Patrones de concordancia de los nombres nās y ʕālam en el árabe de Damasco

MEAH, SECCIÓN ÁRABE-ISLAM [1696-5868] 70 (2021), 3-31. DOI 10.30827/meaharabe.v70i0.15152 AGREEMENT PATTERNS WITH THE NOUNS NĀS AND ʕĀLAM IN DAMASCUS ARABIC Patrones de concordancia de los nombres nās y ʕālam en el árabe de Damasco Carmen BERLINCHES RAMOS [email protected] Universidad de Zaragoza Abstract: This paper examines the agreement patterns with the controllers nās and ʕālam in Damascus Arabic, a question not yet thoroughly studied. The collective nature of these nouns allows two kinds of agreement: strict agreement in the plural and deflected agreement in the feminine singular. The analysis focuses on the variation between strict and deflected agreement, especially as regards semantic and pragmatic factors, and also the morphological structure of the targets. The study is synchronic and based on data collected in the capital of Syria, Damascus. It is divided into two main sections: the first analyzes target-related factors, including target type, word order, and distance; the second focuses on controller-related factors, including quantification, qualification, reflexibility, specificity, and definiteness. In recent years numerous studies on agreement have been conducted and they indicate parallels among different varieties of Arabic, but such studies mainly address non-human controllers. The results of the present study show a prevalence of strict agreement with the nouns nās and ʕālam. Target type is a key factor in determining the kind of agreement, with distance between controller and target also being important. In line with previous studies, variation in agreement is closely connected to the controller’s degree of individualization or collectivity as perceived by the speaker, who ultimately chooses which kind of agreement to use. Resumen: Este artículo examina los patrones de concordancia de los nombres nās y ʕālam en árabe de Damasco, una cuestión aún no estudiada en profundidad. El sentido colectivo de ambos términos controladores permite la concordancia en femenino singular o en masculino plural, por lo que se analiza esta variación atendiendo a factores semánticos y pragmáticos, así como a la estructura morfológica de los términos controlados. La perspectiva del estudio es sincrónica y está basada en un corpus de textos recogidos en Damasco. Se divide en dos partes: (1) Factores relacionados con el término controlado (tipología; posición en la oración; distancia con respecto al término controlador). (2) Factores relacionados con el término controlador (cuantificación; cualificación; reflexividad; especificación; determinación). Asimismo, los estudios sobre la concordancia, particularmente prolíficos en los últimos años, han permitido trazar paralelismos con otras variedades árabes. Los resultados muestran una mayor frecuencia de concordancia en plural. Además, la tipología del término controlador es un factor relevante en la concordancia, así como una cierta distancia entre controlador y controlado. Por último, y de acuerdo con estudios anteriores, la variación en la concordancia está estrechamente relacionada con el grado de individualidad o colectividad del controlador percibido por el hablante quién, en última instancia elige el tipo de concordancia. Key words: Arabic dialectology. Agreement. Nās and ʕālam. Damascus Arabic. CARMEN BERLINCHES RAMOS MEAH, SECCIÓN ÁRABE-ISLAM [1696-5868] 70 (2021), 3-31. DOI 10.30827/meaharabe.v70i0.15152 4 Palabras clave: Dialectología árabe. Concordancia. Nās y ʕālam. Árabe de Damasco. Recibido: 10/04/2020 Aceptado: 10/07/2020 1. INTRODUCTION 1 The term agreement commonly refers to some systematic covariance between a semantic or formal property of one element and a formal property of another 2 . Among these two elements, the controller determines the agreement (say the subject noun phrase) and the target is the element whose form is determined by agreement 3 . In Arabic, the agreement of the target with its controller may be of two types: strict, if some category that is overtly inherently present in the controller (subject or head-noun) and copied in the target (verb, noun-modifier), and deflected, if a plural controller is associated with a feminine singular target 4 . Belnap’s study on Cairene Arabic 5 demonstrated that variation with human controllers is far more extensive in his interviews than the literature would generally lead one to suspect 6 . Although human nouns tend to favor plural agreement, the case of nās and ʕālam, both meaning ―people‖, must be considered apart, because being the most generic terms used to refer to human beings, they have an implicit collective sense 7 and therefore both kinds of 1 . Damascus Arabic phoneme inventory: 1. Consonants. 1.1. Occlusive: ʔ laryngeal, voiceless; b bilabial, voiced; ḅ bilabial, velarized, voiced; t dental, voiceless; d dental, voiced; ṭ dental, velarized, voiceless; ḍ dental, velarized, voiced; k postpalatal, voiceless; q uvular, voiceless; g postpalatal, voiceless. 1.2. Fricative: f labiodental, voiceless; v labiodental, voiced; x velar, voiceless; ġ velar, voiced; h laryngeal, voiceless; ḥ pharyngeal, voiceless; ʕ pharyngeal, voiced. 1.3. Sibilant: s alveolar, voiceless; z alveolar, voiced; ṣ alveolar, velarized, voiceless; ẓ alveolar, velarized, voiced; š prepalatal, voiceless; ž prepalatal, voiced. 1.4. Nasal: m bilabial, voiced; n dental, voiced. 1.5. Lateral: l alveolar, voiced; ḷ alveolar, velarized, voiced. 1.6. Trill: r alveolar, voiced. 2. Semivowels: w bilabial, voiced; y palatal, voiced. 3. Vowels. 3.1. Short vowels: a front, low; e front, mid; ə central, mid; i front, high; o back, mid; u back, high. 3.2. Long vowels: ā front, low; ē front, mid; ī front, high; ō back, mid; ū back, high. 2 . Steele. ―Word order‖, p. 610. 3 . Corbett. Agreement, p. 4. 4 . Ferguson. ―Grammatical agreement‖, p. 9. 5 . Belnap. Grammatical agreement. 6 . Similarly, concerning Damascus Arabic, Ambros (Damascus Arabic, p. 86) pointed out that ―even highly sophisticated rules (which have not been established so far) would fail to cover all instances of agreement with plural subjects as occurring in recorded texts‖. 7 . Cf. Belnap. Grammatical agreement, pp. 173-175, referring to nās in Cairene Arabic. In Modern Standard Arabic (MSA) nās is defined as one of the two plurals of ʔinsān, the other being ʔunās, ―people‖ (Corriente & Ferrando. Diccionario Avanzado, p. 34; Wehr. Arabisches Wörterbuch, p. 48; Wright, A grammar, p. 233). It is, however, marked as collective in Wehr (A Dictionary, p. 30). Regarding ʕālam, Corriente & Ferrando (Diccionario Avanzado, p. 799) take it to mean merely ―world, universe‖, whereas Wehr (Arabisches Wörterbuch, p. 871) relates it generally to ―people (Weltbewohner, Menschen)‖ when it takes the plural ʕalamūn. Barthélemy’s dictionary (Dictionnaire arabe-français), which focused on Levantine, and particularly Aleppine, Arabic, places the term nās under two roots: √ʔns ―man with the general meaning to be human (homme dans le sense général, être humain)‖ (p. 17); and √nws ―men, people (hommes, gens)‖ (p. 855). In both cases nās is a plural. By contrast, ʕālam appears defined by the article (lʕālam), meaning ―the world, the people (le monde, les AGREEMENT PATTERNS WITH THE NOUNS NĀS AND ʕĀLAM IN DAMASCUS ARABIC MEAH, SECCIÓN ÁRABE-ISLAM [1696-5868] 70 (2021), 3-31. DOI 10.30827/meaharabe.v70i0.15152 5 agreement are possible. Besides, recent studies have demonstrated that the degree of individuation or collectivity perceived by the speaker is crucial for determining the kind of agreement, thus linking agreement patterns more to semantic and pragmatic factors than to morphology. The major purpose of this paper is to investigate the agreement patterns of the nouns nās and ʕālam in Damascus Arabic 8 . It will focus on the variation between ―strict agreement‖ (plural — plural) and ―deflected agreement‖ (plural — feminine singular), considering morphological, semantic, and pragmatic factors. The cases in which each kind of agreement tends to occur, and factors which may favor one agreement or the other, will be examined. Results are based on a statistical analysis of the occurrence of nās and ʕālam with the following kinds of targets: verbs, anaphoric pronouns, adjectives, active and passive participles, and personal and demonstrative pronouns. This study is conducted from a synchronic perspective and based on a corpus of texts gathered in Damascus between 2007 and 2008 9 . It consists of about eight hours of interviews (of approximately 63,350 words) produced by fifty informants. Many of the texts describe different places in Damascus, traditions of the society, and personal experiences of the speakers: thus the time used is the present. Other texts, fewer in number, narrate general or personal events in the past. The texts do not contain any conversations. Variation in agreement has been the subject of research in different Arabic varieties, with a noticeable increase in publications during the last years (see the list of references below 10 ). Melanie Hanitsch 11 has studied, particularly in Damascus Arabic, the variation in agreement of non-human plural nouns controlling attributive adjectives 12 . To the best of my knowledge, there is no other piece of research that so far has focused on the subject of this study, except for a few general references available in the main grammars of Damascus Arabic. Among them, Cowell 13 showed the existence of both kinds of agreement with plural and collective nouns, and pointed out individuality vs. collectivity as a main factor in determining agreement. He gave several examples with nās and gens)‖ (p. 547). Thus nās is mainly regarded as plural (or collective), while ʕālam is a formally (masculine) singular which primarily means ―world‖ and may also denote ―(a group of) people‖. 8 . The main linguistic particularities of Damascus Arabic are rigorously exposed in Lentin. Damascus Arabic. Nevertheless, the most relevant features appearing in our examples will be pointed out. 9 . This corpus was gathered for a Ph.D study about the language of the youth in Damascus. Later, it was revised and published as Berlinches. El dialecto. 10 . I am deeply thankful to Simone Bettega for providing me with several of these sources, for reading an early draft of this paper, and for his helpful thoughts and comments. 11 . Hanitsch. ―Kongruenzvariation‖. 12 . Unfortunately, this paper lacks quantitative data. 13 . Cowell. A reference grammar, pp. 423-426. CARMEN BERLINCHES RAMOS MEAH, SECCIÓN ÁRABE-ISLAM [1696-5868] 70 (2021), 3-31. DOI 10.30827/meaharabe.v70i0.15152 6 ʕālam, but unfortunately did not state quantitative results. Ambros 14 briefly referred to the variation in agreement with plural subjects, including two examples showing both kinds of agreement with nās. Ferguson & Ani 15 gave an example in which the verb shows deflected agreement with nās; but otherwise did not treat agreement patterns with nās and ʕālam in any detail 16 . Grotzfeld 17 provided only general information about agreement patterns in Damascus, regarding nās as a plural. Furthermore, some studies on other Arabic varieties include specific references to agreement patterns with the generic term for denoting ―people‖, which usually is nās. Among them are D’Anna (Fezzānī Arabic, southwestern Libya) 18 , Procházka & Gabsi (Urban Tunisian, particularly the variety of the capital) 19 , Ritt-Benmimoun (Bedouin Tunisian, Nifzāwa region) 20 , Feghali (Lebanese Arabic) 21 , Brustad (including data from Moroccan, Egyptian, Syrian, and Kuwaiti Arabic) 22 , Belnap (Cairene Arabic) 23 , and Holes (Bahraini Arabic) 24 . Some remarks about nās in Nadji Arabic are found in Bettega 25 , and for the dialect of the city of Salt (Jordan) in Herin & Al-Wer 26 . 1.1. Individualization vs. collectivity Brustad’s study on agreement in some varieties of Arabic 27 has demonstrated that the lexical items do not determine the agreement patterns but the speaker who chooses it according to semantic or pragmatic factors. She explains that ―if a noun is highly individuated, animate, specific, textually prominent, or quantified, the speaker tends to choose plural agreement 28 ; conversely, if the noun is non14 . Ambros. Damascus Arabic, pp. 72 y 86. 15 . Ferguson & Ani. Damascus Arabic, p. 238. 16 . Ferguson (―Grammatical agreement‖, p. 12) also provides some information about relating nās to deflected agreement. 17 . Grotzfeld. Syrisch-arabische, pp. 97-98 (fn. 1). 18 . D’Anna. ―Agreement‖, pp. 107-110. 19 . Procházka & Gabsi. ―Agreement‖, pp. 244-247 and 253. 20 . Ritt-Benmimoun. ―Agreement‖, pp. 268-272. 21 . Feghali. Syntaxe des parlers, pp. 124-126 and 143-144. 22 . Brustad. The syntax, pp. 53-61. 23 . Belnap. Grammatical agreement, pp. 61-67, 77-78 and 88-89 (this PhD thesis includes numerous references to the noun nās); Belnap. ―The meaning·, pp. 100-102 and 108-109; Belnap. ―A new perspective‖, pp. 173-175. 24 . Holes. Dialect, culture, pp. 326-341. 25 . Bettega. ―Rethinking agreement‖, pp. 136-137. 26 . Herin & Al-Wer. ―From phonological variation‖, p. 67. 27 . Brustad. The syntax, p. 54. 28 . Khan (―Object markers‖, p. 470) adds that ―[a] nominal which refers to a specific entity is more individuated than a generic nominal, which refers to a class of entities.‖ He defines ―individuation‖ as AGREEMENT PATTERNS WITH THE NOUNS NĀS AND ʕĀLAM IN DAMASCUS ARABIC MEAH, SECCIÓN ÁRABE-ISLAM [1696-5868] 70 (2021), 3-31. DOI 10.30827/meaharabe.v70i0.15152 7 specific, collective, and less prominent, the agreement will tend to be feminine singular‖ 29 . Most of the studies on agreement in different varieties of Arabic corroborate this assertion, pointing out that the degree of individuation perceived by the speaker is a key factor in the choice of agreement. The present paper is structured closely on Bettega 30 , with the necessary modifications for treating the controllers’ particularities. It is divided into targetrelated and controller-related factors. The first part analyzes the target type, the distance between controller and target(s), and the word order. The second part treats quantification, qualification, reflexibility, specificity, and definiteness. Our corpus includes 236 instances of nās and ʕālam 31 (nās: 120, ʕālam: 116) controlling 410 targets (266 verbs, 60 anaphoric pronouns, 28 adjectives, 27 active participles, 15 personal pronouns, 9 passive participles, and 7 demonstrative pronouns). Agreement occurs either in masculine plural (i.e. strict) or in third person feminine singular (i.e. deflected) 32 . ANALYSIS 2. TARGET-RELATED FACTORS 2.1. Target type Target type has been stated to be the key factor for determining the kind of agreement 33 . The number of occurrences of each kind of target and the agreement type with the controllers nās and ʕālam provide us with a clear idea of agreement tendencies. Results are given in the following table and analyzed below. Kind of target34 Deflected agreement Strict agreement TOTAL Verbs 104 (39.09%) 162 (60.90%) 266 Anaphoric pronouns 26 (43.33%) 34 (56.66%) 60 ―the distinctness or salience of the nominal from its own background and, with regard to a verb complement nominal, also its distinctness from the subject‖ (p. 470). 29 . Brustad. The syntax, p. 59. Cowell’s grammar (A reference grammar, p. 423) concluded: ―Most inanimate plurals, and some animate plurals and collective, have feminine agreement in the predicate when collectivity or generality is emphasized rather than heterogeneity or particularity‖. 30 . Bettega. ―Agreement with plural‖. 31 . Instances of ʕālam showing masculine singular agreement occur only when ʕālam denotes ―world‖ and not just ―inhabitants of the world (i.e. people)‖: for instance, (1) la-ḥatta nkūn ʕala ṣila maʕ ǝl-ʕālam ǝl-ġarbi ―for us to be in contact with Western world‖ (ġarbi: ADJ.MSG). In the examples in this paper, ʕālam will refer to the generic ―people‖ unless indicated otherwise. 32 . Other types of agreement, like the masculine singular agreement very commonly found in the Jordanian city of Salt (cf. Herin & Al-Wer. ―From phonological variation‖, p. 67), or the feminine plural agreement found in Nifzāwa, southern Tunisia (Ritt-Benminoun. ―Agreement‖) and southwestern Libya (D’Anna. ―Agreement‖, p. 107), do not occur in our data. 33 . Bettega. ―Agreement with plural‖, p. 173. 34 . The kinds of targets are listed in order of number of appearances. CARMEN BERLINCHES RAMOS MEAH, SECCIÓN ÁRABE-ISLAM [1696-5868] 70 (2021), 3-31. DOI 10.30827/meaharabe.v70i0.15152 8 Adjectives 11 (39.28%) 17 (60.71%) 28 Active participles 7 (25.92%) 20 (74.07%) 27 Personal pronouns 0 (0.00%) 13 (100%) 13 Passive participles 4 (44.44%) 5 (55.55%) 9 Demonstrative pronouns 1 (14.28%) 6 (85.71%) 7 2.1.1. Verbs Verbs are the most-frequently used targets in the data, with a total of 266 instances. Strict agreement prevails with a percentage of 60.90% versus 39.09% for deflected agreement. nās ʕālam Total Deflected agreement 64 40 104 Strict agreement 88 74 162 Example of deflected agreement: (2) fi ʕālam bǝtfaḍḍǝla, hiyye ktīr ṭayybe ―there are people who prefer it, it is very tasty‖ (bǝtfaḍḍǝl-: VB.3FSG) Example of strict agreement: (3) fi nās byǝʕǝmlu laban b-ǝxyār ―there are people who make yogurt with cucumber‖ (byǝʕǝmlu: VB.PL) According to Feghali 35 , in Lebanese Arabic the verb agrees indifferently in plural (strict agreement) or in feminine singular (deflected agreement) when controllers, including the nouns nās and ʕālam, refer to human collectives. But this is not the case for Damascus, where, at least in clauses where the controllers nās and ʕālam are involved, strict agreement is more likely to occur. In Bahrain, cases of verbs showing deflected agreement with the noun nās have been associated with descriptions of general actions rather than to what happened at a specific point in the past 36 . Our data suggest the same, though the number of examples of this kind is limited—perhaps because most of our texts are descriptive, focused on different aspects of the daily life in Damascus as well as on the customs and traditions of Damascene society. Therefore generic actions clearly prevail. Holes has related deflected agreement to a particularly common use of nās as an indefinite distributive (i.e. ―some others‖) 37 , which is also true for Damascus, as the following example shows: 35 . Feghali. Syntaxe des parlers, p.124. 36 . Holes. Dialect, culture, p. 331. 37 . Ídem, p. 333. AGREEMENT PATTERNS WITH THE NOUNS NĀS AND ʕĀLAM IN DAMASCUS ARABIC MEAH, SECCIÓN ÁRABE-ISLAM [1696-5868] 70 (2021), 3-31. DOI 10.30827/meaharabe.v70i0.15152 9 (4) fi ʕālam btǝʔʕod 38 ʕal-karāsi w ṭāwlāt, fi ʕālam ǝbtǝšrabon ʕal-wāʔef ―there are people who sit on the chairs and tables, and there are people (=some others) who drink it standing up‖ (btǝʔʕod: VB.3FSG) 39 . 2.1.2. Adjectives Our data contain 28 instances of adjectives controlled by nās or ʕālam. In 60.71% of the cases agreement is strict and in 39.28% deflected 40 . nās ʕālam Total Deflected agreement 10 1 11 Strict agreement 12 5 17 Example of deflected agreement: (5) kānet ǝn-nās bən-nǝsbe ʔǝli ktīr ġarībe ―the people were very strange for me‖ (ġarībe: ADJ.FSG) Example of strict agreement: (6) əl-ʕālam yaʕni hēke wadūdīn ―the people are, I mean... like this, warmhearted‖ (wadūdīn: ADJ.PL) Strict agreement clearly prevails, representing almost two-thirds of the total. According to Brustad 41 , the reason for these results is the higher degree of individuation associated with modified nouns (see section 3.2.1.). Moreover, as can be seen in the next table, no significant differences in agreement type concerning the attributive or predicative nature of the adjectives are observed in the data 42 . nās ʕālam Attributive Predicative Attributive Predicative Deflected agreement 8 2 1 0 Strict agreement 9 3 1 4 TOTAL 17 5 2 4 38 . The verbal modifier for the general present is b-, which generally turns into m for the first person plural, cf. Cowell. A reference grammar, p. 180. Moreover, in Damascus q is realized as a glottal stop (ʔ), except in borrowings of Classical Arabic, cf. Lentin. Damascus Arabic, p. 546. 39 . In these cases the existential particle fi ―there is‖ usually occurs before the controllers. 40 . This example from Cowell (A reference grammar, p. 500), nās ġǝšǝm ―ignorant people‖, also shows strict agreement. 41 . Brustad. The syntax, p. 61. 42 . Only predicative adjectives show a higher number of cases of strict agreement with ʕālam (4 cases of deflected agreement against none of strict agreement). But we consider the number of instances insufficient for any conclusion. CARMEN BERLINCHES RAMOS MEAH, SECCIÓN ÁRABE-ISLAM [1696-5868] 70 (2021), 3-31. DOI 10.30827/meaharabe.v70i0.15152 10 Despite the prevalence of strict agreement, our results indicate that in Damascus Arabic cases of adjectives showing deflected agreement with the controllers nās and ʕālam occur with more frequency than in other Arabic varieties. For example, in Urban Tunisian, Procházka & Gabsi have detected only a few cases of adjectives triggering deflected agreement with nās and ʕbād, both meaning ―people‖ (e.g. nās bāhya ―good people‖) 43 . They consider these cases idiomatic collocations, thus not controlled by pragmatic discourse structures, and they affirm that deflected agreement of adjectives —with human controllers— is even more restricted than with verbs and pronouns. At the same time, Feghali, focusing on Lebanese Arabic, claimed that the agreement of the adjective with controllers referring to human collectives is always strict 44 . In Bahraini Arabic, strict agreement with adjectives clearly prevails 45 . However, Holes also recorded several cases of deflected agreement, usually when the reference was generic or collective. By contrast, Bettega’s findings in Omani Arabic are very interesting since they show a high percentage (94.3%) of attributive adjectives attracting deflected agreement 46 . These results are, however, based on non-human controllers and the author points out that his data about the noun nās is insufficient for statistical analysis 47 . Our results agree with Belnap’s 48 , whose hierarchy of deflected agreement particularly focused on nās places predicative adjectives on the top of the list 49 . According to Brustad 50 , verbs trigger deflected agreement more than adjectives, explained by the higher degree of individuation of modified nouns. This statement is nevertheless not focused on the controllers nās and ʕālam. Our data show exactly the same percentage of deflected agreement for both verbs and adjectives. Adjectives and their connection with the controller’s qualification are further analyzed in section 3.2.1. 43 . Procházka & Gabsi. ―Agreement‖, p. 247. 44 . Feghali. Syntaxe des parlers, p. 143. The author provides two examples of this kind. He attributes exceptions to this rule to the influence of the Classical language, or to the need to rhyme in popular songs (p. 144). These statements definitely do not agree with our findings, which show almost 40% of deflected agreement. We wonder how the situation could change so much in a period of less a century, especially given that Lebanese Arabic is a Levantine variety very close to Damascus Arabic. 45 . Holes. Dialect, culture, pp. 327-329. 46 . Bettega. ―Agreement with plural‖, p. 168. For predicative adjectives the percentage is 55.6%. 47 . Ídem, fn. 17. 48 . Belnap. Grammatical agreement, p. 88. 49 . Belnap (Grammatical agreement, pp. 66-67) moreover mentions that in Abdel-Massih et al. (A reference grammar, p. 22) all the speakers used deflected agreement with nās and ʕālam to some degree. 50 . Brustad. The syntax, p. 61. AGREEMENT PATTERNS WITH THE NOUNS NĀS AND ʕĀLAM IN DAMASCUS ARABIC MEAH, SECCIÓN ÁRABE-ISLAM [1696-5868] 70 (2021), 3-31. DOI 10.30827/meaharabe.v70i0.15152 11 2.1.3. Participles Our data include 36 participles controlled by nās and ʕālam. Among them, 27 are active and 9 passive. Due to their different grammatical nature and behavior, they will be treated separately. a) Active participles Active participles show a clear inclination for strict agreement with a very high percentage of occurrences (74.07%): nās ʕālam Total Deflected agreement 3 4 7 Strict agreement 13 7 20 Example of deflected agreement: (7) ǝn-nās rāyḥa ʕala šǝġla ―the people are going to their job‖ (rāyḥa: AP.FSG) Example of strict agreement: (8) fi nās mu šāyfīn ―there are people who do not see‖ (šāyfīn: AP.PL) The prevalence of strict agreement could be explained by its role as a verb form 51 . However, in 17 of our 20 examples of strict agreement two other features of specification or individuation may have had an impact on the results: the presence of the relative pronoun in 6 cases, and a distance greater than four words from the controller in 11 cases 52 . Concerning the small number of occurrences of deflected agreement (7 occurrences — 25.92%), one is clearly influenced by MSA, being in a religious context with some MSA loans, and three have terms emphasizing collectivity 53 . Therefore, only three cases of deflected agreement are not influenced by any special factors. The limited data about the agreement patterns of these participles with both human and non-human controllers in Syrian Arabic, as well as in other varieties of Arabic, do not allow us to determine whether or not our findings are a regular tendency in agreement. This is definitely a point for further research. b) Passive participles On the other hand, passive participles show a nearly equal proportion of both kinds of agreement, 55.55% strict and 44.44% deflected. Nevertheless, these 51 . In our data verbs show a relatively high percentage of strict agreement (60.90%: see section 2.1.1). 52 . These factors are further examined in sections 3.2.2 and 2.3, respectively. 53 . I.e. fi ktīr, baʕǝḍ, and kǝll, as discussed in sections 3.1.1-2 and 3.3. CARMEN BERLINCHES RAMOS MEAH, SECCIÓN ÁRABE-ISLAM [1696-5868] 70 (2021), 3-31. DOI 10.30827/meaharabe.v70i0.15152 18 The first target (bǝtšūf: VB.3FSG) is separated by one word from its controller and shows deflected agreement, while the two second targets (biḥāwlu: VB.3PL, yʔalldu: VB.3PL) are separated from their controller by nine and ten words, respectively, and show strict agreement. However, in other cases deflected agreement is maintained through the whole sentence despite distance: (16) bass ǝl-ʕālam ma bǝtḥǝss ḥāla mǝṭṭarra tǝḥmǝla ―but the people don’t feel obliged to carry it‖. In this case the consecutive sequence of the targets (bǝtḥǝss: VB.3FSG, -a: PRON.3FSG, mǝṭṭarra: PP.SG, tǝḥmǝl-: VB.3FSG) seems to result in deflected agreement for all of them. (17) fi nās bǝtkūn žāye ʔabl ǝl-kǝll ʕal-madīne 89 , zāyre r-rasūl u baʕdēn bǝtrūḥ ʕala makke ―there are people who first come to Medina, visit the Prophet, and afterwards go to Makkah‖. bǝtrūḥ (VB.3FSG) is separated by nine words from its controller, a distance which normally would have resulted in strict agreement. This example was possibly influenced by MSA because the topic of the conversation was the Pilgrimage and several features of MSA can be observed in it. (18) nnās bǝtkūn bibyūtha bakkīr w-btǝži ʕašīye mǝn ʔašġālha ―the people are early at home and come back from work in the evening‖ In this example, extracted from Bloch & Grotzfeld 90 , all the targets (bǝtkūn: VB.3FSG, -ha: PRON.3FSG, btǝži: VB.3FSG, -ha: PRON.3FSG) show deflected agreement. Finally, Procházka & Gabsi have found particularly common patterns of mixed agreement in which the first clause (V-S) shows deflected agreement and the second strict agreement (with the same controller) 91 . Unfortunately our data only contain the following example with similar structure, which might be due to the limited number of V-S clauses: (19) bass kānet mǝtǝl-ma ʔǝlt-ǝllek yǝžu n-nās kǝlla yǝžtǝmʕu ―but, as I told you, all the people came to meet‖ (kānet: VB.3FSG, yǝžu: VB.3PL, yǝžtǝmʕu: VB.3PL). bətḥəbbi ―you (F) like‖ > bətḥəbbī ―you (F) like him‖. Whereas when a word ends in a consonant, the pronoun is -o, for instance: bēt ―house‖ > bēto ―his house‖. 89 . The shift a > e in final position (madīna > madīne) is known as final ʔimāla and takes place after non-pharyngeal, laryngeal or velarized consonants (i.e. b, t, ž, d, z, s, š, f, k, l, m, n, w, y) and sometimes after r (cf. Cowell. A reference grammar, p. 138-139). 90 . Bloch & Grotzfeld. Damaszenisch-arabische, p. 80. Translation by the present author. 91 . Procházka & Gabsi. ―Agreement‖, p. 246. AGREEMENT PATTERNS WITH THE NOUNS NĀS AND ʕĀLAM IN DAMASCUS ARABIC MEAH, SECCIÓN ÁRABE-ISLAM [1696-5868] 70 (2021), 3-31. DOI 10.30827/meaharabe.v70i0.15152 19 Although our data focuses only on clauses involving the subjects nās and ʕālam, results suggest that this pattern is not very frequent in Damascus Arabic. On the other hand, cases of targets showing the shift from strict to deflected agreement with the same controller are not found in our data. Finally, our data include several instances of coordinated clauses, each with two different controllers and targets. They have similar grammatical structure but show different kinds of agreement. Apparently, there is no criteria for explaining it, as the following examples show: (20) mǝnḥǝbb 92 ǝl-ʕālam ǝt-tānye, fi ḥǝbb lal-musāʕade lan-nās ǝl-ʔāxarīn ―we like other people; there is love for helping (= we like helping) others‖ This example includes two coordinated sentences with parallel semantic construction —ǝl-ʕālam ǝt-tānye and n-nās ǝl-ʔāxarīn— but targets (tānye: ADJ.FSG, ʔāxarīn: ADJ.PL) showing different agreement. The reason for this is opaque. (21) fi masalan ʕālam byǝtbarraʕu b-maṣāri, masalan ʕālam ǝbtǝḥtāž ʕamaliyyāt fi ʕālam byǝtbarraʕu masalan, fi ḥatta ʕālam ǝb-dubay ǝbtǝbʕat ǝtyāb ―there is for example people who donate money; for example (if) people need an operation, there are people who donate, for example, even people in Dubai who send clothes‖ The four targets in this passage, in order of appearance, are: byǝtbarraʕu (VB.3PL), btǝḥtāž (VB.3FSG), byǝtbarraʕu (VB.3PL), and btǝbʕat (VB.3FSG). The four verbs appear after their four respective controllers, which is ʕālam in all four cases. The less agentivity of the verb btǝḥtāž may explain the deflected agreement, but based on this argument we are not able to explain the deflected agreement of the fourth verb. The variation must be due to the speaker’s criteria. 3. CONTROLLER-RELATED FACTORS 3.1. Quantification Previous studies have demonstrated that quantifiers play an important role in determining the type of agreement 93 . These sources particularly confirm that most of the numerical quantifiers strongly favor strict agreement; however, the uncountable nature of the controllers nās and ʕālam does not allow numerical quantification. Our data include the following quantifiers: ktīr ―many‖, kǝll ―all‖, ġālibiyye ―majority‖, ʔaktar ―majority‖, and baʔiyye ―remainder‖. 92 . m is the first person plural verbal modifier for the present, after the assimilation b > m (see fn. 38). 93 . Cf. for instance Belnap. Grammatical agreement. p. 68, and Bettega. ―Agreement with‖, p. 157. CARMEN BERLINCHES RAMOS MEAH, SECCIÓN ÁRABE-ISLAM [1696-5868] 70 (2021), 3-31. DOI 10.30827/meaharabe.v70i0.15152 20 3.1.1. Quantifier ktīr and the co-occurrence of fi ktīr The quantifier ktīr ―many, a lot‖ followed by an indefinite noun fosters deflected agreement because it denotes low individuation 94 . Our data include only one example of ktīr attracting deflected agreement but 7 examples attracting strict agreement, though 5 of the 7 cases of strict agreement show other features of specification, which perhaps explains the higher percentage of strict agreement. This particular question should be further studied by examining more occurrences of ktīr in sentences involving nās and ʕālam and different human and non-human controllers. Examples of deflected agreement: (22) ktīr nās ǝbtǝrʔoṣ dabke ―many people dance Dabka‖ (btǝrʔoṣ: VB.3FSG) Examples of strict agreement: (23) hiyye ʔǝrne mrattabe w ǝktīr ʕālam sāknīn fiyya ―it is a well-organized place and many people live there‖ (sāknīn: AP.PL) However, our data indicate that the co-occurrence of the existential particle fi ―there is/are‖ and the quantifier ktīr clearly promotes deflected agreement. In this construction ktīr may precede or follow the controller (e.g. fi ktīr nās/ʕālam or fi nās/ʕālam ǝktīr). Among our examples, the agreement is deflected in 8 cases and strict in 3, one of the latter involving features of specification. Although studies focused on other varieties of Arabic have indicated that fi is an emphasizer of generality and collectivity, as in Urban Tunisian 95 and southwestern Libya 96 , in Damascus Arabic it seems that this is true only when fi accompanies ktīr, at least in clauses involving the controllers nās and ʕālam 97 . Examples of deflected agreement: (24) fi ʕālam ǝktīr ǝbtǝži ʕala hāda l-ʔǝžtimāʕ ―there are a lot of people who come to this meeting‖ (btǝži: VB.3FSG) (25) fi ktīr ʕālam ʔāxde fǝkret kīf kānet ʕāyše l-masīḥiyye ʔabǝl ―there are a lot of people who have an idea of how the life of the Christians was before‖ (ʔāxde: AP.FSG) Example of strict agreement: 94 . Procházka & Gabsi. ―Agreement‖, p. 244, y p. 246, referring to the form baṛša ―many‖ in Urban Tunisian. Also, example (5) in Ritt-Benmimoun. ―Agreement‖, p. 266 (Tunisian Bedouin), also involving baṛša, is consistent with this. 95 . Procházka & Gabsi. ―Agreement‖, p. 245. 96 . D’Anna. ―Agreement‖, p. 108. 97 . Results which corroborate that fi does not trigger only one kind of agreement are: fi nās fi ʕālam TOTAL Deflected agreement 19 16 35 Strict agreement 22 8 30 AGREEMENT PATTERNS WITH THE NOUNS NĀS AND ʕĀLAM IN DAMASCUS ARABIC MEAH, SECCIÓN ÁRABE-ISLAM [1696-5868] 70 (2021), 3-31. DOI 10.30827/meaharabe.v70i0.15152 21 (26) fi ʕālam ǝktīr byǝlǝbsu ḥalaʔ ǝkbīr ―there are a lot of people who wear big earrings‖ (byǝlǝbsu: VB.3PL) 3.1.2. Quantifier kǝll The quantifier kǝll ―all‖ emphasizes collectivity and generality, hence triggers deflected agreement. Our data show 21 cases of deflected agreement in which kǝll precedes or follows the controllers (i.e. kǝll nās/ʕālam or ǝn-nās/ʕālam kǝll-a/on), and 7 cases of strict agreement, two of which can be explained by the distance between the controller and the target or by the presence of features of specification. The results confirm the clear tendency to collectivity fostered by kǝll – thus deflected agreement prevails. The same has been observed in other varieties of Arabic, as in Urban Tunisian 98 and Bedouin Tunisian 99 , both studies specifically referring to the controller nās, and in Omani Arabic 100 . Examples of deflected agreement: (27) kǝll ǝn-nās ʔaʕdet ǝb-bēta ―all the people stayed at home‖ (ʔaʕdet: VB.3FSG) (28) kǝll ǝl-ʕālam mawžūde bǝž-žāmeʕ ―all the people are at the mosque‖ (mawžūde: PP.FSG) Example of strict agreement: (29) bǝtšūfi kǝll ǝl-ʕālam wāʔfīn ʕaž-žabal u ʕam-byǝdʕu rabbon ―you see all the people standing upon the mountain and imploring God‖ (wāʔfīn: AP.PL) 3.1.3. Other quantifiers In addition to ktīr and kǝll, our data include three more quantifiers: ġālibiyye ―majority‖, ʔaktar ―majority‖, and baʔiyye ―rest‖. All of them seem to favor strict agreement. Ġālibiyye appears in the following two clauses: (30) bǝtlāʔi ʔǝnno ġālibīt ǝl-ʕālam ǝlli sāknīn byǝštǝġlu tǝžžār, bǝl-ʕamāra, ġālibīton byǝštǝġlu bǝl-mǝhan ǝl-yadawiyye ―you find that most of the people who live (there) are traders; in the ʕAmāra 101 most of them are craftsmen‖ (sāknīn: AP.PL, byǝštǝġlu: VB.3PL, ġālibīton: NOUN-PRON.PL, byǝštǝġlu: VB.3PL). In this case, the controller is clearly specified by the relative pronoun ǝlli. 98 . Procházka & Gabsi. ―Agreement‖, p. 245. 99 . Ritt-Benmimoun. ―Agreement‖, p. 272, p. 283. 100 . Bettega. ―Agreement with plural‖, p. 163. On the other hand, in southwestern Libya kǝll does not seem to attract deflected agreement (D’Anna. ―Agreement‖, p. 108, specifically refering to nās). 101 . Name of a traditional neighborhood in Damascus. CARMEN BERLINCHES RAMOS MEAH, SECCIÓN ÁRABE-ISLAM [1696-5868] 70 (2021), 3-31. DOI 10.30827/meaharabe.v70i0.15152 22 (31) kamān ġālibīt hal-ʕālam bǝtḥǝssiyyon ʔǝnno hǝnnen la-hallaʔ šǝġlon kamwaẓẓafīn ―also, most of the people, you feel that up to now they have worked as employees‖ (bǝtḥǝssiyy-on: VB-PRON.PL, hǝnnen: PRON.3PL, šǝġl-on: NOUNPRON.PL, mwaẓẓafīn: PP.PL) ʔaktar occurred twice in our data, both times in clauses which show strict agreement: (32) fiyya nās bass ʔaktaron sūriyyīn ―there are people in it; but most of them are Syrians‖ (ʔaktar-on: ADV-PRON.PL, sūriyyīn: NOUN.PL) Finally, baʔiyye appears just once, in a sentence showing strict agreement: (33) ǝl-baʔiyye nās ʕādiyyīn ―the rest are common people‖ (ʕādiyyīn: ADJ.PL) 3.2. Qualification 3.2.1. Adjectives Khan stated that a nominal specified by a qualifier is more individuated than one not so specified 102 . Among the qualifiers which specify the controllers nās and ʕālam, adjectives are the most numerous and attract strict agreement with more frequency (60.71%; see section 2.1.2.). Examples of strict agreement: (34) ʔaktar mǝn ǝl-nās ǝl-ʕādīn ―more than the common people‖ (ʕādīn: ADJ.PL) (35) ǝl-nās ḥabbābīn u kǝllon laṭīfīn ―the people are affectionate and all (of them) charming‖ (ḥabbābīn: ADJ.PL, kǝll-on: ADV-PRON.PL, laṭīfīn: ADJ.PL) Strict agreement is maintained despite the presence of kǝll, which triggers deflected agreement (see section 3.1.2.). Examples of deflected agreement: (36) ṭabʕan mǝnḥǝbb ǝl-ʕālam ǝt-tānye ―of course we love the other people‖ (tānye: ADJ.3FSG) (37) fi nās ġarībe birǝššu l-mǝlǝḥ ʔāxer ʔāxer ǝš-ši ―there are strange people who add the salt at the very end‖ This is a case of mixed agreement in which the first target (ġarībe: ADJ.FSG) shows deflected agreement and the second target (birǝššu: VB.3PL) strict agreement. Interestingly, both targets appear one after the other. 3.2.2. Relative pronoun (yǝ)lli The presence of the relative pronoun in a clause clearly emphasizes specification, thus attracting strict agreement. Our data has 25 instances of this, against 2 which show deflected agreement. Examples of strict agreement: 102 . Khan. ―Object markers‖, p. 470. AGREEMENT PATTERNS WITH THE NOUNS NĀS AND ʕĀLAM IN DAMASCUS ARABIC MEAH, SECCIÓN ÁRABE-ISLAM [1696-5868] 70 (2021), 3-31. DOI 10.30827/meaharabe.v70i0.15152 23 (38) ǝn-nās ǝlli ʕandon ǝrfīʔ ṣāru yǝsharu ―the people who have a partner, started going out in the night‖ (ʕand-on: PREP-PRON.PL, ṣāru: VB.3PL, yǝsharu: VB.3PL) (39) ḥābeb ʔǝnno n-nās ǝlli barra bi-ʔōrǝbba yšūfūha ―I would like that the people who are outside, in Europe, see it‖ (yšūfū-: VB.3PL) Examples of deflected agreement: (40) nǝḥna ʕanna nās ǝlli ʕam-yǝtʕallam ǝl-ʔǝslām 103 ―we have people who are studying Islam‖ (yǝtʕallam: VB.3FSG) (41) lāzem ǝtsāʕdi n-nās yǝlli ma maʕa maṣāri ―you must help the people who have no money‖ (maʕ-a: PREP-PRON.3FSG) The prevalence of strict agreement in relative clauses has been observed in other varieties of Arabic, such as that of southwestern Libya 104 and in Urban Tunisian 105 . Therefore qualification (by an adjective or a relative pronoun) favor strict agreement, in contrast to Omani Arabic, where qualified controllers show a preference for deflected agreement 106 . 3.3. Reflexibility 3.3.1. Reflexive pronoun baʕḍ Khan pointed out that reflexive complements are a feature of nonindividuation 107 . Our data show 6 cases of deflected agreement and 2 of strict agreement. Perhaps these two cases of strict agreement are because the targets follow a target already in the plural. As was stated above, the shift strict > deflected is unusual in Damascus Arabic (see section 2.3.). Examples of deflected agreement: (42) ǝl-ḥāra š-šāmiyye kānet kǝll ǝn-nās taʕref baʕḍa ―in the Damascene neighborhood all the people knew each other‖ (taʕref: VB.3FSG, baʕḍ-a: PRONPRON.3FSG) (43) masalan ǝn-nās ʕalāqāta maʕ baʕǝḍha kīf ―for example, how is the people’s relation to each other‖ (ʕalāqāt-a: NOUN-PRON.3FSG, baʕǝḍ-ha: PRONPRON.3FSG) 103 . Usually the relative pronoun is only used when its antecedent is definite (cf. Aldoukhi et al. Lehrbuch, p. 217). Since nās is indefinite, this example is an exception to the rule. 104 . D’Anna. ―Agreement‖, p. 108. 105 . Procházka & Gabsi. ―Agreement‖, p. 252. 106 . Cf. Bettega ―Agreement with plural‖, p. 161. 107 . Khan. ―Object markers‖, p. 470; Brustad. The Syntax, p. 55. However Brustad proves with the following example extracted from Cowell (A reference grammar, p. 425) that deflected agreement does not always occur in these cases: hal-ʔalwān mā bināsbu baʕḍon ―these colors don’t go together‖. Notice, however, that this example shows a non-human controller. CARMEN BERLINCHES RAMOS MEAH, SECCIÓN ÁRABE-ISLAM [1696-5868] 70 (2021), 3-31. DOI 10.30827/meaharabe.v70i0.15152 24 Examples of strict agreement: (44) ǝn-nās yaʕni mǝnfǝtḥīn, mǝnfǝtḥīn ʕala baʕḍon ―the people are open-minded; open-minded to each other‖ (mǝnfǝtḥīn: AP.PL, baʕḍ-on: PRON-PRON.PL) 3.4. Specificity 3.4.1. Personal pronouns The high level of individuation provided by personal pronouns seems to be the reason why in our data they only show strict agreement with their controllers. But in 4 of the 14 occurrences of personal pronouns the relative pronoun, which favours strict agreement, was also present (see section 3.2.2.). Personal pronouns are placed at the end of Corbett’s hierarchy for deflected agreement 108 . Examples of strict agreement: (45) hǝnnen ǝn-nās ma byǝšrabu mn ǝn-nahǝr mubāšaratan ―they, the people, do not drink directly from the river‖ (hǝnnen: PRON.3PL, byǝšrabu: VB.3PL) (46) bǝl-manṭiʔa bēn l-ǝtnēn bǝtlāʔi ʔǝnno ʕālam hǝnnen ʔaṣlan mānon šwām ―in the area between both, you find people who are not originally Damascene‖ (hǝnnen: PRON.3PL, mān-on: NEG-PRON.PL, šwām: ADJ.PL) 3.4.2. Demonstrative pronouns Demonstrative pronouns prominently individualize their controllers, hence almost all the instances in our data (6/7) show strict agreement. The only example of deflected agreement involves the short form hal-, which does not show gender and number. The strong deictic character of the demonstratives provides a high degree of prominence to the controller, as observed in Urban Tunisian 109 . Examples of strict agreement: (47) lēš hadōl ǝn-nās hēke byǝʕǝmlu? ―why do these people do so?‖ (hadōl: DEM.PL, byǝʕǝmlu: VB.3PL) (48) hadōl ǝn-nās yǝlli ma biḥǝṭṭu ǝl-ḥijāb ―these people who don’t wear the hijab‖ (hadōl: DEM.PL, biḥǝṭṭu: VB.3PL). In this example, individuation is prominently marked by the co-existence of the relative pronoun yǝlli (see section 3.2.2.). Example of deflected agreement: (49) hāda bikūn rabb ǝl-ʕālamīn ʔǝnno ġafar-lon la-han-nās kǝlla ―this is God who absolved all these people‖ (kǝll-a: ADV-PRON.3FSG). 108 . Cf. Corbett. Hierarchies, targets, p. 207. 109 . Procházka & Gabsi. ―Agreement‖, p. 252. But the authors add that this is not the only possible reason for explaining the strict agreement: it could also result from morpho-syntactic principles. AGREEMENT PATTERNS WITH THE NOUNS NĀS AND ʕĀLAM IN DAMASCUS ARABIC MEAH, SECCIÓN ÁRABE-ISLAM [1696-5868] 70 (2021), 3-31. DOI 10.30827/meaharabe.v70i0.15152 25 This is the only example of deflected agreement found in the data. It perhaps can be explained by the presence of the quantifier kǝll (see section 3.1.2.), which triggers deflected agreement. Examples of demonstrative pronouns prominently marking the controllers and therefore attracting strict agreement are found in other varieties of Arabic, like Urban Tunisian 110 . Bettega’s findings specifically concerning nās in Omani Arabic suggest the same conclusion 111 . 3.4.3. Locative constructions In many cases where there are references to local places, strict agreement prevails. This could be explained by the fact that such references provide specific information about people in a definite place, thus clearly individualizing the group. Examples: (50) ǝl-ʕālam ʔaḥyānan barra ʔaw hēk ši bifakkru ʔǝnno ṣ-ṣalā ši bixawwef ǝlʕālam ―the people sometimes, outside (Syria) or so, think that (Muslim) praying is something that scares the people‖ (bifakkru: VB.3PL) The locative phrase barra ʔaw hēk ši specifies that not everyone thinks in the way the sentence describes, only those abroad. (51) fi ʕālam bǝš-šām biʔūlu l-mara btǝʕmel ǝḥžāb la-žōza ―there are people in Damascus who say that the woman makes an amulet for her husband‖ (biʔūlu: VB.3PL) The locative phrase bǝš-šām specifies that the sentence refers only to people in Damascus. A few examples of this kind include the preposition ʕand ―with, at‖: (52) bǝdda ktīr nās ʕandon byǝʔdru yǝtʕāmalu maʕ ǝl-ʕālam u yǝḥku maʕ ǝlʕālam u hēk ―many people there can mix with the people and talk to the people and so‖ (ʕand-on: PREP-PRON.PL, byǝʔdru: VB.3PL, yǝtʕāmalu: VB.3PL, yǝḥku: VB.3PL) The preposition ʕand followed by the anaphoric pronoun -on literally means ―at that place, at the place where they (= these people) are‖. However, the following example does not show strict agreement despite the occurrence of hnīke ―there‖—perhaps due to the double presence of the quantifier kǝll, which attracts deflected agreement: 110 . Ídem, p. 243. 111 . Bettega. ―Agreement with plural‖, p. 168, fn. 17. CARMEN BERLINCHES RAMOS MEAH, SECCIÓN ÁRABE-ISLAM [1696-5868] 70 (2021), 3-31. DOI 10.30827/meaharabe.v70i0.15152 26 (53) kǝll ǝl-ʕālam ǝhnīke lābse ʔabyaḍ, kǝll ǝl-ʕālam ―all the people there wear white (clothes), all the people‖ (lābse: AP.FSG) 112 . 3.4.4. Distance & Specification It was mentioned in section 2.3. that mixed agreement clauses usually include a first target which is near to its controller and shows deflected agreement, and a second target distanced from the same controller and showing strict agreement. In many occasions, the distanced target(s) add further information about their controllers; therefore distance is regarded as a frequent factor for specification or individuation 113 . This phenomenon has been registered in other varieties, like Urban Tunisian 114 , Bahrain 115 , and southwestern Libyan 116 . Examples: (54) ǝn-nās ǝtsāfer, trūḥ u tǝži, ṣāret ǝtšūf ġēr ʔašyāʔ, fa-ṭṭarru ʔǝnno ydawwru ʕala luġa tkūn ʔashal bǝt-taʕāmol ―the people travel, go and come, started seeing other things; therefore they were obliged to look for another language easier for the relation between them‖ (tsāfer: VB.3FSG, trūḥ: VB.3FSG, tǝži: VB.3FSG, ṣāret: VB.3FSG, tšūf: VB.3FSG, ṭṭarru: VB.3PL, ydawwru: VB.3PL) The second clause, in which all verbs are in plural, explains the actions taken as a result of the situation mentioned in the first clause, in which all the verbs are in feminine singular. (55) ktīr nās ǝbtǝrʔoṣ dabke wǝd-dabke b-sūrya masalan kǝll nās byǝrǝʔṣuwwa šǝkǝl ―many people dance dabka; and the dabka in Syria, every one dances it in one way‖ (btǝrʔoṣ: VB.3FSG, byǝrǝʔṣuww-: VB.3PL). In this example the first sentence shows deflected agreement, referring to the people as an undefined group; and the defected agreement is reinforced by the presence of the quantifier ktīr (see section 3.1.1.). The second sentence, despite the presence of kǝll, which emphasizes collectivity, shows strict agreement, since it specifically refers to the group of people mentioned in the first sentence, i.e. each of those who dance the dabka 117 . 112 . A further mention of specification given by a locative, with an example of it showing strict agreement, is in D’Anna (―Agreement‖, p. 108), referring to southwestern Libya. 113 . Brustad. The syntax, p. 58. 114 . Procházka & Gabsi. ―Agreement‖, p. 246. 115 . Holes. Dialect, culture, p. 335. 116 . D’Anna. ―Agreement‖, p. 109. 117 . This example could fit in the group of mixed patterns described thus: ―Mixed patterns also occur with S-V word order in sentences that do not refer to past events but express general assertions. Again the first part contains the general statement about people as a collective whole, followed by a sentence which tells us what the individual members of this group are doing‖ (Procházka & Gabsi ―Agreement‖, p. 246) . AGREEMENT PATTERNS WITH THE NOUNS NĀS AND ʕĀLAM IN DAMASCUS ARABIC MEAH, SECCIÓN ÁRABE-ISLAM [1696-5868] 70 (2021), 3-31. DOI 10.30827/meaharabe.v70i0.15152 27 Finally, sometimes distanced anaphoric pronouns provide further specification about the controllers, as in the following: (56) fi ʕālam ǝktīr ǝbtǝži ʕala hāda l-ʔǝžtimāʕ masalan ǝl-ʕālam yǝlli… hallaʔ masalan ʔana bǝnzel maʕon kǝll ǝnhār žǝmʕa ―there are many people who come to this meeting, for example, the people who… for example, I go with them every Friday daytime‖ (btǝži: VB.3FSG, maʕ-on: PREP-PRON.PL). In this case the first target (btǝži) shows deflected agreement, probably fostered by the presence of fi ktīr (see section 3.1.1.), its clause referring to people as a collective group. The second clause shows strict agreement, specifying precisely those at the meeting to whom it refers: that is, those whom ―I go with … every Friday daytime‖. 3.5. Definiteness Definiteness was regarded by Khan as a feature of individuation 118 ; but our data do not show any difference in agreement attending the controllers’ definition or non-definition. The reference to the controllers in many of our examples is definite, though clearly not specific or individual. For instance: (57) mšān hēk yaʕni n-nās bǝtrūḥ ―this is why the people go‖ (bǝtrūḥ: VB.3FSG), (58) hāda lmanẓar ǝl-ʕālam lǝssa mu mʕawwade ʕalē bi-ʔōrǝbba ―in Europe the people are not used to this view yet‖ (mʕawwade: AP.FSG). Therefore, with regard to clauses involving the nouns nās and ʕālam, definiteness does not seem to be a determining factor in kind of agreement. The same phenomenon has been observed in Omani Arabic. Bettega found in his data a slightly higher percentage (8%) of strict agreement when controllers are defined; however these results are based on inanimate controllers 119 . It would be interesting to further study this question with non-human plural controllers in Damascus Arabic. 4. CONCLUTION Agreement patterns with the nouns nās and ʕālam in Damascus Arabic show strict agreement prevailing, being used in 60.90% of the instances. Also, both nouns pattern similarly; and the number of instances of each is about the same (nās: 120 instances, ʕālam: 116 instances). Some factors seem to promote one agreement type over the other; but all types of targets show preference for strict agreement. However, our limited number of 118 . Khan. ―Object markers‖, p. 470. The author, however, points out that definiteness vs. indefiniteness is not the only factor related to the degree of individuation of a nominal. 119 . Bettega. ―Agreement with plural‖, p. 161.