă variantsă ofă aă binaryă variableă need not necessarily be
symmetricallyă balancedă whenă ită comesă toă signală code.ă […]ă Whereasă oneă variantă
cană beă ană importantă varietyă marker,ă theă otheră mayă beă nearlyă unmarked‖ă
(Maehlum 1986:171, my translation).
Living (with) a ‘diffuse’ situation
Leaving the real challenges concerning sociolinguistic interpretation, I will
turn to a different, and very attractive perspective, one in which we need not fear
the unpredictable, nor despair of the high degree of variability! The notion of
‗diffuse‘ăwasăcoinedăbyăRobert LePage (et co.) to the sociolinguistic situation and
varieties in the multilingual and diglossic Creole setting. In their perspective, the
individual speaker´s version of a linguistic variety links to other individual
versions in the community, and as long as these versions are not overlapping, the
varietyăisă‗diffuse‘.ăTheăprocessăofăaccommodatingătoăotherăversionsă– for instance
through genre formation, I add – is a process which may eventually, but not
necessarily, lead to convergence of the individual versions to become a more
‗focused‘ăvariety.ă
Peter Auer (1998) proposes a similarly dynamic view of contact varieties.
Heă positsă aă continuumă ―whichă spansă oută betweenă threeă well-documented cases
[…]:ă alternatională code-switching (CS), language mixing (LM), and fused lects
(FLs).‖ăCSăisăreservedăforăcasesăwhereăalternationăisăperceivedăbyăusersăasă―locallyă
meaningful‖,ămeaningăthatăeachăswitchătoăanotherăcodeăisăintendedăandăinterpretedă
as indexing some aspect of the situation or relationship between participants. In
LM varieties, the combined use of two codes is meaningful, but not every
alternation.ă Thisă seemsă similară toă theă ‘compositeă matrix‘ă ofă Myers-Scotton´s
code-switching model, but is less ambitious with regard to claiming systematicity.
An LM code or variety, however, may become a FL, a stabilized mixed variety.
―Theă transitionă fromă LMă toă FLă isă primarilyă ană issueă foră grammaticală research‖,ă
saysăAuer,ă―essentialăingredientsăofăthisătransitionăareăaăreductionăofăvariationăandă
an increase of rule-governed, non-variableăstructuralăregularities‖ă(1998:1).
I believe we may confortably and impatiently continue the search for
regularities, while keeping in mind the variable, diffuse and only partially
systematic nature of this kind of data – which I reluctantly agree to call a variety
only because its native users perceive there to be one. Conventionalization of
patterns, reduction of variation, is likely to appear first in recurrent situations and
functions, as genre formations. We should therefore be careful to specify the
93
sociolinguistic context of the speech we are investigating, and be very receptive to
nativeăspeakers‘ăperceptionsăofăwhatăcountsăasămeaningfulăvariation.
References
AbūăTāyeh,ă‗Aliyā‘.ă1995.ă―Mustawayātăal-luġaăfīăwasā‘ilăal-‘i‗lām al-mant}ūqaăfīă
al-‘urdunn‖,ă in: Nadiaă Anghelescuă andă Nicolaeă Dobrişană (eds.):ă Proceedings of the
Colloquium on Arabic Linguistics, Part Two. Bucharest: Center for Arab studies, 75-95.
Auer, Peter. 1998. From Code-switching via Language Mixing to Fused lects:
Towrd a Dynamic Typology of Bilingual Speech. In LiSt No.6 [Interaction and Linguistic
Structures]
Badawī,ăElsaidăMuhammad.ă1973.ăMustawayāt al-luġa al-‗arabiyya al-mu‗ās}ira
fī mis}r . Cairo:ăDārăal-ma‗ārif.
Bassiouny, Reem. 2006. Functions of Code Switching in Egypt. Evidence from
Monologues. [Studies in Semitic Languages and Linguistics 46]. Leiden:Brill
Blanc,ăHaim.ă1964ă[1960].ă―Stylistică VariationsăinăSpokenă Arabic:ă Aăsampleă ofă
Interdialectală Educatedă Conversation‖,ă in:ă Charlesă Fergusonă (ed.):ă Contributions to
Arabic Linguistics, Cambridge:Harvard University Press, 79-161.
Blau,ă Joshua.ă 1991.ă ―Theă Statusă andă Linguistică Structureă ofă Middleă Arabic‖,ă
Jersualem Studies of Arabic and Islam 23 (1999), 221-227.
Boussofara-Omar,ăNaima.ă2006.ă―Neither third language nor middle varieties but
diglossică switching‖,ă in:ă Zeitschrift für Arabische Linguistik/Journal of Arabic
Linguistics, 45 (2006), 55-80.
Diem, Werner. 1974. Hochsprache und Dialekt im Arabischen. Untersuchungen
zur heutigen arabischen Zweisprachigkeit. [Abhandlungen für die Kunde des
Morgenlandes XLI,1] Wiesbaden: Franz Steiner (Deutsche Morgenländische
Gesellschaft).
Eid,ă Mushira.ă 1988.ă ―Principlesă ofă codeă switchingă betweenă Standardă andă
EgyptianăArabic‖,ăin:ăAl-‗arabiyya 21:51-79.
Holes, Clive. 1987. Language variation and change in a modernising Arab state:
the case of Bahrain. [Library of Arabic Linguistics. Monograph 7] London: Kegan Paul
International.
Lentin, Jérôme. 2002. ―Variantesă dialectalesă‘objectives‘ă etă‘subjectives‘:ă l‘écart
entre différences de forme et différences de statut sociolinguistique, et ses implications
poură l‘enquêteă dialectologique‖,ă in:ă Abderrahmană Youssiă etă al.ă (eds.): Aspects of the
dialects of Arabic today. Rabat: Amatril, 43-54.
LeăPage,ăRobert.ă1992.‖Youăcanăneverătellăwhereăaăwordăcomesăfrom‖:ălanguageă
contact in a diffuse setting. In: Ernst Håkon Jahr (ed.): Language Contact. Theoretical
and Empirical Studies. Berlin/New York: Mouton de Gruyter.
Maehlum, Brit. 1986. Språklige variasjonsmønstre hos innflyttere i Oslo. Oslo:
Novus.
94
Mazraani, Nathalie. 1997. Aspects of Language Variation in Arabic Political
Speech-Making. Surrey: Curzon.
Meiseles,ă Gustav.ă 1977.ă ―Restitutionă ofă ‗Word-Endings‘ă ină Modernă Literaryă
Arabic‖,ăin:ăIsrael Oriental Studies 7 (1977), 173-195.
Mejdell, Gunvor. 2006. Mixed styles in spoken Arabic in Egypt. Somewhere
between order and chaos. [Studies in Semitic Languages and Linguistics 48]. Leiden:
Brill.
Owens, Jonathan and Raslan Bani-Yasin.ă1987.ă―Theălexicalăbasisăofăvariationăină
Arabic‖,ăin:ăLinguistics 25 (1987), 705-38.
Petersen,ăJennifer.ă1988.ă―Word-internal code-switching constraints in a bilingual
childﺍsăgrammar‖,ăin:ăLinguistics 26 (1988), 479-493.
Salib, Maurice. 1979. Spoken Literary Arabic: Oral Approximation of Literary
Arabic in Egyptian Formal Discourse. Unpublished Ph.D. thesis, University of California
at Berkely.
Schmidt, Richard Wilbur. 1974. Sociolinguistic variation in spoken Egyptian
Arabic: A re-examination of the concept of diglossia. Unpublished Ph.D. dissertation,
Brown University.
Schulz, David Eugene. 1981. Diglossia and Variation in Formal Spoken Arabic
in Egypt. Unpublished Ph.D. dissertation, University of Wisconsin at Madison.
al-Šāyib,ă Muh}a mmad.ă 1976.ă ―Al-‗arabiyyaă al-wust}āă – wa-māă naﻣa‘aă fī-hāă mină
tadāxul bayn al-fus}hā} ă wa-l-da:rija‖,ă in:ă Revue Tunisienne des Sciences Sociales 46: 4766.
95
96
Grundvokabular und idiomatische Struktur:
Arabisch in Nigeria und Nordafrika
Jonathan Owens, Maryland
Fadila Brahimi, Bayreuth
Ein Großteil der linguistischen Forschung zu den genetischen
Beziehungen zwischen Sprachen geht davon aus, dass Sprachen als Einheit zu
verstehen sind, die die syntaktische, morphologische oder semantische Ebene
gleichermaßen umfasst (z.B. Thomason und Kaufman 1988). Nur eine Minderheit
in der Forschung (z.B. Allen 1953) geht davon aus, dass es sinnvoll sein kann,
einzelne linguistische Komponenten wie Morphologie, Syntax oder Semantik als
Basis eines Vergleichs zu sehen.
Bislang ist in der Arabistik die Sprachkontaktforschung anders als in der
Afrikanistik oder allgemeinen Linguistik (Chambers et al 2002) nur rudimentär
etabliert (Arnold und Behnstedt zu Aramäisch – Arabisch, 1993). Häufig steht
man diesem Ansatz grundsätzlich eher skeptisch gegenüber (Diem 1979) und
neigt dazu, Sprachveränderungen durch Vereinfachung oder internen
Sprachwandel zu erklären (Diem 1971). Ein Grund dafür liegt sicherlich darin,
dass in bestimmten Bereichen, insbesondere der Morphologie, kaum
Veränderungen stattgefunden haben. Vergleicht man beispielsweise ein Verb aus
dem nigerianischen Arabischen mit der in Usbekistan verwendeten Entsprechung,
so ist trotz der erheblichen geographischen Entfernung und einer über 1200 Jahre
getrennten sprachlichen Entwicklung kaum ein Unterschied zu erkennen.
Western Sudanic Arabic (imperfect) Uzbekistan (imperfect)a-ktubă―Iăwrite‖,ăetc
a-ktibta-ktub ti-ktibta-ktub-i
tiktib-iini-ktub
i-ktibta-ktub ti-ktib
Zugleich legen die Ergebnisse unserer Forschung zum Arabischen in
Nigeria und Nordafrika die Vermutung nahe, dass ein kompositionelles Modell,
dass das Augenmerk auf einzelne Aspekte der Sprache legt und wie in diesem Fall
die idiomatische Struktur in den Fokus stellt, bestimmte Aspekte genauer zu
beschreiben in der Lage ist. Dieser Ansatz kann nach unserer Auffassung
insbesondere einen Beitrag dazu leisten, die in der Arabistik bislang wenig
ausgeprägte Erforschung des Sprachkontakts zu forcierten und in diesem Kontext
97
die infolge von Migrationsbewegungen entstehenden sprachlichen Variationen
detaillierter und umfassender darstellen zu können.
Diese Arbeiten basieren auf dem Ansatz von Owens (1996), und stehen in
der sich in jüngster Zeit entwickelnden Tradition von Johannson (2002), Ross
(1996), Bakker (2003) sowie Backus (2003), wonach durch Sprachkontakt
bedingte Änderungen auch einzelne Komponenten der Sprache betreffen können.
Karte 1, Die arabischsprachige Welt
Ausgehend von Owens Untersuchungen zum Arabischen in Nigeria (1993,
1998) wurde versucht, die idiomatische Struktur des Arabischen in den
verschiedenen Gebieten zu vergleichen und deren Besonderheiten zu beschreiben.
Das Ziel war es festzustellen, in welchen semantischen Bereichen und in welchem
Grad semantischer Wandel zu beobachten ist und in wie weit dieser Wandel dem
Sprachkontakt zuzurechnen ist. Dieser Überblick soll zeigen, welche
semantischen Bereiche sich trotz einer langen Trennung von verwandten
Dialekten als relativ resistent gegen den Einfluss von Kontaktsprachen erwiesen
98
haben, bei welchen semantischen Bereichen dies nicht der Fall ist und in welchen
Bereichen (z.B. Verwandtschaftsterminologie) beide Phänomene parallel
auftreten.
Eine systematische Beschreibung des Arabischen in Afrika aus einer
vergleichenden historischen Perspektive steht bislang aus, obwohl die
Migrationsbewegungen, durch die sich Arabisch verbreitet hat, bekannt sind.
Oberägypten nimmt darin eine Schlüsselrolle ein (Lewis 1970, Garcin 1976), da
sich von hier aus arabische Stämme sowohl in den sudanesischen Raum
(Braukämper 1993, Zeltner 1993) als auch nach Nordafrika (Brett 1978)
ausdehnten.
Owens (2003) hat in Anschluss an diese Forschung dargelegt, dass die
arabische Varietät sowohl in Nordafrika als auch im Tschadsee-Gebiet, in dem die
Untersuchung des Arabischen in Nigeria angesiedelt ist, durch
Wanderungsbewegungen etabliert wurde, deren Ausgangspunkt in beiden Fällen
in Ägypten anzunehmen ist. Es ist mithin von einer weitgehend kohärenten
semantischen Basis auszugehen, die in den folgenden Jahrhunderten durch soziale
und sprachliche Einflüsse überformt und transformiert wurde. Da die arabische
Sprache im Tschadsee-Gebiet in einem unterschiedlichen sozialen und
sprachlichen Kontext stand und in der Gesellschaft des Tschadsee-Gebietes einen
zum Teil anderen gesellschaftlichen Stellenwert als in Nordafrika eingenommen
hat, können aus der Beschreibung der Verbreitung, Aneignung und Variation von
semantischen Systemen Rückschlüsse auch auf die Wechselwirkungen zwischen
der Funktion von Sprachen als Mehrheits- oder Minderheitensprache oder dem
sozialen Status der Sprache gezogen werden.
1. Arabisch im Tschadsee-Gebiet
In der ersten Fallstudie (Owens) wurde untersucht, wie sich das Arabische
der ägyptischen Immigranten im Tschadseegebiet Nigerias bzw. Nordafrikas vom
ursprünglichen Arabisch des Ausgangsgebietes unterscheidet. Die geographische
und kulturelle Einbettung in das neue nigerianische Umfeld lassen eine deutliche
Afrikanisierung sprachlicher Strukturen des lokalen Arabischen vermuten. Das
besondere Augenmerk der Studie liegt auf eventuellen Bedeutungsverschiebungen
und afrikanischen Lehnbedeutungen vor allem im Bereich der Idiomatisierung.
Die Kanuri sind die numerisch größte Gruppe des Gebietes und die Ethnie,
die zugleich politisch und kulturell dominant ist. Die Hausa sind zwar eine
Minderheit in Maiduguri, ihre Sprache ist jedoch in den letzten 50 Jahren zur
dominanten Sprache sowohl in Maiduguri als auch allgemein in Nordnigeria
geworden. Hausa hat sich im Zuge dieser Entwicklung als Zweitsprache etabliert.
Die Schuwa-Araber sind eine relative große Minderheit, die wie andere
Sprachminderheiten im Gebiet typischerweise mehrsprachig sind. In Maiduguri
99
Karte 2, Arabisch und andere Sprachen im Tschadseegebiet
können fast alle Araber auch Hausa, und viele können auch Englisch und Kanuri,
während eine kleine Minderheit Hocharabisch beherrscht.
Zunächst wurde am Beispiel der Mehrsprachigkeit in Maiduguri in NordostNigeria untersucht, wie Arabisch von der sprachlichen Konstellation beeinflusst
wird, in die es eingebettet ist bzw. umgekehrt auf die sprachliche Umgebung
wirkt.. Zunächst wurde die Hierarchie der Sprachen erfasst, die sich aus dem
Sprachgebrauch ableiten lässt.
Tabelle 1
Sprachhierarchie
1. Hausa, Englisch (Hocharabisch)
2.
Kanuri
3. Arabisch
Sprachen der ersten Ebene sind die dominanten Sprachen. Englisch ist als
Nationalsprache Nigerias und wichtigste Bildungssprache, Hausa als allgemeine
Verkehrssprache der Region von zentraler Bedeutung. Trotz der Tatsache, dass es
von der zahlenmäßig größten ethnischen Gruppe gesprochen wird, ist Kanuri
zunehmend Sprache nur einer Ethnie. Dies zeigt sich einerseits daran, dass Kanuri
100
regelmäßig Hausa lernen, Hausa dagegen nur bedingt Kanuri, andererseits daran,
dass nur einige Araber Kanuri lernen, während alle Araber von Maiduguri Hausa
lernen. Somit ist Hausa neben Englisch die wichtigste Sprache der Stadt.
Von besonderer Bedeutung für unsere Analyse ist jedoch die Tatsache,
dass Kanuri aus einer historischer Perspektive betrachtet die dominante Sprache
im Gebiet war und die Araber seit 1400 unter einer kanurischen Hegemonie gelebt
haben. Bis zum Anfang des 20. Jahrhunderts stellte sich die Sprachhierarchie wie
folgt dar.
Kanuri
Arabisch ... andere kleine Sprachen
Hocharabisch als gesprochene Sprache spielt nur unter den gebildeten
Arabern eine Rolle und ist für unseren Kontext deshalb nur von untergeordneter
Bedeutung. Allerdings ist zu erkennen, dass Hocharabisch unter den Muslimen im
allgemeinen ein hohes Prestige genießt.
Insgesamt hat sich gezeigt, dass sich zwar eine Hierarchisierung der
Sprachen ergibt, die jedoch zumindest in dem hier beschriebenen Kontext durch
ein funktionelles Nebeneinander der Sprachen begleitet ist, das den Fortbestand
dieser Sprachen in der urbanen Landschaft sichert.
2. Arabisch in Nordafrika
Die arabische Sprache hat seit Beginn der Arabisierung im 8. Jahrhundert
die Funktion der religiös legitimierten Hochsprache und Sprache der sozial
dominanten Mehrheit. Seit der Unabhängigkeit der maghribinischen Staaten
Marokko, Tunesien und Algerien ist arabisch Nationalsprache. Neben Arabisch
fungiert französisch als Sprache der früheren Kolonialmacht Frankreich als
Hochsprache und Schriftsprache, die je nach sprachlicher Domäne bevorzugt
wird. Ihr Einfluss konnte trotz vielfältiger Arabisierungsbestrebungen bislang
nicht wesentlich zurückgedrängt werden. Als Umgangssprache dient neben Arabe
dialectal, das sich zum Teil erheblich von Hocharabisch unterscheidet, Berberisch.
Die Berber bilden in Algerien und vor allem in Marokko eine relativ große
Minderheit, ohne dass ihre Sprache eine wesentliche Rolle im öffentlichen Leben
einnimmt. Erst in jüngster Zeit haben Bestrebungen, Berberisch im
Bildungsbereich eine stärkere Geltung zu verschaffen, größeren, wenngleich lokal
begrenzten Erfolg. Während die Berber, die in Marokko und Algerien einen
Anteil von bis zu 20 oder gar 40% an der Gesamtbevölkerung stellen, praktisch
ausnahmslos Arabisch und Französisch sprechen, beherrscht nur eine kleine
101
Minderheit der arabischstämmigen Marokkaner oder Algerier Berberisch. Frühere
Untersuchungen zur Sprachsituation in Algerien (Brahimi 1993), in denen vor
allem der Status der Sprachen bzw. Sprachvarianten sowie die Untersuchung von
Sprachattitüden und Entlehnungen im Vordergrund standen, haben gezeigt, dass
ein Einfluss der berberischen Sprachen auf Arabisch nur in verschwindend
geringem Ausmaß festzustellen ist. Während Syntax, Phonologie und
Morphologie des Arabischen praktisch nicht tangiert werden, sind Entlehnungen
aus dem Berberischen allenfalls regional begrenzte Einzelfälle. Umgekehrt lässt
sich beobachten, dass eine Vielzahl von Entlehnungen aus dem Arabischen im
Berberischen zu finden sind (Brahimi, 2000). Es war aus diesem Grund zu
erwarten, dass eine vergleichende Untersuchung der Semantik des Arabischen
keine wesentlichen Einflüsse anderer Sprachen erbringen würde, zumal auch
Französisch als Sprache der Kolonialmacht und aufgrund der in historischer
Perspektive kurzen Zeit des Einflusses in dieser Hinsicht kaum Wirkung entfaltet
haben dürfte.
Tabelle 2
Sprachhierarchie in Algerien
1.
Hocharabisch, Französisch
2.
Algerisches Arabisch, Berberisch
3. Aufbau der Untersuchung
Ausgehend von der Sprachsituation in Nigeria wurde durch die
Auswertung von Tonaufnahmen und Texten ein Korpus entwickelt, der zum einen
den Grundwortschatz und zum anderen idiomatische Wendungen erfasst. Darüber
hinaus wurde die Verwandtschaftsterminologie ausgewertet.
Auf dieser Grundlage wurden die idiomatischen Wendungen extrahiert und
vergleichend u.a in Marokko (Fès) sowie in Südalgerien (Ouargla), Westalgerien
(Oran) und Ostalgerien (Annaba) erhoben.
- eine Basis-Wortliste (100-Wort-Liste bzw. Swadesh 200-Wort-Liste)
- etwa 170 idiomatische Ausdrücke, die auf der Grundlage der Arbeiten von
Owens (1993) im Tschadsee-Gebiet entwickelt wurden.
- Verwandtschaftsterminologie 1
1
Die statistische Auswertung der Verwandtschaftsbezeichungen steht noch aus. Es zeigt sich
jedoch, dass die grundlegenden verwandtschaftlichen Bezeichnungen im nigerianischen Arabisch
und Nordafrika bzw. Ägypten identisch sind. Zugleich ist jedoch auch zu erkennen, dass es für
einige Verwandtschaftsbeziehungen im nigerianischen Arabisch Bezeichnungen gibt, die etwa in
Algerien oder Marokko nicht existieren.
102
Die Auswertung der Basis-Wortliste erlaubt es festzustellen, wie weit sich
das nigerianische Arabisch von Kairo bzw. Algerien entfernt hat. Sie dient
zugleich als Maßstab für einen Vergleich der idiomatischen Ausdrücke in diesen
Gebieten.
Der Vergleich der Idiome umfasst neben den arabischen Varietäten, auch
verschiedene afrikanischen Sprachen, unter denen sowohl Sprachen aus dem
Tschadsee-Gebiet als auch aus anderen Regionen Nigerias sind. Auf diese Weise
kann festgestellt werden, ob sich NA ausschließlich an Kanuri anpasst hat oder ob
es sich um ein allgemeines Sprachkontaktphänomen handelt.
4. Ergebnisse:
4.1. Untersuchung des Basiswortschatzes
Die Untersuchung des Grundvokabulars, das Funktionswörter, häufig
verwendete Verben oder Farben umfasst, zeigt eine relativ deutliche
Übereinstimmung zwischen nigerianischem Arabisch auf der einen Seite und
kairinischen und algerischem Arabisch auf der anderen Seite:
Abbildung 5
Basis: nigerianisches Arabisch (100 Wörter)
Kairinisches
Arabisch
Algerisches
Arabisch
gesamt
Übereinstimmung
78
69
147
Keine Übereinstimmung
22
31
53
Gesamt
100
100
200
Insgesamt zeigt sich, ein hohes Maß an Übereinstimmung zwischen dem
Grundwortschatz des NA mit dem kairinischen Arabisch. Zudem ist zu
berücksichtigen, dass Synonyme nicht als Übereinstimmungen gewertet wurden.
Soă istă z.B.ă istă imă NAă dasă Wortă füră „Berg―ă hajar, ursprünglichă „Stein―ă (Kair.ă
Gabal), eine Bezeichnung, die ein charakteristisches Merkmal des Gebirges
bezeichnet. Das Wort entspricht im Kairinischen ħagar „Stein―.ă Obwohlă deră
UrsprungădesăNAă„Berg―ăunmittelbarăzuăăerkennenăist,ăwurdenăinăderăAuswertungă
derăSwadeshăListeădieăWorteăfüră„Berg―ăalsăunterschiedlichăklassifiziert.
Eine parallele Untersuchung der 200-word-list für kairinisches Arabisch,
marokkanisches sowie algerisches Arabisch, wobei zwischen südalgerischem,
103
westalgerischem und ostalgerischem Arabisch unterschieden wurde ergab eine
sehr starke Übereinstimmung.
Kair.Arab.
Übereinstimmung
Keine Übereinst.
Gesamt
191
15
207
Marok.Arab.
Westalger.
189
18
207
188
19
207
Ostalger.
185
22
207
Südalger.
183
24
207
4.2. Untersuchung der idiomatischen Wendungen
Es ist zu beobachten, dass sich im Bereich der idiomatischen Ausdrücke
NA in vieler Hinsicht an Kanuri angepasst hat. So auch in den folgenden drei
Beispielen idiomatischer Wendungen, wo NA eine genaue Lehnübersetzung des
Kanuri ist.
1. akal-o amaanti ―theyăateămyătrust‖ăţădeceivedămeăţăKanuriăamân-a zú-wú-na
2. an-naar akalat il-hille ―theăfireăateătheătown‖ăţăburnedăităupăţK.ăkánnú beladéya zú-wú-na (fire town-ACC ate)
3. akal an-naas ―heăateătheăpeople‖ăţăcheatedăthem‖ăţăK.ăkam-a zú-wú-na
Allerdings zeigt sich, dass nicht nur NA, sondern auch andere Sprachen des
Tschadseegebiets von Kanuri beinflusst wurden. Der Ausdruck, ein Auto
auszuschaltenăheisstăimăNAă„einăAutoătöten―.ăDerăgleicheăAusdruckăfindetăsichăină
Kanuri, Hausa, und in anderen Sprachen des Gebiets.
Nigerian Arabic
Kanuri
Hausa
aktul moota
(töte den Auto)
móto yej-é
ƝăăăăăăăăƝ
káshée wútá-r móotaa
ƝăăăăăăăăƝ
Um festzustellen, wie stark die Ähnlichkeiten zwischen den Sprachen des
Tschadseegebiets hinsichtlich der idiomatischen Ausdrücke ausgeprägt sind,
wurden 150 idiomatische Ausdrücke verglichen. Ausgangssprache ist dabei
immer NA, das paarweise mit jeder Sprache auf der Liste verglichen wird. Wenn
die Ausdrücke des NA völlig mit der zweiten Sprache identisch sind, wird dieses
Idiomă mită eineră „2―ă gekennzeichnet.ă Imă obenă genanntenă Beispielă füră „töteă den
Wagen―ă erhaltenă Kanuriă undă Hausaă jeweilsă eineă „2―,ă weilă dieă beidenă Sprachenă
genau dem NA entsprechen. Wenn keine Ähnlichkeit festzustellen ist, wird eine
„0―ă eingetragen.ă Z.ă B.ă hată Deutschă keinenă Ausdruckă wieă „töteă denă Wagen―ă füră
„denă Motoră ausschalten― undă wirdă deshalbă mită eineră „0―ă gekennzeichnet.ă
hnlichkeiten,ădieănichtăalsăidentischăzuăklassifizierenăsind,ăwerdenămităeineră„1―ă
klassifiziert. Neben verschiedenen Sprachen des Tschadseegebiets, dem
kairinischen. Arabisch sowie drei Dialekten Algeriens und einen Dialekt aus
104
Marokko, werden als Kontrollgruppe auch Englisch, Deutsch und Yoruba
berücksichtigt.
Wäre NA völlig mit einer anderen Sprache identisch, würde sich eine
Summeăvonă„300―ăergebenă(150ăidiomatischeăAusdrückeămultipliziertămită2ăergibtă
300).
Numeric Rank
1.00
2.00
Cairene Arabic
24
20
44
Algerian Arabic
34
36
70
Kanuri
11
232
243
Hausa
25
168
193
Fulfulde
25
178
203
Bura
15
190
205
Glavda
24
136
160
Kotoko
17
180
197
Yoruba
37
54
91
eastern Algerian
14
48
62
southern Algerian
17
46
63
Moroccan Arabic
9
26
35
Kabylisch
11
56
67
8
34
42
American English
57
8
65
German
47
4
51
378
1416
1794
southern French
Total
Total
105
Im Vergleich zur Untersuchung des Basisvokabulars zeigt sich bei den
idiomatischen Wendungen ein starker Einfluss der sprachlichen und sozialen
Umgebung auf die Ergebnisse. Das Arabisch des Maiduguri-Gebietes wird
deutlich von Kanuri beeinflusst. Hier ist eine signifikante Einflussnahme zu
erkennen, die die Idiomatik des Schuwa-Arabisch in relativ hohem Maße
bestimmt. Die soziokulturelle Kontext der Schuwa-Araber als ethnische
Minderheit spiegelt sich deutlich in einer stark gewandelten Idiomatik wider, die
sich in diesen Bereichen weit von der Idiomatik des Arabischen in Nordafrika
entfernt hat (vgl. auch Owens/Brahimi 2000).
Dass Kanuri den grössten Einfluss auf Arabisch ausgeübt hat, kann auch
dadurch festgestellt werden, dass NA einen hohen Grad an Ähnlichkeiten auch
mit Bura hat. Bura ist jedoch eine Sprache, mit der die Araber in diesem Gebiet
historisch nie eine Grenze gehabt haben. Trotzdem ist die Ähnlichkeit mit NA
hoch. Genau wie die Araber im Gebiet wurden auch die Bura in früheren Zeiten
stark von den Kanuri dominiert und beinflusst. Da sie offenbar im Kontext dieses
Kontakts in ähnlicher Weise idiomatische Ausdrücke von den Kanuri
übernommen haben, ergeben sich große Ähnlichkeiten auch zwischen NA und
Bura.
Dass es vor allem die Sprachen im Tschadseegebiet sind, die diesen
idiomatischen Sprachbund teilen, zeigt sich darin, dass Yoruba, eine Sprache, die
im Südwesten Nigerias gesprochen wird, nur wenige Gemeinsamkeiten mit NA
aufweist. Besonders beindruckend ist die Tatsache, dass NA in dieser Beziehung
kaum größere Ähnlichkeiten mit Kairinisch oder mit algerischem Arabisch hat,
als mit Deutsch oder Englisch.
5. Fazit
Es zeichnet sich ein Modell des Sprachkontakts und Sprachwandels ab, in
dem sich bestimmte Module der Sprache viel stärker als andere ändern (s. Blau
1981). Während Grundelemente der Sprache wie Phonologie oder Morphologie
konstant bleiben, weisen andere Teile wie bestimmte semantische Gebiete
erhebliche Veränderungen auf. Angesichts des starken Übertragungsdrucks
bestimmter semantischer Merkmale von einer Sprache auf eine andere könnte
106
man von einer Koinisierung der Semantik sprechen. Ursache dieser Änderungen
ist der Sprachkontakt und die in diesem Kontext entstehende Mehrsprachigkeit.
Die einzelnen Sprachen nehmen Merkmale an, die sprachübergreifende
Gemeinsamkeiten aufweisen. Es ergeben sich so Umrisse eines Modells, das die
uneingeschränkte Bedeutung von dominanten Sprachen und die gleichzeitig
erkennbare Vitalität von lokalen Sprachen in Einklang zu bringen vermag, wie sie
etwa in der lokalen semantischen Überformung der dominanten Sprache zum
Ausdruck kommen kann.
Literaturliste
Allen, W. 1953. Relationship in Comparative Linguistics. Transaction of the Philological
Society 1953:52-108. Oxford Blackwell.
Arnold W./Behnstedt, P. 1993. Arabisch-Aramäische Beziehungen in Qalamun.
Wiesbaden: Harrassowitz.
Backus,ă Ad.ă 2003.ă „Can a mixed language be conventionalized alternational
codeswitching? In: Y.Matras, P.Bakker: The Mixed Language Debate, New
York. Mouton de Gruyter, 2003. 237-270.
Bakker,ăPeter.ă2003.ă―Mixedălanguagesăasăautonomousăsystems.ă In. Y.Matras, P.Bakker:
The Mixed Language Debate, New York. Mouton de Gruyter 2003. 107-150.
Brahimi, Fadila. 1993. Spracheinstellungen in mehrsprachigen Gesellschaften. Das
Beispiel Algerien. Frankf./M.: Lang.
_______ă2000.ă―Loanwordsăină AlgerianăBerber‖.ăIn:ăOwensă(Hg.)ă Arabic as a Minority
Language. Berlin. New York.Mouton:371-382.
_______,ă undă Owens,ă Jonathan.ă „Language legitimisation: Arabic in multiethnic
contexts‖.ă Ină Owensă (Hg.)ă Arabic as a Minority Language. Berlin. New York:
405-446.
Blau, Joshua. 1981. The Emergence and Linguistic Background of Judeo-Arabic.
Jerusalem: Ben_Zvi Institute for the Study of Jewish Communities in the East.
Braukämper, Ulrich. 1993. Notes on the Origin of Baggara Arab Culture with
Special Reference to the Shuw‖.ăSpracheăundăGeschichte in Afrika 14.13-46.
Brett, Michael. 1978. The Arab Conquest and the Rise of Islam in North Africa. In Fage,
J.D. (Hg.) The Cambridge History of Africa, Vol.2, Cambridge University Press.
490-555.
Chambers, Jack, Trudgill, Peter, Schilling-Estes (Hg.). The Handbook of Language
Variation and Change. Oxford: Blackwell.
Diem, Werner. 1973. Hochsprache und Dialekt im Arabischen. Wiesbaden 1973
________ 1978. Divergenz und Konvergenz im Arabischen. In: Arabica 25: 128-147.
Garcin, Jean-Claude. 1976. Un Centre Musulman de la Haute-Egyptemédié-vale: Quus.
Cairo: Institut Français dàrchéologie Orientale du Caire
Johannson, Lars. 2002. Structural Factors in Turkic Languages. London: Curzon.
107
Lewis,ăBernard.ă1970.ă―EgyptăandăSyria‖.ăInăHoltăetăal.ă,ăeds.,ă175-230.
Owens, Jonathan. 1993. Nigerian Arabic in Comparative Perspective. Sprache und
Geschichte in Afrika (SUGIA) 14:85-176.
_________1996. Idiomatic Structure and the Theory of Genetic Relationship.
Diachronica 13: 283-318.
_________ 1998. Neighborhood an Ancestry: Variations in the Spoken Arabic in
Maiduguri, Nigeria. Amsterdam: Benjamins.
_________ 2003. Arabic Dialect History and Historical Linguistic Mythology. In: Journal
of the American Oriental Society 123.4 (2003): 715-740).
Ross 1996. ―Contact-induced change and the comparative method: Cases from Papua
Newă Guinea‖.ă Ină M.ă Durieă andă M.ă Rossă (eds.).ă The Comparative Method
Reviewed: Regularity and Irregularity in Language Change. Oxford: OUP.
Thomason, Sarah, Kaufman, Terence 1988. Language Contact, Creolization and Genetic
Linguistics. Berkeley: Univ. Of California Press.
Zeltner, Jean-Claude. 1993. Premières migrations arabes du Fezzan au Kanem. Une
brève communication. SIGIA 14 :81-84.
108
Does geographical periphery imply linguistic periphery?
The examples of the Arabic dialects of Cilicia and Urfa in
Southern Turkey
Stephan Procházka
University of Vienna
Austria
In the introduction to his valuable book Arabic as a Minority Language,
Jonathană Owensă wrote:ă ―Allă aroundă theă fringesă ofă theă present-day Arabic world
are or were found enclaves of Arabic-speaking peoples who identify themselves
asă ‗Arabs‘‖. 1 These speech-enclaves are usually what is meant by the term
―peripheralăArabicădialects‖.ăCharacteristicăofănearlyăallătheseădialectsăisăthatătheyă
are isolated from other Arabic speaking communities and that the otherwise
typical diglossia of spoken dialect versus written standard language is replaced by
theă dichotomyă ―Arabică dialectă versusă non-Arabic, mostly even non-Semitic,
language‖,ătheălatterădominatingăinăeducationăandăeconomics. 2
These are indisputable facts and even a glance into such grammars as Otto
Jastrow‘să ă Daragözü, Alexanderă Borg‘să Cypriot Arabic, oră Jonathană Owens‘ă
Nigerian Arabic leave no doubt that these are varieties of Arabic which have
undergone immense changes during their long history in isolation. This
impression is even stronger when one reads texts in these dialects, which are
commonly unintelligible to someone who has no knowledge of the adstratum
language in question.
However, my intensive research into two different varieties of Arabic
spoken within the borders of present-day Turkey has made me question how close
the correlation between geographic and linguistic periphery really is. As the
subtitle of this paper implies, these two varieties are the Bedouin-type dialects of
Urfa and Harran and the Syrian-type dialects of the Cilician Plain, called
Çukurova in modern Turkish. The latter are strongly related to the Arabic spoken
inătheăTurkishăHatayă(Antioch)ăprovince,ăwhichăbecameă―periphery‖ăonlyăină1939ă
1
2
Owens 2000:6.
For the situation in southern and southeastern Turkey cf. Procházka 1999 and Lahdo 2004.
109
when the so-called Sanjak of Alexandrette was taken by the French mandatory
authoritiesăfromăSyriaăandăpresentedăasăaă―gift‖ătoăTurkey. 3
Beside other things responsible for my doubts were the facts that the Arabs of
Urfa listen with great pleasure to cassettes of Arabian epics narrated by Bedouins
from the Syrian steppe, and that the Arabs of Adana and Tarsus understand the
humour of the popular Syrian film Rasāyil šafahīye, whose actors speak in the
dialect of the Syrian coast. Neither would be the case if these two peripheral
dialectsăandătheirărespectiveăSyriană―cognates‖ăwereămutuallyăunintelligible.
Ină oneă ofă hisă manyă persuasiveă articles,ă ―Divergenză undă Konvergenz im
Arabischen‖,ă Werneră Diemă postulatedă ină theă 1970să thată peripherală dialectsă areă
chiefly characterized by two things: first, by an accumulation of developments
intrinsic to the Arabic language; and second, by linguistic phenomena which are
the results of convergence with a (non-Semitic) adstratum language. 4 Alexander
Borg in his Glossary of Cypriot Arabic (p. 59) put a similar concept into these
well-chosenăwords:ă―Theăformalăprofilesăofătheseă[peripheral]ăvernacularsăpresentă
a Janus-like character being, on the one hand, integral microcosms of Arabic
linguisticăevolutionă[…]ăand,ăonătheăother,ăexemplarsăofăidiosyncraticăevolutionaryă
paths representing the outcome of a far-reaching adaptation to their non-Arabic
linguistică andă culturală milieu.‖5 Borg (op. cit., p.61) additionally stresses – and
with good reason – that many of these peripheral dialects have preserved
archaisms, particularly in the realm of the lexicon.
In search of a reasonably convincing answer to the question formulated in
the title of this paper I took the following into consideration:
First: Periphery cannot be taken alone, as every periphery by definition has
a centre. This means that we first have to determine a reference dialect to which
our alleged peripheral dialect can be regarded as peripheral. Such a non-peripheral
reference dialect must of course be chosen from an area which is dialectologically
related to the hypothetically peripheral dialect, since a comparison with a prestige
variety of another dialect group would lead to useless results. After all, compared
with Damascus or Cairo, every Algerian and Moroccan dialect would turn out to
be peripheral, and vice versa!
Thus I decided to consider three or four dialects from each dialect group:
first, the dialect whose linguistic ―peripherality‖ă isă underă study; second, one
3
For the dialects of that region see Arnold 1998.
Diem 1978:133.
5
It goes without saying that all Arabic dialects spoken in the Levant and in Mesopotamia had and
partially still have close contacts to foreign languages – e.g., Palestinian Arabic with Modern
Hebrew. What makes the difference is the intensity of such contacts peripheral dialects are
exposed to (cf. also Borg 2004:64).
4
110
related dialect which represents the linguistic centre; and third, if any is available,
a dialect whose peripheral character has been unanimously agreed upon in
previous studies. To make the results more reliable I also included in this study
the third group of Arabic dialects spoken in Turkey, namely the qəltu-dialects of
south-eastern Anatolia. Hence we are considering a total of three dialect groups:
1. The Syrian group with Aleppo as reference dialect 6, Cilician Arabic as a
possible peripheral dialect, and the Kormakiti Arabic of Cyprus as the
proven peripheral dialect.7
2. The second group consists of the Bedouin dialects of the šāwi- (or
šawāya-) type, with the dialect of ar-Raqqa as reference dialect and the
dialects of the Urfa-Harran region as alleged peripheral dialects. Hitherto no
dialect of the šāwi group has been clearly identified as peripheral. However,
to have another point of reference besides ar-Raqqa we have included some
data on Kuwayti Arabic, which is related to the šāwi branch.8
3. The third group are the qəltu-dialects of northern Mesopotamia, for which
I have chosen Mossul as reference dialect. According to Otto Jastrow9, the
so-called Mas}lāwiăisăcharacteristicănotăonlyăforătheăTigris group but for the
whole of Iraq – evenă―moreăIraqi‖ăthanăChristianăBaghdadiăArabic.ăHenceăită
isă theă ideală pointă ofă reference.ă Unfortunately,ă otheră thană Jastrow‘să 1979article, data on Mossul is very limited. 10 As unquestioned examples of qəltu
peripheral dialects I took the dialects of Daragözü and Hasköy spoken near
Lake Van. For a dialect whose peripheral status is not so extreme I have also
included Mardin in the group, in part because Grigore 2007 offers abundant
and reliable data on this dialect.
As can be seen on the map below, the distances between the points of
reference are not quite equal, but they are not so different that the results would be
distorted. Group 1: Aleppo –Adana 180 km, Aleppo – Cyprus 370 km; group 2:
ar-Raqqa – Urfa 150 km; group 3: Mossul – Mardin 220 km, Mossul – Daragözü
260 km, Mossul – Hasköy 280 km.
6
Aleppo is geographically and dialectologically sufficiently close to Cilicia for our purposes; and
as the dialect ofăSyria‘săsecondălargestăcityăităcannotăcalledă―peripheral‖.
7
Borg 1985:154-159 and 2004:24-52 has claimed that Cypriot Arabic is not a direct offshoot of
what we call Syro-Lebanese Arabic but exhibits many features which link it to the qəltu-dialects
(e.g., the umlaut-imāla).ăButăasăthisăisăalsoătrueăforăCilicianăArabică– and to a certain extent also for
the dialect of Aleppo – we do not see a methodological problem in grouping them together.
8
Forăanăoverviewăofăthisăgroupăseeătheăarticleă―shāwiya‖ in The Encyclopaedia of Islam, vol.ăıx.
9
Jastrow 1979:54f.
10
Some additional material I owe to Mrs Sonja al-Dulayme, a student in Vienna who was born and
raised in Mossul.
111
^
Map of the region and dialects dealt with
The next step was to determine which linguistic features should be used
for comparison. This was not so difficult for phonology, in which unusual sound
shifts and other striking developments are relatively easy to detect and data is
abundant (Tables 1 and 2). As for morphology, I decided to restrict my
investigation mainly to some parts of the closed-list classes: personal pronouns,
interrogatives, and prepositions (Tables 3A and 3B). In addition, I included verbal
modifiers, the use of the article, and the numerals (Table 4). However, syntax
(Table 5) and lexicon (Tables 6A-C), are much harder to deal with. Thorough
syntactical studies on Arabic dialects are still rare. Word order, for instance, a
subject which would certainly be useful for our purposes, has been studied only in
regard to the qəltu-dialects.11 Lexicon is an immense field and it is very difficult
to select lexical items which are more convincing than a random sample.
Moreover, there is insufficient lexical material available for some of the dialects
here regarded, in particular for the Bedouin group. Thus, I considered only some
veryă frequentă verbs,ă suchă asă ―toă make,ă toă go‖, etc., since differences in these
items, representing a kind of lexical core of a dialect, would suggest an unusual
lexical development in that dialect.
Dahlgren 1998 found that the Anatolian qəltu-dialects exhibit a significantly higher rate of S-VO word order than the other dialects investigated by him.
11
112
To facilitate comparative analysis of the data, I established a kind of
―peripheralityăindex‖ăforăeveryădeviatingăfeature,ăaddingăeitherăoneăorătwoăpoints.ă
The criteria for assigning points will be explained in the following paragraphs: in
the tables they are indicated by 会 or 解. It should be stressed that the following
remarks should be seen as a first attempt at finding a way to objectively measure
theă degreeă ofă ―peripherality‖ă ofă aă givenă dialect.ă Iă amă fullyă awareă thată theă
methodology needs refinement and a broader corpus of data than presented here.
Tables 1-2 deal with phonology. A good indicator for the influence of
foreign languages is the emergence of completely new phonemes unknown to Old
Arabic. These are shown on Table 1 A. It remains an open question how to deal
with such phonemes if they are the result of an internal development. For
example, if Cilician Arabic has a phoneme /č/ because of Turkish influence, for
our purposes this must be handled completely differently from a Bedouin Arabic
/č/ that reflects the Old Arabic /k/ in the vicinity of front vowels. As this problem
plays only a marginal role in the material discussed here12, we have neglected it.
But for further studies of the present sort, this problem certainly needs a solution.
Inămeasuringătheă―peripheralityăindex‖ăIăgaveă2ăpointsătoăeachădeviationăină
phonology from the reference dialect, except for two situations for which I gave
only 1 point: (1) Cases where the deviating phenomenon is widely found in the
same dialect area elsewhere (e.g. for Cilician Arabic /q/ in contrast to Aleppine /‘/,
as /q/ is typical for most rural dialects in northern and north-western Syria13); and
(2) cases where the deviation is only partial (e.g.ăforătheăpartialăshiftăfromă/‗/ to /h}/
in Daragözü).
It is worth mentioning with regard to consonants that the most numerous
deviations are found in the dialects of Cyprus and Hasköy. In the vowel system it
is striking that neither Urfa Arabic nor the Anatolian dialects show deviations
from their respective Raqqa and Mossul reference dialects.
For deviations in morphology, I gave 2 points for each item of completely
different origin – e.g.ă―what‖:ăinăMossulăaš, in Daragözü št}a ba – or if the word in
question is etymologically related but has been completely restructured – for
instance,ă―she‖:ăinăMossulăhīya, in Hasköy yā. Otherwise I gave 1 point. No points
were given for only slight differences (e.g. nih}na, nəh}ne etc.). Among the
prepositions it is noticeable that several dialects exhibit the temporal use of
reflexes of quddāma ―inăfrontăof‖. At least with regard to the features considered
12
13
E.g.,ăCypriotăArabică/ž/ăŢă/ğ/,ă/v/ăŢă/w/.
Cf. Behnstedt 1997: map 9.
113
in this study, Cilician and Cypriot Arabic exhibit almost the same index of
―peripherality‖.
In the realm of syntax (see Table 5), the construction of determined
adjectival attribute phrases like an id}āfa – i.e. the use of the construct state for the
substantive (type bayt al-kabīr) – is most probably an archaic feature since it is
frequently found in pre-Classical Arabic and in Middle Arabic. 14 The use of a
copula, however, seems to be a typical innovation under the influence of IndoEuropean languages such as Kurdish and Greek, and perhaps of Altaic Turkish
where the use of a copula is obligatory in all but the third persons. As mentioned
above, it would be very interesting to measure the impact of the adstratum
languages on word order but as of yet an insufficient number of studies are
available on this subject to do so.
Lexical differences are much harder to compare and to measure than those
in other fields. Therefore I shall briefly mention only two tendencies which are
typical for the periphery. First is the preservation of archaic words. This is,
however, true not only for Arabic dialects spoken outside the Arabic world but for
the periphery in general and, therefore is found, for example, in Yemen and
Oman. 15
In the Syrian group we can mention the non-diminutive forms kayyis and
kaes ină contrastătoă ―modern‖ă Aleppineă kwayyes, the Cilician sahal ―toă ask‖,ă andă
theăwordsăforă―hair‖,ăCilicianăğimmi (< Old Arabic ğumma) and Cypriot šaxve (<
Old Arabic ša‗afa) which has a cognate in the Anatolian šah}fe.16 In Anatolia an
important archaic word is ar }a ―toă see‖,ă which is used in Mardin. Idiosyncratic
internal developments seem to prevail in Daragözü and Hasköy where we find
wordsăforă―toăsee,ătoăspeak,ăandătoăask‖ăwhichăareănotăattestedăelsewhere,ăatăleastă
not outside Anatolia (see Table 6 B).
The second tendency of peripheral lexicons, the high percentage of
loanwords, is without doubt typical for all these dialects. According to Vocke &
Waldner‘să glossary,ă oneă thirdă ofă Daragözü‘să vocabularyă isă notă Arabic 17 and this
may be the case for Cypriot Arabic as well. However, as our tables show, foreign
words rarely replace core elements of the lexicon. With the exception of the
Cilician tqōnaš ―toă speak‖ă (Ţă Turkishă konuşmak) and the Hasköy ādaš / ītəš ―toă
14
In the dialects treated in this study, such a construction is very frequent in Cilicia (Procházka
2002a:152f.) andă ină ƥzəxă (Wittrichă 2001:148).ă Blancă 1964:161ă reportsă theă sameă foră Jewishă andă
Christian Baghdadi.
15
Cf. Behnstedt & Woidich 2005:143-147, who maintain that, if two marginal regions exhibit the
same forms deviating from the centre, it can be supposed that they have preserved the older forms.
16
Cf. Borg 2004:292.
17
See Vocke & Waldner 1982:XLIII.
114
see‖,ă whichă isă perhapsă ofă Armeniană origin, 18 there is no loanword among the
verbs for the most important actions.
The following table contains the figures for the six geographically
peripherală dialectsă investigatedă hereă showingă theiră ―peripheralityă indices‖ă foră
phonology, morphology, and syntax.
Syrian Group
Bedouin
Anatolian Group
Cilicia Cyprus Urfa-Harran
Daragözü Hasköy Mardin
Phonology
Morphology
Syntax
5
20
3
28
21
3
0
8
0
14
24
4
17
32
4
5
12
4
Sum
28
52
8
42
53
21
Lexicon aside, the table presents a picture which was to a certain extent
expected. Nevertheless it displays a couple of new and interesting details. First of
all it is very striking that the dialects of the Urfa-Harran region have a
―peripheralityăindex‖ăofăonlyă8.ăHenceădeviationăfromătheăreferenceădialectăRaqqaă
can be regarded as minor. Even if we would compare Urfa Arabic with Kuwayti
Arabic, 1200 km away, the differences would not be as big as those between
Aleppo and Cilicia, separated by a distance of only 180 km. Thus we can state
with good reasons that the Arabic of the Urfa-Harran area cannot be labelled as
linguisticallyă ―peripheral‖,ă althoughă ită isă spokenă ină aă non-Arabic country in a
mixed Turkish-Kurdish-Arabic milieu. And this is true despite the fact that since
Turkey gained its independence in 1923 personal contacts between the Arabs of
Urfa and their fellow tribes-people in Syria have been very difficult, if not
impossible, to maintain. As for the second group, Cilician Arabic as compared to
the Aleppine dialect, the numbers clearly indicate a peripherality only slightly
greater than that of Mardin Arabic as compared to Mas}lāwi.ă Bută thereă isă aă
significantă gapă betweenă Cilicia‘să ―peripheralityă index‖ă andă thată ofă Cyprusă oră
Hasköy.
Consequently we can say that the mere fact that a given Arabic dialect is
spoken outside of the Arabic world proper and among a non-Semitic majority
does not necessarily lead to a high number of idiosyncratic developments.
18
According to Shabo Talay, p. c.
115
Geographical location on the fringe of the Arabic speaking world is important, no
doubt, but it is only one of several factors for extreme linguistic developments.
More significant seems to be isolation, since so-called language islands (Cyprus,
Cilicia, Daragözü, Hasköy) show more deviation from the respective reference
dialects than those regions which are separated from the Arabic world only by
political borders (e.g. Hatay, Urfa, and Mardin).
Nor does time seem to play such an important role in linguistic deviation.
Some of the Bedouin Arab tribes of Urfa and its surroundings reached the region
probably in the first century of the Hijra19 – that is, more than a millennium ago.
The lands of their dwellings have always been situated at the edges of the Arab
world and for centuries they have had close contacts with Kurds and Turks.
Despiteă this,ă theă ―peripheralityă index‖ă ofă theiră dialect is quite low. The main
reasons for that may be primarily explained by their social and religious identity.
The Arabs of Urfa and Harran still have strong tribal structures, and in spite of the
difficulties contacting their relatives in Syria, remain very aware of these ties.
They are proud to be Arabs and consider themselves of more noble blood than the
Turks, and especially the Kurds around them. The awareness of being part of the
greată ―Arabă nation‖,ă togetheră withă aă lackă ofă everydayă contactsă withă non-Arabs
outside of the town of Urfa itself, may be the principle reasons for the few
typically peripheral features in their Arabic. Until the collapse of the Ottoman
Empire these two factors were certainly reinforced by the semi-nomadic way of
life that enabled close contacts among the tribes of the entire Syrian steppe. The
surprising uniformity of the šāwi Bedouin dialects spoken at the edges of the
Fertile Crescent, including the region treated here, is certainly the result of the
Bedouin‘să mobilityă thată enabledă frequentă migrationă ofă wholeă tribesă oră ofă subtribes throughout the Syrian Steppe from the slopes of the mountains of Kurdistan
in the north to the shores of the Persian Gulf in the south.
The Arabs of Cilicia, by contrast, did not immigrate to their present
homeland before the late 18th and early 19th centuries, but their dialects exhibit a
significantly higher periphery parameter than the sedentary vernacular of the
people of Mardin, who settled there no later than the 8th century. 20 One reason for
that may be that in Cilicia the sense of being Arab is at present very weak. 21 The
main point of self-identity for the Arabic speakers of Cilicia is their Alawi faith. It
can be assumed that their adherence to this religion, which is highly despised by
19
See Oppenheim 1939:222 and Procházka 2003:76f.
Time sometimes can be an important factor, however, as the different degree in peripheral
development of Cilician and Cypriot Arabic may largely be the result of the much longer isolation
of the latter.
21
Cf. Procházka 1999.
20
116
their Turkish compatriots, has resulted in fewer contacts with the majority
population than if they were of the same faith. I cannot prove it but perhaps the
muchă higheră ―peripheralityă index‖ă ofă theă otheră isolatedă groupsă – Cyprus,
Daragözü, Hasköy – is to a certain extent explicable by the fact that these three
groups have been living mainly among people of the same religion, 22 with whom
they therefore more freely intermingle than do the Alawis with the Sunni Turks
around them. Evidences for this assumption are the facts that in the Anatolian
qəltu-dialects the percentage of loanwords from the Christian Aramaic is far
below 1% 23, and that loanwords from Armenian, to which at least Hasköy and
Daragözü must had been exposed, are extremely rare if not practically nonexistent.
In conclusion we can say that most Arabic dialects spoken at the
geographical periphery of the Arabic world unquestionably exhibit linguistic
peculiarities. However, a closer study is necessary to determine how far-reaching
these changes from mainstream Arabic really are. Especially the high number of
non-Semitic loanwords in these dialects often leads one to jump to the conclusion
that they have developed significantly away from their Arabic sisters. But such an
assertion should be handled with caution. A few years ago Otto Jastrow (1997)
convincingly demonstrated that even the Arabic of Uzbekistan is much more
Arabic than was maintained by its first describers. And to answer the question in
the title of this paper: Cilician Arabic belongs – like Mardin Arabic in the East –
to a group of Arabic dialects which we could perhaps label as marginal or semiperipheral. That means they show some features typical for the periphery but still
widely conform to related vernaculars in the centre with significant deviations
only in the lexicon, where the high number of non-Arabic words certainly
impedes mutual intelligibility. As for the Arabic of Urfa, a thorough comparison
with other dialects of the šāwi-type shows that this dialect is in no sense
peripheral with regard to linguistic structures.
References
Arnold, Werner. 1998. Die arabischen Dialekte Antiochiens. Wiesbaden: Harrassowitz.
Barthélemy, Adrien. 1935-1954. Dictionnaire arabe-français. Paris: Geuthner.
Behnstedt, Peter. 1997. Sprachatlas von Syrien: Kartenband. Wiesbaden: Harrassowitz.
Behnstedt, Peter. 2000. Sprachatlas von Syrien: Volkskundliche Texte. Wiesbaden:
Harrassowitz.
22
Although Jastrow in this volume mentions the former (?) existence of Christian Arabs in the
regions of Kozluk and Sason.
23
See Vocke & Waldner 1982:XLIII.
117
Behnstedt, Peter & Woidich, Manfred. 2005. Arabische Dialektgeographie: Eine
Einführung. Leiden: Brill.
Blanc, Haim. 1964. Communal Dialects in Baghdad. Harvard University Press.
Borg, Alexander. 1985. Cypriot Arabic. Stuttgart: Steiner.
Borg, Alexander. 2004. A comparative glossary of Cypriot Maronite Arabic: (Arabic English). Leiden: Brill.
Cantineau, Jean. 1936-37.ă―Étudesăsurăquelquesăparlersădeănomadesăarabesăd‘Orient‖, in:
Annales de l‘Institut d‘Études Orientales 2, 3, 1-118 ; 119-237.
Dahlgren, Sven-Olof. 1998. Word Order in Arabic. Göteborg: Acta Universitatis.
Diem, Werner. 1978. ―Divergenz und Konvergenz im Arabischen‖, in: Arabica 25, 128147.
Grigore, George. 2007. L‘arabe parlé à Mardin – monographie d‘un parler arabe
périphérique. Bucarest :ăEdituraăUniversitĒįiiădinăBucureşti.
Jastrow, Otto. 1973. Daragözü - eine arabische Mundart der Kozluk-Sason-Gruppe (SOAnatolien), Nürnberg: H. Carl.
Jastrow, Otto. 1978. Die mesopotamisch-arabischen qəltu-Dialekte. Vol. 1. Wiesbaden:
Steiner.
Jastrow, Otto. 1979.ă―Zur arabischen Mundart von Mossul‖, in: ZAL 2, 36-75.
Jastrow, Otto. 1997. ―Wie arabisch ist Uzbekistan-Arabisch?‖, in: E. Wardini (ed.): Built
on Solid Rock. Studies in Honour of Professor Ebbe Egede Knudsen. Oslo:
Novus, 141-153.
Lahdo,ă Ablahad.ă 2004.ă ―Turkishă influenceă onă theă Arabică dialectă ofă Tillo‖,ă in:ă Turkic
Languages 8 (2004), 122-137.
Levin,ăAryeh.ă1987.ă―TheăParticleăla as an object marker in some Arabic dialects of the
Galilee‖,ăin:ăZeitschrift für Arabische Linguistik 17.31-40.
Oppenheim, Max Freiherr von. 1939. Die Beduinen, vol. I, Leipzig.
Owens, Jonathan. 1993. A grammar of Nigerian Arabic. Wiesbaden: Harrassowitz.
Owens, Jonathan. 2000. Arabic as a Minority Language. Berlin: Mouton de Gruyter.
Procházka, Stephan. 1999.ă―FromăLanguageăContactătoăLanguageăDeath:ăTheăExampleăofă
theăArabicăSpokenăinăCiliciaă(SouthernăTurkey)‖,ăin:ăOrientalia Suecana 48
(1999), 115-125.
Procházka, Stephan. 2002a. Die arabischen Dialekte der Çukurova (Südtürkei).
Wiesbaden: O. Harrassowitz.
Procházka, Stephan. 2002b. ―DieăAlawitendialekteăderăÇukurovaă– Vergleichsmaterialien
zu Peter Behnstedts Sprachatlas von Syrien‖, in: WZKM 92.91-109.
Procházka,ăStephan.ă2003.ă―TheăBedouinăArabicăDialectsăofăUrfa‖ăin:ăI.ăFerrandoă&ăJ.J.ă
Sanchez Sandoval: AIDA: 5th Conference Proceedings, Cádiz, 75-88.
Roth,ă Arlette.ă 2004.ă ―Leă parleră arabeă maroniteă deă Chypre:ă observationsă àă proposă d‘ună
contactălinguistiqueăpluriséculaire‖,ăin:ăInternational Journal of the Sociology of
Language 168 (2004), 55-76.
Sabuni, Abdulghafur. 1980. Laut- und Formenlehre des arabischen Dialekts von Aleppo,
Frankfurt a. Main.
118
Sasse, Hans-Jürgen. 1971. Linguistische Analyse des arabischen Dialekts der Mh}allamīye
in der Provinz Mardin (Südosttürkei). München, Diss.
Talay, Shabo. 2001. ―Der arabische Dialekt von Hasköy I‖, in: ZAL 40, 71-89.
Talay, Shabo. 2002. ―Der arabische Dialekt von Hasköy II‖, in: ZAL 41, 46-104.
Vocke, Sibylle & Waldner, Wolfram. 1982. Der Wortschatz des anatolischen Arabisch.
Erlangen.
Wittrich, Michaela. 2001. Der arabische Dialekt von ƨzəx. Wiesbaden: Harrassowitz.
119
Table 1 A – Phonology I: Consonants – New Phonemes
120
1. Consonants
Syrian Group
Aleppo Cilicia
A.New
phonemes1
č
g
č
g
č
g
p
—
—
v
—
ž
—
v
—
Cyprus
č2
Bedouin Group
Urfa
Raqqa
Kuway
t
Anatolian Group
Mossul Daragözü
Hasköy
Mardin
—
č<k
g< q
č<k
g< q
č<k
g< q
č
g
č
g
č
g
č
g
p
—
—
—
p
p
p
p
v< w
—
—
—
—
v
v
v
ž<ǧ
—
—
—
—
ž
ž
ž
Sources: Aleppo: Sabuni 1980:5-14 – Cilicia: Procházka 2002a:20-22 – Cyprus: Borg 1985:11ă(č,ăž),ă24ă(p),ă40ă(v)
-Urfa: Procházka 2003:78f.– Raqqa: Behnstedt 1997: maps 2 and 9, – Kuwayt: Johnstone 1967:4-7
– Mossul: Jastrow 1979:68ă(čǝngāl),ă75ă(čāy),ă60ă(zangīnīn), 62 (glāsāt) – Daragözü: Jastrow 1973:15-16
– Hasköy: Talay 2001:73-76 – Mardin: Grigore 2007:37-41.
1
Due to foreign influence or internal development.
Borgă1985:11ăwritesă/tﻣ/ăandăinterpretsăthisăsoundăasăbiphonemically;ăităappearsămostlyăinăloansăfromăCypriotăGreekăandăTurkish,
e.g. tše ír ―grass‖ăŢăCypr.ăGreekăčaíri.
2
Table 1 B – Phonology I: Consonants – Diachronic developments3
Cilicia
ḏ>d
Cyprus
ḏ
Aleppo
ḏ>d
Raqqa
ḏ
Kuwayt
ḏ
Mossul
ḏ
Daragözü
Hasköy
ḏ>t
Urfa
ḏ
ḏ>z
ḏ>z
Mardin
ḏ
ṯ
ṯ>t
ṯ>t
ṯ
ṯ
ṯ
ṯ
ṯ
ṯ>s
ṯ>s
ṯ
>ḍ
>ḍ
>ḏ
>ẓ
> ẓ/z
ġ
ġ
ġ
ġ>
ġ>q
ġ>q
ġ
ġ
ġ
ġ
ġ
ḥ
ḥ
ḥ
ḥ>x
ḥ
ḥ
ḥ
ḥ
ḥ
ḥ/h/
do not
ḥ
121
contrast
in
many positions
ʿ
>ḥ
> ø4
only partially
q
q>
q
q> k
q > g/ǧ
q > g/ǧ
q> g
q
q
q > x/ġ
q
r
r
r
r
r
r
r
ġ
r
r
r
y
y
y
VyV > ø
y
y
y
y
y
y
y
Develarization
—
—
yes
—
—
—
—
—
frequent
—
3
4
Excluding developments that have led to completely new phonemes which are treated in Table 1 A.
Except in initial position; the details are quite complicated, see Talay 2001:75.
Devoicing of
nonfinal
consonants
—
—
b
>
p,
—
—
—
—
—
—
d> t
122
Sources: Aleppo: Sabuni 1980: 13-27 – Cilicia: Procházka 2002a:16-20
– Cyprus: Borg 1985:26 (interdentals), 28 (devoicing), 31 (de-velarization),ă35ă(ăqă>ăk),ă36ă(ġă>ăʿ,ăḥă>ăx),ă43ă(VyVă>ăø)
– Urfa: Procházka 2003:78f. – Raqqa: Behnstedt 2000:429-432. – Kuwayt: Johnstone 1967:4-20.
– Mossul: Jastrow 1979:38-39 – Daragözü: Jastrow 1973:15-19 – Hasköy: Talay 2001:73-76.
– Mardin: Grigore 2007:37-43.
—
Table 2 – Phonology II: Vowels5 and Phonotactics
Syrian Group
Aleppo Cilicia
Cyprus
—
—
ā/ī/ūă>ăa/i/u
stressed a > i
—
—
-āC#ă>ă-ōC#
—
a>i
kalb >
kilp
šams > šims
rās > ṛōs —
sīqān
>
sīqōn6
Two
subsequent
vowels possible
—
—
shortening
long
vowels
123
5
6
of
Bedouin Group
Urfa
Raqq
a
some patterns:
laban > liban
―buttermilk‖
yes – pitsaíon —
―sheă
makesă
them‖
Kuwayt
Anatolian Group
Mossul Daragözü
—
—
Hasköy
Mardin
like
Urfa
—
The widespread so-called Umlaut-imāla has not been regarded as unusual.
Only in the region of Tarsus.
yes – tǝsǝn —
―sheă makesă
them‖
—
partiallyă ūă >ă —
ǖ,ăōă>ăȫ
gahawaSyndrom
yakūn
>
ykǖn
dawr > dōr
> dȫr
—
yes
yes
yes
Sources: Cilicia: Procházka 2002a:29, 37, 39 – Cyprus: Borg 1985:52, 48; Borg 2004:279 (pitsaíon)
– Urfa: Procházka 2003:77f.
– Raqqa: Behnstedt 2000:427, Behnstedt 1997: map 70 – Kuwayt: Johnstone 1967:6 – Daragözü: Jastrow 1973:66.
124
Table 3 A – Morphology I – Personal Pronouns, Interrogatives
Syrian Group
Bedouin Group
Anatolian Group
Aleppo
Cilicia
Cyprus
Urfa
Raqqa
Kuwayt
Mossul
Daragözü
Hasköy
Mardin
3. Sg. m.
hūwe
hūwi
uo
huwwe
hūwa
hū
hīyu
yū
hūwe
3. Sg. f.
hīye
hīyi
hiyye
hīya
hī
hīya
inte
inte
int
ǝnta
ǝnt
2. Sg. f.
ǝnte,
ǝnt
ǝnti
yā
ǝnt
hīye
2. Sg. m.
ie
int
huwwa
/ hīnu
hīya
inti
inti
inti
inti
ǝnti
ǝnte
ǝnte
ǝnti
1. Sg. c.
ana
hinti
ana
ana
āni
āni
ana
ana
ina/īnā/inā
ana
3. Pl. m.
hǝnne
hinni
innen
humma
humma
hum
hīyǝm
nā
hīyǝn
īyen/ǝnnen
hǝnne
hinne
henna
intum
intu
ǝntem
ǝnto
ǝnto
ǝntǝn
intin
iḥna
ḥinna
nǝḥna
naḥne
na na/naḥna
nǝḥne
mīn
min
mǝn
mǝni
ande
mǝn
A. Pers.
pronouns
125
3. Pl. f.
2. Pl. m.
2. Pl. f.
1. Sg. c.
B.
Interrog.
who?
hint
ǝntu
hintu
intu
nǝḥne
niḥna
naxni
mīn
men
mǝn
mēn
/
intu
intan
iḥna
minu
minha
/
ǝnt
126
what?
ēš
šū
ayš/aš
šinu
šinu
ēš
aš(nu)
šṭaba
šǝne / ašne
ǝšne
ayš ~ aš
how?
šlōn
škīf / šiklu
kíf
šnōn
šlōn
šlōn
áššōn
ǝšma
šǝmde
ášwan
where?
wēn
fēn
wēn
ayn/an
aṣṣōb
wēn
?
ēṣab
ǝmma
amma
ayn
when?
ēmat
ēmtan/īmtan
mitten
ēmat
ēmat
?
ēmta
čīčax
aččāx
áymate
how
much?
š add
ašqad
ašqa
ašqay
áška
bēš
šgadd
šgadd
ašqad
čaqa
qāč
ašqad
/
/
Table 3 B – Morphology I – Prepositions
C.Prepositions
for (the sake of)
Aleppo
la-
behind
waṛa
before
abl
Cilicia
Cyprus
ala
pšan
Urfa
l-
Raqqa
l-
Mossul
la-
xalf
xalf
waṛa
waṛa
qafa
waġa
qiddām
kintám
gabəl
?
qabǝl
/
Daragözü
Hasköy
Mardin
ša-
ša-/šān
b-šān
mqafa
/waṛa
qafa / xalf
mǝġdām
qabǝl
qfā
wara
?
/
inside of
ǧǝwwāt
ǧōfāt
ǧūwāt
/
žava
b-galb
?
ǧawwa
?
ǧuwa
ǧawwāt
127
Sources: Aleppo: Sabuni 1980:68 (pers. pron.), 76-79 (interrog.), 196-200 (prep.) – Cilicia: Procházka 2002a:64 (pers. pron.),
70f., 134f. (interrog.), 137-142 (prep.) – Cyprus: Borg 1985:133 (pers. pron.), 146-148 (interrog.), Borg 2004:282, 225, 382,
189 (prep.) – Urfa: Procházka 2003:79 (pers. pron.), 80 (interrog.), own data (prep.) – Raqqa: Behnstedt 1997:maps 250-257
(pers. pron.), maps 284-290 (interrog.), Cantineau 1937:206f. (interrog.), Cantineau 1937:208f. (prep.), Kuwayt:
Johnstone 1967:66 (pers. pron.), 68f. (interrog.) – Mossul: Jastrow 1979:42 (pers. pron.), 44-45 (interrog.), own data (prep.)
- Daragözü: Jastrow 1973:37 (pers. pron), 42f. (interrog.), 97f. (prep.) – Hasköy: Talay 2001:76 (pers. pron.), 80 (interrog.,
qāč Talay 2002:54), Talay 2001:82 (prep., 2002:72 ǧuwa),
– Mardin: Grigore 2007:227f. (pers. pron.), 237f., 249f. (interrog.). 270-282 (prep.).
Table 4 – Morphology II – Verbal Modifiers, Phrasal Verbs, Article, Numerals
D.
Verbal
modifier
progressive
present tense
128
E.
Phrasal
verbs
F. Usage of
the
definite
article
Syrian Group
Aleppo
Cilicia
abma-
—
frequent:
sawwa
ṣoxbat
―toă
talk‖ă sawwa
idāra ―saveă
(money)‖
Cyprus
pi-
Bedouin
Urfa
Raqqa
ǧā ad ǧā id
—
restricted
savayna tokka
―weă
shookă
hands‖
under
influence of
Greek
different
usage: l-ixpír
páyt ―theă bigă
house‖
—
Anatolian Group
Mossul
Daragözü
qaqa- k- nā kāqǝllǝk
yfattlūnu
‖I‘mă tellingă
bǝssōq
you‖
―they
are kībko ―they are
leading them crying‖
around the
market‖
—
sawa ǧōt ―toă
plough‖
sawa
mǝnāqāšā ―toă
quarrel‖
Hasköy
Mardin
ø
ø
īsī īdāra ―makeă
aăliving‖
īsīr qābūl ―ită
will
be
accepted‖
sawa
tǝlǝfōn ―to
make
a
telephone
call‖
article
is the article ǝl- is
usually
not only
used
used in subject sporadically
position: w kalb
ǧā ―andă theă dogă
came‖
G. Numerals
a.
Cardinal
numbers
no
changes
only
percentage: bil-mī
tis īn
―90%‖
Arabic
numerals used
up to 10 only,
rest is Greek
ašṛa w wa de
11
sab īn w sǝmǝn
78
> 100 only
Kurdish: ṣaṭṭ,
dūṣaṭṭ …hazār
b.
Ordinal
numbers
no
changes
awwil, tēni
rest Turkish
avel
Greek
no data,
probably
analogous
cardinal
numbers
rest
129
Sources: Aleppo: Behnstedt 1997:map 161: – Cilicia: Procházka 2002a:113, 197f., 201, 186
– Cyprus: Borg 1985:76, Borg 2004:279f, Borg 1985:145, 126, 129
– Urfa: Procházka 2003:83 – Raqqa: Behnstedt 1997:map 161 – Mossul: Jastrow 1979:47
–Daragözü: Jastrow 1973:50, 91, 101; Vocke & Waldner 1982:215f. (sawa…)
– Hasköy: Talay 2002:74, 79; Talay 2001:81, 81f. – Mardin: Grigore 2007:157-159 .
but
to
ašṛa sǝtte 16
no data, but
probably
analogous
to
cardinal numbers
Table 5 – Syntax
A. Analytic
accusative
130
B.
Adjectival
attributes
constructed
like
iḍāfa
C. Copula
Syrian Group
Aleppo
Cilicia
restricted very
frequent
Cyprus
very frequent
Bedouin Group
Urfa Raqqa Kuwayt
—
—
—
Anatolian Group
Mossul
Daragözü
very
no
frequent
Hasköy
no
Mardin
no
no
yes
no
no
no
no
no
no
no
no
—
—
interposed
—
copula:
náxni
naxni
mpsallín
―weă
areă
educated‖;ă áḏa o
xáyti ―thisă isă myă
brother‖
—
—
—
ǝmmi lbayt-i Dar-xāṣ
ana gbīr
―myă motheră isă šǝmde-ya ana ―I
―howă isă am great‖
at home‖
Hasköy?
‖
Sources: Cilicia: Procházka 2002a:158, 152 – Cyprus: Borg 1985:138; 134f. – Mossul: Jastrow 1979:49 – Daragözü: Jastrow 1973:40
– Hasköy: Talay 2001:78 – Mardin: Grigore 2007:290.
Table 6 A – Lexicon: Syrian Group
Percentage of loanwords (estimations except for Anatolia and Cilicia7):
low: < 5
medium: 5-10
high: 10-20
very high: > 20
131
percentage of loanwords
to go
to make
to see
to speak
good, pretty
hair
to ask
to understand
to fall
to finish
to hide oneself
Syrian Group
Aleppo
medium
rāḥ / yrūḥ
sawwa / sāwa
šāf / yšūf ǝše / yǝ ša
ḥaka
kwayyes
ša r
sa al
fǝhem
wa a / wu e
xalaṣ
xǝfi/staxfa staxba / sxabba
Cilicia
very high
rāḥ / yrūḥ
sawwa / ysawwi
šāf/yšūf qiši / yiqša
laqaš / tlaqwaš tqōnaš
kayyis
ǧimmi
sahal
ftaham
waqa nitir
hayya
xtafa
Sources: Aleppo: Barthélemy 1935-54 – Cilicia: Procházka 2002a – Cyprus: Borg 2004 s.r.
7
Vocke & Waldner 1982:XL - XLIV, Procházka 2002a:187f.
Cyprus
very high
rax / pirúx
sava / pisáy
kiš e / pkyákša a
xka
kaes
šaxve < ša afa
sael > sāyal
?
vak a
?
sxabba
Table 6 B – Lexicon: Bedouin and Anatolian Group
132
Bedouin Group
Urfa
ar-Raqqa
medium
low
šī
bī
māmin
mā bi
zahīd
šwayye
sōlaf
ḥača
sawwa
sawwa
anṭa
inṭa
percentage of loanwords
there is
there is not
a little
to speak
to make
to give
percentage of loanwords
to go
to make
to see
to speak
to ask
to understand
8
Anatolian Group
Mossul Mardin
medium high
rāḥ
rāḥ
sawwa
sawa
qaša
aṛa
ḥaka
ḥaka / štaġal
sa al
sa al
ftaham ftaham
Only used in present tense.
Daragözü
very high
rāḥ
sawa
ṭašš / ayyǝn
ǝštaġǝl
ǝstaġbǝr
ǝftahǝm
The lexical data available for the Bedouin dialects in question is
too weak for a depth analysis (Sources: Procházka 2003:85f.;
Behnstedt 1997: map 366, Cantineau 1936:32,
Behnstedt 1997:maps 367 and 363).
Hasköy
very high
mǝši - īmši / īmme8
sā
ādaš ītǝš
štaġal
?
?
Sources: Mossul: own data – Mardin: Vocke &
Waldner 1982, s.r. – Daragözü: Jastrow 1973:
53f, 72f. – Hasköy: Talay 2002, s.r.
Quelles nouvelles perspectives s’ouvrent avec l’exploration et
la description des dialectes arabes dits périphériques ?
Arlette Roth
CNRS, Paris
France
Ma contribution à la thématique de ce Colloque consacré aux variétés dites
périphériquesă deă l‘arabeă seă fondeă sură ună inventaireă nonă exhaustifă deă travauxă
réalisés au cours des trente dernières années et engage une réflexion critique sur
les traits proposés pour les identifier et les caractériser. La question qui se pose est
la suivante : « peut-on, au-delàă deă l‘étiquetteă ‗dialecteă périphérique‘,ă repéreră desă
problématiques et élaborer empiriquement une notion, à la fois cohérente et
complexe, qui rende compte de leur spécificité, en partant de traits définitoires
dégagés par certains auteurs ».ă Ouă formuléeă enă d‘autresă termes : « les dialectes
arabes périphériques ainsi diagnostiqués constituent-ils un groupe homogène ? »
1. Inventaire sélectif des sources considérées
Laă curiositéă etă l‘intérêtă poură lesă variétésă dialectalesă arabesă parlées aux
margesă deă l‘aireă arabophoneă sontă anciensă etă diversă aspectsă duă contactă ontă étéă
envisagés par le passé, privilégiant généralement les points de vue des influences
et du mélange linguistiques. Les métaphores de type géographique se sont
imposées progressivement dans les trois dernières décennies. Les études sur les
dialectes appréhendés comme périphériques ont été stimulées par la découverte et
laădescriptionădeăl‘arabeămaroniteădeăChypreăetăsaămiseăenărelationăavecăleămaltais,ă
pară l‘explorationă linguistique récente et systématique des variétés arabes de
Turquieăetăparălaărelanceădesăétudesăsurăl‘arabeăparléăauăKhorasan,ăenăOuzbékistană
etăenăAfghanistan.ăLaădécouverteăduăstatutăcomplexeădeă l‘arabeăparléăenă Afriqueă
subsaharienne centrale (Soudan, Tchad et Nigeria), sans parler de celle des
pidgins créoles à base arabe, complique et enrichit la problématique de la
périphérie dialectale en ce qui concerne les dialectes arabes de cette aire.
Sansă négligeră l‘apportă totalementă indispensableă enă matériauxă etă enă
analyses linguistiques livré par les monographies descriptives traditionnelles
portantă sură cesă dialectes,ă quiă élargissentă laă connaissanceă deă l‘aireă arabophoneă
contemporaine,ă monă attentionă s‘estă surtoută focaliséeă sură desă articlesă quiă ontă
cherché à mettre en évidence les particularités des dialectes désignés comme
133
périphériques. Ces études, je les ai principalement relevées dans les contributions
de dialectologues arabisants participant aux Conférences organisées par
l‘Associationă Internationaleă deă Dialectologie Arabe (désormais AIDA) et dans
quelques autres publications apparentées. Ainsi, dans les Actes de AIDA parus en
2000 et 2004, une section spéciale a-t-elle été réservée aux études portant sur
l‘arabeăpériphérique,ărespectivementăsousăleătitreă« Peripheral dialects » (AIDA 3)
et « Studies on peripheral dialects and contact phenomena »ă (AIDAă 5).ă Ilă s‘agită
doncă d‘ună échantillonă d‘étudesă exprimantă laă visionă d‘ună cercleă d‘arabisantsă
dialectologuesă engagésă concrètementă dansă l‘étudeă linguistiqueă desă dialectes en
question1. Même si cette discussion envisage, dans un premier temps, la prise en
compte des principaux facteurs externes qui définissent des variétés dites
périphériques, et plus précisément le facteur géographique, les travaux de ces
spécialistes font une grande place au linguistique et visent à établir des
discriminantsăpropresăàăfonderăcetteăcatégorisation.ăDansămaăsélection,ăj‘aiăretenuă
les études qui mentionnent au moins trois des traits supposés caractériser les
variétés dites périphériques.
En 1973, dans un article qui traite de la variation, David Cohen, qui a
souventăinsistéădansăsonăenseignementăsociolinguistiqueăpourăqueăl‘onăattribueădesă
qualificatifsă auxă groupesă deă locuteursă plutôtă qu‘auxă langues,ă mentionneă (p.ă 239)ă
des « groupesă d‘arabophones extrapériphériques, coupés du centre de
l‘arabophonie », pratiquant des variétés « investies par des langues de structure
totalement différente » ; et il cite, par exemple, les Maronites de Kormakiti en
Chypreăetălesăgroupesăarabophonesăd‘Asieăcentrale.ăEnă1994,ădansă l‘Introductionă
aux Actes des premières journées internationales de dialectologie arabe de Paris,
àăproposădeăl‘importanceăduăstatutădansăleătraitementăsociolinguistiqueădesăvariétésă
dialectales, il écrit (p. 16) « Les statuts diffèrent pour un groupe périphérique,
vivant aux confins, un groupe extrapériphérique inclus dans un domaine
linguistique hétérogène ou un groupe en expansion sur un terrain en voie de
conquête ». Ces distinctions déterminent bien évidemment des dynamiques
différentes. Bienă qu‘ilă neă leă préciseă pas,ă onă peută imagineră qu‘ilă s‘agită poură leă
premierăcasădeăl‘arabeăparléăenăMauritanieăouăenăArabieăduăSud,ăpourăleădeuxièmeă
deăl‘arabeăparléăenăTurquieăouăenăAsieăcentrale,ăetăpourăleătroisième,ădeăl‘arabeăenă
usage, voire en expansion, en Afrique subsaharienne centrale. On retiendra ici
pour notre propos, dans la formulation donnée à la combinaison : groupe de
locuteurs et territoires, la distinction nécessaire entre « groupe périphérique »
versus « groupe extrapériphérique ».
Celaăconcerneăuneăbonneăvingtaineădeăspécialistesăetăuneăcinquantaineădeăcontributions.ăIlăn‘estă
pas indifférent que ces spécialistes soient des linguistes de terrain.
1
134
En 1977, dans un compte rendu proposant une reconstruction du système
deăl‘arabeămaroniteădeăChypreăàăpartirădesăfaitsălivrésăparăuneăpremièreădescriptionă
duă parleră évaluéeă deă façonă trèsă critique,ă Ottoă Jastrowă utiliseă l‘appellationă « die
äussersten Randregionen des arabischen Sprachraums », qui englobe aussi
« l‘Afriqueă centrale,ă l‘Ouzbékistan,ă leă Sud-Est anatolien, Malte ou le Sud de la
Péninsule arabique ».ă Cetă ensembleă àă laă margeă extrêmeă s‘opposeă auxă grandesă
régionsă arabophonesă formantă ună ensembleă d‘ună seul tenant (zusammenhängende
Regionen). Cet inventaire ne distingue pas encore entre ce qui est externe aux
frontièresă duă mondeă arabophoneă etă ceă quiă estă périphériqueă àă l‘intérieură deă l‘aireă
(par exemple, le Sud de la Péninsule arabique dans la citation).
Dans la monographie que O. Jastrow a consacrée aux dialectes arabes
mésopotamiens de type qəltu (1978),ă lesă variétésă deă l‘Anatolieă duă Sud-Est sont
désignées comme Randdialekte. La distribution des dialectes du type qəltu se fait
àă l‘intérieură etă àă l‘extérieură deă l‘aireă arabeă proprementă dite.ă Partantă deă laă
qualification de Randdialekte,ă l‘auteură leură adjointă l‘arabeă enă Ouzbékistan,ă qu‘ilă
présente comme une Sprachinsel,ă l‘arabeă maroniteădeăChypre,ă leă maltais,ă l‘arabeă
au Tchad. Á première vue, chacun des composants de cet ensemble constitue
géographiquement et linguistiquement un cas particulier, le dénominateur
communăétantăuneăpositionăexcentréeăparărapportăàăl‘aireăarabophone.
En 1980, le Handbuch ajoute à cet ensemble de variétés périphériques
l‘arabeă parléă auă Khuzistan, en Afghanistan, une rangée de Sprachinseln en
Turquie,ăl‘EstăduăSoudanăetă« certaines parties » du Tchad. Ces localisations sont
évidemmentă tributairesă deă laă documentationă disponibleă àă l‘époqueă etă s‘ilă estă faită
allusionăàăl‘arabeăparléădansăl‘Afrique centrale subsaharienne, le Handbuch ne lui
consacre pas de chapitre établissant ses caractéristiques, comme il le fait pour les
autresă grandesă régionsă deă l‘aireă arabophone.ă Lesă deuxă déterminationsă deă typeă
géographique qui interviennent sont toujours Randdialekte et Sprachinseln.
En 1983, dans la revue Zeitschrift für arabische Linguistik, O. Jastrow
présenteălaăsituationădeăl‘arabeăenăTurquieăetădéfinitătroisăaires,ădifférenciéesădansă
l‘espaceă etă dansă laă caractérisationă dialectale,ă auxquellesă estă appliquée la
désignation Sprachinseln.ă Ilă s‘agită d‘Estă enă Ouestă de :ă a)ă l‘ensembleă Diyarbakır,
Siirt et Mardin, de type qəltu, relié aux dialectes qəltu de la Mésopotamie, b) de la
provinceă d‘Urfa,ă avecă ună dialecteă deă typeă bédouină gilit, relié à la péninsule
arabique, c) de la province du Hatay (Antioche), dont les dialectes sont
historiquementă etă linguistiquementă enă relationă avecă l‘ensembleă syro-libanopalestinien.
En 1990 se tient à Madrid le « Congreso internacional sobre interferencias
linguisticas arabo-romances y paralelos extra-iberos » dont Les Actes sont parus
en 1994 à Saragosse. On y relève en particulier :
135
1. la première tentative de systématisation des traits constitutifs de la
notionădeăpériphérieădansă leăchampădeă laădialectologieăarabe,ăsousălaăplumeăd‘A.ă
Borg, spécialiste du maltaisă etă deă l‘arabeă maroniteă deă Chypre,ă deuxă dialectesă
périphériques insulaires,
2. une relation établie entre la périphérie dialectale arabe et la formation de
pidgins et créoles arabes en Afrique subsaharienne centrale, sous la plume du
regretté Alan S. Kaye,
3. la première mise en relation historique et linguistique argumentée de
deuxăvariétésăarabesăpériphériques,ăl‘arabeămaroniteădeăChypreăetăleămaltais,ăavecă
une comparaison systématique des facteurs externes et internes qui ont donné à
ces deux projectionsă trèsă anciennesă deă l‘arabeă enă Méditerranéeă leursă profilsă
spécifiques (A. Borg).
On retrouve également dans ces Actes des contributions qui portent sur les
dialectes qəltu constituant des Sprachinseln en territoires turcophone et
kurdophone (O. Jastrow) etă sură l‘arabeă d‘Afghanistan,ă définiă commeă uneă fringe
variety of Arabic (B. Ingham),ăl‘accentăportantăprincipalement,ădansăcetteădernièreă
contribution, sur les effets linguistiques du contact et de la pratique du
plurilinguisme. Une mise en garde, à proposă deă l‘arabeă deă Mauritanie,ă inviteă àă
distinguer entre périphérie au sens géographique et au sens linguistique (C. TaineCheikh, « Le h}assāniyya de Mauritanie : un dialecte non-marginal de la
périphérie »)2.
En 1995, dans les Actes de la 2ème ConférenceădeăAIDA,ăonărelèveăl‘entréeă
en scène de nouveaux dialectes arabes parlés en Turquie, à côté de ceux déjà
repérés.ăIlăs‘agitădesădialectes de Cilicie (S. Procházka). O. Jastrow, à propos de
l‘arabeă d‘Ouzbékistan,ă fournită auxă dialectologuesă arabisantsă uneă manièreă deă
feuilleă deă routeă poură l‘exploitationă linguistiqueă desă matériauxă livrésă pară lesă
dialectes périphériques dans une perspective résolument historique et comparative
interneă àă l‘arabe.ă Elleă privilégieă laă miseă enă évidenceă duă noyauă arabeă etă desă
affiliationsădesădialectes.ăElleăsouligneălaăpartăd‘archaïsmesăetăd‘innovations,ădansă
une proportion qui semble typique des dialectes périphériques. Pour le maltais, M.
Mifsudăattireăl‘attentionăsurăleăfaităqueăsiăleălexiqueădeăbaseădeălaăvieăquotidienneă
estă d‘origineă arabe,ă ceă traită n‘estă pasă représentatifă dansă l‘ensembleă duă systèmeă
linguistique maltais contemporain. Il propose une analyse fonctionnelle des
constituantsăarabesăàălaăbaseădeăcetteălangue,ăd‘uneăpart,ăetădesăcomposantesănonă
sémitiques,ăd‘autreăpart,ămettantăenăreliefăuneă forteăsynergieăentreălesădeuxătypesă
de morphologie. Ce faisant, il identifie les forces qui ont assuré linguistiquement
Le questionnement sur le bien-fondé deă l‘équivalenceă « périphérie géographique » et
« marginalité linguistique » est repris par deux participants au présent Colloque, C. Taine-Cheikh
et S. Procházka.
2
136
laă survieă etă laă continuitéă duă maltais,ă àă savoiră l‘efficienceă sémantiqueă (semantic
usefulness)ă etă l‘adaptabilitéă formelleă (formal adaptability). Son étude institue un
axe de recherche prometteur.
En 2000, dans les Actes de la 3ème Conférence de AIDA, sousă l‘intituléă
« Arabe périphérique »,ăonărelèveăuneăcontributionăsurăl‘arabeădeăQashqadaryaă(G.ă
Chikovani),ă laă présentationă d‘ună projetă d‘explorationă intituléă « Border Area
Arabic »ăciblantă l‘Asieă centraleă(B.ăIsaksson).ăL‘arabeădeă Cilicieă(S.ăProcházka),
ună témoignageă écrită d‘arabeă chyprioteă duă XVIIème siècle (J. Lentin) et le Juba
Arabic (C. Miller) sont également représentés. J. Ferrando pose le problème de
l‘arabeă andalouă deă laă périodeă post-islamiqueă dansă leă Nordă deă l‘Espagneă etă
l‘identifie,ă dansă cetă article,ă nonă pasă commeă deă l‘arabeă périphériqueă maisă commeă
du moyen arabe3.
En 2002, dans les Actes de la 4ème Conférence de AIDA, pour la Turquie,
desă complémentsă d‘informationă portentă sură lesă variétésă d‘Adanaă etă deă Mersină
(S. Procházka)ăetăcellesăd‘Antiocheăetădeăsaărégionă(W.ăArnold).ăL‘accentăestămisă
sur le contact en synchronie, avec des études analysant divers types de codeswitching. Des contributions ciblant ponctuellement des faits linguistiques ou
sociolinguistiques concernent des dialectes précédemment déjà étiquetés
périphériques,ăcommeăleăJubaăArabică(C.ăMiller),ăl‘arabeădeăKormakitiăenăChypreă
(A.ăRoth).ăUnăisolatălinguistiqueăd‘apparitionătrèsărécenteăs‘introduităàălaămargeădeă
l‘inventaire,ă soită uneă variétéă parléeă dansă laă Républiqueă deă Djiboutiă (M.-Cl.
Simeone-Senelle).ă Onă voită dansă cetteă étudeă deă cas,ă commentă l‘homogénéitéă
ethnique et sociale du groupe et la brièveté du contact – par comparaison avec les
autres variétés – limitentăl‘impactădeăl‘afarăchezălesălocuteursăbilingues.
En 2003, les Actes de la 5ème Conférence de AIDA font la part belle à
l‘arabeă parléă enă Asieă mineureă etă centraleă (G.ă Chikovani,ă B.ă Ingham).ă
L‘explorationăsystématiqueădeăl‘arabeăenăTurquieăcontinue,ămenéeăprincipalementă
parăl‘écoleădeădialectologieăallemandeă(etăautrichienne !).ăS‘ajoutentăauxădialectesă
précédemment repérés par O. Jastrow,ădesădonnéesăinéditesăsurăl‘arabeăduădistrictă
de Sason présenté comme isolated [Arabic] border et qualifié de variété de
minoritéă (B.ă Isaksson),ă sură l‘arabeă deă Təllo (A. Lahdo), sur celui de Hasköy
(S. Talay),ă pointă leă plusă septentrională deă laă présenceă deă l‘arabe,ă variétéă
linguistiquement proche du dialecte de Daragözü décrit par O. Jastrow (1973), et,
enfin, sur la variété arabe de type bédouin shāwi deă laă plaineă d‘Urfaă
(S. Procházka). À propos de cette variété très comparable à ses congénères
Dansă J.ă Owensă (éd.)ă (2000),ă I.ă Ferrandoă traiteă duă casă deă l‘arabeă deă laă minoritéă chrétienne
mozarabeă deă Tolède,ă dontă l‘emploiă seă perpétuaă pendantă deuxă sièclesă aprèsă laă chuteă deă laă ville.ă
Perte de souveraineté et passage au statut de langue minoritaire sont évidemment liés. Ce qui est
intéressant,ăc‘estăqueădansălesădeuxăcasăétudiésăparăI. Ferrando,ăl‘usageăoralăaădisparuăavantăceluiă
deăl‘arabeăécrit,ăqueăleăgroupeădeălocuteursăfûtămusulmanăouăchrétien.
3
137
bédouinsă syriensă toută prochesă ouă arabiquesă plusă éloignés,ă l‘auteură déploreă deă
façon concise et humoristique : « Ină contrastă toă Jastrow‘să discoveryă ofă theă
Anatolian qəltu-dialects, nothing radically new is to be learned from the Bedouin
dialects of the same region ».ă Affligeanteă conformitéă pară rapportă àă l‘exotismeă
excitant des dialectes de sédentaires extraterritoriaux ! Toute plaisanterie mise à
part,ăcetteăobservationăattireăl‘attentionăsurăleămoindre degré de perméabilité et de
réactivité des dialectes de type bédouin au contact de langues non sémitiques. On
trouve sur ce point dans le champ de la périphérie des exemples contradictoires et
ce thème exige un traitement particulier.
Leătableauădeă l‘arabe parlé en Turquie ainsi complété se présente comme
ună ‗microcosmeă périphériqueă trèsă diversifié‘ 4 ,ă oùă l‘onă peută distingueră ună
continuumă deă ‗périphérique‘,ă géographiquementă parlant,ă etă ună continuumă deă
‗marginalité‘,ă linguistiquementă parlant.ă Lesă auteursă proposent des listes de
discriminants spécifiques pour comparer entre eux les sous-groupes participant de
chaque ensemble, à côté de ceux déjà connus qui caractérisent les dialectes de
type qəltu. Les problématiques de la dialectologie historique comparée, de la
typologie et de la périphérie dialectales, des contacts en diachronie et en
synchronieăavecădesălanguesănonăsémitiquesăs‘entrecroisent.ă
Dans les Actes de la 6ème Conférence de AÏDA, figurent des sujets ciblés
concernant des variétés périphériques déjà inscritesă dansă l‘inventaireă (George
Grigoreă poură l‘arabeă deă Mardin,ă A.ă Rothă poură leă parleră arabeă deă Kormakiti,ă
Martin R. Zamitte pour le maltais).
On peut faire quelques observations à propos de cet inventaire.
Desă flottementsă ontă existéă auă débută pară rapportă àă l‘inclusionă deă certainsă
dialectes arabes dans le groupe des variétés dites périphériques (Mauritanie,
Yémen,ă Arabieă duă Sud,ă pară exemple)ă quandă laă distinctionă n‘étaită pasă clairementă
faite entre périphérie interne et périphérie située au-delàă deă laă frontièreă deă l‘aireă
arabe, ou, dans les termes choisis par D. Cohen, dans la désignation des groupes
de locuteurs, entre « extrapériphérique » et « périphérique ».
L‘intégrationă deă certaines variétés arabes africaines a aussi connu des
flottements,ă dansă laă mesureă oùă seulsă certainsă usagesă deă l‘arabeă pară certainesă
populations peuvent relever de la périphérie. Pour le Tchad, ces hésitations sont
dues, semble-t-il,ă auă faită queă l‘arabeă seă présente à la fois comme langue
maternelle de tribus arabes ou de groupes linguistiquement arabisés
(principalement au nord du 13ème parallèle) et aussi comme langue véhiculaire sur
l‘ensembleăduăterritoireătchadienăchezădesăpopulationsăpratiquantăcommeăpremièreă
langueădesăvernaculairesăafricains.ăL‘usageăvéhiculaireăestăutiliséăpratiquementăetă
4
Dans certaines variétés on relève des usages différenciés entre sédentaires, sédentarisés et
bédouins,ăd‘uneăpart,ăetăentreăalaouites,ămusulmansăetăchrétiens,ăd‘autreăpart.
138
officiellementăcommeălangueădeăl‘administrationăetădeălaăcommunicationăpolitiqueă
(cf. P. Jullien de Pommerol 1997). Le Soudan présente aussi une situation
complexe.ăIlăs‘agităd‘unăpaysăoùăl‘arabeăcoexisteăavecădesălanguesăafricainesămaisă
dontă l‘arabeă estă laă langueă officielle.ă L‘arabeă parléă dansă leă Sudă aă ună statută etă ună
fonctionnementă trèsă particuliers.ă L‘existence,ă dansă cetteă zone,ă deă pidginsă etă
créoles à base lexicale arabe rend le tableau encore plus complexe. Les travaux de
J.ă Owens,ă sură l‘arabeă africaină n‘excluentă pasă leă traită périphériqueă maisă leă
subordonne apparemment à la perspective « arabe de minorité » et « arabe
minoré » (ou « arabe dominé ») (cf. Arabic as a minority language, 2000).
Onă remarqueă queă dansă laă dernièreă décennie,ă poură uneă majoritéă d‘auteursă
cités,ăleăfacteurăgéographiqueăneăseăprésenteăpasăseulămaisăfaităintervenirăd‘autresă
interprétants. On verra ci-dessous quelles approches sont parallèlement envisagées
pour traiter des dialectes à la marge et quelles orientations thématiques ont été
esquissées.
2. L’arabe dit périphérique : de la métaphore à l’élaboration d’une
notion complexe ?
Laă métaphoreă géographiqueă deă laă périphérieă s‘estă imposéeă poură qualifieră
desă dialectesă situésă àă laă margeă deă l‘aireă arabophone,ă au-delà de la frontière
linguistique, par opposition aux dialectes assignés à un centre (mainland). Aux
dialectes dits périphériques sont opposés des dialectes centraux (mainstream),
parfois aussi nommésă ‗normaux‘ă (entreă guillemets).ă Laă différenceă s‘exprimeraită
enă termesă d‘écartă pară rapportă auxă structuresă sémitiquesă originelles.ă Leă cœură deă
l‘aireă arabophoneă devraită théoriquementă êtreă représentéă àă égalitéă pară lesă paysă
officiellement et politiquement dits arabes, mais en réalité, traditionnellement, ce
centre symbolique est localisé de façon préférentielle au Proche ou au MoyenOrient. On pourrait dire par boutade que le repérage des dialectes situés aux
marges est paradoxalement plus aisé que la déterminationă d‘ună centreă unique.ă Ilă
estă certaină queă lesă grandesă régionsă ouă subdivisionsă deă l‘aireă arabophoneă
définissent leurs propres centres par rapport auxquels se situent des périphéries.
Les désignations relevées chez les auteurs (périphérie, marges,
Randdialekte, border area, fringe variety)ădénotentă l‘éloignement,ătandisăqueă lesă
termes Sprachinsel (speech island) ou enclave linguistique signalent la
discontinuitéă dansă l‘occupationă duă territoire 5 . Les deux représentations
(éloignement et discontinuité) impliquentă l‘isolement.ă Auă filă desă (re)découvertesă
Onăpeutăremarquerăqueălaădispersionăetăl‘occupationădiscontinueăontălongtempsăcaractériséăaussiă
laădistributionădeăgroupesăetădeătypesădialectauxăàăl‘intérieurădeăl‘aireăarabophone,ăenărelationăavecă
des facteurs socio-historiques. Cf., par exemple, la distribution géographique des dialectes dits préhilaliens, villageois et hilaliens au Maghreb.
5
139
et des descriptions, des rapprochements entre variétés dialectales à la marge
s‘esquissentă etă ună ensembleă deă caractéristiquesă s‘érigeă empiriquementă enă notionă
plus complexe, avec un contenu dépassant la dimension purement géographique.
Commeă onă l‘aă déjàă signalé,ă A.ă Borgă enă aă leă premieră systématiséă l‘expressionă
(Borgă 1985,ă 1994,ă 2004).ă Lesă traitsă définitoiresă qu‘ilă aă sélectionnésă ontă presqueă
tousăétéămentionnésăparăd‘autresăauteurs,ăsansăêtreăsystématisés.ăAinsiăsont cités :
1.ăl‘implantationătrèsăancienneăliéeăauxăpériodesăhistoriquesădeăl‘expansionă
islamique (du 7ème au 15ème siècles),ă2.ăl‘éloignementăgéographique,ăet,ăsurtout,ă3.ă
la distanciation culturelle 6,ă 4.ă l‘absenceă deă contactă avecă l‘arabeă standardă etă lesă
autres dialectes arabes centraux (mainstream) contemporains, 5. un
fonctionnement essentiellement oral, 6. un contact linguistique au long cours avec
des langues non sémitiques, générant le bilinguisme, voire le plurilinguisme, et la
transformation progressive et profonde des structures linguistiques. A. Borg
insisteă sură ceă qu‘ilă nomme :ă l‘acculturationă linguistiqueă etă leă critèreă
d‘inintelligibilité.ă
C‘estăleălongătermeădéterminantălesămodalitésăetălaănatureăduăcontactăavecă
des langues non sémitiques et les formes de bilinguisme qui est fortement
souligné et qui incite A. Borg (1994) à vouloir maintenir distincts le cas des
dialectes périphériques et celui des pidgins et créoles à base arabe. On a pu
évoquer pour ces derniers une modalité de contact-choc distincteă deă l‘impactă auă
long cours. La question de la relation entre dialectes arabes périphériques et
pidgins et créoles à base arabe reste encore ouverte.
Certains des traits cités pour définir empiriquement un dialecte
périphérique ont une valeur quasi absolue. Il en est ainsi : a) de la coupure
culturelleă d‘avecă leă mondeă arabophone,ă b)ă duă fonctionnementă purementă orală desă
variétésă concernées,ă doncă deă l‘absenceă deă l‘arabeă standardă etă duă coupleă
diglossique traditionnel7, c) du contact avec des langues non sémitiques et de la
pratique généralisée du bilinguisme.
La très grande majorité des variétés périphériques contemporaines figurant
dansă l‘inventaireă obéissentă àă cesă troisă critères.ă Seulă leă maltaisă quiă fonctionneă àă
l‘orală etă àă l‘écrită faită exceptionă sură le pointă b),ă aprèsă qu‘aă étéă entrepriseă dèsă leă
15ème siècleă l‘élaborationă deă sonă propreă registreă écrită littéraire.ă Leă statută
d‘autonomieădeăcertainesă variétésăpériphériquesăaăpuăêtreăaffirméăouărevendiqué 8
Leăfacteurăd‘éloignementăgéographiqueăn‘aăpasăleămêmeăpoidsăqueăceluiădeălaăcoupureăculturelle.ă
Cette circonstance est diversementă appréciéeă etă commentée.ă Alorsă queă d‘aucunsă yă voientă ună
facteurăpositifăpourăl‘étudeădeădialectesăainsiăpréservésădeăl‘influenceăjugéeănivelanteădeălaălangueă
standard,ăd‘autresăyăvoientăunăfacteurăaggravantădeădévalorisationămenantăàăl‘extinction.ăLesădeux
observations sont vraies au plan où elles se situent.
8
Ainsi,ăauănomădeăl‘écart,ăG.ăChikovaniăécrit-ilă«ă…theăArabicădialectsăofăCentralăAsiaădifferăsoă
greatly from all known Arabic dialects (as regards both external influences and their main semitic
6
7
140
etă leă sujetă faită l‘objetă deă controversesă entreă tenants du primat de la parenté
génétiqueăetădeăl‘unitéădeăl‘arabeăetălinguistesăseăréclamantăstrictementăduăprimată
duă fonctionnementă linguistique,ă horsă deă toută contexte.ă C‘estă égalementă uneă
question ouverte.
On ajoutera que le statut de minorité pour les groupes de locuteurs, et de
dialecte minoré ou dominé en référence aux usages linguistiques, a été mentionné
cursivementăparăplusieursăauteurs.ăIlăestăintéressantăd‘observerăqueădansăl‘ouvrageă
édité par J. Owens, trois types de minorités et partant de variétés arabes minorées
sont cités. Les territoires de celles qui sont désignées comme « anciennes et en
survie » (op. cit. : Introduction p. 6), coïncident, à quelques exceptions près, avec
ceux des variétés figurant dans notre inventaire et il est important de souligner que
toutes les variétés périphériques relevées sont des variétés parlées par des
minorités. Ce critère, au plan synchronique et sociolinguistique, se présente
également comme un critère qui fonctionne de façon absolue. Extrapériphérie et
statut minoritaire vont de pair. Il semble donc normal que les inventaires
respectifs se recoupent9.
Leă facteură deă l‘éloignementă géographiqueă etă celuiă deă l‘anciennetéă deă
l‘implantationă peuventă varier.ă L‘arabeă chyprioteă aă étéă implantéă àă proximitéă desă
côtes syriennes etăc‘estălaăruptureăsocio-historique et culturelle qui est à la base de
sa marginalité structurale et fonctionnelle. Les dialectes arabes de Turquie, dans
leură variété,ă sontă limitrophesă ouă plusă ouă moinsă éloignésă deă l‘aireă arabophone.ă
Certainsăsontăd‘implantationătrèsăancienneăetăd‘autresăsontăliésăàădesădéplacementsă
de populations récents10. Mais la rupture historique là aussi a pleinement joué son
rôle.ă Enfin,ă leă caractèreă d‘inintelligibilitéă meă sembleă personnellementă aussiă ună
critère relatif.
body) that they form an altogether independent group, occupying, like Maltese, a special place
among other known dialects of the Arabic language » (2004 : 187). Statut spécial ou autonomie,
toută leă problèmeă estă là.ă L‘analyseă etă laă comparaisonă desă données linguistiques, la mesure,
l‘appréciationăetălaăpondérationădesăindicesărelevésăquiăpermettraientăpeut-être de trancher au plan
linguistiqueăn‘enăsontăqu‘àăleurădébut.
9
Ună auteură commeă I.ă Ferrandoă aă enă quelqueă sorteă testéă poură l‘arabeă andalouă – échantillon
historique – les trois perspectives offertes :ăleărattachementăàăl‘arabeăpériphérique,ăl‘identificationă
au moyen arabe, la prise en compte déterminante du critère « arabe de minorité ». Il semblerait que
ceă soită leă rattachementă auă moyenă arabeă quiă l‘emporteă dansă l‘étată actuelă duă dépouillementă deă laă
documentation. De façon plus générale, la relation entre moyen arabe et (extra)périphérie reste à
explorer.
10
On peut se demander si, du point de vue linguistique, la nature et la portée des écarts et des
innovations relevées dans les différents dialectes arabes de Turquie sont comparables entre elles,
quelleă queă soită l‘anciennetéă deă l‘implantationă etă siă ellesă seă rapportentă plutôtă auă processusă deă
marginalisationăouăauxămodèlesăd‘évolutionăproposésăouăimposésăparăles langues en contact. On a
vu,ăavecăl‘exempleăduădialecteădeătypeăbédouinădeălaăplaineăd‘Urfa,ăqueălaăcatégorieăarabeăbédouină
peut limiter les effets du contact.
141
Enfin, dans laăperspectiveăhistoriqueăetăcomparativeăinterneăàăl‘histoireădesă
dialectesă arabesă etă plusă généralementă deă laă langueă arabe,ă l‘identificationă commeă
relic area desă variétésă périphériquesă d‘implantationă ancienneă fondeă laă valeură deă
leur apport dans la perspective diachronique.
3. Sous l’enseigne du contact
Onă aă évoquéă jusqu‘ici,ă parmiă lesă traitsă constitutifsă deă laă notionă deă
périphérie,ă ceuxă quiă renvoientă plusă spécifiquementă àă l‘espace :ă l‘éloignement,ă laă
discontinuitéă etă l‘isolement.ă Maisă laă géographieă enă tant que discipline a aussi
vocationăàăêtreă―humaine‖ăetăăpériphérieăetăcontactăsontăindissociablementăliés.ăIlă
convientă doncă d‘examineră rapidementă quelsă sontă lesă thèmesă conjointementă
abordésăparălesăauteurs,ăqu‘ilsărelientăplusăspécifiquementăàălaănotionădeăcontact.
Deuxă façonsă d‘envisageră leă contactă ontă étéă relevées : la première
s‘intéresseăenăprioritéăauxătracesăqueăleăcontactădansăleălongătermeăinscritădansălesă
structures linguistiques ;ă elleă metă enă évidence,ă commeă onă l‘aă déjàă rappelé,ă laă
coexistenceăd‘archaïsmesăetăd‘innovations.ăLaăsecondeăprivilégieăl‘observationăduă
contact sur le vif en synchronie, dans sa dynamique de transformation,
d‘expansionă ouă deă récession.ă Avecă lesă thèmesă qu‘ilsă ontă reliésă auă contact,ă lesă
auteurs ont évoqué un ensemble de points qui suggère des situations comparables
et des évolutions relativement convergentes.
Áă proposă deă l‘oppositionă expansion vs récession et du binôme fatal
récession —> extinction11, le processus est illustré par le syndrome des 4 (ou 3)
générations,ădansă l‘acquisitionăetălaăperte.ăLaăgénérationăestăl‘unitéădeămesureăduă
changement dans le temps court ou en période de crise 12 .ă Dansă l‘expansion
africaineă illustréeă pară leă passageă I)ă deă l‘emploiă ponctuelă etă restreintă deă l‘arabeă
(marché,ă armée)ă àă l‘usageă commeă II)ă langue véhiculaire dans des emplois en
extensionă ouă àă l‘usageă commeă III)ă langueă maternelle,ă d‘emploiă exclusifă ouă non,ă
cela se réduit à trois générations (Abbéché ou Sud Soudan). En récession, S.
Procházkaă fournită desă exemplesă poură l‘arabeă deă Cilicie : le grand-père parle
uniquementă l‘arabe,ă leă père connaît bien le turc, ego connaît mieux le turc que
l‘arabe,ă sesă enfantsă neă connaissentă queă leă turc.ă Àă Mardin,ă uneă autreă formuleă
expliciteă lesă statutsă respectifsă desă languesă enă contactă etă laă ‗dévalorisation‘ă deă
11
Certains auteurs nuancent le côté irréversible de la récession et parlent de « survie » (B. Ingham)
ou de « résistance »ă(A.ăRoth).ăLeăprincipeădeăl‘accommodationăouădeălaămiseăenăconvergenceădesă
systèmes sémitiques et non-sémitiques (M. Mifsud) peut linguistiquement sinon assurer la survie
du moins ajourner la disparition. Celle-ci dépend en dernier ressort exclusivement des facteurs
externes.
12
Laăcriseădansăleăcontactăauălongăcoursăsurvientăprincipalementăavecăl‘irruptionădeălaămodernité.ă
L‘arabeă neă peută plusă assureră l‘expressionă desă nouveauxă besoinsă deă communication,ă ilă estă
dévalorisé et ne se transmet plus aux enfants (par exemple, A. Lahdo 2003).
142
l‘arabe : ana tərki ana, abūy kərdi-we, əmmi ‗arabiyye-ye « je suis turc, mon père
est kurde, ma mère est arabe » (George Grigore).
En ce qui concerne le bilinguisme, dans le contact au long cours,
l‘hypothèseă d‘ună bilinguismeă ‗équilibré‘ă ouă duă moinsă ‗stabilisé‘ă pendantă une
longueă périodeă aă étéă avancéeă poură l‘arabeă deă Kormakiti.ă L‘arabeă etă leă grecă
chypriote sont supposés avoir couvert en gros les mêmes emplois pendant des
sièclesăetăleursăstatutsădevaientăêtreăplusăouămoinsăéquivalents,ăjusqu‘àăl‘irruptionă
de la modernité,ălaămultiplicationădesăcontactsăetăleădéveloppementădeăl‘école.
Au thème de la récession ont été associés celui de la langue en péril
(endangered language) et de l‘extinction, celui de la langue de minorité, celui de
l‘identité ou du lien identitaire que symbolise ou non la pratique du dialecte
périphérique, celui de la dévalorisation qui entraîne la rupture de la transmission.
Sur le plan proprement linguistique, des études ponctuelles ont porté sur la
réduction ou la simplification relative des structuresă formellesă deă l‘arabe.ă
Toujours au plan linguistique, un autre thème majeur lié aux effets du contact
concerne les structures, les fonctionnements et la réactivité des constituants arabes.
Plusieursă auteursă ontă abordéă l‘étudeă deă laă stratégieă deă coexistence des systèmes,
parlantă ainsiă d‘accommodation ou de mise en convergence par transferts de
normes (ainsi D. Cohen, M. Mifsud, A. Borg).ăCeăthème,ădeăquelqueă façonăqu‘ilă
ait été abordé et nommé, a donné lieu à des études et à des propositions
prometteuses (A. Borg,ă poură leă systèmeă phonologiqueă deă l‘arabeă maronite,ă M.ă
Mifsud pour le fonctionnement morphologique du maltais). On peut aussi
considéreră commeă relevantă deă l‘accommodation,ă lesă stratégiesă d‘arabisationă desă
emprunts (M. Mifsud, G. Grigore). De larges et fructueuses perspectives de
recherche sur le thème de la mise en convergence sont ouvertes.
Pourăl‘exploitationăetălaămiseăenăvaleurădesămatériauxălinguistiquesăfournisă
par les dialectes périphériques, diverses orientations ont été explorées. On a
évoqué ci-dessus (p. 03) les options retenues par O. Jastrow et par M. Mifsud
(AIDAă 2).ă A.ă Borgă s‘inscrită dansă leă sillonă tracéă pară O.ă Jastrow,ă àă laă foisă dansă laă
perspective historique et la quête des affiliations par rapport au groupe des
dialectes qəltu,ămaisălaăréflexionăqu‘ilăproposeăconjointementăsurăl‘arabeămaroniteă
deă Chypreă considéréă commeă isolată linguistiqueă méditerranéenă ouvreă d‘autresă
perspectivesăquiăfontăintervenirăl‘anthropologie.ăS. Procházka (1999) expérimente
une autre voie, en spécifiant les facteurs internes et externes qui déterminent
l‘extinctionăprévisibleădesălanguesămenacées.ăIlălesăsoumet,ăd‘uneăpart,ăàăl‘auneădeă
la « typologie géographique », modèle construit par John Edwards en 1992 et les
organise,ăd‘autreăpart,ăenăréférenceăà une typologie des facteurs de mise en danger
(« a typology of language endangerment »)ăarticuléeăsurăleăconceptăd‘écologie,ăenă
référence à E. Haugen et ses successeurs (cf. op. cit. Table 1). Cette construction
articulée joue sur la combinaison du géographique et du social. En tant que
143
construction, elle transcende sur ce point précis le côté descriptif et énumératif
qu‘onă peută releveră dansă lesă associationsă proposéesă dansă notreă inventaire.ă Elleă
faciliteălaăcomparaisonăglobaleăavecăd‘autresălanguesămenacées.
Conclusion
Siă l‘onă enă revientă auxă deuxă questionsă poséesă dansă l‘introduction,ăonă peută
dire brièvement que les critères, de valeur absolue ou relative, dégagés par les
différentsă auteursă s‘appliquentă àă laă majoritéă desă dialectesă désignésă commeă
périphériques. Ces critères sont plus descriptifs que conceptuels, en tout cas moins
conceptualisés que dans les modèles translinguistiques. Ils contribuent
incontestablement,ă enă l‘état,ă àă uneă catégorisationă interneă àă l‘arabe,ă quiă intègreă laă
perspective historique et la tient pour indispensable. Les dialectes posés comme
périphériquesăontădeănombreuxătraitsăcommuns,ămêmeăs‘ilsăneăconstituentăpasăuneă
classe parfaitement homogène. Il reste à déterminer à quel niveau se situe les
convergences linguistiques et quelle est leur ampleur. Tout le matériau
linguistique proposé dans les études retenues ne pouvait apparaître dans
l‘inventaire,ă enă particulieră l‘arsenală desă discriminantsă proposés.ă Leă champă
d‘explorationăestălargementăouvert.
Enfin, il semble que les recherches stimulantes sur les dialectes
périphériquesă soientă ună lieuă quiă favoriseă l‘ouvertureă àă l‘interdisciplinarité,ă auxă
approchesă etă auxă problématiquesă développéesă pară lesă spécialistesă d‘autresă airesă
linguistiques, des linguistes généralistes ou des typologues.
Références
Association Internationale de Dialectologie Arabe (fondée en 1993, en abrégé AIDA ou
AÏDA)
Actes des premières journées internationales de dialectologie arabe de Paris, 27-30
janvier 1993, Paris :ăPublicationsăLangues‘O.
Proceedings of the 2nd international Conference of L‘Association internationale pour la
dialectologie arabe, 10-14 september 1995, University of Cambridge.
Proceedings of the third international Conference of AÏDA, Malta, 29 March–2 April
1998, Paris-Malte : AIDA.
Aspects of the dialects of Arabic today. Proceedings of the 4th Conference of AÏDA, in
Honour of Professor David Cohen, Marrakesh, Apr. 1-4. 2000, Rabat :
Amapatril.
144
AIDA 5th Conference Proceedings, Cádiz, september 2002, Cádiz : Universidad, Servicio
de publicaciones.
L‘arabe dialectal : enquêtes, descriptions, interprétations.ă (Actesă d‘AIDAă 6),ă Travauxă
offerts au Professeur Taïeb Baccouche,Tunis : CERES.
Arnold, Werner . 1995. « Arabian dialects in the Turkish province of Hatay », in :
Proceedings of the 2nd international Conference of AIDA, 10-14 september 1995,
University of Cambridge, 1-10.
—
2002. « Code switching and code-mixing in the Arabic dialects of Antioch », in :
Abderrahim Youssi et alii (eds) : Aspects of the dialects of Arabic today.
Proceedings of the 4th Conference of AÏDA, Marrakesh, Apr. 1-4. 2000, 163-168.
Borg, Alexander. 1985. Cypriot Arabic. A historical and comparative investigation into
the phonology and morphology of the Arabic vernacular spoken by the Maronites
of Kormakiti village in the Kyrenia district of North-Western Cyprus, Stuttgart :
Franz Steiner.
–
1994. « Some evolutionary parallels and divergences in Cypriot Arabic and
Maltese », in : Jordi Aguadé, Federico Corriente y Marina Marugán (eds.) : Actas
del congreso internacional sobre interferencias lingüisticas arabo-romances y
paralelos extra-iberos, Madrid 10-14 de decembre de 1990, Zaragoza, 21-40.
Repris avec corrections in : Mediterranean Language Rewiev, 8 (1994), 41-67.
—
2004. A comparative glossary of Cypriot Maronite Arabic (Arabic-English), with
an Introductory Essay, Leiden-Boston : Brill.
Cohen, David.
1973. « Variantes, variétés dialectales et contacts linguistiques en
domaine arabe », in : Bulletin de la Société de Linguistique de Paris, tome
LXVIII, 1 (1973), 215-248.
—
1994. « Préface », in : Dominique Caubet et Martine Vanhove (éds) : Actes des
premières journées internationales de dialectologie arabe, Paris, 27-30 janvier
1993, 9-19.
Chikovani, Guram. 2000. « The Qashqadarian Arabic dialect of Central Asia », in :
Manwel Mifsud (ed.) : Proceedings of the third international Conference of
AÏDA, Malta, 29 March–2 April 1998, 189-194.
—
2002. « The verb in the Arabic dialects of Central Asia », in : Abderrahim Youssi
et alii (eds) : Aspects of the dialects of Arabic today. Proceedings of the 4th
Conference of AÏDA, Apr. 1-4. 2000, 179-188.
—
2004. « Central Asian Arabic dialects. The main lexical peculiarities », in :
Ignacio Ferrando y Juan José Sánchez Sandoval (eds.) : AIDA 5th Conference
Proceedings, Cádiz, september 2002, 1-12.
Ferrando, Ignacio. 2000. « Andalūsī Arabic in postislamic North of Spain : the language
ofăAragon‘săMudejarsăandăMoriscosă»,ăin : Manwel Mifsud (ed.) : Proceedings of
the third international Conference of AÏDA, Malta 29 March–2 April 1998, 195200.
145
—
2002. « L‘arabeă andalouă etă laă classificationă desă dialectesă néo-arabes », in :
Abderrahim Youssi et alii (eds) : Aspects of the dialects of Arabic today.
Proceedings of the 4th Conference of AÏDA, Marrakesh, Apr. 1-4. 2000, 189-199.
Fischer, Wolfdietrich / Otto Jastrow (eds.). 1980. Handbuch der arabischen Dialekte,
Wiesbaden : Harrassowitz.
Grigore, George. 2002. « ku- ună préfixeă temporelă dansă l‘arabeă mardinienă »,ă in :
Abderrahim Youssi et alii (eds) : Aspects of the dialects of Arabic today.
Proceedings of the 4th Conference of AÏDA, Apr. 1-4. 2000, 374-380.
—
2004. « L‘assimilationă desă empruntsă pară leă systèmeă paradigmatiqueă deă l‘arabeă
mardinien », in : Ignacio Ferrando / Juan José Sánchez Sandoval (eds.) : AIDA 5th
Conference Proceedings, Cádiz, september 2002, 13-20.
—
2006. « L‘alternanceă codiqueă cheză lesă Arabesă deă Mardină (Turquie) », in : Salah
Mejri (éd.) : Arabe dialectal : enquêtes, descriptions, interprétations (Actes
d‘AIDAă6),ă159-169.
Ingham, Bruce. 1994. « The effect of language contact on the Arabic dialect of
Afghanistan », in : Jordi Aguadé, Federico Corriente y Marina Marugán (eds.) :
Actas del congreso internacional sobre interferencias lingüisticas araboromances y paralelos extra-iberos, Madrid 10-14 de decembre de 1990, Zaragoza,
105-123.
—
2003. « Language survival in isolation : the Arabic dialect of Afghanistan », in :
Ignacio Ferrando / Juan José Sánchez Sandoval (eds.) : AIDA 5th Conference
Proceedings, Cádiz, september 2002, 21-37.
Isaksson, Bo. 2000. « Iranian and Turcic influence on border area Arabic dialects », in :
Manwel Mifsud (ed.) : Proceedings of the third international Conference of
AÏDA, Malta 29 March – 2 April 1998, 201-206.
—
2002. « Report from a research project and a brief account of the dialect of
‗Amūda », in : Abderrahim Youssi et alii (eds) : Aspects of the dialects of Arabic
today. Proceedings of the 4th Conference of AÏDA, Marrakesh, Apr. 1-4. 2000,
374-380.
—
2003. « Sason : New data from a nearly forgotten Arabic minority in Turkey »,
in : Ignacio Ferrando / Juan José Sánchez Sandoval (eds.) : AIDA 5th Conference
Proceedings, Cádiz, september 2002, 39-49.
Jastrow, Otto. 1977. « Gedanken zum zypriotischen Arabisch », in : Zeitschrift der
deutschen Morgenländischen Gesellschaft, Band 127, 2 (1977), 258-286.
—
1978. Die mesopotamisch-arabischen Qəltu-Dialekte, Band I : Phonologie und
Morphologie, Wiesbaden : Franz Steiner.
—
1983. « Beobachtungen zum arabischen Dialekt von Adana (Türkey) », in :
Zeitschrift
—
für arabische Linguistik 11 (1983), 72-79.
—
1994. « The qəltu-dialects of Mesopotamian Arabic », in : Jordi Aguadé,
Federico Corriente y Marina Marugán (eds.) : Actas del congreso internacional
sobre interferencias lingüisticas arabo-romances y paralelos extra-iberos
Madrid 10-14 de decembre de 1990, Zaragoza.
146
—
1995. « Towards a reassessment of Uzbekistan Arabic », in : Proceedings of the
2nd international Conference of L‘Association internationale pour la
dialectologie arabe, 10-14 september 1995, University of Cambridge, 95-103.
Jullien de Pommerol, Patrice. 1997. L‘arabe tchadien. Émergence d‘une langue
véhiculaire, Paris : Karthala.
Kaye, Alan S. 1994. « Peripheral dialectology and Arabic pidgins and creoles », in : Jordi
Aguadé, Federico Corriente y Marina Marugán (eds.) : Actas del congreso
internacional sobre interferencias lingüisticas arabo-romances y paralelos extraiberos Madrid 10-14 de decembre de 1990, Zaragoza, 125-137.
Lahdo, Ablahad. 2003. « The Arabic dialect of Təllo », in : Ignacio Ferrando / Juan José
Sánchez Sandoval (eds.) : AIDA 5th Conference Proceedings, Cádiz, september
2002, 51-60.
Lentin, Jérôme. 2000. « A 17th century document on Cypriot Arabic », in : Manwel
Mifsud (ed.) : Proceedings of the third international Conference AÏDA
Association internationale de dialectologie arabe, Malta 29 March – 2 April 1998,
207-211.
Mifsud, Manwel. 1994. « International pluralization in Maltese : continuity and
innovation », in : Dominique Caubet et Martine Vanhove (éds) : Actes des
premières journées internationales de dialectologie arabe de Paris, 27-30 janvier
1993, 91-105.
—
1995. « The productivity of Arabic in Maltese », in : Proceedings of the 2nd
international Conference of L‘Association internationale pour la dialectologie
arabe, 10-14 september 1995, University of Cambridge, 151-160.
Miller,ăCatherine.ă2000.ă«ăTheăgrammaticalizationăofătheăverbă‗toăsay‗ăinăJubaăArabică»,ă
in : Manwel Mifsud (ed.) : Proceedings of the third international Conference
AÏDA Association internationale de dialectologie arabe, Malta 29 March – 2
April 1998, 213-218.
—
2002. « Juba Arabic as a way of expressing a Southern Sudanese identity in
Khartoum », in : Abderrahim Youssi et alii (eds) : Aspects of the dialects of
Arabic today. Proceedings of the 4th Conference of AÏDA, Marrakesh, Apr. 1-4.
2000, 114-122.
Owens, Jonathan. (ed.) 2000. Arabic as a minority language, New-York-Berlin :
Mouton/de Gruyter.
Procházka, Stephan. 1995. « The Arabic dialects of Cilicia (Southern Turkey) », in :
Proceedings of the 2nd international Conference of AIDA, 10-14 september 1995,
University of Cambridge, 189-195.
—
1999. « From language contact to language death : the example of the Arabic
spoken in Cilicia (Southern Turkey) », in Orientalia Suecana vol. XLVIII (1999),
115-125.
—
2000. « Some morphological and syntactical characteristics of the Arabic dialects
spoken in Cilicia (Southern Turkey) », in : Manwel Mifsud (ed.) : Proceedings of the
third international Conference of AÏDA, Malta 29 March – 2 April 1998, 219-223.
147
— 2002. « Contact phenomena, code-copying, and code-switching in the Arabic dialects of Adna and Mersin
(Southern Turkey) », in : Abderrahim Youssi et alii (eds) : Aspects of the dialects of Arabic today. Proceedings
of the 4th Conference of AÏDA, Apr. 1-4. 2000, 133-139.
—
2003. « The bedouin Arabic dialects of Urfa », in : Ignacio Ferrando /Juan José Sánchez Sandoval
(eds.) : AIDA5th Conference Proceedings, Cádiz, september 2002, Cádiz, 75-88.
Roth, Arlette. 2002. « Quelques observations concernant un parler arabe en éviction
(Kormakiti, Chypre) », in : Abderrahim Youssi et alii (eds) : Aspects of the
dialects of Arabic today. Proceedings of the 4th Conference of AÏDA, Marrakesh,
Apr. 1-4. 2000, 133-139.
—
2003. « Le conditionnel dans le parler arabe de Kormakiti (Chypre) », in :
Ignacio Ferrando/Juan José Sánchez Sandoval (eds.) : AIDA 5th Conference
Proceedings, Cádiz, september 2002, 89-101.
—
2004. « Le parler arabe maronite de Chypre :ă observationsăàă proposă d‘ună contactă linguistiqueă
pluriséculaire », in : International Journal of the Sociology of Language, The Sociolinguistics of
Cyprus I, Dionysis Goutsos and Marilena Karyolemou (eds.), 55-76.
2006. « Les localisateurs spatiaux dans le parler arabe de Kormakiti (Chypre) »,
in : Salah Mejri (éd.) : Arabe dialectal : enquêtes, descriptions, interprétations
(Actesăd‘AIDAă6),ă395-409.
Simeone-Seneelle, Marie-Claude. 2002. « L‘arabeă langueă maternelleă deă citoyensă
djiboutiens du nord de la République de Djibouti », in : Abderrahim Youssi et
alii (eds) : Aspects of the dialects of Arabic today. Proceedings of the 4th
Conference of AÏDA, Marrakesh, Apr. 1-4. 2000, 140-150.
Taine-Cheikh, Catherine. 1994. « Le h}assāniyya de Mauritanie : un dialecte non-marginal de la périphérie »,
in : Jordi Aguadé, Federico Corriente y Marina Marugán (eds.) : Actas del congreso internacional
sobre interferencias lingüisticas arabo-romances y paralelos extra-iberos Madrid 10-14 de
decembre de 1990, Zaragoza, 173-199.
Tsiapera, Maria. 1973. « The Maronite Arabs in Cyprus : a linguistic discussion », in : Micheline
Galley and David R Marshall (eds.) : Actes du premier congrès d‘études des cultures
méditerranéennes d‘influence arabo-berbère, Alger : SNED, 136-139.
Talay, Shabo. 2003. « Some grammatical remarks on the Arabic dialects of Hasköy (East
Anatolia) », in : Ignacio Ferrando / Juan José Sánchez Sandoval (eds.) : AIDA 5th
Conference Proceedings, Cádiz, september 2002, 119-129.
Vanhove, Martine. 2003. « Bilinguisme et alternance bedja-arabe au Soudan », in :
Ignacio Ferrando / Juan José Sánchez Sandoval (eds.) : AIDA 5th Conference
Proceedings, Cádiz, september 2002, 131-142.
Versteegh, Kees. 1997. The Arabic Language, Edinburgh : Edinburgh University Press
(Chapter 13 « Arabic as a minority language »).
Zammit, Martin R. 2006. « Unrecognizable Arabic-Maltese :ă theă innovativeă ‗Maltese‘ă
element in the Maltese language », in : Salah Mejri (éd.) : Arabe dialectal :
enquêtes, descriptions, interprétations.ă(Actesăd‘AIDAă6),ă487-498.
148
Beda u Qabad: The Maltese Inchoative / Ingressive
Thomas Stolz & Andreas Ammann
University of Bremen
Germany
1. Introductory remarks
Our contribution is meant as the starter of a series of studies dedicated to
the system of aktionsarten and related categories of Maltese. Following the line of
reasoning of the late Reinhold Kontzi (2005), we eventually aim to determine
what in the grammatical system of the Maltese VP is Semitic, what is Romance
and what is idiosyncratically Maltese. Inchoatives / ingressives are probably
universal as an aktionsart or lexical aspect: consider the English item fall asleep,
which singles out the beginning of a more encompassing semantic concept as
coded by the unmarked verb sleep. It unequivocally renders the first stage of the
sleeping as a whole, but it is not created by applying a productive pattern. For
instance, one cannot *fall awalk. (For a study of lexical aspect in the Cairene
variety of Arabic, see Eisele 1990.) Among the categories which are closely
related to the aktionsarten, we find the so-called phasal verbs, which usually serve
the purpose ofă highlightingăsomeă segmentă(―phase‖,ăseeăBinnickă1991:ă197-207)
of the process (also event, more rarely state) described by the lexical verb. In
English, the relevant phasal verbs would be begin and start, which do combine
rather freely with gerunds. Maltese – both spoken and written – makes ample use
of phasal verbs too. Incidentally, we start our investigation of Maltese phasal
verbs with a category which is commonly labelled inchoative (sometimes also
ingressive) whose task it is to indicate that a process has begun. The inchoative
focuses on the initiation of the process whereas what actually happens later – e.g.
whether or not the process is continued or successfully terminated – remains
unspecified.
We have opted for the inchoative because Maltese displays certain
characteristics which set it apart not only from its Indo-European neighbours but
also from its closest relatives in the Semitic phylum. Inchoative phasal verbs
usually correspond to English
begin/start, French commencer, German
anfangen/beginnen, etc. In Maltese, however, we find several candidates for the
status of inchoative phasal verbs which still have lexical meanings of a different
kind, namely beda ‗toăbegin‘ă(cognateăofăModernăStandardăArabicăbada‘a), qabad
‗toă grasp‘ă (~MSAă qabad}a ), telaq ‗toă leave‘ă (~MSA talaqa), rema ‗toă prepare‘ă
149
(~MSA ramā), fetaħ ‗toă open‘ă (~MSAă fatah}a ) and ħabat ‗toă knock‘ă (~MSAă
h~a bat}a ) (Borg/Azzopardi-Alexander 1997: 233). These six verbs compete in
peculiar ways whose systematicity still awaits an in-depth description. Vanhove
(1993: 241-64)ădiscussesătheăparadigmăofăinchoativeă―auxiliaries‖ăonătheăbasisăofăaă
corpus analysis of written and spoken Maltese. According to her description,
some of the phasal verbs are severely restricted in their use (e.g. they occur only
in the literary register or only in combination with a few verbs [and thus behave
likeă collocations]).ă Vanhove‘să chapteră VIIă providesă aă handyă startingă pointă foră aă
thorough-going analysis of inchoativity in Maltese. Since beda is by far the most
frequent of the above phasal verbs, we will turn to this candidate first and then
add some observation on qabad. The other verbs will be studied in more detail in
a follow-up investigation.
2. What could be noteworthy about Maltese beda?
Morphosyntactically, beda behaves very much like its Indo-European equivalents
as it functions as the syntactically higher verb to which the lexical verb is
―subordinated‖:ă beda precedes the lexical verb which in turn cannot be in the
perfective if the phasal verb is in the imperfective. Except for this restriction,
there are no overt means of subordination (conjunctions or the like) – a fact which
fits the general structure of the VP in neo-Arabic varieties. However, what lets the
Maltese inchoative stand out is the at times astonishingly high frequency with
which it is employed in all registers of the language. This extraordinary frequency
is said to exert influence also on the non-native English spoken on Malta because
speakers tend to translate Maltese constructions directly into their version of
English (Ray Fabri p.c.). To give the reader an idea of the ubiquitous nature of
beda, we quote a passage from literary Maltese. This passage is drawn from page
33ăofăƳinoăMuscatăAzzopardi‘sănovelăAnġli tan-Niket published quite recently in
2002.
(1) Maltese1
Kien
għadhom
be.PERF.3SG.M still:3PL
Bonaparti
għall-Eġittu
m‘għaddewx
żewġt iġmiegħi
mill-mawra
NEG‘pass.PERF:3PL:NEG two week.PL from:DET-departure of
li
fost
il-Maltin
ma
bediex jitqanqal
1
The typographic convention used in examples, glosses and translations will be as follows: nonitalicized items in examples are ones we wish to highlight, likewise italicized items in glosses and
translations. Thus, italics are not found in the original sources, but were introduced by us for the
sake of reader-friendliness.
150
ta‘
Bonaparte
for:DET-Egypt that among DET-Maltese:PL
NEG start:NEG
3:PASS:stir.IMPER
tgemgim
u
tfantis qawwi mħabba
t-tibdil
li
beda
grumbling
and sulking heavy
because_of
DET-change that
start.PERF.3SG.M3:become.IMPER
f‘kull
ħaġa
li
tolqot
lin-nies
ta‘
dawn il-gżejjer.
in‘every
thing
that 3.F:hit.IMPER
to:DET-people
of
these
DET-island.PL
jsir
Sakemm bdew
ħerġin
mill-Kummissjoni
tal-Gvern tagħrifiet
until start.PERF:3PL
come_out.AP:PL
from:DET-commission
of:DET-governmentinformation:PL
li
l-Belt Valletta,
il-Birgu, Bormla
u l-Isla
jibdew
jissejħu
that DET-City Valletta
DET-Birgu Bormla and DET-Isla3:start.IMPER:PL 3:PASS:call.IMPER:PL
bl-isem
waħdieni
ta‘
Belt ta‘
Malta, sakemm
ħarġu
tnedijiet
with:DET-name unique
of
City of
Malta until come_out.PERF:3PL
threat:PL
għal dawk li
kienu Kavalieri sabiex
ifittxu
jbewġu
minn
for those that
be.PERF:3PL Knight:PL in_order_to 3:seek.IMPER:PL 3:leave.IMPER:PL from
dawn il-gżejjer,
il-Maltin
ftit
li
xejn
ħaduha
bi
kbira,
these
DET-island.PL DET-Maltese:PL
little that nothing
take.PERF:3PL:O3.F with big:F
iżda meta
bdew
jaraw
li
dawk li ħalla
warajh
but
when
start.PERF:3PL
3:see.IMPER:PL that those
that leave.PERF.3SG.M behind:3SG.M
Bonaparti kienu
ħasbu
biex
iwettqu,
u
mill-aktar fis,
Bonaparte
be.PERF:3PL think.PERF:3PL so_that
3:realise.IMPER:PL
and from:DET-more soon
l-Ordnijiet
kollha minnu
maħruġa
sewwa
fuq
il-knejjes u
DET-Order:PL all
from:3SG.M PP:come_out.PL
truth
about
DET-church.PL and
sewwa
fuq
ħwejjeġ oħra
li
jolqtu
fil-laħam
il-haj
truth
about thing.PL other.PL that 3:hit.IMPER:PL
in:DET-flesh
DET-living
il-jeddijiet
tagħhom,
bosta bdew
joqomsu
u
jitnaffru
DET-right:PL
of:3PL
many
start.PERF:3PL
3:rebel.IMPER:PL
and 3:PASS:displease.IMPER:PL
u ma
naqsux
li juru
lil min ma
kellux moħħ
and NEG fail.PERF.3PL:NEG
that 3:show.IMPER:PL
to who NEG have.PAST:3SG.M:NEG brain
biżżejjed
sabiex
jifhem,
dak kollu
li,
fi ftit
xhur,
enough
in_order_to 3:understand.IMPER
that all
that in little month:PL
kien
sejjer
ikollu
jġarrab
il-ġens
Malti.
be.PERF.3SG
go:AP
3:have.FUT:3SG.M
3:endure.IMPER
DET-people
Maltese
Il-Bandu tal-25
ta‘ ƶunju sabiex
ix-xgħir
li
jiġi minn
barra
DET-edict of:DET-25 of June
in_order_to
DET-cereal that 3:come.IMPER from
outside
jibda
jħallas
skud
kull
mod
ħarrex
3:start.IMPER 3:pay.IMPER twenty_pence every measure harvest
mhux
ftit
lil
kull
min kellu
nitfa
ta‘
bhima
NEG:he:NEG little that every who have.PAST:3SG.M a_bit
of
cattle
u
jrid
jaqla‘
x‘jiekol
biha.
and 3:want.IMPER
3:gain.IMPER what‘3:eat.IMPER with:3SG.F
‗Notă evenă twoă weeksă afteră Bonaparte‘să departureă foră Egypt,ă muchă
grumbling and a strong feeling of displeasure were created among the Maltese
because of the changes which began to affect all areas which had to do with the
151
people of these islands. Until information came from the Governmental
Commission that Valletta, Birgu, Bormla and Isla should be called henceforth
with the common name City of Malta, until threats were directed at the former
Knights to leave these islands, the Maltese did not take things seriously, but when
they saw that those whom Bonaparte had left behind had thought of realising as
soon as possible all the Orders given by him not only about the churches but also
about other things which affected their vital interests, many started to rebel and
became discontent and they did not fail to show to those who did not have enough
brains to understand all the things that in a few months the Maltese people would
have to endure. The edict of 25 June according to which for every imported
measure of cereal an amount of twenty pence had to be paid was nothing
negligible for everybody who had a head of cattle and wanted to gain a living with
it.‘2
This longish passage contains only three lengthy sentences but seven
occurrences of forms of beda. In the English translation of the examples, the use
of phasal verbs corresponding semantically to beda makes sense only in two
instances, marked with italics in our translation. There is no need for the English
translator to use begin/start every time beda is used in the Maltese original. Of
course, employing begin/start more often would not render the English translation
ungrammatical. Nevertheless, the strict adoption of the Maltese pattern would
have a strange stylistic effect – namely one of too much repetitive information.
This is an interesting fact of contrastive grammar: what speakers of one language
consider necessary information is superfluous for those who speak a different
language. As it seems, speakers of English are freer in their choice. They may or
may not mark the inchoative overtly by a phasal verb and very often they refrain
from using a phasal verb. In contrast, speakers of Maltese are less parsimonious
with their inchoative constructions whose high frequency may be understood as
an indicator of a higher degree of grammaticalisation of the inchoative –
grammaticalisationă ină theă senseă ofă aă higheră degreeă ofă ―obligatorification‖ă
(Lehmann 1995: 139). The use of an inchoative phasal verb is perhaps not strictly
compulsory (yet) but speakers tend to use beda rather than making do with
contextual or semantically inherent information.
3. Inchoatives in (translations of) Le Petit Prince
The preference for overt marking of the inchoativity we exemplified in (1) is by
no means a stylistic mannerism of the above author. Quite the opposite is true:
2
All translations into English are ours.
152
high frequency of beda is a commonality in Maltese texts of all genres 3. This
observation also holds for translated literature. Again in accordance with
principles established by Kontzi (2005), we have checked a sample text in a
variety of languages in order to determine whether or not the Maltese inchoative
deviates from the inchoative of other languages. For this purpose, we have chosen
Antoine de Saint-Exupéry‘săLe Petit Prince, for which translations into more than
150 languages and regional varieties worldwide are presently available. There are
two Maltese translations (one by Tusè Costa and another one by Toni Aquilina,
referred to as M1 and M2 respectively). The use made of beda and qabad in these
is compared to inchoative constructions in the French original (F) and the
translations into Modern Standard Arabic (MSA) and Tunisian Arabic (TA)4. Our
point of departure is the two Maltese texts: wherever one of the two Maltese
translators has used a form of beda or qabad, we have checked for markers of
inchoativity in the corresponding sentences in the other languages. The outcome
of the comparative study is telling. Table 1 summarizes our counts.
‗begin‘
‗take‘
total
F
commencer
11
prendre: 3
14
5
M1
: beda: 24
qabad: 2
26
M2
beda: 12
MSA
bada‘a6: 6
TA
bada‘a: 16
qabad: 10
22
ah~a da 7: 10
16
./.
16
Table 1: Candidates for overt marking of inchoativity in Le Petit Prince
Whatădoătheseănumbersătellsăus?ăFirst,ăifăweăconflateă‗begin‘- andă‗take‘-verbs, the
Maltese inchoatives are more numerous than what we find in any other language
in our small sample. More specifically, the French original is the text with the
largestănumberăofăsentencesăwithăaăprocess/stateăneitheră markedăwithă‗begin‘ă noră
withă‗take‘,ăandăthereăisănoăquantitativeădifferenceăbetween MSA and TA: they are
both in between F and M1/M2 in terms of frequency.
The most conspicious difference in the distribution of phasal verbs is the
relationăbetweenă‗begin‘- andă‗take‘-verbs. MSA is the only variety in the sample
3
We emphasise that the use of beda and its competitors is by no means a phenomenon restricted to
the spoken language as was suggested during the discussion in Bucharest. On the contrary, the
higheră theă literaryă ambitionsă ofă theă authoră theă higheră theă numberă ofă attestationsă ofă ―surplus‖ă
inchoative markers.
4
We wish to thank Dana Hanoun (Cologne) for supplying us with the TA text.
5
The figure includes two attestations of recommencer ‗startăagain‘.ă
6
One instance of the nominal form bad‘ ‗beginning‘ăwasăalsoăcounted.
7
Aside from nine inflected verbal forms, we also counted one active participle āh~id.
153
that shows a preference for ah~a da ‗take‘,ăaăverbăthatăhasănoăetymologicalăcognateă
in the other Semitic varieties in its inchoative use and not even a semantic
equivalent in TA. This may provide the explanation for the fact that TA bada‘a
outranks its cognate beda in M2 – though the difference between M1 and M2
remainsă ană openă question.ă Foră theă semantică overlapă betweenă ‗take‘ă andă ‗begin‘,ă
cf. English collocations like take sick ‗becomeăsick‘ăorătake a walk ‗setăoutătoăgoă
forăaăwalk‘.
Semantically, a native speaker of German isănotăsurprisedătoăseeăaă‗begin‘verb where there is commencer, and we suspect a translation with start or begin
would be just as unspectacular for a native speaker of English. On the other hand,
our material from the Maltese translations of Le Petit Prince also provides
additional examples of what we saw in (1), i.e. inchoative verbs that do not
contribute a discernable inchoative meaning. By contrast, almost all uses of
bada‘a in the MSA version translate commencer. We illustrate this with (2) and
(3).
(2) M1: 39
Hija ħaġa li
ntesiet,
beda
volpi.
3SG.F thing REL forget.PERF.PASS.3SG.F start.PERF:3SG
fox
‗‗Thisăisăaăthingăwhichăgetsăforgotten‘,ăsaidătheăfox.
jfisser
il-
3:declare.IMPER DET-
(3) MSA : 36
fa-bada‘a
rih}lata-hu bi-ziyārati-hi [...]
and-begin.PERF:3SG:M
journey-his
with-visit-his
‗Andăheăbegan his journey withăhisăvisită[toăthemă…]‘
Sentence (2) translates the French C‘est une chose trop oubliée, dit le
renard (F: 68), with no sign of inchoativity in the original, unlike (3), which
stands for Il commença donc par les visiter [ ...] (F : 36).
In the cases where Maltese uses qabad with a bleached meaning 8, as in (4),
butănoă‗begin‘-verb is present in the French original, we are in the same semantic
gray zone exemplified with (1). The same goes for ah~a da in (5), and in fact for
prendre ‗take‘ăină(6).
(4) M2: 68
8
In a few of the examples, which are not particularly interesting, it retains the lexical meaning of
‗seize‘ăoră‗hold‘.
154
Madanakollu qbadna
mexjin.
all_the_while grasp.PERF:1PL go.AP:PL
‗Inătheămeantime,ăweăgotăgoing‘
(5) MSA: 80
hadihi l-ba‘ara fa-ah~adat
[...] fa-innanā īqāz}nā
ġinā‘
and-S:1PL
wake_up.PERF:1PL that
DET-well and-take:PERF:3SG:F
DET-singing
‗[...]ăandăweăhaveăwokenăupăthisăwell,ăandăităhasăstarted toăsing‘
fī lin
(6) F: 88
Et il
me prit par la
main.
and he me took by DET:F hand
‗Andăheătookămyăhand‘
Sentence (4) translates Cependent nous nous mîmes en marche (F: 77)
[lit.:ă‗weăputăourselfă inăwalking‘],ăandăoneă mightăasăwellărenderăthatăsimplyăwithă
‗andăweăleft‘,ăbutăoneăcouldăargueăthatăthereăisăaălittleămoreăsemanticăsubstanceăin
nous nous mîmes en marche. (5) stands for nous réveillons ce puits et il chante...
(F:ă80),ăi.e.ătheăFrenchăoriginalăhasăaăsimpleălexicalăverbă‗ităsings‘.ăOurătranslationă
of (6) indicates that prendre cană beă renderedă withă transitiveă ‗take‘ă ină English,ă aă
case where, if we want to spot inchoativity, it is a lexical property of prendre, viz.
where the semantics of the verb is understood as the first phase of a longer period
ofă‗holding‘.
4. Phasal verbs in other Arabic varieties
Of the verbs MSA dictionaries offer as translational equivalents of start or
begin, we have not observed ibtada‘a ‗(intransitive)ă begin‘,ă theă so-called VIIIth
stem of bada‘a, in our data, nor is it attested in the M or TA data9. For MSA, this
may simply reflect an idiosyncratic preference of the translator in favor of the Ist
stem bada‘a. As for Neo-Arabic varieties, we should note that the texts collected
by Brustad (2000: 377-420) do not contain a single form of (a cognate of) bada‘a.
Brustad´s texts, from Moroccan, Egyptian, Syrian and Kuwaiti varieties, differ
from Le Petit Prince in being shorter and oral texts, so we should not jump to
conclusions about how significant it is that bada‘a occurs 16 times in our TA data.
The structural model of combining phasal verbs with lexical ones to
generate inchoative (or inchoative-like) meanings is certainly nothing unusual for
9
Mitchell & El-Hassan (1994: 36) do report ibtada ‗begin‘ăforăEgyptianăArabic.
155
Neo-Arabic. The material from Brustad (2000) along with her glosses and
translations yields the following candidates:
(I)
(II)
(III)
Egyptian ‘ām(it) ‗upăand‘ă~ăMSA qāma ‗getăup‘,ă
Moroccan nād} ‗upăand‘,ăsbh}a ‗begin‘ă~ăMSA as}bah}a ‗become‘,ă
Syrian s}ār ‗start,ăbegin‘ă~ăMSAăs}āra ‗become,ăbegin‘,ăalsoănotedăbyăMitchellă
& El-Hassană(1994:ă87)ăasăbeingă―essentiallyăaăchangeăofăstate‖ăindicatorăandă
―primarilyăLevantine‖
FromăHoles‘ă(1990:ă196)ădescriptionăofăGulfăArabicăweăcanăaddăğā ‗come‘ă
(as well as s}ār and gām). Whether this means ex negativo that cognates of beda
and qabad are unknown in these varieties, we are in no position to evaluate.
For inchoatives in Standard Arabic, we can refer the reader to
Anghelescu (2004: 413-420), where she approaches the semantic field from an
aspect-theoretical framework, referencing both modern authors and the views of
Arab grammarians. She compares the function of bada‘a ‗start‘ă toă otheră verbsă
indicating a change of state. Inchoative constructions with bada‘a ‗start‘ă and
ah~a da ‗take‘ă areă alsoătreated,ă withă severală transliteratedă examples,ă byă Badawiă et
al. (2004: 427-429).
5. To sum up
Our understanding of the topic at hand is preliminary – only beginning to
evolve – in at least the following respects:
(I) quantity of data: we have observed the phenomenon so astonishingly often that
we took notice, but in what so far can only be called a limited amount of Maltese
data
(II) quality of data: the numerical results we have gathered here are taken from
translations of a text into Maltese, Standard Arabic and Tunisian Arabic from the
original, written in French
(III) tense-aspect theory: we have not aimed to make any strong claims on the
nature of the Maltese tense-aspect(-aktionsart) system in general (for which see
Borg 1981, Ebert 2000 and Fabri 1995) or how beda and qabad and their
competitors fit into that picture
(IV) areal linguistics: we have not investigated systematically how Sicilian
varieties of Italian behave in this respect, for example
156
What we do argue is that there is an explanandum: the use of Maltese beda
and qabad in contexts that, as far as we can see, do not require an explicit marking
of inchoative in Standard Average European languages, Modern Standard Arabic
or Tunisian Arabic, at least as a matter of a higher frequency of these inchoativelooking constructions, even in parallel translations of one and the same text.
It is still somewhat premature to claim that the higher density of overt inchoative marking in
Maltese texts is evidence of a difference between Maltese verbs and those of Standard Average
European (SAE) languages – a difference on the conceptual level in such a way that Maltese verbs
tend to be only weakly contoured on the left boundary and thus do not allow for an inchoative
reading of the lexical verb itself, at least not as easily as the average SAE verb depicting a process. To
substantiate this claim, many more data, and, above all, additional psycholinguistic tests are
necessary. For the time being, it suffices to show that Maltese deviates markedly from the usage to
which the inchoative is put to in SAE languages. We are planning to come back to this issue later this
year already: in Stolz & Ammann (in preparation) where we intend to give a full account of the
paradigm of inchoative phasal verbs as employed in contemporary Maltese and on earlier stages of
this language.
Abbreviations
1
3
AP
DET
F
FUT
IMPER
M
NEG
O
PASS
PERF
PL
PP
REL
S
SG
first person
third person
active participle
determiner
feminine
future
imperfective
masculine
negator
object
passive
perfective
plural
passive participle
relative marker
subject
singular
Primary Sources
de Saint-Exupéry, Antoine. 1946. Le Petit Prince, Paris: Gallimard. (=F, French)
de Saint-Exupéry, Antoine. 1982. Il-Prinƭep iż-żgħir [Translated by Tusè Costa], Luqa:
Toussaint Costa. (=M1, Maltese)
de Saint-Exupéry, Antoine. 2000. Iƭ-Ƭkejken Prinƭep. [Translated by Toni Aquilina], LImsida: Mireva. (=M2, Maltese)
157
de Saint-Exupéry, Antoine. s.d. al-amīr as}-s}aġīr [Translated by Youssef Ghoussoub],
Beirut: al-manﻣūrātăal-‗arabīya. (=MSA, Modern Standard Arabic)
de Saint-Exupéry, Antoine. 1997. al-amīr as}-s}ġayyar. Le Petit Prince [Translated by
Hédi Balegh], Tunis. (=TA, Tunisian Arabic)
Other references
Anghelescu, Nadia. 2004. La langue arabe dans une perspective typologique, Bucharest:
EdituraăUniversitĒįiiădinăBucureşti.
Badawi, Elsaid; Michael G. Carter & Adrian Gully. 2004. Modern written Arabic. A
comprehensive grammar, London: Routledge.
Binnick, Robert I. 1991. Time and the verb. A guide to tense and aspect, New York:
Oxford University Press.
Borg, Albert. 1981. A study of aspect in Maltese, Ann Arbor: Karoma Press.
Borg, Albert & Marie Azzopardi-Alexander. 1997. Maltese, London: Routledge.
Brustad, Kristen E. 2000. The syntax of spoken Arabic. A comparative study of
Moroccan, Egyptian, Syrian, and Kuwaiti dialects, Washington, D.C.: Georgetown
University Press.
Ebert,ăKarenăH.ă2000.ă―AspectăinăMaltese‖,ăin:ăÖstenăDahlă(ed.):ăTense and aspect in the
languages of Europe, Berlin: Mouton de Gruyter, 753-785.
Eisele,ăJohnăC.ă1990.ă―AspectualăclassificationăofăverbsăinăCaireneăArabic‖,ăin:ăMushiraă
Eid & John McCarthy (eds.): Perspectives on Arabic Linguistics II, Amsterdam:
Benjamins, 192-233.
Fabri,ă Ray.ă 1995.ă ―Theă tenseă andă aspectă systemă ofă Maltese‖,ă in:ă Rolfă Thieroffă (ed.):ă
Tense systems in European languages II, Tübingen: Niemeyer, 327-343.
Holes, Clive. 1990. Gulf Arabic, London: Routledge.
Kontzi, Reinhold. 2005. Sprachkontakt im Mittelmeer. Gesammelte Aufsätze zum
Maltesischen, Tübingen: Narr.
Lehmann, Christian. 1995. Thoughts on grammaticalization, Munich: LINCOM Europa.
Mitchell, T[erence] F. & S[hahir] A. El-Hassan. 1994. Modality, mood and aspect in
Spoken Arabic. With special reference to Egypt and the Levant, London: Kegan
Paul.
Stolz,ă Thomasă &ă Andreasă Ammann.ă Ină preparation.ă ―Newsă onă theă Malteseă inchoative‖,ă
Poster to be presented at the conference Maltese Linguistics / Lingwistika Maltija,
University of Bremen (Germany), October 18-20, 2007.
Vanhove, Martine. 1993. La langue maltaise. Etudes syntaxiques d‘un dialecte arabe
‗périphérique‘, Wiesbaden: Harrassowitz.
158
Périphérie géographique et perméabilité aux contacts.
Le cas du Maghreb
Catherine Taine-Cheikh
LACITO - CNRS (Villejuif)
France
La notion de périphérie
La notion de périphérie est géographique, mais elle est aussi employée en
économieăetăenăpolitique.ăSiăl‘onădéfinitălaăpériphérieăparărapportăauăcentre,ăceăquiă
semble a priori nécessaire,ă l‘éventailă desă choixă apparaîtă encoreă plusă clairement.ă
S‘agit-il de centres historiques comme le foyer de la langue arabe (la Péninsule
arabique)ă ouă lesă lieuxă d‘émergenceă duă pouvoiră politiqueă etă culturelă (Damasă etă
Bagdad)ă ?ă S‘agit-il de centres quasi permanents comme lieux de référence du
religieuxă (laă Mecque)ă ouă duă culturelă (leă Caire)ă ?ă S‘agit-il des centres
contemporains des pouvoirs économiques et politiques, tant au plan transnational
(les pays arabes les plus peuplés et / ou les plus riches par rapport aux pays moins
peuplésă etă /ă ouă moinsă riches)ă qu‘auă plană natională (lesă grandesă villesă – en
particulier les capitales – par rapport aux arrières-paysăàăl‘économieăvillageoiseăouă
pastorale) ?
Lorsqu‘ilă s‘agită deă faireă uneă présentationă globaleă des dialectes arabes, il est
clair que – vuălaădifficultéădeăl‘entrepriseăetălaămultiplicitéădesălignesădeăfractureă–
l‘importanceă relativeă desă différentesă communautésă influeă sură laă descriptionă desă
faitsă linguistiques.ă Onă neă peută pasă traiteră sură ună piedă d‘égalité les dialectes
irakiens (ou le domaine mésopotamien) et les dialectes libyens, ni le parler du
Caire comme celui de Rabat. Le poids historique, démographique, économique,
politique et culturel de chaque groupe linguistique rejaillit tout naturellement sur
laăplaceăqu‘onăattribueăàă leursă variétésă langagières.ăCelaă meă frappeăd‘autantăplusă
queă jeă m‘intéresse,ă deă pară mesă recherchesă personnelles,ă auă dialecteă arabeă deă
Mauritanie. Étant amenée à considérer la place qui est faite dans la dialectologie
arabe au h}assāniyya,ă jeă meă rendsă compteă queă celle-ci est plus variable que ce à
quoiăonăpourraităs‘attendre.ă
Le h}assāniyyaăestăgénéralementăclasséă – comme il se doit, mais souvent assez
succinctement (cf. Fischer & Jastrow 1980 et, plus encore, Durand 1995) –
comme un dialecte de nomades du Maghreb occidental. Il lui arrive cependant
d‘êtreăplusămalătraité,ăyăcomprisăparăd‘excellentsălinguistesădialectologues.ăIlăpeută
159
enă effetă êtreă complètementă oublié,ă alorsă mêmeă qu‘ună paragrapheă particulieră estă
consacré aux parlersă ma‗qiliensă duă Sud-Maroc (cf. Palva 2006). On peut aussi
réécrire son histoire en le rajeunissant (il devient alors classable plus
commodément dans une aire sub-saharienneă àă l‘arabisationă récente 1 ) ou en le
comptant artificiellement « selon sa géographieăd‘origineăàăl‘instarăduăTouaregăduă
Hoggar,ă c‘est-à-dire comme un dialecte oriental » (cf. Elmedlaoui 2000 : 181,
note 82).
Tout ceci serait assez anecdotique – etă deă peuă d‘intérêtă enă dehorsă deă laă
Mauritanie et du petit cercle de ses spécialistes – si la façon de traiter le
h}assāniyyaă n‘étaită pasă symptomatiqueă d‘uneă difficultéă :ă celleă d‘avoiră uneă visionă
globaleădesăvariétésădeăl‘arabeăoùăchaqueătypeădeădialecteăpourraităêtreăcaractériséă
de manière relativement simple et sans ambiguïté possible. On aimerait en effet –
c‘estă làă l‘ună desă objectifsă desă étudesă dialectologiquesă – pouvoir distinguer
linguistiquementăunătypeăd‘arabeămaghrébinăetăunătypeăd‘arabeăorientalăouăunătypeă
d‘arabeădeănomadesăetăunătypeăd‘arabeădeăsédentaires.ăCetteăunité,ăonăsouhaiterait
la retrouver pour tous les traits qui, au moins une fois, peuvent être associés à des
variations linguistiques et tendent, par là même, à fournir un principe de
classement : parlers des chrétiens ou des juifs par opposition aux parlers des
musulmans (eux-mêmes distingués en parlers de druzes, de chiites, de sunnites,
etc.), pidgins ou créoles par opposition aux autres parlers, parlers marocains par
rapport aux parlers algériens, parlers de femmes par rapport aux parlers
d‘hommes,ăetc.3
L‘énoncéămêmeădeătousăcesăclassementsăpossiblesămontreălesălimitesăd‘uneătelleă
attenteăpuisqueăchaqueădialecteărelèveăenă mêmeătempsădeăplusieursătypesăetăqu‘ilă
n‘estă pasă possibleă deă procéderă dansă l‘absoluă àă ună classementă hiérarchiséă desă
différents traits extralinguistiques 4.ăIlăn‘estăpasăvainătoutefoisădeăseăposerăchaqueă
1
« Sub-Saharan Bedouin dialects. There are Bedouin dialects found in Sub-saharan areas, e.g.
Sudan, Chad, Nigeria, and Mauritania, whose speakers arrived there apparently within the last 300
years » (Rosenhouse 2006 : 260). EnăfaitălesăBanūăMa‗qilă(auxquelsăseărattachentălesăBanūăH}assānă
et leur ancêtre éponyme H}assānă– qui a donné son nom au dialecte h}assāniyyaădeăMauritanie)ăseă
rattachentă auxă Banūă Hilāl,ă arrivésă dansă leă Maghrebă occidentală auă XII e siècle. Que leur descente
vers le sud soit plus tardive (plutôt à partir des XIII-XIVe)ăneăchangeărienăauăfaităqu‘onăneăsignaleă
pasădeănouvellesămigrationsăsignificativesădeăpopulationăpostérieuresăàăcellesădesăBanūăMa‗qil.
2
Elmedlaouiă s‘inspireă iciă duă classementă deă Durandă quiăavaităraisonnéă enă fonctionă duă faită queăleă
h}assāniyya,ădialecte arabe du Sud-Ouest, était « post-hilalien ».
3
Surăl‘inexistenceăd‘ună« judéo-arabe »ăàăproprementăparler,ămaisăl‘existenceădeătraitsăparticuliersă
distinguantă localementă l‘usageă desă juifsă deă celuiă desă musulmans,ă cf.ă D.ă Cohenă (1978).ă Sură laă
grande opposition entre dialectes qəltu et dialectes gilit en Mésopotamie et sa relation avec les
appartenances religieuses, cf. Blanc 1964 et les travaux de Jastrow.
4
Sansă parleră desă différencesă d‘usage,ă d‘ordreă pragmatiqueă ouă énonciatif,ă tellesă queă lesă variétésă
oralesăouăécritesăparăoppositionăauxăvariétésă‗moyennes‘.
160
foisă laă questionă cară touteă donnéeă différenteă auă plană deă l‘écologieă socialeă estă
susceptibleă d‘avoiră uneă influenceă directeă ouă indirecteă sură lesă réalisationsă
linguistiques.ă Siă l‘onă espèreă cependantă aboutiră àă un résultat, il me semble
nécessaire de prendre deux précautions.
La première serait de définir le mieux possible le facteur externe dont on veut
évaluerălesăeffets.ăC‘estălaăraisonăpourălaquelleăjeăprendraiălaănotionădeăpériphérieă
dans son emploi le plus simpleă :ă celle,ă spatiale,ă deă pourtoursă d‘uneă région,ă enă
l‘occurrenceă ceuxă deă l‘aireă linguistiqueă arabe 5. Dans cette optique, il me semble
qu‘uneă définitionă trèsă approximativeă duă centreă (leă nordă deă laă Péninsuleă arabiqueă
ou même, plus largement, le Proche-Orient) ne soit pas vraiment un obstacle à la
réflexionă:ăilăs‘agiraădeăsavoirădansăquelleămesureălaăpériphérieăgéographiqueăest,ă
ou non, associable à un type dialectal spécifique ou à un type de changement
linguistique particulier.
La seconde serait de ne pas exclure a posteriori – et sans réelle discussion –
certainsă dialectes,ă sousă prétexteă qu‘ilsă s‘intègrentă mală dansă leă tableauă général,ă
alorsă mêmeă qu‘ilsă répondentă auă départă auă critèreă retenu 6. Je pense, après mon
maître David Cohen, que les exceptions sontăsouventăinstructivesăetăqu‘onăseăpriveă
d‘élémentsă intéressantsă enă rejetantă uneă partieă desă données.ă C‘estă d‘ailleursă l‘uneă
desă raisonsă qu‘onă aă d‘étudieră lesă dialectesă périphériquesă puisqu‘onă s‘attend,ă enă
grande part, à trouver là des variétés originales, s‘écartantă plusă qu‘ailleursă desă
normes communes tout en présentant éventuellement quelques conservatismes
remarquables.
Eloignement géographique et contacts de langues
Les langues connaissent des développements très différents. Certaines, comme
le basque, semblent se transmettre au cours des siècles sans beaucoup changer de
territoire.ă D‘autres,ă commeă l‘anglaisă ouă l‘espagnol,ă neă cessentă deă gagneră deă
nouveaux locuteurs, y compris sur des terres très éloignées de leurs lieux de
naissance. Parmi les langues chamito-sémitiques,ă l‘extensionă duă domaineă deă
l‘arabeăestăassurémentăl‘unădesătraitsă lesăplusăétonnantsădeăleurăhistoire,ămêmeăsiă
sonălienăavecălaădiffusionădeăl‘islamălaărendăassezăfacilementăcompréhensible.ă
L‘aireă linguistiqueă deă l‘arabeă aă connuă certains mouvements de reculs –
notammentă dansă l‘Europeă duă sud,ă enă Espagneă etă enă Sicileă – mais surtout des
Lesă notionsă deă ‗minoration‘ă etă deă ‗stigmatisation‘ă meă semblentă plusă appropriéesă queă celleă deă
‗périphérique‘ă poură traiteră desă variétésă parléesă pară desă communautésă qu‘onă pourraită qualifieră
d‘économiquement ou de politiquement périphériques.
6
C‘estăleăsoupçonăqueăjeănourris,ăàăvoirăleătraitementăaccordéăbienăsouventăauăh}assāniyya.
5
161
mouvementsă d‘extensionă etă ilă enă connaîtă encoreă aujourd‘hui,ă pară exempleă enă
Afrique sub-saharienne.ă Enă général,ă l‘idéeă prévaută queă leă déracinementă favoriseă
lesă changements,ă maisă onă peută seă demanderă siă ceă n‘estă pasă làă uneă illusionă duă
‗centre‘ă (uneă formeă bienă communeă deă ‗centro-centrisme‘,ă siă l‘onă meă permetă ceă
néologisme).ă Lorsqu‘ună françaisă deă métropoleă entendă parleră leă françaisă duă
Québec,ă n‘aura-t-il pas spontanémentă l‘idéeă queă c‘estă leă secondă quiă aă subiă desă
transformations considérables depuis les XVII e et XVIIIe siècles, alors que le
premier n‘auraităconnuăqueădesăchangementsălimitésă?7 Il est peut-être trop tôt, de
toute façon, pour généraliser en affirmant queă l‘évolutionă estă – dans toutes les
familles linguistiques – plusă importanteă àă laă périphérieă qu‘auă centre.ă Laă thèseă
inverse, peut-êtreă moinsă probable,ă n‘estă d‘ailleursă pasă nonă plusă catégoriquementă
exclueă car,ă siă l‘innovationă naîtă auă centre,ă elleă peută seă diffuser moins largement
dans les aires latérales8.
S‘agissantăenătoutăcasăduădomaineăarabe,ăoùăbonănombreădeălinguistesătendentăàă
opposer les dialectes de la Péninsule arabique à tous les autres9,ăonăs‘attendăplutôtă
àă ceă queă laă situationă d‘ună dialecteă àă laă périphérieă deă l‘aireă linguistiqueă neă fasseă
qu‘accentueră cetteă propensionă auă changementă etă soită doncă ună facteură deă fortesă
différenciations.ăC‘estălaăraisonăpourălaquelle,ălorsăduăcolloqueădeăMadridădeă1990ă
sură lesă interférencesă linguistiques,ă ilă m‘avaită paruă importantă d‘insisteră sură leă faită
qu‘ună dialecteă commeă leă h}assāniyyaă pouvaită êtreă (géographiquement)ă
périphérique tout en étant (linguistiquement) « non-marginal » (cf. Taine-Cheikh
1994). Plutôt que de reprendre les éléments de mon exposé – que je referais plus
ouămoinsăàă l‘identiqueă malgréă lesă annéesăécouléesă –, je voudrais revenir ici plus
généralementă sură lesă effetsă attendusă deă l‘éloignementă géographiqueă etă sură lesă
facteurs secondaires qui se combinent souvent avec le facteur proprement
géographique pour en décupler les effets.
•ă Lesă languesă sontă susceptiblesă d‘évolueră indépendammentă deătouteă influenceă
externe, mais les échanges entre locuteurs parlant des langues ou des variétés
différentesăsontăuneăcauseăreconnueădeăchangement,ăqu‘elleăagisseădirectement ou
indirectement. De ce point de vue, il semble clair que les parlers de la Péninsule
arabique – en particulier ceux des nomades du centre – présentent une situation
particulière par rapport aux autres parlers du monde arabe : au moins une partie
7
Cf. les recherches en créolistique qui soulignent le fait que le français (ou plutôt : les français)
parlé(s)ă àă l‘époqueă desă coloniesă étai(en)tă bienă éloigné(s)ă duă françaisă actuelă etă qu‘ilă faută doncă
impérativement en tenir compte dans la comparaison (cf. Mufwene 2006).
8
Pourălesălanguesăbantoues,ăilăsembleăqueăceăsoitădansăleăfoyerăd‘origineăqueăl‘évolutionăaăétéălaă
plus forte – mais il est vrai que ce foyer est relativement périphérique (au nord-ouest) par rapport à
laătotalitéădeăl‘aireăactuellementăconcernéeăparăl‘ensembleădeălaăfamille.
9
Cf., dans la lignée de Ferguson (1959) et Blau (1969), les divers développements sur le néoarabe.
162
d‘entreă euxă n‘aă euă qu‘ună contactă trèsă limitéă avecă desă variétésă nonă arabes.ă Poură
tous les autres (a fortiori àăl‘extérieurădeălaăPéninsuleăarabique),ăs‘estăposéeăàăună
momentă ouă àă ună autreă deă l‘histoire,ă outreă laă questionă desă contactsă avecă desă
variétés non-arabes (parlées par les communautés conquérantes, conquises ou
simplementă voisines),ă celleă deă l‘impactă duă substrată sură laă langueă arabeă – chaque
fois que des populations non-arabophones se mettaient à adopter une variété
dialectaleă deă l‘arabeă commeă langueă deă communication. Il me semble que Zima
(2007)ă aă raisonă d‘insisteră sură laă questionă deă l‘aspectă intensifă /ă quantitatifă etă
qualitatif des contacts, mais je me concentrerai ici sur la question des substrats,
adstrats et superstrats.
•ăLorsqueădesălocuteursăparlentădesăvariétésălinguistiquesă‗transplantées‘ă(néesă
sură d‘autresă territoiresă queă leă leur),ă ilă yă aă deă fortesă probabilitésă poură queă lesă
moyensă d‘expressionă offertsă pară cesă variétésă s‘avèrent,ă auă moinsă dansă certainsă
domaines, insuffisants pour satisfaire les besoins langagiers de la communauté.
Cela peut se produire dans les champs abstraits de la communication (notamment
religieux) mais cela concerne plus fréquemment le référentiel de la vie
quotidienneă :ă l‘environnementă naturel,ă lesă activitésă humainesă particulières qui y
sont liées ou encore les us et coutumes conservés du passé. Une solution très
simple consiste alors à emprunter le vocable à une langue où cette réalité est
nommable.ă Siă laă nativisationă deă l‘arabeă seă faită dansă ună contexteă deă bilinguismeă
plus ou moins large et que le terme existe dans la langue-substrat,ă c‘estă bienă
évidemmentă àă cetteă sourceă queă lesă locuteursă puiserontă chaqueă foisă qu‘ilsă aurontă
l‘impressionădeăneăpasătrouverăunăéquivalentăenăarabe.
Sur ce point, le h}assāniyyaăn‘estăcertainementăpas exceptionnel, mais il illustre
bien le principe précédemment énoncé. Alors que la totalité du vocabulaire
‗général‘ăestăpratiquementăd‘origineăarabe,ăilăestăfacileăd‘énumérerătouteăuneălisteă
de lexèmes, en particulier parmi les nominaux, empruntés au substrat.ă L‘origineă
berbèreă quiă neă semblaită souventă queă probable,ă sură laă baseă deă laă forme,ă s‘estă
trouvéeă largementă confortéeă pară l‘étudeă récenteă duă zénagaă (cf.ă Taine-Cheikh
1997). La comparaison des lexiques a également montré que le transfert du
berbère à l‘arabeă seă faisaită quasimentă sansă aucună changementă deă sensă etă qu‘ilă
portait régulièrement sur des unités lexicales de compréhension riche mais
d‘extensionăréduite.ăExemplesă:ăzén.ăi‘gəgi « variétéăd‘arbreăépineux,ăCombretum
aculeatum » > h}ass. i(i)kīk ; zén. to‘gugum}t / täwgugum}t « mouche tsé-tsé » >
h}ass. tagūgāmət ; zén. a‘gäyäf « vent charriant de la poussière (notamment avant
lesăpluiesăd‘hivernage) » > h}ass. āgäyāv ; zén. täš(š)äġyi‘d « louche, cuiller (faite
163
dans une courge) » > h}ass. tāšəġlīt ; zén. äšugdäy « socleă(deăl‘enclume) » > h}ass.
äšəgdäl10.
Onăpourraităreleverăunăcertainănombreăd‘empruntsăconcernantăl‘élevageăbovin,ă
alors que celui des camelins correspond, en h}assāniyya,ă àă ună lexiqueă d‘origineă
arabe,ămaisăilăneăfaudraităpasăenăconclureăqueăleăzénagaăn‘aăpasădeălexiqueăpropreă
pour les chameaux : simplement, comme il existait déjà en arabe, le dialecte arabe
n‘aăpasăeuăbesoinădeăl‘emprunterăauăberbère.ăAuătotal, leăpourcentageăd‘empruntsă
faits par le h}assāniyyaăauăsubstratăberbèreăn‘estăsansădouteăpasănégligeable,ămaisăilă
n‘estă pasă nécessairementă trèsă élevéă comparéă àă d‘autres 11 : il existe trop de
similarité entre le mode de vie des Maures nomades et celui des bédouins de la
Péninsule pour que la majorité du vocabulaire ne soit pas transposable (même si
celaăposeăparfoisăquelquesăproblèmesăd‘adaptation,ăcommeăpourălaăroseădesăvents,ă
cf. Taine-Cheikh 1991).
•ă Dansă ună certaină nombreă deă cas,ă l‘éloignementă géographique a finalement
abouti à une coupure avec le reste du monde arabe. En ce qui concerne les
populations du Sahara ouest-atlantique, on peut penser – vuăleăpeuăd‘informationsă
dontăonă disposeă sură ellesă pendantă plusieursă siècles,ă quasimentă jusqu‘auă débută duă
XVIIIe siècle – qu‘ellesăn‘ontăeu,ădurantăuneătrèsălongueăpériode,ăqueădesăcontactsă
très limités avec le reste du monde (monde arabe compris, en dehors des régions
méridionales du Maroc). Cependant, même si les liens furent, pendant des siècles,
très distendus avec le Moyen-Orient et la Péninsule arabique, on ne pourra pas
parleră deă véritableă coupureă cară laă pratiqueă d‘uneă mêmeă religionă etă leă partageă deă
références culturelles largement communes ont fortement contribué à préparer ces
populationsă àă l‘intégrationă dans une grande ‘umma réunissant toutes les
populations musulmanes, a fortiori s‘agissantădesăpopulationsăarabophones.ăAlorsă
queă laă meră deă sableă freinaită lesă contactsă sansă lesă empêcher,ă c‘estă laă Meră
Méditerranée qui a finalement constitué, pendant cette même période, une barrière
autrement plus efficace. Parmi les régions qui avaient été arabisées lors de la
conquêteămusulmaneădesăpremiersăsièclesădeăl‘Hégire,ăcellesăquiăétaientăsituéesăduă
côté européen ont, pour la plupart, été reconquises par les langues indoeuropéennes latines (espagnol ou italien). Il reste toutefois le cas particulier de
Malteăoùăl‘arabeăaăperduré,ădevenantămêmeăleăseulă‗dialecte‘ăarabeăpromuăauărangă
10
Les emprunts de verbes sont moins fréquents (et moins faciles à détecter). Signalons : zén.
yäšša‘yäm « raboter » > h}ass. šäyläm ; zén. yuġräž « utiliseră ;ă employeră (deă l‘argentă ouă deă laă
nourriture) » > h}ass. ġrəž ; zén. yugräž « attacher ensemble » > h}ass. grəž.
11
Il faudrait pouvoir comparer le lexique du h}assāniyyaă avecă desă lexiquesă ‗traditionnels‘ă deă
dimensions comparables – avantă queă l‘introductionădeă laămodernitéăneă bouleverseă leărapportă desă
sociétés bédouines et rurales avec leur environnement.ă Laă proportionă duă lexiqueă ‗arabe‗ă dansă leă
dictionnaireă Colină d‘arabeă marocaină (éditéă pară Iraqui-Sinaceur 1993) est en tout cas plutôt
inférieureăàăceăqu‘elleăestăenăh}assāniyya.
164
deălangueăofficielleăetăemployéăàăl‘écrităcommeăàăl‘oral.ăDansălaămesureătoutefoisă
où ilă s‘agit,ă àă l‘instară notammentă deă l‘arabeă chyprioteă deă Kormakiti,ă deă variétésă
d‘arabeă parléesă dansă ună domaineă linguistiqueă hétérogèneă (enă contactă étroită avecă
des langues indo-européennes), en rupture à la fois avec la culture arabe classique
(duă faită deă l‘insertionă étroiteă dansă leă mondeă deă laă chrétienté)ă et,ă poură l‘essentiel,ă
avec les autres communautés arabophones12, il semble judicieux de parler alors
d‘arabesă « extra-périphériques » plutôt que de variétés simplement
« périphériques ».
•ăDavidăCohen,ăàăqui je reprends cette distinction, propose également de traiter
à part un troisième groupe : « Les statuts diffèrent pour un groupe périphérique
vivant aux confins, un groupe extra-périphérique inclus dans un domaine
linguistique hétérogène ou un groupe en expansion sur un terrain en voie de
conquête. Et des phénomènes linguistiques sont propres à chacun, en partie en
raisonă deă ‗l‘idéal‘ă linguistiqueă quiă estă leă sien,ă imposantă uneă normeă quiă peută
différeră selonă leă statută duă groupeăouă sonă modeă d‘expression » (D. Cohen 1994 :
16).ă Laă questionă mériteraită d‘êtreă approfondie,ă s‘agissantă enă particulieră desă
variétés dites créoles – qui sont peut-être plus précisément des pidgins créolisés 13
–, mais il me semble effectivement important de pouvoir faire un sort particulier
aux pidgins et aux arabes véhiculaires en voie de constitution. On peut en effet
s‘attendreă àă ceă queă leursă conditionsă particulièresă deă naissanceă soientă plusă
importantesă queă leă faită qu‘ilsă seă produisentă trèsă souventă sură lesă pourtoursă duă
domaine arabe.
Inversement,ăceăn‘estăpasăparceăqueăl‘aireăhassanophoneăseădéploieăauxămargesă
deă l‘aireăarabeăetăqueă lesă voisinsădesăMauresăsont,ădepuisădesăsiècles,ădesă négroafricains islamisés14, que le h}assāniyyaăaădesăemploisăd‘arabeăvéhiculaireă(saufăcasă
particuliers,ă jusqu‘àă présentă fortă limités),ă niă qu‘ilă aă faită quelqueă empruntă notableă
aux langues africaines non chamito-sémitiques. Contrairement à une idée assez
répandue, le h}assāniyyaă neă doită pratiquement rien à ses voisins du sud au plan
linguistique. Cela se comprend dans la mesure où la société maure a développé
uneăcultureăconstruiteăautourăd‘unădoubleăhéritageăarabeă(celuiăduădialecteăetăceluiă
deă l‘arabeă classique).ă Elleă aă entretenuă desă échanges (notamment économiques)
avec ses voisins et partagé avec eux certaines affiliations religieuses (en
D‘oùă l‘absenceă d‘innovationsă queă l‘onă peută constateră pourtantă d‘ună boută àă l‘autreă deă l‘aireă
linguistiqueă arabe,ă ainsiă l‘emploiă deă šāv comme expression du verbe « voir » (notamment en
remplacement de r }a (a)) dansălaăplupartădesădialectesăarabes,ăàăl‘exceptionădeăcommunautésăplusăouă
moins isolées (cf. Kaye 1986).
13
Du moinsăsiăl‘onăconsidère,ăcommeăMufweneă(2006)ănousăinviteăàăleăfaire,ăqueălaănaissanceădeă
véritables créoles est liée à une écologie sociale très spécifique (celle des colonies de plantation).
14
Ils ont pour langue maternelle le pulaar, le wolof, le soninké ou le bambara.
12
165
particulieră confrériques),ă maisă elleă s‘estă forgéă ună idéală qui,ă sură bienă desă pointsă
(surtoutăconcernantălaălangueăetă‗l‘arabité‘ăgénéralogique),ăleurătourne le dos.
Les dialectes maghrébins et l’influence du substrat berbère
Dans le cas des variétés « extra-périphériques » comme dans celui des variétés
« en expansion », les contacts linguistiques exogènes semblent jouer un rôle
déterminant,ămêmeăs‘ilăs‘agit plutôtăd‘adstrată(ouăsuperstrat)ăpourălesăpremièresăetă
de substrat pour les secondes 15 .ă Qu‘enă est-il pour le troisième groupe défini
comme seulement « périphériques » ? Afin de répondre à cette question, je me
proposeă deă prendreă l‘exempleă deă l‘arabeă maghrébin (maltais non compris) et
d‘essayeră deă voiră dansă quelleă mesureă leă substrată berbèreă peută êtreă associéă auxă
changements observables16.
On ne peut pas, certes, limiter les contacts au seul substrat, notamment dans le
nordădeăl‘Afriqueăoùălesăhabitantsăn‘ont cesséăd‘avoirădesărapportsăplusăouămoinsă
durables – mais souvent inégaux – avec les populations du pourtour de la
Méditerranée (des Latins aux Italiens, sans oublier les Turcs, les Espagnols et les
Français) 17 .ă Ilă n‘enă resteă pasă moinsă que,ă enă dehorsă desă emprunts lexicaux,
l‘évolutionă deă l‘arabeă maghrébină – jusqu‘àă uneă périodeă asseză récenteă – semblait
marquée,ăpourăl‘essentiel,ăparăsesărelationsăétroitesăavecăle(s)ăberbère(s),ămêmeăsiă
les autres langues ont occasionnellement laissé des traces18.
Un inventaire rapide point par point nous permettra de voir si le substrat peut
être relié à certains changements, partiellement ou en totalité. Il ne faut pas
s‘attendreăenăeffetăàăceăqu‘onăpuisseă isoleră facilementălesăcausesăresponsablesădeă
changement. On verra que le principe des causes multiples, valable en
créolistique, se vérifie très souvent dans les dialectes maghrébins : « Substratum
always works beautifully when it reinforces what develops via normal
15
Voir notamment, dans la publication des actes du colloque de Madrid (1994), les articles de
BorgăetăInghamăd‘uneăpart,ădeăKayeăd‘autreăpart.
16
Le maltais fait partie du groupe dialectal « occidental »ă(maghrébinăetăandalou).ăSiăjeăl‘exclus ici
(toută commeă l‘arabeă andalouă ouă sicilien),ă c‘estă d‘abordă enă tantă qu‘arabeă « extra-périphérique »,
mais aussi parce que le berbère – quiăn‘estăpas,ădansăceăcas,ăunăsubstrată– semble avoir eu sur lui
un impact très limité.
17
Sur les métissages linguistiquesăquiăontăjalonnéăcetteălongueăhistoire,ăcf.ăl‘ouvrageăcollectifăéditéă
par Dakhlia (2004).
18
Ainsi le suffixe -əš deădiminutifă(quiăexisteăaussiăenăberbère)ădontăl‘origineăpourraităremonterăàă
la terminaison latine -us ou le suffixe -ži desă nomsă d‘artisan – ex. qahwa « café » > qahwāži
« cafetier » – qui vient du turc (cf. Marçais 1977 : 114).
166
evolutionaryă ‗laws‘ă andă tendanciesă andă viceă versa » (Kaye 1994 : 132)19. Dans
bienădesăcas,ălesămêmesăfaitsăontăétéărelevésăenăd‘autresăpointsăduădomaineăarabe,ă
ceăquiăpourraităsemblerăexclureăuneăactionăduăsubstrat.ăSiăl‘onăretientăcependantăleă
principeădesăcombinaisonsădeăfacteurs,ăl‘influenceăduăsubstratăresteăenvisageable,
même si elle devient particulièrement difficile à cerner.
Par ailleurs, on ne peut oublier le fait que, parmi toutes les langues avec
lesquellesă l‘arabeă estă enă contact,ă certainesă sontă sémitiquesă (ainsiă l‘araméenă ouă
l‘hébreu)ăetăd‘autresăsontăchamito-sémitiquesă(c‘estăleăcasăduăcopteăetăduăberbère).ă
Des ressemblances typologiques importantes sont attestées entre les différentes
rameauxădeăl‘ensembleăetăpeuventăseăretrouverăenădiachronie,ăpuisqueălesălanguesă
apparentées sont susceptibles de présenterădesăchangementsăsimilaires.ăL‘arabeăaă
donc a priori plus de facilité de faire des échanges limités – etă d‘évolueră enă
parallèle – avec des langues appartenant à la même famille ou au même phylum,
qu‘ilă n‘enă aă deă leă faireă s‘ilă entretientă desă contactsă étroits avec des langues
structurellement très différentes (telles les langues indo-européennes et ouestatlantiques).
1. Phonétique et phonologie
Les systèmes sont variés et les réalisations, multiples. Quelques traits semblent
plus intéressants à discuter queăd‘autresăduăpointădeăvueădeălaăquestionăduăsubstrat,ă
maisă laă comparaisonă arabeă /ă berbèreă estă d‘autantă plusă complexeă queă lesă parlersă
berbèresă sontă aussiă différenciésă (siă ceă n‘estă plus)ă queă lesă parlersă arabesă
maghrébins.
•ă Laă réalisationă duă qāf estă l‘un des critères de classement des dialectes, si ce
n‘estă leă seulă véritablementă général.ă Laă questionă deă l‘influenceă éventuelleă duă
substrat ne concerne que le choix des réalisations sourdes. En effet, alors que les
parlersă deă nomadesă maghrébinsă (‗hilaliens‘ă au sens large) ne connaissent guère
que la réalisation occlusive de la sonore – à la différence des parlers de nomades
orientaux,ă oùă laă palatalisationă etă l‘affricationă duă g sont fréquentes –, les parlers
‗non-hilaliens‘ă enă connaissentă auă moinsă troisă différentes, toutes occlusives :
vélaireă[q],ălaryngaleă[‘]ăetăpostpalataleă[k}], plus postérieure que le [k] normal (cf.
Marçais 1977 : 11). La réalisation laryngale (largement attestée en dehors de
l‘Afriqueă duă Nord,ă queă ceă soită àă Malte,ă enă Egypteă ouă auă Liban)ă se rencontre au
Maghreb,ănotammentădansădesăparlersăcitadins,ămaisăonăaăd‘autantămoinsădeăraisonă
deă l‘associeră auă substrată berbèreă queă laă laryngaleă occlusiveă estă pratiquementă
inconnue en berbère (en dehors du zénaga de Mauritanie). La réalisation [k}], plus
19
Voir par exemple les discussions autour de certains traits caractéristiques du Ki-Nubi et du Juba
Arabicădontăl‘articleăd‘Avramă(2004)ămontreătousălesăenjeuxă– telăl‘emploiăduăverbeăgale « dire »
comme complémentiseur.
167
fréquente semble-t-ilă auă Maghrebă qu‘ailleursă (cf.ă Taine-Cheikh 1998-1999) peut
parăcontreăavoirăétéăfavoriséeăparăleăsubstratăberbère.ăC‘estăuneăhypothèseăavancéeă
parăLévyă(1998ă:ă21)ăquiăsembleăcohérenteăavecălesăprincipauxălieuxăd‘attestationă
de cette réalisation20.ă Elleă s‘appuieă sură leă statută ambiguă duă qāf en berbère, qui
apparaît essentiellement comme la correspondante géminée ou « tendue » du ġayn
(ġġ se réalise toujours [qq]).
L‘alternanceă ġayn simple-qāf géminé, qui se retrouve dans des parlers arabes
(notamment de sédentaires, mais pas seulement), peut être vue comme un fait du
substrat.ăCeătraită‗berbère‘,ăquiăpourraităavoirăpréservéăcertainsăparlersăbédouinsădeă
la mutation du ġ en q21, aurait ainsi contribué indirectement à la constitution, au
seinădesăparlersă‗hilaliens‘,ădeădeuxăgroupes.ăCurieusement, cette ligne de partage
traverse le h}assāniyyaă :ă onă aă d‘uneă partă celuiă deă l‘est,ă quiă neă connaîtă queă laă
réalisation qāf comme la plupart des parlers sahariens, depuis le Sud oranais
jusqu‘àălaăLibyeă(cf.ălesăparlersăAădeăCantineauă1941)ăetăd‘autreăpart les parlers à
alternance ġ / qq, tel le h}assāniyyaă deă l‘ouestă (etă cela,ă bienă queă laă vélaireă ġ soit
rare dans le zénaga actuel22).
•ăL‘autreămacro-discriminant dialectal est celui de la conservation, ou non, des
interdentales (cf. Taine-Cheikh ibid.). Les dialectes à réalisation sourde du qāf les
ont généralement perdues. Certaines aires font cependant exception, soit en
totalitéă (ainsiă l‘aireă mésopotamienneă enă qāf a-t-elle conservé les réalisations
interdentalesă et,ă àă l‘inverse,ă l‘aireă tchado-soudanaise en gāf les a-t-elle perdues),
soit tendanciellement (telles les interdentales massivement maintenues en Tunisie
ou massivement perdues au Maroc), soit même plus ponctuellement (cf. les
interdentalesă deă quelquesă vieillesă citésă algériennes).ă Dansă l‘ensembleă des parlers
maghrébins,ălesădeuxăcasăsontăbienăreprésentés,ăilăn‘yăaădoncăpasălieuăd‘yăvoirăuneă
influence directe du substrat. Tout juste peut-on se demander, la disparition étant
beaucoupă plusă généraliséeă àă l‘ouestă qu‘àă l‘estă (h}assāniyyaă misă àă part),ă s‘il n‘yă aă
pas – poură l‘aireă maghébineă septentrionaleă –, un effet de la position
‗périphérique‘.
Par ailleurs, il faut examiner le cas des interdentales dans le substrat. Pour être
dépourvuesădeăstatutăphonématique,ăellesă n‘enă sontăpasă moinsă attestéesădansăune
partie du domaine berbère comme variantes phonétiques des dentales. Dans les
parlers dits « spirants », réalisations occlusives et réalisations fricatives sont en
20
Parlers des juifs de Debdou, Tafilalt et Coude Draa au Maroc et, en Algérie, parlers de ruraux
(Djidjelliăetăl‘estădeălaăKabylieăd‘uneăpart,ărégionădesăMsirdaăetădesăTraraăd‘autreăpart).
21
Cette mutation est très fréquente dans les parlers bédouins orientaux.
22
Le ġayn est attesté dans des emprunts du h}assāniyyaă auăzénaga,ă ceă quiă sembleă indiqueră queălaă
disparition de la réalisation fricative est relativement récente (cf. h}ass. äġân « corde à puits en cuir
très longue » / (<) zén. a‘ni « corde pour puiser »). Par ailleurs la tendance à remplacer q par ġ
existe dans le h}assāniyyaădeăl‘ouest,ăcp.ăġandîr « bougie » / cl. qandīl.
168
distribution complémentaires. Cela concerne les dentales apicales, mais aussi –
souvent – d‘autres consonnes (bilabiales comme b, palatales comme k et g, ...).
Réalisées comme des occlusives quand elles sont géminées ou en certaines
positionsă (àă l‘initiale,ă aprèsă n, ...), elles sont réalisées spirantes dans les autres
positions,ă notammentă àă l‘intervocalique. En général, que le substrat berbère soit
unăparlerăspirantăouănonăneă sembleăpasă avoirăd‘impactăsură leă maintien,ăenăarabe,ă
des interdentales 23 . Par contre il en a un, très net, sur la présence de diverses
réalisations spirantisées ou affriquées des occlusives du type b, d, k et même, plus
rarement, d} et t. (par contre ts est une réalisation fréquente de t en Algérie et au
Maroc).ă Laă présenceă d‘interdentalesă dansă certainsă parlersă marocainsă duă Nordouest ne reflète donc pas des faits de conservatisme, mais elle a pu, dans un cas au
moins, aboutir à une phonologisation des variantes (cp. la situation à Cherchaouen
etăcelleăd‘AnjraăauădébutăduăXXeăsiècleăd‘aprèsăNatividadă1998ăetăVicenteă1998).
•ăOutreălesăphénomènesădeăspirantisationă(signalésădansă divers parlers comme
celui, algérien, de Djidjelli, cf. Marçais 1956 : 3 et sq.), le substrat berbère a pu
avoiră ună effetă sură lesă emphatiques.ă Lesă parlersă berbèresă connaissentă l‘emphase,ă
maisă ilsă présententă desă particularités.ă Anciennementă ilă n‘existaită que deux
phonèmes emphatiques, toutes deux sonores : la dentale d} et la sifflante z}. Les
sourdes sont des variantes contextuelles des sonores (marginalement
phonologisées). Il est fréquent que d} s‘assourdisseă spontanémentă lorsqu‘elleă estă
‗tendue‘,ămaisăon trouve aussi des s et des t emphatisés du fait du contexte car, en
berbère,ăl‘emphaseăs‘étendărégulièrementăauxăautresăconsonnesăduămot.ăCesădiversă
phénomènesăneăsontăpasăsansăconséquenceăpourăl‘arabe.ăLesăassourdissementsăetă
lesă fluctuationsă deă l‘emphase du t} (et même du s}) semblent dues au substrat (cf.
Natividad ibid. : 110, Marçais ibid. : 6-8, Vicente ibid. : 12524). Par ailleurs, le
substrată aă puă favoriseră laă fréquenteă diffusionă deă l‘emphaseă enă arabeă maghrébină
(même si elle existe aussi ailleurs) et,ă indirectement,ă l‘apparitionă deă nouveauxă
phonèmes emphatiques – sans oublier le cas de z}, qui apparaît souvent dans les
emprunts faits au berbère et, parfois, dans des lexèmes particulièrement expressifs
(cf. Cohen 1963 : 22).
•ăEnăberbère,ălesăparlersămarocains – et notamment la tachelhit – se caractérise
par une raréfaction des éléments vocaliques et une très haute fréquence de
syllabes à double coda. Les successions de plusieurs consonnes admettent les
Cf. leur présence en h}assāniyyaăetăleurăabsence,ăcommeăphonèmesăàăpart entière, dans la plupart
des parlers de sédentaires maghrébins.
24
Mêmeă siă laă positionă deă Zavadovskijă (1981ă :ă 27)ă sură l‘inexistenceă duă traită d‘emphaseă enă
h}assāniyyaă sembleă intenable,ă ilă n‘enă resteă pasă moinsă queă l‘oppositionă s/s} montre quelques
fluctuationsă remarquables.ă Voiră aussiă leă problèmeă deă l‘emphaseă duă t évoqué par Cohen (1963 :
16), même si elle semble, en dehors des analphabètes, surtout caractéristiques des berbérophones
deălaărégionădeăl‘Iguidi.
23
169
combinaisonsăd‘uneătendueăetăd‘uneăsimpleăenăattaque ou en fermeture de syllabe
(ce qui est impossible dans un parler comme le zénaga où, par ailleurs les syllabes
ouvertes sont assez bien représentées – etămêmeăbeaucoupăplusă fréquentesăqu‘enă
h}assāniyya).ă
Ilă sembleă enă faită qu‘onă puisseă établiră globalement un certain parallélisme, au
Maghreb,ăentreălesăstructuresăsyllabiquesădeăl‘arabeăetăduăberbèreăenăfonctionădesă
différentesă airesă régionales.ă C‘estă clairă poură laă Mauritanieă (àă l‘exceptionă prèsă
signalée précédemment) et cela semble vrai du Maghreb. On a observé,ă d‘uneă
façonăgénérale,ăuneădétériorationăcroissanteăduăvocalismeăd‘estăenăouestă:ăl‘arabeă
marocain, plus encore que le berbère marocain, se distingue des variétés plus
orientales par un bouleversement de la structure syllabique. Il est difficile de
savoir, dans la mesure où le berbère méridional (zénaga, mais aussi touareg)
présenteă uneă syllabationă asseză différenteă deă celleă deă l‘ensembleă ‗septentrional‘ă –
spécialement peut-être du berbère marocain –,ăsiăc‘estăl‘évolutionăduăberbèreăquiăaă
entraîné celleădeăl‘arabeăouăs‘ilăneăs‘agităpasăplutôtăd‘uneăévolutionăparallèle.
Quoiăqu‘ilăenăsoit,ălaăruineăduăvocalismeăbrefăpeutăavoirăeuădesăconséquences,ă
non seulement sur la structure syllabique, mais encore sur la labiovélarisation des
dorsales. Elmedlaoui (2002), constatant en effet la présence alternante de g/gw et
k/kw en berbère chleuh et en arabe marocain, voit dans la labiovélarisation
(apparueă d‘abordă enă berbère,ă enă relationă avecă l‘alternanceă presqueă panberbèreă
entre le /w/ simple et un [ggw] géminé) uneăconséquenceădeălaăchuteăd‘uneăvoyelleă
u (p. 169 : « laă récupérationă duă timbreă bémolă d‘ună ‗U’ qui tombe, par la classe
‘K’ ») : berb. « enclume en pierre » *assarug > assargw pl issurag. Il appuie sa
démonstrationă sură laă comparaisonă deă l‘arabeă marocaină avec le classique :
« s‘asseoir » g‗əd Impér. gw‗əd ! (cf. cl. qa‗ad Impér. ‘uq‗ud !).
•ă S‘agissantă desă systèmesă vocaliques,ă onă noteraă qu‘enă berbèreă l‘existenceă
historiqueă d‘uneă oppositionă deă longueură estă loină d‘êtreă prouvée.ă Enă dehorsă desă
anciennes diphtongues et de certains schèmes innovants attestés en particulier en
touareg, la présence de voyelles longues phonologiques pourrait être réservée à
certainsă dialectesă (ceuxă deă l‘ensembleă méridional,ă plusă ceuxă duă Rif).ă Ellesă
semblent compenser la disparition de certaines consonnes comme le r en rifain et,
ailleurs, les laryngales (cf. Taine-Cheikhă 2005).ă L‘existenceă deă parlersă arabesă
dépourvusă d‘oppositionă deă longueură pourraită doncă êtreă miseă enă relationă avecă leă
substratăberbère.ăPourăl‘arabeăcommeăpourăleăberbère,ăc‘estăleăMaghrebăoccidentală
septentrională (leă Marocă etă l‘ouestă deă l‘Algérie)ă quiă sembleă concernéă enă premieră
lieu.ă Onă auraită donc,ă outreă l‘oppositionă entreă parlersă deă nomadesă etă parlersă deă
sédentaires du point de vue des diphtongues (mieux conservées en général dans le
premier groupe) et du point de vue du vocalisme bref (tendance à la confusion de i
et u chez les premiers, de a et i chez les seconds, cf. Cohen 1970 : 176-7), une
secondeăligneădeăpartageăquiăpasseraităcetteăfoisăentreălesăparlersă‗non-hilaliens‘ădeă
170
l‘ouest,ăd‘uneăpart,ăetătousălesăautres,ăd‘autreăpartă(parlersă‗hilaliens‘ădeăl‘ouestăetă
parlersă maghrébinsă deă l‘estă enă général).ă Laă questionă deă laă disparitionă deă
l‘oppositionă deă longueură neă faită pasă l‘unanimité,ă maisă elleă neă faită pasă deă douteă
pour bon nombre de dialectologues.
Cetteă évolution,ă commeă celleă deă laă détériorationă duă vocalismeă bref,ă n‘estă
certainementă pasă sansă lienă avecă laă questionă deă l‘accentă (accentă marquéă
préférentiellementă pară laă durée,ă secondairementă pară l‘intensité,ă dansă ună dialecte
comme le h}assāniyyaă – ce qui contribue à donner une durée extra-longue aux
voyelles longues accentuées –,ăalorsăqu‘ilăsembleămarquéăpréférentiellementăparălaă
hauteur,ăsecondairementăparăl‘intensité,ăauăMaghrebăoccidental,ăcf.ăTaine-Cheikh
1997 : 97 et sq.). En tout cas, si la structure syllabique et le vocalisme contribuent
à caractériser assez nettement les parlers maghrébins dans leur ensemble, ils
semblent présenter une évolution particulièrement nette dans sa partie
périphérique septentrionale.
2. Morphosyntaxe
•ăS‘agissantăduătraităcaractéristiqueădesăparlersămaghrébinsă(laăgénéralisationăduă
n- commeă marqueă deă 1èreă pers.ă deă l‘inaccompliă et,ă deă manièreă concomitante,ă
l‘extensionă duă -u(u) comme marque de pluriel à la 1ère pers.), on peut rejeter
toute influence du substrat puisque, dans tous les parlers berbères, le n- préfixal
est resté la marque spécifique de 1ère pers. du pluriel, comme dans les parlers
arabesăd‘Orient.25
•ă Dansă lesă dialectesă arabes,ă certainsă nominauxă sontă dotésă d‘affixesă berbères
(préfixe a- pour le masculin, affixes du type t- ...-t pour le féminin, pluriel
masculin en i-...-n, pluriel féminin en t-...-n). En principe, ces affixes sont réservés
auxă seulsă nominauxă d‘origineă berbèreă etă leură présenceă exclută souventă celleă deă
l‘article défini (surtout au masculin). Il arrive que ces affixes entrent dans la
formationă deă nominauxă d‘origineă arabe,ă notammentă dansă leă vocabulaireă abstrait,ă
cependant ce phénomène semble surtout attesté dans les dialectes arabes où
l‘empreinteă duă berbèreă estă importante (cf. takəbburt « orgueil » à Djidjelli,
Marçais 1956 : 406).
•ăDansă leăcasăduăduel,ăl‘influenceăduăsubstratăn‘estăpasănetteăcarăleăduelăestăenă
recul dans tous les dialectes arabes : de manière générale pour ce qui est des
marquesă d‘accordă dansă leă verbe,ă l‘adjectif ou les pronoms, de manière
différenciée pour ce qui est des noms. Dans les parlers de nomades comme le
h}assāniyya,ăleăduelăestărestéătrèsăusuelă(duămoinsăenăl‘absenceădeăl‘articleădéfini).ă
Leă faită mériteă d‘autantă plusă d‘êtreă préciséă que les formes niktib-niktibu attestées en Égypte
occidentale peuvent être associées à un afflux ancien de tribus maghrébines (pour plus de détails,
cf. Behnstedt 1998).
25
171
Dansă lesă parlersă ‗non-hilaliens‘,ă l‘usageă estă souventă limitéă à quelques nominaux
commeăceuxădesăpartiesăduăcorps,ăsansăqu‘onăpuisseăaffirmerăqueăc‘estăunăeffetăduă
substrat.ă Leă duelă n‘existeă effectivementă pasă (ouă plus)ă enă berbèreă etă n‘aă doncă puă
joueră qu‘enă faveură deă laă disparitionă duă duelă mais,ă pară ailleurs,ă cetteă tendance
semble universelle.
•ăLaăquestionădesămarquesădeăgenreădansălesăverbesăetădansălesănomsăillustreăună
autre cas de figure. Là encore on observe une moindre différenciation de
l‘oppositionămasculină(M)ă/ăféminină(F),ădeămanièreăgénérale,ădansălesăparlers de
sédentairesă et,ă làă encore,ă onă pourraită seă contenteră d‘évoqueră uneă tendanceă
universelle.ă Siă l‘onă compareă pourtantă leă systèmeă desă pronomsă personnelsă
autonomes dans des dialectes aussi différents que celui de Djidelli et celui de
Mauritanie, on constate des faits intéressants, apparemment contradictoires. Ils se
situent pratiquement aux deux extrêmes : à Djidjelli, la différenciation M / F ne se
faită qu‘àă laă 3e pers. du singulier ; en h}assāniyya,ă elleă seă faită auxă 2 e et 3e du
singulier (sg.) et du pluriel (pl.). Pourtant, dans les deux cas, le substrat berbère
pourrait avoir exercé une influence.
D‘uneă part,ă leă systèmeă berbèreă seă caractériseă pară uneă moindreă différenciationă
M / F à la 2e sg.ă:ăàălaădifférenceădeăl‘arabeăancien,ăelleăn‘existeăpasăpourăl‘indiceă
personnel du verbe et cela a pu jouer un rôle dans la disparition (assez rare en
dehors du Maghreb) de la marque du 2e F sg. à Djidjelli (dans le verbe comme
dans les pronoms, cf. Marçaisă 1956ă :ă 155,ă 435,ă 449).ă D‘autreă part,ă leă systèmeă
berbère se caractérise par la différenciation M / F des 2 e et 3e pl. 26 – d‘oùă laă
présence (assez rare, même en dehors du Maghreb, cf. Fischer & Jastrow 1980 :
79) de formes différenciées en genre au pl. en h}assāniyyaă (cf.ă 3e pl. M hûmä F
hūmâti).
Pourquoi invoquer le substrat, la distinction de genre existant aux 2 e et 3e pl. en
arabe ancien ? Cohen (1963 : 147-8) avait observé que le h}assāniyyaă avaită
renouveléă lesă formesă duă systèmeă :ă c‘estă l‘ajout,ă auxă formes de M pl., du
morphème -āti (qu‘onă retrouveă dansă lesă adverbesă deă lieu)ă quiă permetă laă
différenciationă enă genre.ă Oră ilă seă trouveă qu‘enă berbère,ă l‘oppositionă deă genreă
comporteăaussiăl‘ajout,ăauăféminin,ăd‘unăsuffixeăàădentaleă(-ti en chleuh, -nn / mn
+yät en zénaga). Dans le pronom affixe et dans le verbe, la marque de genre à ces
mêmes personnes est à peine plus réduite en zénaga (-n +yät ) : était-elle à
l‘origineătropă‗légère‘ă(cf.ă-t en chleuh) pour favoriser le maintien de la distinction
en h}assāniyya ? Ces oppositions sont tellement rares dans les dialectes arabes, a
fortiori auă Maghreb,ă queă leură disparitionă n‘estă pasă surprenante.ă Pară contre,ă leă
réemploiă duă suffixeă duă féminină poură l‘étoffementă desă adverbesă deă lieuă (cf.ă
h}assāniyyaăvämm / vämmāti / vämmātiyyä « là ») pourrait être lié au substrat car
26
Elle est même souvent étendue, dans le cas des pronoms personnels autonomes, à la 1 e pl.
172
le lien entre pronoms personnels et adverbes de lieu est récurrent en berbère (cf.
Galand 2002 : 187-191).
L‘absenceădeădifférenciationădeăgenreădansă laăconjugaisonăverbale,ăàălaă2 e sg.,
est intéressante car elle tend à faire correspondre assez nettement périphérie
géographique et périphérie linguistique. En Algérie, le cas de Djidjelli est
relativementăisoléă:ăilăneăpartageălaăconfusionădesăgenresăqu‘avecăquelquesăparlersă
citadins non influencés par les parlersă ‗hilaliens‘ă commeă Nédromaă (cf.ă Marçaisă
E.I.2 : 367).ăAuăMaroc,ăparăcontre,ălesăparlersăatypiquesăsontălesăparlersă‗hilaliens‘ă
(ouăplutôtă‗ma‗qiliens‘)ăquiăcontinuentăàădistinguerălaă2 e sg. M et F, non seulement
dans les pronoms, mais encore dans le verbe (comme celui de Skûra, cf.
Elyaâcoubi 1998 : 153-4 ou le h}assāniyyaăduăSudămarocain).ăLesăautresăparlersăneă
font plus systématiquement la distinction de genre à la 2 e sg.,ă siă ceă n‘estă àă
l‘impératifă :ă niă lesă parlersă deă sédentaires,ă niă leă parleră desă Z‗îră ouă celuiă deă
Casablancaă classésă pourtantă commeă ‗hilaliens‘ă (cf.ă Aguadéă 1998ă :ă 144-527). Ce
quiă estă surprenant,ă c‘estă queă ceă soită laă formeă enă -i réservée normalement au
féminin qui soit utilisée pour les deux genres, passant semble-t-il de certains
parlersă citadinsă auxă parlersă ‗hilaliens‘ă pară l‘intermédiaireă deă laă koinéă (elleă n‘estă
pas généralisée toutefois28).
•ă Cesă effetsă contradictoiresă duă substrată neă sontă peut-être pas limités à
l‘oppositionădeăgenre.ăSiăl‘onăconsidèreăparăexempleăleăsystèmeădesăracines et des
schèmes,ăilăapparaîtăassezănettementăqu‘enăberbère,ăparărapportăauăsémitique,ă« le
schèmeă n‘estă plusă toujoursă ‗parlant‘,ă surtoută dansă leă systèmeă nominal,ă moinsă
fortement structuré que le système verbal » (Galand 1983 : 465). Cette tendance à
la dissolutionă desă schèmes,ă renforcéeă pară lesă effetsă deă l‘évolutionă phonétique,ă
n‘estă pasă limitéeă auxă dialectesă berbères 29 :ă elleă sembleă aussiă s‘êtreă produiteă enă
arabe maghrébin, en particulier dans les parlers de sédentaires. Divers traits
relevés dans ces dialectes montrent que le système des schèmes a évolué en
s‘éloignantăduăsystèmeăancien,ăprincipalementăparăconfusionădeăcertainsăschèmesă
(tel le pl. des quadriconsonantiques : mfātīh} > mfātəh} « clés »), par abandon des
alternances vocaliques au profit des affixes (ainsi pour le pl. des adjectifs : kbār
QueălaădistinctionăperdureăchezălesăZ‗îrăàăl‘impératifămontreăqueăleăreculăestăprogressifă(cf.ăaussiă
leă tableauă contrastifă dresséă poură Fèsă dansă Caubetă 1998).ă Laă pressionă deălaă koinéă estă forteă etă j‘enă
avais aussi observé quelques effets dans la communauté hassanophone de Goulimine (cf. TaineCheikh 1999).
28
Cf.,ădansăleădialecteădeăChefchaouen,ăl‘absenceădeăsuffixeă-i dans la 2e sg. commune, quand bien
même la forme de 2e sg. commune du pronom autonome est, elle, en -i (cf. Natividad 1998 : 1127).
29
Peut-être faut-il nuancer l'affirmation pour le berbère méridional. En touareg, les schèmes
semblent avoir mieux résisté, y compris dans le domaine nominal. En zénaga, ils sont bien vivants
dans le domaine verbal.
27
173
>kbīrīn à Chefchaouen, cf. Natividad 1998 : 118) et par perte des oppositions
distinctives (cf. la non distinction entre les formes réfléchies et les formes
passives). Cette évolution semble particulièrement forte au Maghreb occidental,
commeăleămontreăl‘exempleădesăformesădiminutivesăétudiéesăparăRatcliffeă(2006ă:ă
76-8)ă:ăfinalement,ăparăanalogie,ădeuxăformesăsuffisentăpresqueăpourăl‘expressionă
deă tousă lesă diminutifs,ă l‘uneă poură leă masculină (CCiyəC),ă l‘autreă pour le féminin
(CCiCa).
Le h}assāniyyaă illustreă plutôtă uneă tendanceă inverse,ă poură ceă quiă estă deă laă
profusion des schèmes et de la conservation des oppositions distinctives, et dans
l‘ensembleă ilă s‘agită certainementă d‘innovationsă propresă àă l‘arabeă mauritanien,
mais certains traits – commeă leărôleă importantăjouéăparăl‘affixationă – ne sont pas
sans équivalent en berbère. Les préfixes a- et u-, qui font exception à la règle
générale de chute des voyelles en syllabe ouverte, sont en effet devenus des
marques spécifiques,ădeăl‘élatifăpourăa- (marque étendue par analogie à toutes les
formes dérivées), du passif pour u- (pour toutes les formes dérivées actives, en
complément de n- et -t- pour les verbes nus).
En règle générale, on peut dire que les dérivations causatives et passives sont
asseză productivesă enă berbère,ă àă laă différenceă deă cellesă àă sensă réfléchi.ă S‘ilă yă aă
confusion entre le réfléchi et le passif dans les dialectes maghrébins
septentrionaux, le substrat ne paraît pas pouvoir être mis en cause. Quant au passif
h}assāniyya,ălaămarqueău- ne doit sans doute pas grand chose au substrat, mais on
peutăconstaterăqu‘ilăyăaăbien,ăentreălesăpréfixesăduăh}assāniyyaăetăceluiă(enătyty-) du
zénaga,ăunăgrandăparallélismeădeăvaleurăetăd‘emploiă(cf.ăTaine-Cheikh 2007)30.
S‘agissant du diminutif, le h}assāniyyaă etă leă zénagaă n‘ontă aucuneă formeă enă
commun,ă maisă ilsă offrentă tousă deuxă desă moyensă d‘expressionă particulièrementă
développés (cf. Taine-Cheikh 1988 et 2002). En h}assāniyya,ă leă schèmeă CCäyC
desătrilitèresăs‘estămaintenu,ăcommeădans la majorité des dialectes de nomades, et
l‘emploiădeă l‘infixeă -äy- a même été étendu à des formes inhabituelles (celui des
verbes et celui des élatifs). En zénaga, le diminutif est beaucoup moins fréquent
mais il se singularise, lui aussi, par des formes inhabituelles en berbère. Alors que
leă diminutifă s‘exprimeă seulement,ă dansă lesă parlersă berbères,ă pară l‘emploiă duă
féminin31, le zénaga a développé un système de dérivation qui lui est propre en
grammaticalisant les morphèmes aġ et tyā(d) quiăsignifientăàăl‘origineă« fils (de) »
En h}assāniyya,ăl‘existenceădeăformesăàăpréfixe s- (comme causatives des formes réfléchies en st)ăn‘estă pasă sansărapportăavecă lesă formesă causativesă duă substrat,ă maisălesă formesă elles-mêmes ont
pris en arabe une valeur particulière, qui reflète leur place spécifique dans le système (cf. TaineCheikh 2003).
31
Phénomène qui semble attesté anciennement en chamito-sémitiqueă maisă quiă n‘estă pasă d‘uneă
grande fréquence dans les dialectes arabes maghrébins – saufă sousă l‘influenceă duă berbèreă (typeă
qādûm « marteau » / qādûma « petit marteau », cf. Marçais 1956 : 610).
30
174
et « fille (de) » (cf. aġmah « frère » et tyādmah « sœur », littéralement « fils / fille
de mère »).
•ă Lesă développementsă parallèlesă neă sontă peut-être pas très évidents en
morphologie, surtout quand ils utilisent des matériaux différents. Ils sont par
contre assez nets quand on considère les évolutions qui se sont produites en
syntaxe dans les dialectes de sédentaires et qui ont fini par affecter assez
profondémentă leă systèmeă ancienă deă l‘arabe.ă Lesă trois points les plus importants
sont les suivants : remplacement du syntagme de détermination directe par un
syntagmeăanalytique,ăavecăprésenceăd‘unămorphèmeăjouantăleărôleădeăprépositionă;ă
développement de particules aspecto-temporelles,ă notammentă poură l‘expression
de la concomitance ; renforcement de la particule de négation par recours à un
morphème réassertif (souvent dérivé de « chose »). Ces évolutions sont
fondamentales et elles contribuent à donner aux dialectes modernes un type
nouveauă qu‘onă peută alors avec raison, me semble-t-il, qualifier de « néo-arabe »
(cf. Taine-Cheikh 2002). Il faudrait pouvoir exposer ces changements en détail,
mais en montrant que – si des évolutions comparables se sont produites dans
l‘arabeămaghrébinăetăenăberbèreăseptentrional32 – elles se sont produites aussi dans
les dialectes arabes orientaux de sédentaires – etăqu‘ellesăsontăabsentes,ăparăcontre,ă
des parlers méridionaux (en touareg, en zénaga et en h}assāniyya).ăFauteădeăplace,ă
je ne pourrai pas évoquer la littérature abondante qui traite de ces questions 33 ,
maisă peuă d‘auteursă apportentă laă preuveă qu‘onă aă vraimentă affaire,ă dansă cesă
différentsă cas,ă àă desă phénomènesă d‘empruntsă (deă l‘arabeă auă berbèreă ouă
inversement).ă Ilă sembleă s‘agiră essentiellementă d‘évolutionsă parallèlesă etă de
phénomènesă d‘interférence,ă sansă qu‘onă sacheă auă justeă siă uneă desă deuxă languesă
peută êtreă assignéeă commeă cibleă pară rapportă àă l‘autre.ă Siă l‘onă trouveă quelquesă
exceptions,ă c‘estă seulementă dansă desă casă asseză circonscrits,ă telă l‘usageă ‗berbère‘ă
de la particule d après les noms de nombre, dans les parlers arabes du Maroc ou
comme annonciatif, à Djidjelli (cf. Marçais 1956 : 465 et Galand 2002 : 216-7).
Au terme de ce parcours, un certain tableau se dégage. Au premier abord, on ne
retientă queă lesă traitsă ‗berbères‘ă lesă plusă saillants,ă observablesă dansă lesă petitesă
communautés de sédentaires comme celles de Chefchaouen ou de Djidjelli et on
ne voit, ailleurs, que des évolutions parallèles (pas toujours spécifiques au
Surăl‘émergenceădeălaănégationădiscontinueădansălesădialectesăarabesăetăberbèresăduăMaghreb,ăcf.ă
Chaker et Caubet (éds.) 1996. Sur le génitif analytique en arabe, cf. Eksell Harning 1980.
33
Colină(1935)ăvoyait,ădansăleădédoublementădeăl‘inaccompli (par recours à un préfixe du type ka /
ku)ă etă laă créationă d‘ună articleă indéfini,ă ună développementă parallèleă deă l‘arabeă etă duă berbère.ă Ilă
précisait cependant : « Ilă sembleă doncă bienă queă l‘expressionă deă l‘oppositionă réel ≠ irréel
corresponde, en arabe, à ună besoină deă laă mentalitéă sédentaireă enă général,ă etă qu‘auă Maroc,ă plusă
spécialement,ăelleăsoitădueăàăl‘actionăduăsubstratălinguistiqueăberbère ».
32
175
Maghreb). Puis, en analysant plus précisément chacun des phénomènes en jeu, on
se rend compte que les effets du contact sont importants, mais ne se donnent pas à
voiră directementă car,ă leă tempsă passant,ă l‘empreinteă deă chaqueă langueă sură l‘autreă
estă digérée,ă retravailléeă deă l‘intérieur.ă Ilă yă aă donc,ă effectivement, une possibilité
queă lesă parlersă périphériquesă évoluentă plusă queă d‘autres,ă maisă ilă yă aă surtoută laă
probabilitéă qu‘ilsă s‘engagentă plusă décisivementă dansă desă directionsă originales.ă Ilă
neă faută doncă pasă s‘étonneră siă lesă parlersă marocainsă – où la masse des parlers de
sédentairesă l‘emporteă – vont particulièrement loin dans le sens du changement
typologiqueă(deălaă‗métatypie‘)ăalorsăqueăleăparlerăh}assāniyyaă– où le système des
racines et des schèmes est bien conservé, comme le caractère synthétique des
constructions –, innove sur beaucoup de points, mais sans remettre en cause les
fondements de la langue.
Références
Aguadé, Jordi, Federico Corriente & Marina Marugan (éds.). 1994. Actas del Congreso
Internacional sobre interferencias Arabo-Romances y Paralelos Extra-Iberos
(Madrid,10-14 déc. 1990), Zaragoza.
Aguadé, Jordi, Patrice Cressier et Angeles Vicente (éds.). 1998. Peuplement et
arabisation au Maghreb occidental. Dialectologie et histoire, Madrid / Zaragoza :
Casa de Velázquez.
Aguadé, Jorge. 1998. « Ună dialecteă ma‗qilienă :ă leă parleră desă Z‗īră auă Maroc », en Jordi
Aguadé & al. (éds.) : Peuplement et arabisation au Maghreb occidental, 141-150.
Avram, Andrei A. 2004. « Arabic pidgins and creoles from a comparative perspective »,
Romano-Arabica, New Series III (Arabic Linguistics), 25-40.
Behnstedt, Peter. 1998. « La frontière orientale des parlers maghrébins en Égypte », in
Jordi Aguadé & al. (éds.) : Peuplement et arabisation au Maghreb occidental, 85-96.
Blanc, Haïm. 1964. Communal Dialects in Baghdad, Cambridge, Massachussetts (Mass.)
: Harvard University Press.
Blau, Joshua. 1969. « L‘apparitionă duă typeă linguistiqueă néo-arabe », Revue des Etudes
Islamiques 37/2, 191-201.
Borg, Alexander. 1994. « Observations on some Evolutionary Parallels and Divergences
in Cypriot Arabic and Maltese », in Jordi Aguadé & al. (éds.) : Actas del Congreso
Internacional sobre interferencias..., 21-40.
Cantineau, Jean. 1941. « Les parlers arabes des Territoires du Sud », Revue Africaine 85,
72-77.
Caubet, Dominique. 1998. « Étude sociolinguistique des traits préhilaliens dans un
dialecteă enă voieă d‘urbanisationăàăFès », in Jordi Aguadé & al. (éds.) : Peuplement et
arabisation au Maghreb occidental, 165-176.
Chaker, Salem & Dominique Caubet (éds.). 1996. La négation en berbère et en arabe
maghrébin,ăParisă:ăL‘Harmattan.ă
Cohen, David. 1963. Le dialecte arabe H}assānīya de Mauritanie, Paris : Klincksieck.
176
Cohen, David. 1970. « Le système des voyelles brèves dans les dialectes maghrébins », in
Études de linguistique sémitique et arabe, The Hague - Paris : Mouton, 172-178.
Cohen, David. 1978. Parlers judéo-arabes [= Judéo-arabe. I], in Encyclopédie de l‘Islam,
2e édition, Leyden : Brill, 312-315.
Cohen, David. 1994. « Préface », in Actes des premières journées internationales de
dialectologie arabe de Paris,ăINALCOă(Paris)ă:ăLangues‘O,ă9-19.
Colin, George-S. 1986. « Aperçu linguistique », art. « Al-Maghrib », in Encyclopédie de
l‘Islam 2e édition, Leiden/Paris : Brill/Maisonneuve & Larose, 1193-98.
Colin, George-S. 1935. « L‘oppositionăduăréelăetădeăl‘éventuelăenăarabeămarocain », Bull.
de la SLP 36, 133-40.
Dakhlia, Jocelyne (éd.) 2004 Trames de langues. Usages et métissages linguistiques dans
l‘histoire du Maghreb, Paris : Maisonneuve & Larose.
Durand, Olivier. 1995. Introduzione ai dialetti arabi, Milano : Centro Studi CamitoSemitici.
EALL: voir Versteegh.
Eksell Harning, Kerstin. 1980. The Analytic Genitive in the Modern Arabic Dialects,
Göteborg : Orientalia Gothoburgensia, vol. 5.
Elmedlaoui, M. (2000), « L‘arabeă Marocain,ă ună lexique sémitique inséré sur un fond
grammatical berbère », in Salem Chaker (éd.) : Etudes berbères et chamito-sémitiques,
Paris-Louvain : Peeters, 155-188.
Elyaakoubi, Mohammad. (1998), « À propos de la classification des dialectes du Sud du
Marocă :ă l‘exemple de Skûra », in Jordi Aguadé & al. (éds.) : Peuplement et
arabisation au Maghreb occidental, 151-156.
Ferguson, Charles. 1959. « The Arabic Koinè », Language 35/4, 616-630.
Fischer, Wolfdietrich. & Otto Jastrow, (éds.). 1980. Handbuch der arabischen Dialekte,
Wiesbaden : Harrassowitz.
Galand, Lionel. 1983. « Berbèreă etă‗traitsăsémitiquesăcommunsă?‘ », Comptes rendus du
G.L.E.C.S. (1973-79) XVIII-XXIII/3, 463-93.
Galand, Lionel. 2002. Études de linguistique berbère, Leuven-Paris : Peeters.
Ingham, Bruce. 1994. « The Effect of Language Contact on the Arabic Dialect of
Afghanistan », in Jordi Aguadé & al. (éds.) : Actas del Congreso Internacional sobre
interferencias..., 105-17.
Iraqui-Sinaceur Zakia (ed.). 1993. Le Dictionnaire Colin d‘Arabe Dialectal Marocain,
Rabat-Paris : Al Manahil.
Kaye, Alan S. 1986. « TheăVerbă‗See‘ăinăArabicăDialectsă»,ăinăJoshuaăA.ăFishmană&ă al.
(éds.) The Fergusonian Impact, vol. 2, Berlin : Mouton de Gruyter, 211-222.
Kaye, Alan S. 1994. « Peripheral Dialectology and Arabic Pidgins and Creoles », in Jordi
Aguadé & al. (éds.) : Actas del Congreso Internacional sobre interferencias..., 12540.
Lévy, Simon. 1998. « Problématiqueă historiqueă duă processusă d‘arabisationă auă Marocă :ă
pour une histoire linguistique du Maroc », in Jordi Aguadé & al. (éds.) : Peuplement et
arabisation au Maghreb occidental, 11-26.
177
Marçais, Philippe. 1956. Le parler arabe de Djidjelli (Nord Constantinois, Algérie), Paris
: Lib. Adrien-Maisonneuve.
Marçais, Philippe. 1975. « Algérie. V – Langues », in Encyclopédie de l‘islam, Leiden et
Paris : Brill et Maisonneuve, 384-390.
Marçais, Philippe. 1977. Esquisse grammaticale de l‘arabe maghrébin, Paris : Lib. AdrienMaisonneuve.
Mufwene, Salikoko S. 2005. Créoles, écologie sociale, évolution linguistique, Parisă:ăL‘Harmattan.
Natividad, Emma 1998. « Le dialecte de Chefchaouen », in Peuplement et arabisation au
Maghreb occidental. Dialectologie et histoire, 109-120.
Palva, Heikki. 2006. « Dialects : Classification », in EALL, 604-613.
Ratcliffe, Robert. 2006. « Analogy », in EALL, 74-82.
Rosenhouse, Judith. 2006. « Bedouin Arabic », in EALL, 259-269.
Taine-Cheikh, Catherine. 1991. « Le vent et le devant – Deă l‘orientationă cheză lesă
Maures », Journal asiatique CCLXXIX/1-2, 93-126.
Taine-Cheikh, Catherine. 1994. « Le H}assāniyyaădeăMauritanie,ăunădialecteănon-marginal
de la périphérie », in Jordi Aguadé & al. (éds.) : Actas del Congreso Internacional
sobre interferencias..., 173-99.
Taine-Cheikh, Catherine. 1997. « Les emprunts au berbère zénaga – Un sous-système
vocalique du h}assāniyya », Matériaux arabes et sudarabiques 8, 93-142.
Taine-Cheikh, Catherine. 1998-99. « Deux macro-discriminants de la dialectologie arabe
(la réalisation du qâf et des interdentales) », Matériauxarabes et sudarabiques 9, 11-51.
Taine-Cheikh, Catherine. 1999. « Les hassanophones du Maroc – Entre affirmation de soi
et auto-reniement », Peuples méditerranéens 79 (« Langues et stigmatisations sociales
au Maghreb »), 85-102.
Taine-Cheikh, Catherine. 2002. « Àă proposă deă l‘oppositionă ‗typeă synthétique‘ă vs ‗typeă
analytique‘ăenăarabe », in Abderrahim Youssi & al. (éds.) : Aspects of the Dialects of
Arabic Today, Rabat : Amapatril, 234-243.
Taine-Cheikh, Catherine. 2003. « Les valeurs du préfixe s- en h}assāniyyaă etă lesă
conditions de sa grammaticalisation », in Ignacio Ferrando & al. (éds.) : AIDA 5th
Conference Proceedings, Cádiz : Publicationes de la Universidad de Cádiz, 103-118.
Taine-Cheikh, Catherine. 2005. « Du rôle de la quantité vocalique en morphogénie. Réflexions à partir
deă l‘arabeă etă duăberbèreă deă Mauritanie », in Faits de Langues [Les langues chamito-sémitiques
(afro-asiatiques) vol. 1], Paris : Ophrys, 41-63.
Taine-Cheikh, Catherine. 2007. « Voixă moyenneă etă variationsăd‘actanceă :ă leăréfléchi en arabe et en
berbère (exemples de Mauritanie) », in André Rousseau & al. (éds.) : L‘énoncé réfléchi, Rennes :
Presses Univ. de Rennes, 321-342.
Versteegh, Kees (ed.) 2006. Encyclopedia of Arabic Language and Linguistics (EALL), Leiden-Boston : Brill.
Vicente, Angeles. 1998. « UnătypeădeădialecteămontagnardăauăMarocă:ăleăparlerăd‘Anjra », in Jordi
Aguadé & al. (éds.) : Peuplement et arabisation au Maghreb occidental, 121-130.
Zavadovskij, Ju. N. 1981. Mavritanskij dialekt arabskogo jazyka (xassaniya), Moscou :
Izdatelstvoă‗Nauka‘.
Zima, Peter. 2007. « Why Languages and Contacts ? », Journal of Language Contact,
Thema 1(www.jlc-contact.org), 101-115.
178
The influence of Turkish, Kurdish and other neighbouring
languages
on Anatolian Arabic
Shabo Talay,
Friedrich-Alexander University, Erlangen-Nürnberg
Germany
0: Introductory remarks
Ină thisă contributionă theă termă ―Anatoliană Arabic‖ă isă usedă foră allă dialectsă
spoken in south-eastern Turkey excluding those of the Urfa region. Although
there are some Bedouin dialectsă spokenă betweenă Kızıltepeă andă Urfa,ă allă knownă
dialects of this area belong to the so called qəltu-dialects.ăSoă―AnatolianăArabic‖ă
is equivalent to Anatolian qəltu -Arabic 1 . While in former times the Arabic
speaking community of this area was religiously mixed; there were Jews,
Christians and Muslims – nowadays there is an almost homogeneous Muslim
community speaking Arabic. The Jews and meanwhile all Christians left this area;
onlyă aă smallă numberă ofă Christiansă stillă liveă ină Mardin,ă Bnēbīl,ă Qəlləf, Ma‗sarte
andăsomeăinăƥzəx2. Their number does not exceed one or two hundred persons at
all.ăThat‘săwhyătheăDiyarbakirăgroupăofătheă qəltu-Dialects is now extinct, and in
Nusaybin and Cizre too, there are no Arabic speakers since the 1950 ies at all.
Anatolian Arabic has to be counted as peripheral because since centuries it
is located and spoken in an area where the dominant language is not Arabic. And
since the establishment of the Turkish Republic in 1923 Anatolian Arabic
speakers are fully separated from the Arabic world and completely cut off from
the influence of Modern Standard Arabic and the Arabic multimedia. Actually, in
addition,ă theă Arabică dialectsă ofă Anatoliaă themselvesă don‘tă formă aă homogeneousă
language area, but they rather occur only in form of isolated linguistic enclaves in
the sea of Kurdish and Turkish. Mainly under the influence of these two
languages Anatolian Arabic experienced significant diversifications in all
1
For further information see Jastrow 1978, p. 3ff. and Jastrow 2004.
Wittrichă2001.ăTheăTurkishănameăofăƥzəx is Idil. It is now a district town (Kaza) in the province
ofă Şirnakă withă aboută 20.000ă Kurdishă inhabitants,ă whoă speakă Kurdish.ă Onlyă 10ă Christiană
individuals still live in this originally Christian village. Till the 1970ies Christians constituted the
majority of the population, and in 1978 for the first time a Kurd won the elections for the mayor of
the town.
2
179
language fields and categories, in the sound system as well as in grammar, syntax
and semantics.
The dominant language of daily life of this region is traditionally Kurdish
and the official one is Turkish, which becomes more and more important for the
Arabic speakers as the everyday speech. Beside their own language, almost all
Arabic speakers are fluent in at least one other language, i.e. Turkish or Kurdish,
with the older ones preferring Kurdish and the younger ones Turkish.
In the past there must have been a contact to Aramaic 3 in the southern
region, for example in Mardin and Mh}allami area, but also in Siirt, and Armenian4
in the northern region, for example in the Muş plain.
That‘săwhyăweăcanăfindăaăratherăhighăintensityăofăinfluencesăfromăallătheseă
languages in Anatolian Arabic, Kurdish and Turkish being the dominant exporters
of language material into Anatolian Arabic. Even Persian plays an important role
in exporting material into Anatolian Arabic, but the majority of Persian words
came by the way of Kurdish, Turkish or Aramaic.
I: Influences on the consonant system
Firstă ofă allă someă phonologicală featuresă ofă Anatoliană Arabică catchă one‘să
eyes. These are the phonemesă/p/,ă/v/,ă/č/,ă/ž/ăandă/g/ăwhichăenrichătheăconsonantă
inventory of Anatolian Arabic and are inexistent in Old Arabic. We encounter
these phonemes primarily in borrowings from Turkish and Kurdish 5:
/p/
pār }āt ( apsəs)ă―money‖ă
pūz ( apsəs)ă―nose‖ă
qappa ṭ ―heăclosed‖ă
< Turk. para
< Kurd. pûz
< Turk. kapatmak
/v/
nvīne (ƥzəx)ă―bed‖ă
pēlāv (Hasköy)ă―shoe‖ă
səvīl (ƥzəx)ă―civil‖ă
< Kurd. nvîn
< Kurd. pêlav
< Turk. sivil
/č/
čāra ―possibility,ăremedy‖ă
rēnčbarī (Hasköy)ă―husbandry‖ă
< Turk. çare
< Kurd. rênčberî
3
Beside the lexical material especially in the Mh}allami dialects (Jastrow 2004), the Aramaic
influence appears in forms for the genitive exponent dīl with modifications (s. Jastrow 1978, p.
125f. and Talay 2001, p. 78). The construct state bē (Arab.: bēt < *bayt)ă―theăfamilyăof‖ăinăHasköyă
seems also to be of Aramaic origin as well as the preposition amma ―with,ătogetherăwith‖ă(Talayă
2001, p. 77 Fn.21; p. 82).
4
e.g. ādaš - ītəš ―toăsee‖ăinăDaragözüăandăHasköy.
5
The examples are from Vocke/Waldner 1982 and Talay 2002, those of H}a bsəsăandăƥzəx are from
my own investigations.
180
čātəšma ―struggle‖ă
< Turk. çatışma
/ž/ă
ṭāži ―greyhound‖ă
žāre ―ohădear‖
žīžo (ƥzəx)ă―hedgehog‖
< Kurd. tajî
< Kurd. žarê
< Kurd. žîžo
/g/
gāzǝ́n (ƥzǝx)ă―accusation‖ă
gēčǝ́m ―subsistence‖ă
gōzláke (Hasköy)ă―glasses‖ă
< Kurd. gazin
< Turk. geçim
< Turk. gözlük
The influence of the major languages caused even variants in realisation of
inherited consonant phonemes. Thus, some of them show particularities which are
not to be found in other Arabic dialects, for instance:
a) the phonemes /t/, /k/ have an aspirated and a totally unaspirated allophone,
which must be an influence of Kurdish: so we have a [k] without aspiration in
kalam [Ɯkalam]ă ―cabbage‖ă (Hasköy),ă kalame [kaƜlame]ă (ƥzəx)ă ―thorn‖ă andă ană
aspirated [kh] in kallam [Ɯkhallam]ă ―heă spoke‖,ă kalām [khaƜla:m], pl. kəlēmi
[khəƜle:mi]ă―word‖.
b) in Hasköy, Daragözü and other Kozluk-Sason dialects the Arabic velar or
emphatic phonemes /t}/, /s}/ and z}/d} in many positions lost their phonemic status
and are deemphasized, examples from Hasköy6:
ata ―heăgave‖ă (<*‘at}ā)
tare ―way‖ă (< *t}arīq)ă
sēbi ―boy‖ă
(< *s}abīy)
zarab ―heăhit‖ă (< *d}arab)
This deemphasizing does not mean that thereăaren‘tăanyăemphaticăconsonants,ăbută
under the influence of Turkish and Kurdish in these dialects the phonemic status 7
of the emphatic consonants seems to be at least weakened or abandoned.
c) Again, the weakness of the pharyngeal /‘/8 in Hasköy, but also in Siirt, KozlukSason might be an influence from the neighbouring languages, which do not
possess this phoneme (Examples from Hasköy):
6
Talay 2001, p. 74.
Jastrow 1978, p. 47ff.
8
Jastrow 73, p. 18 and Talay 2001, p. 75.
7
181
abze ―Iăamăafraid‖ă
nəqte ―weăcut‖ă
āyla ―family‖ă
ayd ―feast‖ă
(<*bz‘9)
(< *qt}‘)
(< Turk. ayle < har. *‘ā’ila)
(< Kurd. ayd < har. *‘īd)
d) Finally the interdental fricatives /t/, /d/ and /d}/ show different realizations in
Kozluk-Sason dialects. Sometimes they remained as fricatives for instance in idǝn
―ear‖,ă andă ină nəd}r əbu ―weă hită him‖,ă ină otheră formsă theyă shiftedă toă theiră sibilantă
counterparts, as it is in:
saġīl ―heavy‖
(< *taqīl)
gəzbe ―lie‖ă
(< *kidba)
bayza ―whiteă(f.)‖ă
(< *bayd}a)
fəz}z}a ―silver‖ă
(< *fid}d}a)
But there are also forms with dental stops /t/, /d/ and /d}/ instead of the fricatives:
dəfər ―fingernails‖
(< *d}ifr)
tēnāden ―theyăboth‖ (< *tnēnātən)
This inconsequent treatment of the interdental fricatives has to be seen as a quasiadaptation of the consonant inventory to that of the superstrate languages.
II: Influences on the vowel system
Traditionally Arabic has three short vowels /a/, /i/, /u/, three long vowels
/ā/,ă/ī/,ă/ū/ăandătheătwoădiphthongs /aw/ and /ay/. In opposition to the majority of
the modern Arabic dialects, in Anatolian Arabic the diphthongs are preserved in
mostăcases.ăEvenăthoughătheăvowelăsystemăconsistsăofătheăfiveălongăvowelsă/ā/,ă/ē/,ă
/ī/,ă/ō/ăandă/ū/,ăwithă/ō/ăandă/ē/ăbeingăpredominantly in borrowings from Kurdish
and Turkish, as you may see in
čōl ―desert‖
ğōt ―ploughing‖
ğwāmēr ―knight‖
ğā bkēr ―heăwasăgoodăfor‖ă
Turk. çöl
Kurd. jôt
Kurd. jwamêr
Kurd. b-kêr hat
The older speakers used to integrate the borrowings totally into their own system
In Mardin bəzə‗-yəbza‗ , in other dialects the root of this verb is with /f/ as the first radical, fz‘
cf.: Arbel: fəzə‗-yəfza‗ ―sichă fürchten‖,ă otheră examplesă foră bz‗ and fz‗ in the qəltu dialects see
Vocke/Waldner 1982: 313 (fz‗).
9
182
of vowels. Actually, the vowels in these forms followed the phonological rules of
Anatolian Arabic. But the younger speakers who mostly speak Turkish better than
their Arabic dialect do not adapt the vowels of borrowings into the Arabic system.
Therefore you may find all Turkish vowels realized in original way.
III: Influences on parts of the morphology
a) The copula
One of the most characteristic features of Anatolian Arabic is the existence
of the copula as an enclitic suffix. If we compare the copula of Anatolian Arabic
with the other neighbouring languages, we will see that it exists in all languages
of the area, Turkish, Kurdish and Aramaic. Because the copula is almost unknown
in other Arabic speaking areas, it seems that the development of a full
morphological paradigm for the copula is at least furthered by the neighbouring
languages. Although the copula forms themselves are not imported, the way they
are used in Anatolian Arabic is exactly the same as it is in Aramaic, Kurdish and
Turkish, respectively10.
Kinderib11
malīh}-we
mlāh}-ənne
beautiful
malīh}-ana
malīh}a-ənti
beautiful
Aramaic (Tur.)
šafiro-yo
šafire-ne
Kurdish
spehi-ya
spehi-ne
Turkish
güzel-dir
güzel-dirler
he is beautiful
they
are
šafiro-no
šafirto-hat
spehi-me
spahi-yi
güzel-im
güzel-sin
I am beautiful
you (sg.f.) are
b) Verbal system
1. In a forthcoming article Otto Jastrow12 points out that the tense, aspect
andă moodă categoriesă ofă Anatoliană Arabică ―areă formed under the influence of
Turkishă andă Kurdish‖.ă Ină additionă toă theă twoă morphologicală verbă categoriesă
perfect and imperfect Anatolian Arabic uses the tense markers ku-, tə-, ka-, ka-təprefixed to the imperfect forms and kən- prefixed to the perfect forms in order to
modify the verb concerning its temporal expression. Although all these verb
modifiers seem to be etymologically of Arabic origin the verb categories they
expressăproveăJastrow‘săaboveăquotedăstatement,ăasăcanăbeă seenă inătheă followingă
10
See also Grigore 2007, p. 288ff.
Kinderib forms are from Jastrow 1978, p. 131.
12
Jastrow (forthcoming).
11
183
examplesăfromăƥzəx compared with their Turkish correspondences:
Imperfect:
+ ku+ tə+ ka+ ka-təPerfect:
+ kən-
ƥzəx Arabic13
təfta
ku-təftah}
tə-təfta h}
ka-təfta h}
ka-tə-təftah}
fatah}t
kən-fatah}t
Turkish
açarsın
açıyorsun
açacaksın
açardın
açıyordun
açacaktın
açardın
açtın
açmıştın
―youăopen‖
―youăareăopening‖
―youăwillăopen‖
―youăusedătoăopen‖
―youăwereăopening‖
―youăwereăaboutătoăopen‖
―youăwouldăhaveăopened‖
―youăopened‖
―youăhadăopened‖ă
2. Compound verbs with sawa-ysawi
Another interesting phenomenon is, that there is a relatively high number
of compound verbs constructed with Arabic sawa – ysawi ―toădo‖ăfollowedă byăaă
Kurdish or Turkish noun. In the majority of the cases, the construction is a
complete calque of its Turkish or Kurdish counterparts. Examples:
sawa
sawa
sawa
sawa
sawa
gāzǝ́n ―toăcomplain‖
ğōt ―toămow‖
ğāmērtīye ―toăactăgenerous‖ă
bēnāmūsīye ―toădoăsth.ăforbidden‖
brīndār ―toăinjure‖ă
Kurd. gazin kirin
Kurd. ğôt kirin
Kurd. camêrtî kirin
Kurd. bê namûsî kirin
Kurd. brîndar kirin
sawa
sawa
sawa
sawa
sawa
huğūm ―toăattack‖
muh}āfaz}a ―toăpreserve‖
mnāqaša ―toădiscuss‖
yārdəm ―toăhelp‖ă
yar }a li ―toăinjure‖
Turk. hücüm etmek
Turk. muhafaza etmek
Turk. munakaşa etmek
Turk. yardım etmek
Turk. yaralı etmek
Sometimes there are compound verbs with sawa + Turkish verbal form of
the indefinite past (miş-verb):
sawa gačənməš ―toămanage‖ăăă
sawa qātəl}məš ―toăparticipate‖
sawa banzatməš ―toătakeăso.ăforăso.ăelse‖ă
13
Turk. geçinmiş
Turk. katılmış
Turk. benzetmiş
The examples have been taken from Wittrich 2001, p. 69-78.
184
sawa qārəšməš ―toăintervene‖
Turk. karışmış
Other borrowed verbal forms have been totally assimilated to the Arabic
verb system; these verbs are by the majority formed according to the II. or III.
verbal stem as can be seen in the followingăexamplesăfromăƥzəx:
Turk. kapatmak
qappa t} – īqappət} ―toăclose‖
qayyad – īqayyəd ―toăregister‖ă
Turk. kayıt etmek
III. stem
d}āyan – īd}āyən ―toăbeăpatient,ătoăbearăup‖ăăăăăăTurk. dayanmak
tēlan – ītēlən ―toărob‖ă
Kurd. talan kirin
c) Nouns
1. Some derivational suffixes of Kurdish and Turkish are widely spread in
Anatolian Arabic, for instance:
– The Kurdish diminutive marker -ək/əke:
šətlǝ́ke ―coldăframe‖ (Daragözü)
fərr ǝ́ke ―something‖
ġarzǝ́k ―vinyard‖ă(ƥzəx)
barxǝ́k ―lamb‖
II. stem
– The Turkish suffix for occupational titles -či:
‗ar }abánči ―charioteer‖
baxčaği ―gardener‖ă
takarləkči ―tyreăseller‖ă
bazərganči ―tradesman‖ă
2. In Kozluk-Sason-Hasköy dialects the determination of nouns remains
usually unmarked, as it is the case in Turkish and Kurdish:
Hasköy
mar a} ―theăwoman‖ă
rağəl ―theăman‖ă
mar }a -ma ―aăwoman‖ă
rağəl-ma ―aăman‖
For comparison, here are the Turkish and Kurdish equivalents:
Kurdish:
Turkish
zlam ―theăman‖ă
çocuk ―theăchild‖ă
zlam-ek ―aăman‖
bir çocuk ―aăchild‖
3. The numerals
In Hasköy and Daragözü the numerals from 11 to 19 have been adapted
185
from Turkish; they have the structure with tens + units14:
‗ašr }a sətte
‗ašr }a sab‗a
―sixteen‖ă
―seventeen‖
Turk. on altı
Turk. on yedi
Even the numerals from 21 to 99 in Daragözü have the structure tens +
units as it is in Turkish and Kurdish. In addition in Daragözü from 100 on Kurdish
forms are common practise.
IV: Lexicon
In the lexicon there are chronologically different levels of borrowings, and
according to the time of their import into Anatolian Arabic the sources for the
borrowings are different languages. Some older borrowings seem to be Aramaic,
like gəzīr ―herold‖ă (Ţă fromă Persian),ă magzūn ―sickle‖ă andă someă otheră formsă
published in Jastrow 2004, most Persian borrowings came via Kurdish or Turkish,
like karxāna ―brothel‖,ă čaydān ―teaă kettle‖,ă qastaxāna ―hospital‖.ă Modernă
borrowings concerning rural live are mostly from Kurdish, but all others
especially those related to modern life and technologies are of Turkish origin.
V: To Sum up
Actually, we have seen that most of the characteristic features of Anatolian
Arabic are based on its proximity to languages, which belong to different
language families with different systems of grammar and phonology. In its
phonological system but also in syntax and semantics it now adapted to major
languages Kurdish and Turkish.
Another point to mention is the spreading of Turkish in the last decades in the
eastern and south-eastern regions of the country. Turkish became more powerful
through the language policy of the Turkish Republic and the Turkish media. All
younger speakers are fluent in Turkish both writing and reading, while the
influence of Kurdish is declining. The other languages, like Persian, Aramaic and
Armenian nowadays have no influence on Anatolian Arabic at all.
References
Behnstedt,ăPeter.ă1992.ă―Qəltu-Dialekte in Ost-Syrien‖.ăZALă24.35-59
Fischer, Wolfdietrich und Otto Jastrow (Hrsg.). 1980. Handbuch der arabischen
Dialekte. Wiesbaden: Harrassowitz (= Porta linguarum orientalium N.S. XVI)
14
In some Kurdish dialects this structure is also adapted, so they have: deh yek (11), deh çar (14)
etc.
186
Grigore,ă George.ă 2000.ă ―Kuă – ună préfixeă temporelă dansă l‘arabeă mardinien‖.ă In :
Abderrahimă Youssiă etă …ă (ed.) : Proceedings of the 4th Conference of the AIDA.
Marrakech. 374-380.
Grigore,ăGeorge.ă2003.ă―Quelquesătracesăduăcontactălinguistiqueă dansăleăparlerăarabeădeă
Mardină(Turquie)‖.ăIn:ăRomano-Arabica 3 : 119-134.
Grigore, George. 2007. L‘arabe parlé à Mardin – monographie d‘un parler arabe
périphérique. Bucarest :ăEdituraăUniversitĒįiiădinăBucureşti.
Jastrow, Otto. 1973a. Daragözü. Eine arabische Mundart der Kozluk-Sason-Gruppe
(Südostanatolien). Grammatik und Texte. Nürnberg: Hans Carl (= Erlanger Beiträge
zur Sprach- und Kunstwissenschaft 46)
Jastrow, Otto. 1978. Die mesopotamisch-arabischen qəltu-Dialekte. Bd. I: Phonologie
und Morphologie. Wiesbaden: Steiner (= Abhandlungen für die Kunde des
Morgenlandes 43,4)
Jastrow, Otto. 1981. Die mesopotamisch-arabischen qəltu-Dialekte. Bd. II:
Volkskundliche Texte in elf Dialekten. Wiesbaden: Steiner (= Abhandlungen für die
Kunde des Morgenlandes 46,1)
Jastrow,ă Otto.ă 1989.ă ―Theă Judaeo-Arabică Dialectă ofă Nusaybin/Qāməﻣli‖.ă Studia
linguistica et orientalia memoriae Haim Blanc dedicata. ed. by Paul Wexler,
Alexander Borg, Sasson Somekh. 156-169. Wiesbaden: Harrassowitz
Jastrow,ăOtto.ă1997.ă―Si‗ird:ăArabicăDialect‖.ăEncyclopædia of Islam (s.v.)
Jastrow, Otto. 2003. Arabische Texte aus Kinderib. Wiesbaden: Harrassowitz (= Semitica
Viva 30)
Jastrow, Otto. 2005a. Glossar zu Kinderib (Anatolisches Arabisch). Wiesbaden:
Harrassowitz (= Semitica Viva 36)
Jastrow,ă Otto.ă 2005b.ă ―Weißeă Zähneă (d}rās bīđ̣). Über einige unregelmäßige
Lautentwicklungenă imă Anatolischenă Arabisch‖.ă Sacrum Arabico-Semiticum.
Homenaje al Profesor Federico Corrente, 215-221. Zaragoza
Jastrow, Otto. 2006. ―Anatoliană Arabic‖.ă Encyclopedia of Arabic Language and
Linguistics. Vol. I, ed. by Kees Versteegh, 86-96. Leiden-Boston
Jastrow, Otto. 2004 (appeared 2006). ―AramäischeăLehnwörterăimăarabischenăDialektăvonă
Kinderib (Südost-Türkei)‖.ă Estudios de dialectología norteafricana y andalusí 8.
Homenaje a Peter Behnstedt, 99-103. Zaragoza
Jastrow, Otto (forthcoming): Turkish and Kurdish influences in the Arabic dialects of
Anatolia.
Socin, Albert. 1882-3.ă―DerăarabischeăDialektăvonăMosulăundăMärdin‖. ZDMG 36, 1-53,
238-277; 37, 293-318
Talay,ă Shabo.ă 2001.ă ―Deră arabischeă Dialektă vonă Hasköy.ă I.ă Grammatikalischeă Skizze‖.ă
ZAL 40, 71-89
Talay,ăShabo.ă2002.ă―DerăarabischeăDialektăvonăHasköy.ăII.ăTexteăundăGlossar‖.ăZAL 41,
46-86
Talay,ăShabo.ă2003.ă―SomeăGrammaticalăRemarksăonătheăArabicăDialectăofăHasköyă(Eastă
Anatolia)‖.ă I.ă Ferrandoă and J.J. Sanchez Sandoval (eds.): AIDA 5th Conference
Proceedings. Servicio de publicaciones Universidad de Cádiz. 119-120.
187
Vocke, Sybille und Waldner, Wolfram. 1982. Der Wortschatz des anatolischen Arabisch.
Erlangen: Selbstverlag
Wittrich, Michaela. 2001. Der arabische Dialekt von ƨzəx. Wiesbaden: Harrassowitz (=
Semitica Viva 25)
188
Phonetic features and changes in Andalusi Arabic:
a case study of the laryngeal plosive /’/
Hristina Tchobanova
University of Sofia
Bulgaria
1. Presentation of the study as a target, object and method.
In this paper I am going only to partly illustrate the variable features
chosen for quantitative analysis at the phonetic level of consonantism and
illustrate the results from the detailed registration of the variable laryngeal plosive
in Andalusi Arabic (AA). The aim of the broader study is to unite methods used in
the traditional descriptive dialectology and contemporary quantitative linguistics
in order to make an attempt on a diachronic quantitative analysis of this variety of
Arabic. The written AA form attested by different sources is examined in the
process of its diachronic development and, when possible, in comparison with
contemporary Arabic dialects (CAD).
As far as the quantity of available sources in which was attested the
written form of AA is known, and each one of them has a certain volume of its
own, I consider that the research has been conducted on the basis of a random
(representative) sample from a real finite general totality.
Contemporary Arabic dialectology which has been developing rapidly
since the middle of the twentieth century lacks both in diachronic and comparative
synchronic studies. Dialectologists have understandably directed their prioritized
efforts to generalizing detailed synchronic descriptions of Arabic dialects,
necessary for teaching Arabic as a foreign language and its good command in its
state of diglossia. Until the present moment, except for Ferrando, who has
thoroughly examined in a diachronic, comparative and functional aspect the
binding morpheme – ană ină ААă (Ferrando 2000: 25-46), diachronic studies have
not been conducted in this dialect. In the comparative dialectological aspect, one
study has been conducted so far by Owens, in which the statistical and
comparative grammatical methods have been combined (Owens 2005: 271-308).
Applying simple statistical procedures on data basis consisting of 49 phonological
and morphological features, four geographically and chronologically distant
dialectal variations of Arabic are compared in it and the results are being used to
propose some reconstructions in the historical development of Arabic.
189
In the lack of both diachronic and comparative synchronic studies in the
wider research of my dissertation there is an attempt to make a diachronic
quantitativeă analysisă ofă aă separateă dialect/ă bundleă ofă dialectă variationsă (АА),ă
which has been well synchronically described by Corriente and the Spanish team
of researchers. Statistical procedures have been applied to representative samples
of data mainly from the middle and low register of АА.ă Theyă helpă distinguishă
between the different degrees of development of the different processes of
language change at its phonological and morphological levels. I hope that the
scrupulous application of statistical methods to the samples could create a
sufficiently reliable database, which could be used by other comparative historical
dialect studies as well.
2. Sources used in data excerption.
Part of the samples has been based on the lexis excerpted from the
critically revised sources and summarized in 1997 by Corriente in A Dictionary of
Andalusi Arabic (Corriente 1997), which offers primarily vocabulary with
documented use in the middle and low registers of the Arabic used in al-Andalus
with some examples of the high register and a lot of foreign loan-words.
Al-munğid fī l-luġa wa l-‘a‗lām has been used for the comparison with the
contemporary literary Arabic language (CLAL).
For the purposes of the comparative study at the respective levels, I
collected also samples from separate sources representative of the different stages
inătheădevelopmentăofăАА.ăTheădata,ăwhichă followsăbelowăaboutătheărealizationsă
ofătheăglottalăplosiveăstopă/‘/ăhave been excerpted as representative of the different
stages of the development of the dialect respectively from A Dictionary of
Andalusi Arabic and the texts of:
1. DīwānăIbnăQuzmānă(Corrienteă1980ă&ă1995),ărepresentativeăofăААăinătheă
12th century.
2. Theă proverbă miscellanyă ofă Abūă Yah}yāă az-Zajjālī³ in the edition of
Ahmed-Salem Ould Mohamed Baba (Ould Mohamed Baba 1999). The
book contains 2157 proverbs collected and written down in the 13 th
century.
3. The Miscellany of Ibnă ‗ƥs}im al-Ƴarnāt}īă (1339 – 1426), according to the
edition of Marugan from 1994: El refranero andalusí de Ibn ‗ƨs}im alƶarnāt}ī (Marugan 1994).
3. Statistical tests used to verify the data from the samples.
The simple statistical procedures used in the study help to distinguish the
extent of change of the examined AA units.
190
The volume of the samples has been randomly chosen in the examined
tests.
Whether the random excerpt is sufficiently representative is defined
through the standard statistical test about the definition of a confidential interval
applied in inferential statistics.
When the relative frequency k/n of the examined qualitative feature of the
sample is known, then about it in the excerpt n the following confidential interval
can be drawn:
k/n ± t ġk/n (l – k/ n
n
The quantity t in this formula is a factor, which defines the probability
with which the relative frequency is realized in this interval of the sample.
When t = 1 →ătheăprobabilityăisăaboută68%.
When t = 2 →ătheăprobabilityăisăaboută95%‖
When t = 2.5 →ătheăprobabilityăisăaboută99%
Having verified the representative sample the realization of the respective
feature in it is followed, and conclusions are drawn about the statistical totality on
the basis of the sample.
To verify the conclusions made on the basis of representative samples the
so-called criterion of Pierson is used ( ² ) :
² = (obs – th)²,
th
where obs is the number of the units in which a certain feature is displayed in
the number under consideration;
th – the number of the units in which a certain feature should be displayed in
normal distribution;
and is the sum of the units counted.
This criterion serves to verify the correlation between normal theoretical
distribution (that is theoretically expected realizations of the feature in all of
the units counted), and the distribution of the qualitative features observed (the
number of the observed realizations).
191
TheăconditionăforătheăapplicationăofăPierson‘săcriterionăisăthatătheăexcerptsă
shouldă notă beă feweră thană 40ă andă ită shouldn‘tă haveă theoreticală valuesă feweră
than 51.
4. Variable features examined at the phonetic level of consonantism.
The classification of the basic Arabic dialects according to the consonantal
composition done by Belova using the latest results in the development of
semitology and afrasian linguistics in the past decades, describes the
complicacies of sound change which to a great extent concern the systems of
the Arabic dialects. The end positions in this classification are occupied by the
south-western dialects of Yemen and the urban dialects from the SyrianLebanese area. She summarizes the relative chronology of sound change in the
following sequence: 1. Reduction of the plosive laryngeal /‘/, delateralisazion
of /Ď/ă andă palatalisationă ofă /g/;ă 2.ă Reductionă ofă theă interdentală sounds;ă 3.ă
Glotalisationăofă/q/ă>ă/‘/ă(Belovaă1999:ă96).
In the light of this common dialect chronology and in the light of the
describedăbyăCorrienteă(1977ăandă1992)ăcharacteristicsăofăААăIăhaveăchosenătoă
examine quantitatively at the phonetic level of consonantism the following
variable features as representative of the particular tendencies of sound change
inăАА:
1. Labial and labio-dental: /b/, /p/, /f/, /m/ and the existence of
unconditioned alternation among them from the type: acipipe < azzabīb ―raisins‖,ăfele tamen < bilā taman ―priceless‖.
2. Interdental sibilants: t – d – z}. In contemporary Arabic dialects the
whole triangle t – d – z} has been preserved or totally undergone
parallel changes. Its preservation is a sign of conservatism and can be
observed in the Bedouin dialects on the Arabian peninsula, the dialects
of Southern Arabia, including Yemen, the dialect of Iraq and the
endmost western periphery of the Arabic-speaking area – ААă andă
h}assānīya.
3. The phoneme corresponding to the emphatic lateral /Ď/ă hasă alsoă
preserved its lateral nature as in other endmost peripheral dialects of
the Arab-speaking area as those of Mauritania, Dacina and Dofar
(SouthernăArabia).ăAăreflexăofătheălaterală/Ď}/ has been preserved in AA
and has penetrated Spanish and Portuguese along with the loan-words
from these dialects: Spanish alcalde ―judge‖,ă Portugueseă arrabal,
arrabalde Ţă ААăar-rabad} ―suburbs,ădistrict‖.ăAlsoăattestedăină AAăareă
1
Concerning statistical tests used here see A. Woods, P. Fletcher & A. Hughes. 1986. Statistics in
Language Studies. Cambridge/New York, and also A.A. Ambros.ă ―Elementare statistische
Verfahren in der Philologie‖ (a lecture), as well as Gatev, K. 1995. Vavedenie v statistikata. Sofia.
192
the realizations /Ď/ă> /t}/ > /t/ and /Ď/ă> /z}/ > /t/: az}-z}afāyir and atafera <
ad}-d}a fīra ―plaits‖.
4. Reflexes from the triad of the medio-lingual pre-Arabică*g‘ă/*kă/ă*yăină
АА.
4.1. The pre-palatală/ğ/: has two main unaffricate realizations: /g/, a reflex
from the Yemenite language characteristics: al-mağarra > almigar ―creek‖ă
and /ž/, typical of the dialects from Central Lebanon to Damascus and the
central cities of Syria, the south-western coast of Yemen (Tihama), the
dialects of North Africa, including the Bedouin dialects of southern and
Eastern Algeria and some rural dialects of Algeria: Marzalcadi > marž alqād}ī. Positional and spontaneous changes of /ğ/ > /d/: dāšīš > ğāšīš ―bran‖ă
can be observed, as well as devocalization /ğ/ > /č/, /ğ/ > /ﻣ/, attested by
someă toponymiesă andă graphicallyă registeredă byă Alkala‘să alternationă withă
medio-lingual glide /y/: aliara > al-ğarra ―jar‖.
The post-palatal /k/ shows weak tendencies to vocalization: žabega > šabaka
―net‖.
The pre-palatal semi-vowel /y/ alternates with the glottal stop in initial or
middle position and assimilates the vibrant /r/ in the structure of some
contemporary toponyms: Bujes, Bujarrabal (burğ > buž ―tower‖). Within the
framework of the pre-Arabic triad there is /y/ < /ğ/ in rare cases.
5. The uvular voiceless /q/ shows in AA the realizations /q/, /g/ and /k/.
Its sonorous realization is characteristic until the last stage of the
existenceăofăАА.ăWhatămakesăanăimpressionăisătheăco-existence of the
realizations, unconditioned by position as it can be observed in the
insular Arabic dialects of Afghanistan and middle Asia: nakfiz / kafaz
and aqlīd (Berberiană algelid)ă ină Ibnă Quzmānă foră instance.ă Ită showsă
some tendencies towards decay in final position: zebra > zawraq
―boat‖
6. The uvular /h~/ăandă/ġ/ăareăcharacteristicăofămostăofătheăArabicădialectsă
and yield some change mainly in the peripheral dialects: in ААă theă
sonorousă/ġ/ăisădevoicedăinăcertainăpositionsăasăcanăalsoăbeăobservedăină
some dialects in Sudan and almost totally in the dialects of Chad and
Nigeria and only partly in Yemen. Devocalization itself to /h~/
undergoes a number of changes to /q/ and /g/ and is written in romance
with the grapheme >ch<: Alcoruche > al-h~urūž ―exit‖,ănaqtoc / qatazt
& nagtoz / gatazt ―toădive‖.ăTheăalternationă/ġ/ă>ă/‗/:ă damġa > dam‗a
(dimāġ ―brain‖)ăcană beăobservedătoo.ăTheăchangeă*ġ>‗>‘ă isăknownă ină
some dialects of Dacina and Dofar in Southern Arabia.
7. The pharyngeal /h}/,ă /‗/ă realizeă ină ААă theă spontaneousă phonetică
variations which can be observed in most Arabic dialects:
193
devocalizationă ofă /‗/ă >ă /h}/ mah}h}ā < ma‗hā ―withă her‖ă oră declineă
(possibly through a glottal stop)ă toă aă zeroă phonemeă ină theă ‗>(‘)>Ø:ă
(nazmau ―weă hear‖,ă cora ―villages‖)ă paradigm.ă Ină theă romanceă
transcriptionăofăААăthereăisăcertainăhesitationăinăputtingăaăgraphemeăforă
/‗/.ă Corrienteă supposesă thată thisă decline,ă markedă alsoă graphically,ă hasă
lead to the appearance of pharyngalazed vowels. It is possible that the
veryăcharacteristicăofăfricativenessăhasădevelopedătowardsădecayăinăААă
as the voiceless fricative has variations too: /h}/ă>ă/‗/,ă/h}/ > /h/, /h}/ > /h~/.
8. Laryngeală/h/,ă/‘/:
8.1.As a whole /h/ has beenăpreservedă ină ААăwithăaătendencyătowardsăitsă
decay in final position: fawāki < fawākih ―fruits‖,ăfaqi < faqīh ―doctoră
ofălaw‖ăandăaănon-positional change in the voiceless pharyngeal.
8.2.The realizations of the laryngeal plosive /‘/ăandăitsăhistoricalăchange in
aă Medievală alreadyă extinctă dialectă asă ААă areă amongă theă distinctiveă
features, the most difficult to register and report on quantitatively. That
is why for this presentation I have chosen this phoneme as illustrating
the methods of quantitative study and the inferences form the
registered data. I am going to look at it more closely in point 5.
4. The inferences from the detailed registration of the realizations of the
variableălaryngealăplosiveă/‘/ăinăАА.ă
I use as a data base all radicals registered in A Dictionary of Andalusi
Arabic which contain this phoneme
There are two processes in the historical change of the laryngeal plosive,
conditioned by its double function, analogical to the function of the semi-vowels –
phonematical (the laryngeal plosive is in the composition of root or grammatical
morphemes) and phonetic (the phoneme has a prosodic function – hiding the
initial vowel, dividing two vowels and so on) (Belova 1999: 77-82).
The tendency towards the decline of this phoneme in both its functions
exists in all Semitic languages and the process reaches its end in the Aramaic
languageă (Rabină 1986:ă 236).ă Theă metamorphosesă ofă theă glottală stopă /‘/ă areă aă
prosodicăphenomenonădatingăbackătoătheăancientăArabicădialects.ăTheăplosiveă/‘/ăisă
most commonly the object of reduction until its full decay. The preserved
linguistic data from the pre-Islamic period show that the process of reduction of
this phoneme is more typical of the western Arabic dialects of Hidjaz whereas a
tendency to preserve it exists in the eastern Arabic dialects of Nedjd, expressed
especially in the language of tribe Tamim (Rabin 1986: 236-263; Belova 1999:
10). The compromising combination of its inferior functioning as a morpheme in
the languages of Western Arabia and its more common realization in the classical
eastern Arabic dialects have lead to the absence of an independent grapheme to
reflect this sound in the script of the Koran and the use of the letter sign ><ء
194
above protetical graphemes – one of the anomalies in the standard Arabic spelling
(Beaston‘săhypothesisă1970,ăquotedăbyăHolesă1995:ă74).
Linguistic sources register three groups of change of theăglottalăstopă/‘/ in
the Medieval Arabic dialects: 1) intensified articulation, 2) reduced articulation, 3)
totalădisappearanceăofă/‘/ă(Lebedevă1977:ă31).
In the attempts to write down the contemporary Arabic dialects the
dependent letter sign is predominantly absent as a reflection of the most common
elision of the glottal stop in the speech flow (Tchobanova 2001: 245).
5.1. The survey of the realizations of theăetymologicală/‘/ăasă firstărootă
constituentă ină theă vocabularyă ofă ААă showsă thată itsă intensifiedă
articulation is alien to this Medieval dialect of Arabic. There is no
ablaut from the phenomenon ‗an‗ana, known about the ancient eastern
Arabic dialects (documented for the tribes Kais, Tamim and Asad and
transferred to some contemporary Bedouin dialects (Lebedev 1977:
31). There are very distinct tendencies towards reduction and elision of
theăetymologicală/‘/ăasăaăfirstărootăconstituentăillustratedăinăTable №:1
Table 1
100
80
60
elision
realisation
40
20
0
Etymological /'/ = First Radical
Theă elisionă ofă /‘/ă ină initială positionă ofă firstă rootă constituentă (84,82%),ă
predominatesă ină AA.ă Theă phonetică functionă ofă theă firstă rootă constituentă /‘/ă hasă
been preserved only, but not always, when in the position between the prefix and
the stem (15,18% of the cases) this phoneme divides two vowels. Most often, its
195
preservationăinăthisăpositionăisăregisteredăinăААăbyăVocabulistaăinăarabicoă(XIIă –
XIII c.) and should be referred to the middle and high register of speech.
Despite the factă thată theă etymologicală /‘/ă ină theă positionă ofă firstă rootă
constituentăisăpredominantlyăonătheădeclineă inăАА,ătheăvowelă inătheăfirstăsyllableă
of the root stems has most often been preserved. As compared to Modern Arabic
dialects (Table № 2) following this distinctive feature shows that the process of
prosodic change in the stem type with first root /‘/ in the classical Arabic dialects
andăClassicalăArabicăatătheăstageăofăААăstillădoesănotăreflectătheăfinalăresultăinătheă
greater part of MAD in which there is a change in the very syllabic structure as a
result of the metathesis of the vowel in the stem ‘a-kal > kal > kla: ―toă eat‖ă
(contemporary Omanian, Lybian, Algerian, Tunesian, Moroccan Arabic dialects).
In some contemporary dialects this process leads to lengthening of the vowel in
the stem after metathesis: ‘ibil > ‘ibil > bīl ―camel‖ă (Lybian)ă oră doublingă ofă theă
third root constituent after the decline of the first root /‘/ and the vowel following
it: ‘ibl > ‘ibil > bill ―camel‖ă (Omanian,ă Iraqi,ă Algerian,ă Tunisian,ă Moroccană
dialects).
Table № 2
AA
Yem.
Oman
ibil
ah~ad/
h~ad
akal
amar
‘ibil
‘ah~ad
‘akal
‘amar
(i)bil/bill
h~ād
kal/kla
amar
Saudi
Arabia
& the
Gulf
‘ibil
‘ah~ad
‘akal
‘amar
Iraq
bill
ah~ad
akal
umar
Syria
&
Palestine
ah~ad
akal
amər
Egypt
Lybia
Mauritania
Alg.
Tun
Mor.
hād
kāl
amar
‘ibal/bīl
h~da
kla
āmər
ībil
h~ād
kāl
ām’¢r
bell
h~da
kla
‗am’r/
ām’r
bell
h~da
kla
amar
bəll
h~da
kla
‘amer
From the peripheral contemporary dialects of the Arabic language the
distinctive feature elision/preservation of the glottal stop as a first root radical and
the vowel following it is illustrated in Bucharian Arabic with kalā and with akal in
the Nigerian Arabic dialect.
The etymological first root /‘/ turns into a semi-vowel in initial position.
This change is spontaneous and can be observed as a variation in the classical,
Medieval and contemporary Arabic dialects: waš < ‘aš*ă ―if/whether‖ă
(contemporary Moroccan), yena < ‘anā (Maltese); its reflection are the doublet
forms registered in the dictionary of Classical Arabic such as ‘akkad/wakkad
196
―confirm,ăsupport,ăassure‖,ă‘annas/wannas ―keepă s.b.ăcompany‖,ă‘annab/wannab
―reproach,ă rebuke‖,ă whichă haveă beenă attestedă ină theă dictionaryă ofă АА.ă Thisă
changeă ofă theă glottală stopă asă firstă rootă cană alsoă beă observedă ină ААă ină veryă
insignificant percentage and obviously refers to a more ancient stage in the
developmentăofăАrabic.
Comparatively well-documentedă ină ААă isă theă initială /‘/ with phonematic
functionăfromătheă‘af‗alămodelăofătheăderivedăverbăfromăwhichăthereăareăreflexesăină
20% out of 1715 examined for the derivation of this type of root groups in the
dictionary of the dialect (Tchobanova 2003: 93-105). The preservation of the
vowel after the phonematic laryngeal voiceless in this stem gives evidence of an
early stage of development of the process of prosodic change which in most MAD
reaches a change in the stem structure after the decline of the initial /‘/ and the
following vowel.
Theăetymologicală/‘/ as a second root constituent:
In contemporary Arabic dialects the laryngeal stop is realized variably to a
predominant elision in middle position. As a result of this the following
vowel is lengthened (Belova 1999: 79). The behaviour of the etymological
/‘/ă asă aă secondă rootă andă theă middleă glottală stopă whichă isă chargedă withă
morpho-phonological functions is different.
From theărootsăwithăaăsecondăradicală/‘/,ăforătheăcommonăArabicăliteratureă
vocabulary, included in al-Munğid,ă ină Aă Dictionaryă ofă ААă areă documentedă asă
shown in Table № 3:
Table № 3
First
Radical
bā‘
alMunğid
17
tā‘
tā‘
5
9
ğīm
8
h}ā‘
h~ā‘
dāl
0
0
7
dāl
rā
6
9
al-Munğid
A Dictionary
A Dictionary
b‘b‘,ă b‘b,ă b‘t,ă b‘ğ,ă b‘d,ă b‘r,ă
b‘z,ă b‘s,ă b‘ﻣ,ă b‘t},ă b‘q,ă b‘l,ă
b‘m,ăb‘n,ăb‘h,ăb‘w,ăb‘y
t‘b,ăt‘t‘,ăt‘răt‘q,ăt‘m
t‘b,ă t‘t‘,ă t‘ğ,ă t‘d,ă t‘r,ă t‘t,ă t‘t},
t‘l,ăt‘y
ğ‘b,ă ğ‘t,ă ğ‘ğ‘,ă ğ‘r,ă ğ‘ﻣ,ă ğ‘f,ă
ğ‘l,ăğ‘y
d‘b,ăd‘d‘,ăd‘s},ăd‘z} ,ăd‘l,ăd‘m,ă
d‘w
d‘b,ăd‘r,ăd‘t}, d‘f,ăd‘l,ăd‘m
r‘b,ă r‘t,ă r‘d,ă r‘r,ă r‘s,ă r‘f,ă r‘l,ă
197
3
b‘r,ăb‘z,ăb‘s
2
3
t‘r,ăt‘m
t‘b,ăt‘r,ăt‘l
0
0
0
0
-
1
4
d‘b
r‘s,ăr‘f,ăr‘m,ăr‘y
zāy
12
sīn
9
ﻣīn
7
s}ād
d}ād
6
10
t}ā‘
z}ā‘
‗ayn
ġayn
fā‘
1
3
0
0
8
qāf
kāf
3
9
lām
mīm
7
8
nūn
10
hā‘
wāw
1
10
yā‘
2
r‘m,ăr‘y
z‘b,ăz‘br,ăz‘bq,ăz‘ğ,ăz‘d,ăz‘r,ă
z‘z,ăz‘f,ăz‘k,ăz‘m,ăz‘n,ăz‘y
s‘b,ăs‘t,ăs‘d,ăs‘r,ăs‘s‘,ăs‘f,ăs‘l,ă
s‘m,ăs‘w
‘ﻣb,ă ‘ﻣz,ă ‘ﻣs,ă ‘ﻣf,ă ‘ﻣm,ă ‘ﻣn,ă
‘ﻣy
s}‘b,ăs}‘s}‘,ăs}‘k,ăs}‘l,ăs}‘m,ăs}‘y
d}‘b,ă d}‘d,ă d}‘z,ă d}‘d}, d}‘d}‘,ă d}‘t},
d}‘k,ăd}‘l,ăd}‘n,ăd}‘y
t}‘ăt}‘
z}‘b,ăz}‘r,ăz}‘m
f‘t,ăf‘d,ăf‘r,ăf‘s,ăf‘f,ă f‘q,ă f‘m,ă
f‘w
q‘b,ăq‘q,ăq‘y
k‘b,ă k‘ğ,ă k‘d,ă k‘s,ă k‘s},ă k‘k,ă
k‘l,ăk‘n,ăk‘y
l‘b,ăl‘t,ăl‘z} ,ăl‘k,ăl‘l‘,ăl‘m,ăl‘y
m‘ğ,ă m‘d,ă m‘r,ă m‘s,ă m‘m,ă
m‘n,ăm‘w,ăm‘y
n‘n‘,ă n‘t,ă n‘ğ,ă n‘d,ă n‘r,ă n‘ﻣ,ă
n‘f,ăn‘l,ăn‘m,ăn‘y
h‘h‘
w‘w‘,ă w‘b,ă w‘ğ,ă w‘d,ă w‘r,ă
w‘s},ăw‘t},ăw‘l,ăw‘m,ăw‘y
y‘s,ăy‘y‘
3
z‘b,ă z‘br,ă z‘bq,ă
z‘r,ăz‘q,ăz‘n,ă
s‘r,ăs‘l,ăs‘m
3
‘ﻣm,ă‘ﻣn,ă‘ﻣw
1
1
s}‘b
d}‘nă
1
0
0
0
0
3
t}‘ăt}‘
f‘d,ă f‘r,ă f‘s,ă f‘q,ă
f‘l,ăf‘w
k‘b,ăk‘s,ăk‘l
2
3
l‘l‘,ăl‘m
m‘q,ăm‘n,ăm‘y
1
n‘y
0
1
w‘d
1
y‘s
6
At the acoustic level predominates the impression that the voiceless
phonemeă /‘/ participates in ancient roots with onomatopoeic origin from the
lexical fund of the language. That is why the tendency towards the decline of this
sound in the historical development of the language is not surprising. The
combinationsă betweenă/‘/ăasăaăsecondărootăandătheăuvulară/h~/,ă/ġ/ăandăpharyngeală
/h}/,ă /‗/ă asă firstă rootă constituentă areă notă characteristică ofă theă commonă Arabică
vocabulary fund. Judging by the frequency of realizations, documented in AA, we
can notice that in this Medieval variation of the spoken Arabic, have almost
completelyă declinedă typesă ofă combinationsă betweenă /‘/ă asă aă secondă rootă andă theă
phonemesă /ğ/,ă /d/,ă /z}/, /q/, /h/ (0 realizations) and /d/, /s}/, /d}/, /t}/, /n/, /w/, /y/ (1
198
realization). The number of the roots which still document second radical /‘/ in its
composition is greatly reduced.
Ină percentageă theă elisions/realisationsă ofă /‘/observedă ină theă vocabularyă ofă
AA were illustrated in Table № 4 as follows:
Тable № 4
80
60
40
elision
20
realization
0
Etymological /'/ = second
radical
The analysis of the realization of the glottal stop in certain vocal contexts
leads us to the following conclusions:
The laryngeal plosive shows the strongest tendency to preservation in
homogeneous vocal surroundings /a – a/ and heterogeneous /a – ‘– i/, /a – ‘ă– ï /
and / ï – ‘ă – a/: ra‘aw ―theyă saw‖,ă ra‘īs ―head,ă chief‖,ă ra‘s ―head‖,ă tas‘al ―youă
ask‖.ăWeăcanăagreeăwithăCorriente,ăthată/‘/ăusedătoăfunctionăasăprosodicăseparatoră
between vocals, but this function disappears in later sources: (AL): diēb, (VA)
diāb ―woolfs‖.ăTheăverbăra‘ā which in MAD has completely been replaced by šāf
was still used and well documented in AA.
The cases of elimination of the glottal pause with elision of the following
vocal predominate in AA by 80%: rayt ―I,ăyouăsaw‖.ăInăallăotherăcontextsăthereăisă
eliminationăofă/‘/.ăThisăchangeăhasăbeenăregisteredăorthographicallyăasă/‘/ă>ă/w/ăină
the predominant surrounding /u – ‘ă– a /: nuwaddab ―toăbeăpunished‖ăandă/‘/ă>ă/y/ă
(which was written with >g< in the later sources) in the surroundings /a – ‘ă– i/, /i
– ‘ă– a/): d}amāyir ―thoughts‖,ăradiyya ―bad‖,ă/ï – ‘ă– v/: tas‘al > tasal ―youăask‖,ă
maš‘ūm > mašum ―ominous‖.
In the context /v – ‘ă– ï/ăААăseemsătoăhaveăcontinuedătheăevolutionăofătheă
classical
199
Arabicădialectsăinăwhichătheăchangeă/‘/ă>ă/vv/ has been registered when the vowel
is tonic: rās ―head‖,ăbīr ―well‖.
In rare cases there is germination of the consonant following the glottal
stop: ya‘kul > yākul > yakkul, yuhaddi‘nī > yahaddāni > yahaddani.
The greatest variation in the realization/decline of this phoneme has been
observed in IQ, ZJ and IA: this may be considered as a sign of conservatism of the
language which has been extended at the stage of most consolidate development
of AA (during XII – XIV centuries).
Theăetymologicală/‘/ăasăaăthirdăroot.
The laryngeal plosive declines in final position of the root stem in all
Modern Arabic dialects.
Fromă theă rootsă withă aă thirdă rootă /‘/,ă foră theă commonă Arabică literatureă
vocabulary included in al-Munğid, in A Dictionary of AA are documented as
shown in Table № 5:
Table №ă5
First
alal-Munğid
A Dictionary A Dictionary
Radical
Munğid
alif
1
‘b‘
0
_
bā‘
6
4
bd‘,ăbd‘,ăbr‘.ăbt}‘,ăbk‘,ăbw‘
bd‘,ăbr‘.ăbt}‘,ăbw‘
tā‘
4
tf‘,ătk‘,ătl‘,ătn‘
0
_
tā‘
2
tf‘, tm‘
0
_
ğīm
10
ğ‘ğ‘,ă ğb‘,ă ğr‘,ă ğz‘,ă ğs‘,ă ğ‘ﻣ,ă 4
ğr‘,ăğz‘,ăğ‘ﻣ,ăğy‘
ğf‘,ăğm‘,ăğn‘,ăğy‘
11
h}ā‘
h}t‘,ă hğ‘,ă h}d‘,ă h}z‘,ă h}‘ﻣ,ă h}s}‘,ă 2
h}f‘,ăh}m‘
h}f‘,ăh}k‘,ăh}l‘,ăh}m‘,ăh}n‘
9
h~ā‘
h~b‘,ă h~t‘,ă hğ‘,ă hd‘,ă h~r‘,ă h~s‘,ă 4
h~b‘,ăh~r‘, h~s‘,ăh~t}‘
h~t}‘,ăh~f‘,ăh~l‘
dāl
5
db‘,ădr‘,ădf‘,ădn‘ădw‘
4
dr‘,ădf‘,ădn‘ădw‘
dāl
2
dr‘,ădm‘
1
dr‘
rā
14
rb‘,ă rt‘,ă rt‘,ă rğ‘,ărd‘,ă rz‘,ă r‘ﻣ,ă 5
rğ‘,ă rd‘,ă r‘ﻣ,ă rf‘,ă
rq‘,
rt}‘,ărf‘,ărq‘,ărq‘,ărm‘,ărn‘,ărw‘
zāy
4
zr‘,ăzk‘,ăzl‘,ăzn‘
0
_
sīn
5
sb‘,ăsr‘,ăsl‘,ăsm‘,ăsw‘
1
sb‘
ﻣīn
4
3
ﻣt}‘,ăﻣq‘,ăﻣn‘,ăﻣy‘
st}‘,ăﻣn‘,ăﻣy‘
2
0
_
s}ād
s}b‘,ăs}d‘
3
1
d}ād
d}b‘,ăd}n‘,ăd}w‘
d}w‘
7
t}ā‘
t}r', t}s‘,ă t}f‘,ă t}m‘,ă t}m‘n,ă t}mn, 2 (3)
t}r‘,ă t}f‘,ă *ă t}m‘ă >ă *ă
t}n‘,ăt}w‘
t}m‘nă>ăt}mn
1
0
_
z}ā‘
z}y‘
‗ayn
1
‗b‘
1
‗b‘
200
ġayn
fā‘
qāf
0
7
7
kāf
8
lām
11
mīm
6
nūn
15
hā‘
10
wāw
13
yā‘
0
_
ft‘,ăfğ‘,ăfr‘,ăfs‘,ăft}‘,ăfq‘,ăfn‘
qt‘,ă qr‘,ă qd}‘,ă qf‘,ă qm‘,ă qn‘,ă
qy‘
kt‘ă kd‘,ă ks‘,ă k‘ﻣ,ă kf‘,ă kl‘,ă
kw‘,ăky‘
lb‘,ă lt‘,ă lt‘,ă lğ‘,ă lz‘,ă lt}‘,ă lz}‘,ă
lf‘,ălk‘,ălm‘,ăly‘
mt‘,ă mr‘,ă ms‘,ă ml‘,ă mn‘,
mw‘ă
n‘n‘ă nb‘,ă nt‘,ă nğ‘,ă nd‘,ă nz‘,ă
ns‘,ă n‘ﻣ,ă ns}‘,ă nf‘,ă nk‘,ă nm‘,ă
nh‘,ănw‘,ăny‘
h‘h‘,ă ht‘,ă hğ‘,ă hd‘,ă hr‘,ă hz‘,ă
hm‘,ăhn‘,ăhw‘,ăhy‘
w‘‘ăwb‘,ăwt‘,ăwt‘,ăwğ‘,ăwd‘,ă
wd‘,ă wr‘,ă ws}‘,ă wd}‘,ă wt}‘,
wk‘,ăwm‘
_
0
2
2
_
fr‘,ăfq‘
qt‘,ăqr‘
2
kf‘,ăkw‘
3
lt‘,ălğ‘,ălt}‘
1
mr‘
5
nb‘,ă nt‘,ă ns‘,ă n‘ﻣ,ă
ny‘
2
hd‘,hn‘,
5
wb‘,ă wd}‘,ă wk‘,ă
wm‘
0
_
As a third radical too, the glottal stop participates in ancient roots with
semantic of everyday life, stock-breeding and agriculture. To the productive roots
withă thirdă constituentă /‘/ă ină Classicală Arabică correspond productive roots well
documented in AA too.
Theă etymologicală /‘/ă declinesă ină finală positionă ofă theă rootă ină AA,ă asă ită isă
typical for MAD. She is neither preserved in cases of morphological incorporation
– it is obvious that the process of its change towards elimination has finished its
development in the Middle Arabic. There are no oppositions with the closest
phonemeă/‗/ăasăareăregisteredăinătheăYemeniăArabic.ăTheăpreservationăofătheăglottală
stop is very rare – mostly for reasons of the rhyme and the number of the
syllables. It is not more often that compensation of this elision happens by
gemination of a consonant or the change of the glottal stop: df‘ > dff: daff, ‗b‘ >
‗bb (VA, AC), t}m‘ > t}mn (‘ă>ăn).
Based on a representative sample the elisionsăandătheărealisationsăofă/‘/ăasăaă
third root are illustrated respectively by more than 90% of cases of elision in
Table № 5 as follows:
Table № 5
100
201
80
60
40
elision
20
realisation
0
The etymological /'/ = thid
radical
9. The realisations of the laryngeal plosive in AA. Conclusion.
The processes of reduction of the laryngeal plosive in AA have almost
reached the stage known in the MAD. The innovations arisen as a result of the
decay of the laryngeal have no homophones in the morphological and the lexical
systems of the dialect. In so far as the tendencies to reduction of this consonant
(the weakest one from the three weak /‘/,ă /w/,ă /y/)ă seemă toă haveă germă ină theă
ancient dialects of the Arabian peninsula, considering its decline in AA as due to
substratum influences is not convincing. At the sight of outstanding of its semivowel and vowel reflexes it is not excluded though.
References
Al-munğid fi l-luġa wa l-‘a‗lām. 1967.ăBayrūt.
Ambros,ăArneăA.ă―ElementareăstatistischeăVerfahrenăinăderăPhilologie‖ă(aălecture).
Belova, Anna G. 1999. Ocherki po istorii arabskogo yazika. Moskva: Prosveta.
Corriente, Federico. 1977. A Grammatical Sketch of the Spanish-Arabic Dialect Bundle,
Madrid: Instituto Hispanoárabe de Cultura.
Corriente, Federico. 1980. Gramática, métrica y texto del Cancionero hispanoárabe de
Aban Guzmán: reflejo de la situación lingüística de Al-Andalus tras concluir el
periodo de las Taifas, Madrid: Instituto Hispanoárabe de Cultura.
Corriente, Federico. 1985-6. ―H~asā‘is} kalāmă ‘ahlă al-‘andalusă natran wa naz}man‖ă In:ă
Revista del Instituto Egypcio de Estudios Islámicos 23, 59-68.
Corriente, Federico. 1987.ă ‖Lasă xarajātă enă árabeă andalusíă (aproximaciónă aă suă ediciónă yă
estudioădialectológicos)‖.ăIn:ă―Al-Qant}ara‖ă8,ă203 – 264.
Corriente, Federico. 1988. El léxico árabe andalusí según Pedro de Alcalá. : (ordenado
por raíces, corregido, anotado y fonémicamente interpretado), Madrid :
Departamento de Estudios Árabes e Islámicos, Universidad Complutense.
Corriente, Federico. 1989. El léxico árabe andalusí según el ―Vocabulista in arábico‖.
Madrid : Departamento de Estudios Arabes e Islámicos 2, Universidad
Complutense
Corriente, Federico. 1990. ―JudíosăyăcristianosăenăelăDīwānăibnăQuzmān,ăcontemporáneoă
deă Abrahamă ibnă Ezra‖ă In:ă Abraham ibn Ezra y su tiempo: actas del Simposio
Internacional, Madrid, Tudela, Toledo, 1-8 febrero 1989.
Corriente, Federico. 1991. El léxico árabe estándar y andalusí del Glosario de Leiden.
Madrid: Departamento de Estudios Arabes e Islámicos, Universidad
Complutense.
Corriente, Federico. 1992. Árabe andalusí y lenguas romances, Madrid: Mapfre.
Corriente, Federico. 1993-4. ―Adicionesăprocedentesădeălaă―ă‗uddatăaljalīs‖ăyăotras fuentes
alăcatálogoădeălasă―xarajāt‖ăenăárabeăandalusí‖.ăIn: Revista del Instituto Egypcio
de Estudios Islámicos 26, 10 – 48.
202
Corriente, Federico. 1995. Dīwān ibn Quzmān al-Qurt}ubī: ‘is}ābat al-‘aƶrād fī dikr al‘a‗rād. al-Qāhira.
Corriente, Federico. 1997. A Dictionary of Andalusi Arabic, Leiden: Brill.
Ferrando, Ignacio. 2000. “Le morphème de liaison /-an-/ en arabe andalou. Notes de
dialectologieăcomparée‖ăIn:ăOrienteăModerno,ăXIXă(LXXX),ăn.s.,ă25-46.
Gatev, Kiril. 1995. Vavedenie v statistikata, Sofia: LIA.
Holes, Clive. 1995. Modern Arabic: structures, functions and variables, New York:
Longman.
Lebedev, Viktor. 1977. Pozdniy srednearabskiy yazik (XIII – XVIII vv.), Moskva: Nauka.
Marugán, Marina. 1994. El refranero andalusí de Ibn ‗ƨs}im al-ƶarnāt}ī, Madrid:
Hiperión.
Ould Mohamed Baba, A.-S. 1999. Estudio dialectológico del ―Refranero andalusí de
Abū Yahyá Azzajjālī‖. Zaragoza: Univercidad de Zaragoza.
Owens,ă Johnathan.ă 2005.ă ―Pre-diaspora Arabic. Dialects, statistics and historical
reconstruction‖.ăIn:ăDiachronicaă22:2,ă271ă– 308.
Rabin, Chaim. 1986. Al-lahağāt al-‗arabīya al-ġarbīya al-qadīma. Al-Kuwayt.
Tchobanova,ăHristina.ă2001.ă―Pravopisniăprintsipiăiăfonetichniăzakoniăvăzapisvanetoănaă
savremenniteăarabskiădialekti‖ăIn:ăArabistika i Islamoznanie 1 (Pashova,
Theophanov, Pavlovich (ed.), Arabic and Islamic Studies 1), Sofia: Sofia
University Press St. Kliment Ohridski, 240-252.
Tchobanova, Hristina. 2003. ―Modelată(‘)af‗alăvăglagolnotoăslovoobrazuvane na
araboandalusiyskiya dialect – edinăsinhronenăpogled‖ăIn:ăArabistika i
Islamoznanie 2 (Evstatiev, Tchobanova & Dyulgerov (ed.), Arabic and Islamic
Studies 2), Sofia: Sofia University Press St. Kliment Ohridski, 93 – 105.
Woods, A., P. Fletcher & A. Hughes. 1986. Statistics in Language Studies.
Cambridge/New York: Cambridge University Press.
203
204