Genitive Markers in Omani Arabic (2019)

UNIVERSITY OF BUCHAREST CENTER FOR ARAB STUDIES ROMANO-ARABICA XIX Curses and Profanity in the Languages and Cultures of the Middle East and North Africa editura universităţii din bucureşti ® 2019 Editors: George Grigore (University of Bucharest, e-mail: [email protected]) Laura Sitaru (University of Bucharest, e-mail: [email protected]) Associate Editors: Gabriel Bițunã (University of Bucharest, e-mail: [email protected]) Ovidiu Pietrãreanu (University of Bucharest, e-mail: [email protected]) Editors in charge of this issue George Grigore & Gabriel Bițună Website administrator: Ștefan Ionete (University of Bucharest, e-mail: [email protected]) Cover designer (Graffito in Tunis. Photo taken by the designer) Dennis Bițună Blind peer reviewed Editorial and Advisory Board: Jordi Aguadé (University of Cadiz, Spain) Andrei A. Avram (University of Bucharest, Romania) Ramzi Baalbaki (American University of Beirut, Lebanon) Ioana Feodorov (Institute for South-East European Studies, Bucharest, Romania) István T. Kristó-Nagy (University of Exeter, UK) Hayder Ghadhban Mohsin (University of Babylon, Iraq) Pierre Larcher (Aix-Marseille University, France) Jérôme Lentin (INALCO, Paris, France) Giuliano Mion (University of Cagliari, Italy) Luminiţa Munteanu (University of Bucharest, Romania) Bilal Orfali (American University of Beirut, Lebanon) Stephan Procházka (University of Vienna, Austria) Mehmet Hakkı Suçin (Gazi University, Ankara, Turkey) Shabo Talay (Free University of Berlin, Germany) Irina Vainovski-Mihai (“Dimitrie Cantemir” Christian University, Bucharest, Romania) Ángeles Vicente (University of Zaragoza, Spain) John O. Voll (Georgetown University, Washington, D.C., USA) Published by: © Center for Arab Studies 7-13, Pitar Moș Street, District 1, 010451, Bucharest, Romania Website: http://araba.lls.unibuc.ro © Editura Universitãții din București Șos. Panduri nr. 90-92, 050663 București, ROMÂNIA. Tel./Fax: +40214102384 E-mail: [email protected] Web: http://editura-unibuc.ro; Centru de vânzare: Bd. Regina Elisabeta nr. 4-12, 030018 București, ROMÂNIA Tel. +40213053703; Tipografia EUB: Bd. Iuliu Maniu nr. 1-3 061071 București, ROMÂNIA Tel./Fax: +40213152510. The official website of Romano-Arabica https://romanoarabica.academy ISSN 1582-6953 Romano-Arabica Journal is an A-class Academic Journal, according to the Italian National Agency for the Evaluation of Universities and Research Institutes – Anvur, Italia: http://www.anvur.it/en/homepage/ 4 Contents I. CURSES AND PROFANITY IN THE LANGUAGES AND CULTURES OF THE MIDDLE EAST AND NORTH AFRICA .............................................. Lucia Avallone. Literary Creativity and Curses. A Study Case: ’an takūn ‘Abbās al‘Abd, by ’Aḥmad al-‘Āydī ................................................................................ Basilius Bawardi. ‫ ابن ضعف اإلنسان" عن التجديف من أجل فرح الوجود في شعر نزيه أبو عفش‬..‫ "هللا‬.. Gabriel Bițună. The Tunisian Swearosaurus. Swear Words in the Spoken Arabic of Tunis Meriem Bouzid Sababou. ‫ دالالت االهانة عند الطوارق من خالل طقوس األداء‬......................... Adela Chiru. Gros Mots and Curses in Asghar Farhadi’s Movies ............................ Luca D’Anna. Curses, Insults and the Power of Words: Verbal Strategies in Maghrebi Dialects ............................................................................................ Emanuela De Blasio. The Use of Youth Language and Coarse Words in the Mashreq Are Nino Ejibadze. Cursing and Reviling Formulas in the Egyptian Arabic Dialect ....... Chiara Fontana. Rhetorical Features of Cursing and Foul-Mouthed Speech in Contemporary Masters of Muǧūn: Muẓaffar an-Nawwāb and Naǧīb Surūr ......... Mufleh Hweitat. ‫ دراسة في الرؤية والتشكيل‬:‫عنَيْن‬ ُ ‫بالغة القُبْح في شعر ابن‬................................. Benjamin Koerber. al-Maqāma al-kāfiyya of ʿAlī al-Mūrālī (c. 1950): An Archive of Tunisian Cursing ......................................................................................... Letizia Lombezzi. ˀAḫu Šarmūṭe and his Relatives: Productive Genealogies for Arabic Embodied Curses ................................................................................. Gabriel M. Rosenbaum. Curses, Insults and Taboo Words in Egyptian Arabic: in Daily Speech and in Written Literature ........................................................... Jonas Sibony. Curses and Profanity in Moroccan Judeo-Arabic and What’s Left of it in the Hebrew Sociolect of Israelis from Moroccan Origins ........................ II: MISCELLANEA .................................................................................................... Maurizio Bagatin. La variation linguistique selon Ibn Ḫaldūn ................................. Simone Bettega. Genitive Markers in Omani Arabic ................................................. Irina Vainovski-Mihai, George Grigore. From Dobrudja to Ada-Kaleh: A Bridge between Empires ............................................................................................... II: BOOK REVIEWS .................................................................................................. Luca D’Anna. 2017. Italiano, siciliano e arabo in contatto. Profilo sociolinguistico della comunità tunisina di Mazara del Vallo Centro di Studi Filologici e Linguistici siciliani, Palermo (Cristiana Bozza) ............................................... Arik Sadan, Almog Kasher. 2018. A Critical Edition of the Grammatical treatise Mīzān al-carabiyya by Ibn al-’Anbārī (d. 577/1181). Wiesbaden: Harassowitz Werlag (Ovidiu Pietrăreanu) .......................................................................... Hela Ouardi. 2016. Les derniers jours de Muhammad (Laura Sitaru) ..................... 7 9 19 35 47 59 71 83 93 99 115 129 143 153 189 205 207 223 239 247 249 255 257 II: MISCELLANEA GENITIVE MARKERS IN OMANI ARABIC SIMONE BETTEGA University of Turin Abstract. This paper investigates the role of the particles māl, ḥāl, ḥagg and ḥaqq in the Arabic dialects of Oman. These particles have often been described as markers of possession in Omani Arabic: however, the data presented in this article seem to indicate that these elements are employed to express a wide variety of NPinternal specifications that go well beyond the realm of linguistic possession, and that they can also be used predicatively outside the boundaries of a NP. ḥāl and ḥagg, in particular, should probably not be described as genitive markers at all. The syntactic and pragmatic contexts in which the aforementioned markers can be used will be discussed, along with the range of meanings and semantic categories they can express. Keywords: Possession, Genitive, Oman, Arabic dialectology. 1.Introduction In the course of this paper, I will analyze the use that speakers of Omani Arabic (OA) make of certain lexical elements mostly connected to the expression of possession. It is important to keep in mind that OA is not a homogeneous linguistic entity, but rather, it constitutes a bundle of (more or less tightly) interrelated dialects. All these varieties, at any rate, are at present gravely under-researched. Due to space constraints it is not possible to offer here a comprehensive survey of the existing studies on OA. The reader is therefore referred to Holes (1989 and 2008) and Davey (2016) for an overview. The present article, in particular, is mostly concerned with the Arabic dialects spoken in the northern half of the Sultanate (though references to the southern varieties of Dhofar will be included as well, and integrated with the materials presented in Davey 2012 and 2016). The article is structured as follows: in § 2 I will circumscribe the scope of my research by providing a standard definition of linguistic possession. In § 3 I will offer an overview of the existing literature on the topic of genitive markers in Arabic dialects, focusing in particular on studies dealing with genitive markers in peninsular varieties. Finally, in § 4 I will discuss the results of the analysis that I have carried out on my corpus of data. 2. What is possession? For the present discussion I will adopt a standard definition of linguistic possession, as formulated by Basic Linguistic Theory. Dixon (2010: 262) defines possession as a loose SIMONE BETTEGA term «used to cover a wide range of relationships». Table 11 illustrates the main types of relationships that can be expressed through a possessive construction within a noun phrase across the world’s languages (the first three elements of the list being the most common cross-linguistically). As can be seen, most of these relationships are quite dissimilar in nature: Types of possessive relationships 1) 2) 3) 4) 5) 6) Ownership or temporary possession (John’s car) Whole-part relationship (The door of the car) Kinship relationship (affinal or consanguineal, as in John’s wife or John’s mother) An attribute of a person, animal or thing (John’s temper) A statement of orientation or location (The inside of the car) Association (John’s dentist) In many languages that make use of specific elements to mark possession, these same markers may be employed to signal other kinds of specification such as quantity/collectivity (two cups of tea, a bunch of bananas) or material (a house of straw, the crown of gold). Although these structures may, at surface level, resemble the ones described above, they are to be kept distinct, since they fall outside the boundaries of “possession”, even in its widest interpretation. This is demonstrated by the fact that they cannot be rephrased using a predicative (that is, non NP-internal) possessive construction (for instance, while it is possible to rephrase John’s car in John has a car2, this cannot be done in the case of a cup of tea). Markers of possessive constructions can sometimes also signal the function of a noun phrase within a clause: for instance, many languages employ the same form for genitive (which marks possessive function within an NP) and dative case (which marks function within a clause)3. As we will see, all of the above holds true for several varieties of Arabic, including OA: this is probably the reason that lead to some inconsistencies in the descriptions of possessive relationships in the existing works on the subject, which will be examined in the following paragraphs. 3. Genitive Markers in the Arabic dialects of Arabia In most varieties of spoken Arabic there are normally two ways of expressing an NPinternal possessive relationship: via a synthetic genitive (SG) structure (called ʔiḍāfa in Arabic, but often referred to as construct in the western literature) or by means of a periphrasis. While in the case of the ʔiḍāfa the possessed and possessor (henceforth PD and PR) are simply juxtaposed, the analytic structure (AG) requires an explicit marker4 (glossed 1 2 3 4 Adapted from Dixon (2010: 262-5), as all the examples presented in this paragraph. Obviously, the syntactic function of the constituents changes. NP-internal possessive constructions presuppose a relationship, predicative possessive constructions are used to establish one. This happens because «something that is ‘for X’ (dative) is likely soon to be ‘X’s’ (genitive)» (Dixon 2010: 291). Genitive markers received different labelling from different authors: Belnap (1991) refers to them as possessive adjectives, Ingham (1994) as possessive particles, Davey (2016) as genitive linkers, Eksell-Harning (1980) and Brustad (2000) as genitive exponents. I follow Holes (2008) in calling them genitive markers. 224 GENITIVE MARKERS IN OMANI ARABIC as GEN in the course of this work) to be inserted between PD and PR. Consider for instance examples (1) and (2), as opposed to (3) and (4): (1) dišdāšt-uh dišdāša=PRON.3S.M ‘His dišdāša’ (typical Omani dress) (2) ṣōt al-wēlāt sound ART=wheel-PL.F ‘The sound of the wheels’ (3) əl-gamal ART=camel ‘His camel’ māl-uh GEN=PRON.3S.M (4) əl-maqhā ART=café ‘Burʕī’s café’ māl GEN burʕī burʕī Eksell (2009: 35) writes that the SG is still productive in most dialects, and that the choice that speakers operate between the two alternative constructions (SG or AG) «is a complex process operative on several levels of speech, with multiple set of factors involved» (see also Brustad 2000: 74). The most extensive study of AGs in Arabic dialects existing to date is probably that of Eksell-Harning. About the geographical area that concerns us here, she writes that «in the Eastern half of the Peninsula, including the Persian Gulf, Oman and Dhofar, the use of the AG seems to be more restricted and it is doubtful whether the AG occurs regularly except at isolated locations. There is also a variety of exponents» (Eksell-Harning, 1980: 69). As we will see, our data appear to confirm both statements. As far as studies dealing specifically with southern and eastern Arabia are concerned, Reinhardt (1894: 79) reports the use of boht māl and ḥāl in OA. Holes (1990: 96, 170-1) notes the two particles ḥagg and māl for the dialect of the northern Gulf coast plus ḥāl for OA5. Johnstone (1967: 69, 90-1) as well reports the use of māl and ḥagg in the Gulf. Johnstone’s work in particular contains a number of interesting remarks. First of all, he notes how ḥagg is used only with the meaning of ‘for’, while «it does not mean ‘of’ as it does in many other dialects, though the context might sometimes appear to suggest this»6. He also observes (as does Brustad, 2000: 72) that the “preposition” māl may agree in gender and number with the preceding noun, giving the two forms mālat (F.SG.) and mālōt (PL.), though such agreement is not obligatory and less common than non-agreement. Finally, 5 6 Holes (2008: 484) lists bu as well for OA. I will not discuss this marker in this paper, since it only rarely appeared in my data. It would seem that it was once one of the main genitive markers in northern Oman, but that it is today receding at the expenses of māl. I have very frequently heard it in the towns and villages of the Hajar massif, such as Bahla or Al-Hamra. Cf. as well Brustad (2000: 72-3). Also Johnstone (1982: 584), in his review of Heksell-Harning’s book, refuted her claim that ḥagg could be employed as a genitive marker in Gulf Arabic. 225 SIMONE BETTEGA Johnstone repeatedly insists on the fact that māl «is not used ordinarily where a construct complex [i.e. an SG] is possible». Eksell-Harning (1980: 70), commenting on examples drawn from Johnstone’s materials, notes that «apparently, the māl phrase is used not only to express possession but also qualification». For southern Oman (Dhofar), Davey (2012: 69) observes that both ḥaqq and māl are in use in the local dialect, and that both of them may inflect for gender and number (though, again, this appears to be an optional feature). Unlike its cognate from the northern Gulf, Dhofari ḥaqq would seem to fully deserve the title of genitive marker, as māl does, since «there does not appear to be any functional difference between these elements in everyday usage». Davey (2012: 71) also gives some interesting information about the definiteness of the PR and PD within AGs in Dhofari Arabic, commenting that, althought a definite PD followed by a definite or pronominal PR represents the most common combination, indefinite PDs and/or PRs are attested as well. This appears to contradict both EksellHarning (1980: 74) and Ingham (1994: 58), according to whom asymmetrically definite AGs can be constituted by an indefinite PD and a definite PR, but not the other way around7. Both Davey and Eksell-Harning discuss the type of semantic relationships that can be expressed through an AG: this will be analyzed in more depth in § 4.2.2. One last point which is worth considering here is the motivation behind the use of AGs (rather than SGs). According to Eksell-Harning (1980: 79-81), these motivations are rarely semantic in nature (e.g. connected to the distinction between alienable and inalienable possession), but rather “stylistic” (e.g. the contrast between two or more concepts, the introduction of a new theme, or the climax in a chain of events). Brustad (2000: 76) apparently agrees with this when she writes that «the genitive exponents fulfill specific pragmatic functions that the construct phrase does not. [They place] a focus on the possessing noun (in linear terms, the second noun) not conveyed by the construct phrase». 4. Data Analysis The analysis that I present in this paragraph is based on a heterogeneous corpus of data, which includes both original interviews that I have recorded in Oman during several fieldwork periods in the years from 2014 to 20168, and material drawn from popular Omani TV shows9. In particular, my own recordings consist of one-to-one interviews, elicited examples 10 and voice messages recorded via instant messaging applications such as 7 8 9 10 Here Eksell-Harning is referring to western Arabian dialect, while Ingham to central Arabian ones. My informants were mostly males with university-level education in their 20s or 30s. They were all from the northern towns of Muscat, Nizwa and Ibri, or their immediate surroundings. Though I have never carried out fieldwork research in Dhofar, the television material I have employed included a limited amount of Dhofari speech. Dhofari Arabic genitive markers present a number of lexical and morphological peculiarities that set them apart from those used in the north of Oman. Some examples of Dhofari AGs are discussed in the following paragraphs. Direct elicitation was only sparingly used during the data-gathering phase. It was only after careful examination of my material that I resorted to elicitation, in order to confirm or dismiss some of the hypotheses that I had formulated. I have clearly signaled the few examples that appear in this article that were obtained by means of elicitation. 226 GENITIVE MARKERS IN OMANI ARABIC WhatsApp or Viber. The latter are of particular interest because, although not dialogical in the proper sense of the term, they closely approximate the style of intimate and informal communication between friends (especially among young users)11. In general, I tried to include in my dataset as many text-types as possible, in order to be able to examine the use of AGs in a wide variety of communicative situations and pragmatic contexts. A cursory survey of this dataset reveals three fundamental facts: the first one is that, especially when compared to other dialects, AGs in OA appear to be extremely uncommon, and vastly outnumbered by SGs. This is consistent with the tentative classification provided by Eksell-Harning (1980: 158). The second fact that emerges from a preliminary analysis of the data is that māl is by far the most commonly employed marker, followed by ḥāl. ḥagg is almost entirely absent, with the exception of its variant ḥaqq, which appears to be well-estabished in Dhofari speech (again, this is consistent with Davey 2012 and 2016). Finally, it appears that the semantic scope of all these elements goes well beyond the boundaries of possession or even genitive case. In fact, some of them should probably not be classified as genitive/possessive markers at all. For this reason, in the next subsections I will analyze ḥāl and ḥagg separately from māl and ḥaqq. 4.1 The markers ḥāl and ḥagg The particle ḥagg does never occur in my data, with one exception: one occurrence appears in the speech of a character from a sit-com whom my informants immediately identified as “a Bedouin from the Batinah” (Oman’s north-eastern coast). As we have seen, both Johnstone (1976) and Brustad (2000) remarked how ḥagg is employed in the dialects of the Gulf coast as a preposition meaning ‘to’ or ‘for’. The dialects of the Batinah bear strong similarities with those of the northern Gulf due prolonged contact, and the presence of ḥagg in this area is therefore unsurprising. Note how, in (5), ḥagg appears to mark dative case rather than genitive12: ḏā agall rāṭəb ḥagg əl-marrixiyāt DEM minimum wage PREP ART=martian.PL.F ‘This is the minimum wage for the martians’ (5) As opposed to ḥagg, the particle ḥāl appears to be fairly common in my texts. Interestingly, it is never used in an NP-internal construction. More often than not, it is employed to introduce verbless copula complements in verbless copula clauses, as in examples from (6) to (8), or indirect objects of finite verbs or active participles (examples 9 to 11). It appears to always express a relation of benefaction, where the beneficiary is the complement it introduces: 11 12 These voice notes were not addressed at me, but had been sent or received by some of my informants, who have later been kind enough to share them with me. This is why, in this and the following examples, both ḥāl and ḥagg are glossed simply as PREP(osition) rather than GEN. 227 SIMONE BETTEGA (6) əl-waẓīfa lā ḥāl-ak u-lā ḥāl-uh ART=job NEG PREP=PRON.2SG.M CONJ=NEG GEN=PRON.3SG.M ‘Neither you nor him will get the job’ (lit. ‘the job [is] not for you and not for him’) (7) intu mətakkidīn əs-siyyāra PRON.2PL.M be certain.AP-PL.M ART=car ‘Are you sure [that] the car [is] for me?’ ḥāl-ya PREP=PRON.1SG (8) əl-ʕaṣabiyya mā hī ART=anger NEG PRON.3SG.F ‘The anger [is] not good for you’ ḥāl-ak PREP=PRON.2SG.M (9) ʕēb tgūl ḥāl-brēk shame PRES.2SG.M-say PREP=brēk ‘Shame [on you, that] you say these things to Brēk!’ zēn-a good-SG.F ha-l-ḥarya DEM=ART=speech (10) ḥəlba ḥal ṯalāṯ ayyām ḥəlba PREP three day.PL ‘ḥəlba13 for three days’ (i.e. an amount of ḥəlba that will last for three days) (11) bāġī a-ṣobb ḥal-ī yaʕnī want.AP PRES.1SG-pour PREP=PRON.1SG INTERJ ‘I want to pour myself, I mean, a cup’ ṣaḥla bowl Only one example of non-predicative use of ḥāl appears in my data. This is shown in (12): (12) ǧihāz ḥāl sīdī device PREP CD ‘CD-reader’ (lit. ‘a device for the CDs’) Examples such as this one are probably the reason why this particle has been repeatedly reported to be a genitive or possessive marker in OA. Even in this last example, however, it is clear that ḥāl expresses dative case (again expressing a relation of benefaction: the device is conceived for CD-reading), and not genitive. As already said, genitive and dative are to be kept distinct, since they mark phrasal and clausal relations respectively. I maintain, in conclusion, that both ḥagg and ḥāl are not markers of NPinternal functions in OA, but rather of clausal relations, and – as such – should not be included in a discussion of genitive markers14. 13 14 A traditional sweet soup made of fenugreek (Trigonella foenum-graecum). Obviously, the distinction between genitive and dative is often a tenuous one. I cannot rule out, therefore, the possibility that genuine genitive relations are sometimes expressed by this marker. No such occurrence, however, was present in my material, and the only example of ḥāl Holes (2008: 484) provides is again of the kind exemplified in (12): ġašmar ḥāl ḥīwān ‘sorghum [feed] for farm animals’. 228 GENITIVE MARKERS IN OMANI ARABIC 4.2 The markers māl and ḥaqq As was the case for ḥāl, the particle māl appears to be relatively common in my corpus of texts. ḥaqq, on the contrary, only appears in the television material, and only in the speech of characters from Dhofar. This is consistent with all the existing literature on the subject: no study on Gulf Arabic or northern OA has ever reported the use of ḥaqq in these areas, while according to Davey (2012 and 2016), this marker is fairly common in southern Oman (though not as common as māl: again, my data would seem to confirm this). It is also interesting to note that, while ḥaqq does always inflect for gender and number in my material (4 occurrences out of 4), I have found no occurrence of inflected māl (be it in the television material or in the other texts, neither in the speech of the Dhofari characters nor in that of the other ones, or that of my informants). This contradicts both Davey’s Dhofari data and the accounts of the almost pan-peninsular māl discussed in § 3. One possible explanation for this fact is that the effects of the process of grammaticalization, which turned the noun māl into a purely syntactic linker, are becoming more evident with the passing of time, thus rendering the already infrequent inflected form less and less common (for further confirmation of this point see Rubin, 2004: 330). Apart from this discrepancy, Dhofari haqq appears to be used in the same syntactical contexts and with the same functions of māl. Again, this is consistent with Davey’s (2012: 69) findings 15. 4.2.1 Definiteness of the PR/PD and implicit PDs As far as the definiteness of both PD and PR is concerned, the vast majority of the occurrences of ḥaqq and māl which appear in my data follow a definite PD and precede a definite PR (examples 13 to 15). Both pronominal and non-pronominal PRs appear to be very common, though the former are almost always associated with definite PDs (13 and 14; note the occurrence of inflected ḥaqq in the latter): (13) ǧiddām əl-maqhā in front of ART=café ‘In front of his café’ māl-ah GEN=PRON.3SG.M (14) əṣ-ṣūra ḥaqqt-uh ART=picture-SG.F GEN-SG.F=PRON.3SG.M ‘Its picture’ (lit. ‘The picture of it’) (15) əl-awrāg māl əs-sandwikāt ART=paper.PL GEN ART=sandwich-PL.F ‘The wraps of the sanwiches’ 15 Admittedly, Davey hints at the possibility that, among older generations of Dhofari speakers, a distinction still exists between a more specific ḥaqq and a more general māl. However, he himself notes how «given the infrequency of the AGC within C[oastal] D[hofari] A[rabic], it is difficult to advance an analysis of this contrastive use of ḥaqq and māl further» (Davey 2012: 80) and that «the collection of a much larger corpus of data would be required to further the analysis of such features». 229 SIMONE BETTEGA Symmetrically indefinite PD/PR couplets appear as well, although more rarely. As can be clearly seen in examples (16) and (17), in these cases māl does not express possession as we have defined it in § 2, but rather a specification of quantity: (16) ṣaḥan māl tamar tray GEN date.COLL ‘A tray of dates’ (17) fī noṣṣ glāṣ aqūl-l-ak EXS half glass PRES.1SG-tell=PREP=PRON.2SG ‘There’s half a glass of tea, I’m telling you’ māl GEN šāy tea Even rarer are asymmetrical pairs. Indefinite PDs with a definite PR do sometimes occur, as in (18) and (19). Note that (inflected) ḥaqq in (18), as well, does not express possession, but rather a different kind of NP-internal specification: (18) mgall-āt ḥaqq-ōt əṭ-ṭabax magazine-PL.F GEN-PL ART=cooking ‘Magazines about cooking and make-up’ u-t-tagmīl CONJ=ART=make-up (19) qṣāym māl əṣ-ṣāḥəb coupon.PL GEN ART=owner ‘Coupons from (of) the owner’ It would seem that the definiteness/indefiniteness of the PR can have important repercussions on the semantics of the whole construction. When I asked my informants to disambiguate for me the meaning of the sentence qaṣʕa māl dxūn (example 32 below), they translated it for me as “a box of frankincense” (as in “a boxful of frankincense”), and contrasted it with qaṣʕa māl əd-dxūn, with definite PR, which on the contrary would be used to refer to a box for frankincense (that is, a box which is used to contain frankincense). Finally, neither Ingham nor Eksell-Harning admit the possibility of an indefinite PD with a definite PR. In my data, however, this combination occurs once (example 20). My informants maintain that this use is acceptable, and provided me with another (elicited) example (21). It would seem that, at least in the context of an NP expressing quantification, this construction is possible in OA: (20) l-maʕāš māl xamsa šuhūr ART=salary GEN five month.PL ‘Five month’s worth of salary’ (lit. ‘The salary of five months’) (21) ət-tamar māl ṯalāṯa nxīl ART=date.COLL GEN three palm.tree.PL ‘Three palm-tree’s worth of dates’ (lit. ‘The dates of/from three palm trees’) 230 GENITIVE MARKERS IN OMANI ARABIC One last structure that is worth discussing in this paragraph is not related to the definiteness/indefiniteness of the PD, but rather to its utter absence. The data show quite clearly that possessive constructions involving māl do sometimes feature “implicit” PDs, the PD being omitted when the speaker reckons that no ambiguity could possibly arise from this omission. (22) abā-k t-xabbər māl əl-amānāt PRES.1SG-want=PRON.2SG.M PRES.2SG.M-inform GEN ART=security-PL ‘I want you to inform the security’ (lit. ‘I want you to inform [the people or the office] of the security’) (23) āxrat-ha ṭalaʕt māl əl-muxaddirāt end-PRON.3SG.F turn out-PAST.2SG.M GEN drug-PL.F ‘In the end you turned out to be a drug-dealer’ (lit. ‘In the end of it you turned out to be [a dealer] of drugs’) Note that, in these examples, māl cannot in any way be interpreted as a preposition introducing an indirect object (also because both xabbər and ṭalaʕ in these sentences are transitive verbs). On the contrary, an implied maktab, ‘office’, and an implied tāgir, ‘dealer’, are to be assumed preceding māl in the first and second example respectively. This construction is probably to be regarded as a means of speeding up the flux of words in dialogical interaction, since the genitive marker would not be there should the omitted noun be restored in its place (but a SG would rather be used, i.e. tāgir əl-muxaddirāt, not *tāgir māl muxaddirāt). This construction appears to be quite common in OA, and my informants spontaneously provided more examples when I questioned them on the subject: (24) rūḥ t-šūf māl barīd go.IMP PRES.2SG.M-see GEN post ‘Go and check at the post office!’ (lit. ‘Go see at [the office] of the post’) Compare, also, the following two examples (25 has been elicited), where the PD is formally mentioned without, however, being specified: (25) ǧīb haḏāk māl ḥammām bring.IMP DEM.DIST GEN bathroom ‘Make the plumber come!’ (lit. ‘Bring that of the bathroom’) (26) hāḏa māl maḥaṭṭat əl-betrōl DEM GEN station ART=gas ‘The guy from the gas station’ (lit. ‘That of the gas station’) 231 SIMONE BETTEGA 4.2.2 Functions and meaning of the markers In general terms, māl and ḥaqq can be said to be most commonly used to express possession as defined in § 2. This is valid for around two thirds of the occurrences. Concrete possession is by far the most common type of possession expressed by the markers (examples 27 and 28), followed by abstract possession (29 and 14 above), association (30 and 31) and wholepart relationship (15 above)16: (27) baġā yištarī want.PAST.3SG PRES.3SG.M-buy ‘He wants to buy your car’ s-siyyāra ART=car māl-ak GEN=PRON.2SG.M (28) l-āy bād ART=i-pad ‘My i-pad’ (29) iḏa ʕand-ak ən-nəmra ḥaqqt-uh COND PREP=PRON.2SG.M ART=number GEN=PRON.3SG.M əttṣal fi-h call.IMP PREP=PRON.3SG.M ‘If you have his number, call him!’ (30) ənta l-arbāb17 PRON.2SG.M ART=employer ‘You are his employer’ (31) əl-musāʕidīn ART=helper-PL.M ‘My helpers’ 16 17 māl-ī GEN=PRON.1SG māl-uh GEN=PRON.3SG.M māl-ī GEN=PRON.1SG This neatly matches Eksell-Harning’s (1980: 75-6) and Davey’s (2012:74-7) lists of the different semantic relationships that the AG can express, except for the fact that both these authors include human relationships among them. In my data, this specific semantic field is scarcely represented, and it never involves kinship relationships. In the rare cases where māl is used to denote the relationship which exists between two human beings, this is always hierarchical in nature (examples (30) and (31), see also Brustad 2000: 80), and is better described as association. Note also that both Eksell-Harning and Davey include in their lists items which cannot be considered examples of proper possession. Brustad and Davey group these under the rather vague heading of “classification” and “qualification”, respectively, without however discriminating clearly between what can be considered linguistic possession and what not. arbāb is the (Arabic) word that Asian workers in the Gulf commonly employ to refer to their employer. Although the example in (30) comes from a conversation between two native speakers of OA, an influence of Gulf Pidgin Arabic in this context cannot be excluded (māl being extensively used in GPA to mark possession, see Naess 2008: 61). 232 GENITIVE MARKERS IN OMANI ARABIC As far as examples of māl and ḥaqq not expressing possession are concerned, these normally involve other kinds of NP-internal qualification, such as quantity (32)18, material (33) or, less specifically, some category the preceding PD belongs to (see example 34, but also 18, 23 and 25 above). (32) qaṣʕa māl dxūn jar GEN frankincense ‘A jar of frankincense’ (33) ṯlāṯ ṣḥān three dish.PL ‘Three clay dishes’ māl GEN fuxār clay (34) barnāmaǧ māl ḥawādiṯ program GEN accident.PL ‘A TV-show about car accidents’ Very few examples of māl not expressing possession nor any kind of NP-internal specification appear in my corpus. These are also the only examples of māl being used outside of the boundaries of a NP19. In the next two examples we see māl being used to introduce a verbless copula complement: (35) əntaw mā māl əl-musābaqāt PRON.2PL.M NEG GEN ART=competition-PL.F ‘You’re not [made] for the competitions’ (36) ḥəlwa bas mā māl rəḥlāt beautiful CONJ NEG GEN trip-PL.F ‘Nice [car], but it’s not [made] for the long trips’ As can be seen, in (35) and (36) māl does not express possession, but rather a relation of benefaction (which can also be interpreted as some type of partitive: an X which is made for Y, thus distinguishing it from other Xs of the same kind that, however, are not suitable for that specific purpose/activity). It seems possible, then, that also māl is, to an extent, used to express the dative case, as ḥāl and ḥagg are. The difference is that, while in the case of the latter this appears to be the only use of the particle, in the case of māl only a minority of examples of it carrying such a value were found (and all non-NP-internal). One last remark is due on the categories of alienable and inalienable possession. All instances of ḥaqq and māl which occur in my data are examples of the former. In particular, in my material kinship relationships and ownership over a specific part of the body are always expressed through an SG. It has to be noted that, although both Davey and EksellHarning have reported the use of genitive markers in association with body parts, their 18 19 On example (32), see also the discussion in § 4.2.1 above. No such use of ḥaqq occurs in my data. Davey (2012 and 2016), however, provides several examples of it. 233 SIMONE BETTEGA examples are not entirely convincing: Davey (2012: 75) specifies that a marker can be used in such a context only with reference to a limb which has been detached from the body, while Eksell-Harning (1980: 76) provides two examples of a dialect from Hadramawt, the first of which refers to the hump of a bull (a body part, then, but not of a human body), and the second to a portion of a body part (the toe of a foot). Both Davey and Eksell-Harning, on the other hand, concord on the fact that in the Dhofari and Yemeni dialects genitive markers can be used to refer to a parent-child relationship (when the reference is to an absent third party). I found no such example in my data, and my informants almost categorically rejected the possibility of such a use, with three possible exceptions: a) if the intended use is openly depreciative; b) in reference to a newborn or an infant; c) in babytalk, to convey intimacy and affection (this last point being obviously connected to the previous one). In light of all this, while I agree with Brustad’s and Eksell-Harning’s view that the specific will of expressing alienable possession is not what prompts a speaker to use an AG, it seems to me that the expression of true inalienable possession is nonetheless precluded to AGs, at least as far as northern OA is concerned. 4.2.3 Motivatations for the use of the AG Motivations for the use of the AG in OA can be formal or pragmatic. Several authors20 have highlighted how an AG is often employed in place of a SG in order to avoid excessively cumbersome or possibly ambiguous constructions: this is often the case when more than three nouns are involved in the construct, or when the PR and/or PD are either accompanied by a modifier (example 33 above) or constituted by a chain of conjoined elements (examples 18 and 26 and above). Overall, however, these syntactically heavy constructions are often avoided in the spoken language, and it seems to me that formal reasons such as the ones just listed can rarely be held responsible for the presence of an AG in my texts21. A different formal factor appears to exert a stronger influence on the use of genitive markers. Brustad (2000: 74) has noted how, when the PD is either a loanword or a word ending in a long vowel, then the AG tends to be preferred over the SG. This is probably because such elements «in general cannot take pronoun suffixes and do not readily fit into Arabic morphosyntactic patterns». Eksell-Harning (1980: 70) seems to agree with this. Examples of this phenomenon abound in my material: see for instance examples (37) to (39), and also (28) and (29) above: (37) əl-kāfītīryā māl mwassasa ART=cafè GEN company ‘The company’s cafeteria’ (38) əl-maqhā ART=café ‘His café’ 20 21 māl-ah GEN=PRON.3SG.M Eksell-Harning (1980: 78-9), Ingham (1994: 58), Brustad (2000: 74). Davey (2012: 71-3) as well noted how these complex constructions appeared rarely in his Dhofari data. 234 GENITIVE MARKERS IN OMANI ARABIC (39) kaʔan-nī sūpērmān CONJ=PRON.1SG superman ‘[It is] like [I am] her superman’ māl-ha GEN=PRON.3SG.F Pragmatic factors also seem to play an important role in determining whether or not, in a given context, a speaker will opt for using an AG. In particular, a specific DEMONSTRATIVE + PD + GEN + PR structure appears to exists in OA which closely mirrors the English “DEM + PD + of + PR” (e.g. “This house of yours”, “that dog of his”, etc.). As its English counterpart, this expression conveys a sense of mild contempt or sarcasm on the part of the speaker, mostly directed at the PD. Consider for instance the following examples, both drawn from an Omani sit-com. In (40), a husband is angry at his wife because she believes all the fictitious news she reads on her BlackBerry. In (41), a man has been injured by a ram he has been raising, which proved to be aggressive and dangerous on more than an occasion: his friend, then, urges him to get rid of the animal by selling it on Facebook. (40) kəll hāḏa min fōq ha-l-bībī all DEM PREP over DEM=ART=bb ‘All of that [comes] from that BlackBerry of yours!’ (41) māl-iš GEN=PRON.2SG.F tḥuṭṭ iʕlān ʕan hāḏa t-tēs PRES.2SG.M-put advertisement PREP DEM ART=ram māl-ak fi-l-fēsbūk GEN=PRON.2SG.M PREP=ART=facebook ‘Put an advertisement for that ram of yours on Facebook!’ As we have seen, among the reasons which might prompt speakers to employ an AG construction Brustad and Eksell-Harning note textual prominence and contrastive focus. This as well might be the reason that lies behind the presence of the marker in certain sentences that appear in my data. Consider yet another example drawn from the same TV show: here, an employee is puzzled by the fact that, after all the phone numbers in the company where he works have been re-assigned to the various offices and sub-sections, the director of the company ended up with the number which was once that of the cafeteria. Thus, the man exclaims: «I can understand that they changed the numbers, but they gave the director the number of the cafeteria!?». Here, the contrastive focus that the speaker wants to put on the last word (cafeteria) is evident, and the use of the AG appears clearly motivated: (42) yaʕṭū r-rāys raqm 22 māl PRES.3-give-PL.M ART=director number GEN ‘They gave the director the number of the cafeteria!?’ 22 l-kāfītīryā cafeteria Note that the word raqm in this sentence is not preceded by a definite article, though in theory it should be. This is not an isolated example in my data, which include a number of AGs whose definite PDs are however formally indefinite. Due to space constraints it is not possible to discuss the phenomenon here, though it would clearly deserve further investigation. 235 SIMONE BETTEGA Before moving on to the concluding paragraph of this paper, one last reason which can prompt the use of genitive markers in OA has to be mentioned. The particle māl is commonly used by speakers whenever the need to ascertain the ownership of an object arises. In such a situation, a construction is employed which involves only a deictic element followed by the marker plus a suffix pronoun, such as hāḏa māl-ak?, ‘Is this yours?’ (lit. “[is] this GEN your”; note that possessive pronouns such as “yours” or “ours” do not exist in OA). This type of structure is of extremely common use23, although it is highly unlikely that it is ever recorded in the course of an interview (purely deictic reference being rare outside the context of dialogic interaction). 5. Concluding remarks Summing up all that has been said in the preceding paragraphs, a number of conclusions emerge from the data presented. Firstly, māl appears to be the real genitive marker in OA (along with the competing form ḥaqq in southern Oman). ḥagg and ḥāl, on the contrary, are to be regarded as markers of dative case. While ḥāl appears to be specifically Omani, māl and ḥagg/ḥaqq seem to have almost pan-Peninsular diffusion. In my data, Dhofari ḥaqq does always show gender and number agreement with its PD, while māl never does (despite the fact that several authors admit this possibility: this is probably the consequence of an ongoing process of grammaticalization). Dative ḥagg, finally, can only appear in a single, invariable form. Why is it that the same (or etymologically related) particles appear, in different parts of the Peninsula, with markedly different morphological and syntactical properties, is an interesting question which awaits further research, and which might help to shed light on the evolution and diffusion of the Arabic dialects of Arabia. As we have seen, though AGs are commonly employed in OA, they remain overall much less common than SGs. Several reasons can prompt a speaker to use a genitive marker rather than a synthetic construct. Some are formal in nature, and span from the avoidance of syntactically cumbersome structures to the systematic association of genitive markers with PDs consisting of a loanword or a word ending in an etymological long vowel. Other reasons can be connected to pragmatic factors, such as the will to place a special focus on the PR, or to express a non-neutral (negative) attitude towards the PD. In general, the markers can appear in a wide varieties of syntactic contexts. They are more commonly employed when both the PD and PR are definite, but all combinations of definite/indefinite PDs and PRs are actually possible. With a definite PD, the markers normally express actual possession, while indefinite PDs are normally connected to other NP-internal specifications such as quantity, material, or qualification. An interesting structure which has not been described in previous studies on the subject is the use of a 23 Cf. Johnstone (1967: 90) for Kuwaiti: “māl is used […] in a genitival complex where the thing owned is not explicitly mentioned, having been already specified or understood», and Johnstone (1967: 106) about Bahraini: «māl is not frequent in occurrence in comparison with a construct phrase except where the object owned is not specified, as hāḏa māl-ī, ‘this is mine’»” (transcription adapted). 236 GENITIVE MARKERS IN OMANI ARABIC genitive marker with an implicit (omitted) PD. This omission is only possible when the speaker presumes that the reference will be unambiguously understood by the hearer. References Belnap, K. 1991. Grammatical Agreement Variation in Cairene Arabic. PhD thesis. Philadelphia: University of Pennsylvania. Brustad, K. 2000. The Syntax of Spoken Arabic: a Comparative Study of Moroccan, Egyptian, Syrian, and Kuwaiti dialects. Washington, D.C.: Georgetown University Press. Davey, R. 2012. “Analytic Genitive Constructions in Coastal Dhofārī Arabic”, Eades, Domenyk (ed.), Grammaticalization in Semitic. Oxford: Oxford University Press. 67-81. Davey, R. 2016. Coastal Dhofārī Arabic: a sketch grammar. Leiden: Brill. Dixon, R. 2010. Basic Linguistic Theory, Vol. II. Grammatical Topics. Oxford: Oxford University Press. Eksell, K. 2009. “D/L-Particles in Arabic Dialects: A Problem Revisited”, Retsö, Jan, & Watson, Janet, Relative Clauses and Genitive Constructions in Semitic. Oxford: Oxford University Press. 35-50. Eksell-Harning, K. 1980. The Analitic Genitive in Modern Arabic Dialects. Gothenburg: Acta Universitatis. Holes, C. 1989. “Towards a Dilaect Geography of Oman”, in Bulletin of the School of Oriental and African Studies LII. 446-62. Holes, C. 1990. Gulf Arabic. London & New York: Routledge. Holes, C. 2008. “Omani Arabic”, Versteegh, Kees, et als. (eds.), Encyclopedia of Arabic Language and Linguistics, Vol. III. Leiden: Brill. 478-91. Ingham, B. 1994. Najdi Arabic: Central Arabian. Amsterdam & Philadelphia: John Benjamins Publishing Company. Johnstone, T. 1967. Eastern Arabian Dialect Studies. Oxford: Oxford University Press. Johnstone, T. 1982. “Review of ‘The Analytic Genitive in the Modern Arabic Dialects’ by Kerstin Eksell Harning”, in Bulletin of the School of Oriental and African Studies XLV. 584. Naess, G. 2008. “Gulf Pidgin Arabic”: Individual Strategies or a Structured Variety? A study of some Features of the Linguistic Behaviour of Asian Migrants in the Gulf Countries. MA thesis. Oslo: University of Oslo. Reinhardt, M. 1894. Ein arabischer Dialekt gesprochen in ‘Omān und Zanzibar. Stuttgart & Berlin: Spemann. Rubin, A. 2004. “Notes on the Genitive Exponents of some Modern Arabic Dialects”, Folia Orientalia XL. 327-336. 237