The Status of Religion/Sect-Based Linguistic Variation in Tartus, Syria: Looking at the Nuances of Qaf as an Example

Abstract

This study investigates the social and historical dynamics affecting the religion/sect-based linguistic distribution and associations of the Qaf variants, namely [q] and [ʔ]. The findings are based on the data gathered through interactions with 93 Arabic speakers from Tartus Center, Syria. The descriptive analysis reveals the presence of religion/sect-based linguistic distribution, with [q] being dominant mainly among Alawites and [ʔ] being dominant among Sunnis and Christians and increasingly becoming popular among Alawites of the urban regions. The paper highlights the emergence of [q] as a religiously, socially, and symbolically marked Alawite variant during the war, contrary to the [ʔ], which became a supralocal and religiously neutral variant that speakers of marked [q] backgrounds can resort to at times of tension and social pressure.

1. Introduction

Large-scale social, economic, and political changes in the Arab World resulted in migratory movements that altered the demographics within and across many of its countries. These movements brought various groups together. The contact that emerged varied in intensity and pace from one setting to another, depending on various factors such as migration patterns, political situation, power dynamics, and rural and urban cleavage. It is not uncommon for these moving waves to have sociolinguistic repercussions on the various communities involved, namely the migrants, the original, and destination societies (Kerswill 2006, pp. 2271–72). In the Arabic context, these changes complicated the sociolinguistic situation and resulted in the emergence of new linguistic patterns that were not present before (Holes 1995, p. 270; Al-Wer 1999, p. 39; Miller 2004, p. 178; Abdel-Jawad and Abu-Radwan 2013, p. 1).

The religious and sectarian diversity that Arab societies enjoy has been among the factors complicating such linguistic and social changes. Religion plays a defining role in constructing individual and group identities in these communities (Bassiouney 2017, p. 37; Al-Wer et al. 2022, p. 106), where the linguistic space can become a site for social differentiation. In such contexts, individual features, whole dialects, and even languages can emerge as distinctive cues that people rely on to differentiate and label others from several religious and sectarian communities (Abdel-Jawad 2006; Abu-Haidar 1991; Amara 2005; Holes 1983; Germanos and Miller 2014; Al-Wer et al. 2015; Bassiouney 2017).

Various cases of linguistic variation correlating with religious or sectarian affiliation have been reported in many cities in the Arab World, with the studies by Cantineau (1946) and Blanc (1953) on the Druze; Blanc (1964) on Muslims, Christians, and Jews; and Holes (1983, 1986) on Shiʿas and Sunnis being among the earliest accounts, as will be highlighted below. However, such a correlation (between religion/sect and linguistic variation) is neither straightforward nor intrinsic. Instead, it is greatly influenced by other mediating factors that largely shape its emergence, maintenance, and future scope (Germanos and Miller 2014; Al-Wer et al. 2015, 2022).

The next section highlights some of the main factors that have been essential to the emergence, maintenance, and potential tapering off of religion/sect-based linguistic variation in a number of studies in the Arabic context.

2. Literature Review

2.1. Factors Affecting the Emergence and Maintenance of Such Variation

Various historical factors have been reportedly responsible for the emergence of religion/sect-based linguistic variation in the Arabic World. The “genealogical origin” (Al-Wer et al. 2015, p. 333) or “migratory movements” have been essential in the “reallocation” process in which geographic variants become social ones in the new social or geographic setting. The geographic factor involves population movements, which are essential in triggering this variation (Germanos and Miller 2014, p. 92). These changes can bring diverse groups into the same geographic region where various factors would interact and shape the degree of contact and its linguistic outcomes leading to the emergence of religion/sect-based linguistic variation.

In some contexts and despite being present in the same geography, “social (and often spatial) segregation” (Germanos and Miller 2014, p. 93) or “social barriers” limit “the amount of contact between different religious groups, thus preventing natural processes of koinéization” which can result in dialectal maintenance (Al-Wer et al. 2015, p. 69). This is not an uncommon practice in the Arabic context, where “religious minorities tend to live in specific areas and to develop different linguistic models” (Miller 2004, p. 190). Such spatial segregation and distribution has significantly contributed to the emergence and further maintenance of the Druzes’ cultural identity and dialectal features in Horan, where the [q] variant, for example, emerged as one of the community’s distinctive features (Cantineau 1946, cited in Al-Wer et al. 2022, pp. 109–10).

The marriage tradition has been an important factor influencing the dynamics of religion/sect-based linguistic differentiation in such communities. In Salt City in Jordan, Al-Wer et al. (2022, pp. 119–21) highlight the linguistic practices of Christian speakers who are more consistent in preserving the distinctive features of their Horani dialect. In addition to the region’s political turbulence that increased the Muslim population, the absence of inter-communal marriage seemed to play a role in preserving such features. Such endogamy influences the linguistic dynamics within the families by making Christian children exposed to their traditional dialects at a time when Muslims are exposed to variation due to the diverse dialects of the Muslim groups. The role of such endogamy in preserving the original features and maintaining the distinction from others has been highlighted among other communities such as Druzes in Syria (Al-Wer et al. 2022, p. 111) and Sunna and Shiʿa in Bahrain (Holes 1995, pp. 272–73). It is evidently one of the mediating factors applicable in the context of Tartus, as the paper argues in the following discussion.

The political power and prestige of the respective groups are essential to such dynamics as well. In Baghdad, for example, Christians and Jews tended to maintain their dialects before the political rule came into the hands of Muslims. Such a tendency was attributed to the social isolation of the minorities and the lack of symbolic power and social prestige of the Muslim dialects. However, this changed during the 20th century, when the Muslim group emerged on the political scene, and their dialect started to acquire power and prestige (Blanc 1964, p. 170). Consequently, “the new urban Muslim-based koine developed and expanded in cities and countries with former religious distinctions” (Miller 2004, p. 191).

Bahrain has been an exemplary situation of how various historical, social, political, and economic factors shape the dynamics of linguistic variation in relation to sect. Holes (1983) analyzed variation and highlighted the direction of change among the two main communal groups, namely Sunnis and Shiʿas. He contends that the two groups differed in the dialects they spoke and the types of jobs they occupied. They further varied on the national level. On the one hand, the Sunnis dominated the political and commercial scene and were liberal and more open to foreign influences. The Shiʿas, on the other hand, suffered from religious persecution at the hands of their Sunni counterparts (Holes 1983, p. 434). These elements led to the emergence of physical segregation, where speakers of each group tended to establish social contacts based on sect and kinship. Holes contends that such factors played a significant role in maintaining the dialectal differences between the two. This eventually led to establishing the ʿArab dialect (i.e., the Sunni dialect) as the local standard for the educated Baharna (i.e., Shiʿas), unlike the uneducated population who maintained the older patterns of dialectal, social, and religious demarcation (Holes 1986, p. 50).

2.2. Factors Countering the Emergence and Maintenance of Such Variation

Despite factors pushing toward the emergence and further maintenance of the religion/sect-based linguistic differentiation, other considerations can emerge and counter such segregation and dialectal cleavage. In Bahrain, Holes speaks of other dynamics that work hand-in-hand in opposing such dialectal division and maintenance, including mixed education, mixed workspaces, rapid industrialization, and the spread of media (Holes 1983, p. 434). These dynamics worked toward the emergence of what Holes refers to as “sectarianly neutral forms” that are distinct from the communal affiliation of the groups in question (Holes 1986, p. 51). In a more recent work, Al-Quz (2009, cited in Holes 2011, pp. 103–4) examined the speech of Bahraini school children (aged between 5 and 17) belonging to both dialectal groups. The findings of Al-Quz revealed a significant shift in the speech of those children toward the dialectal speech of the Sunni group: many linguistic features which characterized the speech of the Shiʿa group have become marginal or disappeared from the speech of young school children in favor of the socially dominant dialect. For Holes, despite the presence of bi-dialectal young adults in Bahrain, Al-Quz’s study is indicative of the direction of change and the role of mixed education and peer pressure on such dynamics.

2.3. The Role of Conflict in These Situations

These emerging and present tendencies toward maintenance or shift can become complicated during tension and conflict. Consequently, the religious/sectarian affiliation of the minority and majority groups and their members is likely to interact with the political situation and power dynamics within the country or the region where the conflict occurs, as referred to before. This can result in either leveling the community’s marked features toward the national standard or maintenance of the community’s specific linguistic features. Overall, the linguistic behavior of the groups involved and the pace of change can be influenced by the symbolic associations of the dialects or the features of the dialects involved.

A case of leveling of the marked linguistic features has been reported in Al-Ahsa, in Saudi Arabia, where Al-Bohnayyah (2019, cited in Al-Wer et al. 2022, pp. 63–65) highlights the sociolinguistic outcomes of political tension on the sect-based linguistic variation between the Sunnis and Shiʿas. Before the Yemen war, a state of self-segregation was reported in various cities and villages in that region, which was later changed due to the creation of mixed localities. The relationship between the various groups was good until tension arose during the Yemen war. Only then did things become complicated as the Sunna–Shiʿa conflict in Yemen started to represent a proxy of the regional conflict between Saudi Arabia and the allies and Iran. Al-Bohnayyah’s study focused on the /a:/ variants ([ɒː]–[ɑː]) and the feminine ending ones, namely raised feminine ending [e] and unraised feminine ending [a]. Al-Bohnayyah’s analysis revealed age, gender, and sect each as statistically significant variables. Specifically, the results show a varying degree of ongoing change in progress in both communities. Al-Bohnayyah speaks of different types of pressure on each of the communities. As for Sunnis, their tendency to adopt the supralocal Saudi variants emerged from their unwillingness to be mistaken for being members of the Shiʿa community. Moreover, the Shiʿa tendency has been explained in the light of their willingness to adhere to the Saudi national identity over the sectarian one. Furthermore, while both have been adopting the supralocal variants, variation emerged between them in the pace of such adoption. While the [ɒː] variant was stigmatized for its sectarian association, the other variant of the feminine ending was not. That is why Sunnis were faster in abandoning the stigmatized [ɒː] variant and slower in the case of the traditional feminine ending. Eventually, such variation may cease to exist if change continues like this.

2.4. The Qaf in Syria

A growing number of recently conducted studies have examined the distribution of Qaf variants and the relevance of religious and sectarian identity for the variants’ distribution. The Qaf is commonly realized as [q] in the traditional speech of Druze, who historically inhabited the mountainous region known as Jabal Al-ʿArab or Jabal Al-Druze (Cantineau 1946, cited in Al-Wer et al. 2022, p. 109). This geographic distribution has significantly contributed to maintaining the Druze’s distinct cultural identity and dialect. Other practices include strict endogamy and the inability to convert to and away from the faith (Al-Wer et al. 2022, p. 111).

Such dialectal differences were reported between Muslims and Christians in Aleppo. Unlike the Muslim population, the glottal stop [ʔ] was common among Christians. This variation was present until the early 20th century but later began to be blurred (Behnstedt 1989, pp. 43–44, cited in Palva 2006, p. 610). Another study that dealt with variation in the speech of Christian migrants in two non-adjacent localities in the City of Hims is Habib (2010). The first locality is Al-Hameeddieh, an old urban residential area in the center of the city where the majority of its inhabitants are native Christian Himsi. The urban and prestigious [ʔ] is commonly used among its native inhabitants. The second is Akrama, a newly developing suburb inhabited by rural migrants, largely Alawites, who are [q] speakers. The study revealed the statistical significance of the residential areas and their dominant features on the speech of rural Christian migrants who are [q] speakers, especially concerning [ʔ]’s usage. The instances of [q] among speakers in Akrama were 68% compared to 45% in Al-Hameeddieh. However, the usage of [ʔ] was 55% in Al-Hameeddieh compared to 32% in Akrama.

This variation is present in other cities in Syria. The Alawites of Tartus and Lattakia Governorates are largely [q] speakers, especially in the rural regions. Ismaʿilis in Tartus Governorate are also known for their [q] usage. Variation exists in Safita District in Tartus Governorate between Alawites who use [q] and Christians who are [ʔ] speakers. Many speakers reported the dominance of [q] among Christians in the City of Muhardah in Hama Governorate. The same is with Christians in Al-Quniyah village in the City of Jisr Al-Shughur in Idlib Governorate who speak with [q]. Generally, people in Idlib Governorate and those residing in the urban center of Deir Al-Zour in Eastern Syria are primarily [q] Sunni speakers.

During the last century, Syria’s political, social, and economic transformation led to waves of migration from the rural regions to the urban center of Tartus. These waves brought together people from various social and religious/sectarian origins and eventually altered the city’s demographics. Thus, examining the sociolinguistic repercussions of such changes on the Qaf variable in this context would be revealing. Therefore, the study seeks to answer the following question:

Does the political, social, and economic transformation influence the linguistic distribution of [q] and [ʔ] in the colloquial speech of speakers of different religious/sectarian backgrounds in Tartus Center?

3. Background of the City and People

3.1. Location

The current study was conducted in the urban center of Tartus District, one of five other districts that constitute the Governorate of Tartus, namely Safita, Al-Shaykh Badr, Duraykish, and Banias. The Governorate is situated on the west coast of Syria and borders Lebanon to the south, Hims to the southeast, Hama to the east, and Lattakia to the north (see Figure 1).

Tartus District consists of an urban administrative center referred to as “Tartus Subdistrict” (henceforth, Tartus Center) and a rural range that includes six subdistricts (henceforth, the rural regions of Tartus) that span the coastal mountains and plains. These include Arwad Island to the west; Al-Sawda to the north; and Khirbet Al-Maʿzzah, Al-Safsafah, Al-Hamidiyah, and Al-Karimah to the south.

3.2. People, Politics, and Religion

Despite Syria being the home for various religious and sectarian groups throughout history, their interaction was not always significant. During the Ottoman Rule (1516–1918), religion emerged as an important factor that divided the empire into separate identities. Like other regions under Ottoman rule, Syria was composed of a number of “closed communities” or “communal identities,” such as the Sunnis, Druzes, Jews, Christians, Ismaʿili, and Alawites. This resulted in a religious/sectarian cleavage which was accompanied by discrimination and persecution against religious minorities (Faksh 1984, p. 138). The sense of communal identity was further accentuated at the time of the French Mandate (1921–1945), which played a significant role in encouraging the independent identities of the various religious minorities, including Alawites (Faksh 1984, p. 139).

Despite being the numerical minority in Tartus Governorate, Muslim Sunnis and Christians politically, economically, and socially ruled the Center and rural regions for decades (Balanche 2015, p. 82). Muslim Alawites, on the other hand, were the numerical majority in the rural regions and were “disadvantaged and exploited vis-à-vis other communities” (Faksh 1984, p. 139), and many of them worked for the Sunni and Christian landlords. The discrimination and lack of contact between Alawites and others were not only in the rural regions but also extended to the city centers. Faksh (1984, p. 138) contends that Alawites “were treated with contempt and were not allowed to partake in the city life. Only minimal contact, if any, existed between the Alawis and other religious groups in the urban centers.” The various groups lived in different worlds that “touched but did not mingle” (Hourani 1947, p. 22).

The political transformation in Syria, which gradually started during the French Mandate and escalated after independence (1946), led to significant changes between these groups. This change brought social and economic reforms mainly in the 1960s through the advent of Al-Baath Social Party and its positive role in encouraging social equality and empowerment at its earlier stages (Balanche 2015, p. 105). Such transformation culminated in a “dramatic shift” (Pipes 1989, p. 429) that brought Hafez Al-Asad, one of the Alawite community members, into the rule of Syria in February of 1971, who was succeeded by his son Bashar Al-Asad after he died in 2000. Hafez Al-Asad’s rise to power marked the end of the traditional Sunni rule and a dramatic shift from urban politics, where urban Sunnis ruled, to rural politics, as rural Alawites took over the power (Khoury 1991, p. 1374). Consequently, the rural population, including Alawites, started to migrate in large numbers toward the nearby plains (Balanche 2015, p. 86), where better opportunities for work and education exist. In our context, Tartus Center was a major attraction for them, leading to a transformation and dramatic demographic change, as the discussion below shows.

3.3. The Demographic Change and the Transformation of Tartus Center

Between the early and mid-20th centuries, the now urban center of Tartus was a small town (see Figure 2) that consisted of two main neighborhoods, namely “Khrab Masihiyah” (as Christians largely inhabited it) and “Khrab Islam” (as Muslims mainly inhabited it), which included the old city that borders and faces the sea shore. Migration to the town from the rural regions, including Arwad Island, varied in density throughout the past century, and the first migrants largely resided in these two main neighborhoods.

This in-migration led to a gradual increase in population and expansion of the town without substantially transforming the older patterns of geographic and social segregation between the various groups. The migratory patterns in Tartus Center were largely unsupervised. They were not different from other regions in Syria, where migrants tend to settle in the outskirts or localities close to earlier migrants or urban people of the same religious or rural backgrounds (Batatu 1981, p. 337). This clustering trend has been observed among the Alawite migrants in the city of Hims. Aziz Nakkash (2013, p. 1) comments on this tendency by saying that “It would not be an exaggeration to state that the Alawites have moved their village into the city.” This was not community-specific, as other groups also showed similar trends. As already reported by Habib (2010), the residential area of Akrama, which Alawites largely inhabit, received more Alawite migrants compared to Al-Hameeddieh, which is a Christian residential area.

The same has primarily happened in Tartus Center, where the rural migrants tended to reside in localities inhabited by people from the same rural backgrounds. This was not peculiar to Alawites, as Sunnis and mainly those who come from the rural Island of Arwad showed such tendencies. This migratory behavior led to the emergence of localities inhabited by people of the same rural, social, and religious/sectarian backgrounds.

One early reported neighborhood to emerge was Al-Berraniyah (i.e., the outer space) to the northeast of the old city. As people from different regions and religious backgrounds shifted to this new locality, it became mixed and still enjoys this diversity, though Muslim Sunnis constitute the majority. Another neighborhood was Al-Mina (or the port), which ranges along with the current port of Tartus. This locality was the destination for many Sunni families who dominate it now. To the south of it emerged other neighborhoods that were the destination of many rural Alawite migrants, such as Al-Jameʿ (named after a mosque that means Jameʿ in Arabic), Al-Raml (lit. the sand which is likely due to its sandy land), and Al-Ghamqa (after a river named Al-Ghamqa that passes through it and pours into the sea). Until now, these localities are largely dominated by Muslim Alawites (see Figure 3).

The construction of the port of Tartus, to the north of the old city and next to Al-Mina locality, in the early 1960s raised the need for much labor that the small town could not secure. Such a project and other emerging job opportunities encouraged more people from rural regions to migrate to the city in pursuit of employment.

With the continuation of migration and transformation of the Center, more localities emerged, such as Al-Malʿab Al-Baladi and Al-Rabiyah, which Muslim Sunnis still dominate. Al-Hamrat was another neighborhood that enjoyed the presence of Christians and other communal groups but with less presence of Muslim Sunnis compared to the old city and the previous two localities. The city’s expansion continued and resulted in new mixed localities such as Al-Mahatta-1 and Al-Mahatta-2 in the last few decades.

These continuous waves increased the urban region’s population from 15,000 in 1960 to 150,000 in 2010 (Balanche 2015, p. 88). Moreover, the religious/sectarian distribution within the city center was altered after being dominated by Muslim Sunnis and Christians. Consequently, Alawites became the majority in Tartus’ urban and rural regions (Balanche 2018, p. 6; Khaddour 2015). While no statistics are available about the religious/sectarian distribution within the urban center, the available data from 2011 (Qutrib 2016, p. 11) shows that Alawites constituted 69% of its population compared to Sunnis (18%), Ismaʿilis (7%), and Christians (6%) (see Figure 4).

These rapid changes transformed Tartus from a “small town” that “sleeps with sunset and wakes up before sunrise” to one of “the most important and largest Arab ports in the Mediterranean basin” (Nasif 1969, p. 80). In 1966, Tartus became an independent governorate after it was considered part of the Governorate of Lattakia.

4. Data Collection

This research is part of a larger project investigating language variation and change. It is based on interactions with 93 participants randomly selected from Tartus Center in Syria. These interactions happened between July and September of 2019. They lasted for an average of 14 min. With children, however, they were usually shorter, and the researcher had to look for multiple ways to extend them, such as asking participants to tell jokes and stories or talk about their favorite cartoons. This inconsistency is not uncommon in situations where naturally occurring speech is sought (Kiesling 2011, p. 37).

The researcher is native to Tartus; he was born in a village but migrated to the urban center in the early 90s at the age of five. He left Syria in 2013 and undertook his fieldwork in 2019. Being an insider, on the one hand, and speaking about daily matters in groups, on the other, have been shown in the literature to reduce the effect of the observer’s paradox toward producing more natural speech (Al-Wer et al. 2022, pp. 13–16). Thus, in all interactions, the researcher was accompanied by friends or friends of friends of different genders other than the targeted speakers. These interactions happened in groups of at least three, and only the targeted speaker’s speech was analyzed. Further, the interactions were largely unstructured and about daily matters, as the variants in question are likely to appear frequently in natural speech. However, toward the end and unless the dialectal issue was brought on its own, the researcher would throw in some questions about dialectal differences to elicit social evaluations and attitudes about the dialects and the Qaf variants.

Speakers’ religious/sectarian affiliation is a sensitive issue in the Syrian context, especially in public contexts. However, all participants and their caretakers agreed to include their religious background, and those who did not were excluded. The sample is not balanced. It includes 71 Muslim Alawites, 10 Christians, and 12 Muslim Sunnis (see

Table 1. Distribution of participants across religious/sectarian groups in the urban region.

Religious/Sectarian Group	Participants
Muslim Alawites	71
Christians	10
Muslim Sunnis	12
Total	93

).

It is important to mention that the binary logistic regression test has been performed on the data regarding gender (Mohamad 2023), age (Mohamad 2022a), and religion/sect (Mohamad 2022b). Only age and gender emerged as significant in the urban region. While the unbalanced number of participants of different religions/sects could have contributed to this insignificance, we also tend to believe that this could partially be due to the variation among the Alawites between the use of [q] and [ʔ], despite being dominant [q] speakers. What matters here is where the [q] is more dominant, and the actual distribution seems to reflect this, as Christians and Muslim Sunnis are mainly [ʔ] speakers. In contrast, [q] is primarily used among Alawites, along with the [ʔ].

The Variable of this Study: The Qaf

Despite the wide range of studies that targeted the Qaf variable in the Arabic context (Al-Wer and Herin 2011, p. 50), the interest in carrying out research on it emerged from its salience (Al-Khatib 1988, p. 80), emblematic nature (Suleiman 1999, p. 16), and its increasing association with Muslim Alawites’ speech in the region under study, despite it being used by others elsewhere.

The primary reflexes of the Qaf variable in our context are the voiceless uvular plosive [q] (or Qaf) and the voiceless glottal stop [ʔ] (or ʔaf)3. The [q] is the reflex used in Standard Arabic (SA) but is also used in many Arabic dialects (Suleiman 2004, p. 98).

In this investigation (and due to the significant level of interaction between the colloquials and SA forms in addition to the presence of colloquial [q] speakers), the researcher eliminated the lexical words that are believed to be more commonly regarded as SA rather than dialectal. This was carried out by resorting to the judgment of four students with a background in linguistics who have a good command of both SA and colloquial Arabic. These four students were equally selected based on gender and their rural and urban backgrounds. On the one hand, one male and female were from the urban region, and one male and female were from the rural ones. The list of all the transcribed words that contain [q] instances was given to them, and they were asked to mark the lexical words that they believed were mainly standard. Their judgments were largely based on whether or not a word follows the morphosyntactic and morphophonemic rules of SA. Moreover, a word would be labeled as SA if it has an equivalent commonly used in vernacular speech. An example of this is the word “village,” which can be realized in its SA form as [qərije] or [qərja], and its dialectal form as [dˤəjʕa], [dˤeʕa], or [dˤi:ʕa]. Names of people and places were not reported as SA according to a common agreement between the four speakers (see Appendix A).

5. Results

Of the 1725 tokens in which Qaf occurs, 1038 (i.e., 60%) are realized as [ʔ], and 687 (i.e., 40%) are realized as [q]. The data reveal that [q] mainly occurred in the speech of Muslim Alawites. The [ʔ], however, is present among Muslim Alawites and dominant among Christians and Sunnis (see Appendix B).

The Qaf Distribution among Religious/Sectarian Groups in the Urban Center

The data show that the number of [q] speakers among Muslim Alawites in the urban region was 32, with 28 of them using the [q] solely and 4 using it as follows: 63%, 71%, 84%, and 91% (

Table 2. Intra-speaker usage of [q] and [ʔ] among [q] Alawite speakers.

Speakers	% of [q]	% of [ʔ]
1	63	37
2	71	29
3	84	16
4	91	9
5	100	0
6	100	0
7	100	0
8	100	0
9	100	0
10	100	0
11	100	0
12	100	0
13	100	0
14	100	0
15	100	0
16	100	0
17	100	0
18	100	0
19	100	0
20	100	0
21	100	0
22	100	0
23	100	0
24	100	0
25	100	0
26	100	0
27	100	0
28	100	0
29	100	0
30	100	0
31	100	0
32	100	0

). As for [ʔ], 39 participants were labeled as [ʔ] speakers. While 5 of them used the [ʔ] ubiquitously, the rest showed variation between [ʔ] and [q] and used the [ʔ] as follows: three between 50 and 59%, eight between 60 and 69%, twelve between 70 and 79%, seven between 80 and 89%, and four between 90% and 99% (see

Table 3. Intra-speaker usage of [q] and [ʔ] among [ʔ] Alawite speakers.

Speakers	% of [q]	% of [ʔ]
33	48	52
34	43	57
35	41	59
36	38	62
37	36	64
38	36	64
39	35	65
40	33	67
41	33	67
42	33	67
43	32	68
44	30	70
45	29	71
46	27	73
47	25	75
48	25	75
49	25	75
50	24	76
51	24	76
52	24	76
53	22	78
54	21	79
55	21	79
56	19	81
57	17	83
58	17	83
59	15	85
60	14	86
61	12	88
62	12	88
63	10	90
64	10	90
65	9	91
66	3	97
67	0	100
68	0	100
69	0	100
70	0	100
71	0	100

).

However, all Christians and Muslim Sunnis in the urban region were labeled as [ʔ] speakers. Five of the ten Christians in the urban region used the [ʔ] ubiquitously. The rest used the [ʔ] as follows: 87%, 88%, 89%, 91%, and 94% (see

Table 4. Intra-speaker usage of [q] and [ʔ] among [ʔ] Christian speakers.

Speakers	% of [q]	% of [ʔ]
72	13	87
73	12	88
74	11	89
75	9	91
76	6	94
77	0	100
78	0	100
79	0	100
80	0	100
81	0	100

). As for the twelve Muslim Sunnis, two speakers showed 100% instances of [ʔ]. The rest showed a high percentage of [ʔ] tokens: 78%, 86%, 86%, 88%, 88%, 88%, 89%, 89%, 91%, and 97% (see

Table 5. Intra-speaker usage of [q] and [ʔ] among [ʔ] Sunni speakers.

Speakers	% of [q]	% of [ʔ]
82	22	78
83	14	86
84	14	86
85	13	87
86	12	88
87	12	88
88	11	89
89	9	91
90	3	97
91	2	98
92	0	100
93	0	100

).

Thus, as far as distribution is concerned, the data show that Muslim Alawites vary between [q] and [ʔ] in the city center. However, the [ʔ] is dominant among Muslim Sunnis and Christians (see

Table 6. Dominance of [q] and [ʔ] among speakers of each religious group.

Religion/Sect	Qaf
	[q]	[ʔ]
Muslim Alawites	32	39
Christians	0	10
Muslim Sunnis	0	12

).

6. Discussion

Such religious/sectarian-based distribution of [q] and [ʔ] can be explained by examining the various historical and contemporary factors these communal groups lived through. The discussion is thus divided into two parts, namely before and after the 1970s transformation. It then brings together the varying associations of [q] and [ʔ] and discusses their role in such distribution.

6.1. Before the 1970s Transformation

As highlighted above, the three religious groups evolved differently. Thus, until the 1970s, such differentiation marked Alawites as a political, economic, and social minority despite being the numerical majority (Rabinovich 1979, pp. 693–94). The geographic exclusion of the Alawites into the rural regions and the concentration of Muslim Sunnis and Christians in the urban center led to an overlap between the urban/rural and religious/sectarian lines (Faksh 1984, p. 134). Consequently, the urbanization of Alawites was significantly delayed compared to other groups (Balanche 2015, p. 88).

These factors significantly limited the chances of contact in society and influenced the distribution of Qaf variants in the urban and rural regions. As [ʔ] spread among the urban regions of Syria in the early decades of the 20th century—including Tartus Center—the [q] was largely maintained among the rural population (Behnstedt 1997, map 9, cited in Palva 2006, p. 608). The persistent social segregation and lack of contact prevented urban features from spreading to the rural regions, especially among Alawites.

This distribution can be gleaned from a few reports on the [q] and [ʔ] by speakers in the rural regions and Tartus Center. In a work entitled “Notes on Jabali: The Arabic Dialect Spoken by the Alawis of Jebel Ansariye,” Lewin (1969, p. 8) says that the [q] is a rural feature and that Alawites are mainly [q] speakers. He further contends that his “informant ridiculed the ‘weak’ hamza-speakers [i.e., ʔ speakers] of [the urban] Hama” (Lewin 1969, p. 8). His data were based on a 1963 visit to a village called “Al-Zayne” in the so-called Alawites’ mountains, which he describes as “an ʿAlawi village within easy reach of Hama and situated about three kilometers north-west of Maṣyaf on the main road from Hama to Lattaqie”, Lewin (1969, p. 3). Not dissimilarly, speakers in Tartus Center apparently had their takes on [q] and its speakers as well. In this respect, Habib (2010, p. 79) reports an incident of ridicule that a Christian [q] speaker originating from the rural village of Al-Oyoun in Hims experienced in Tartus Center in 1934. This incident can indicate the negative social status of [q] and its speakers in that Center and the possible dominance of [ʔ] in this geography, as the research argues in the following section.

6.2. After Transformation

The early years of transformation did not immediately alter the previous dynamics of segregation between these groups. Gradually, however, things started to change with two competing factors. While the first group plays a role in extending the past dynamics and original links, leading to the maintenance of [q], the other group helps encourage contact with [ʔ] speakers in Tartus Center and different urban dialects, leading to leveling toward [ʔ].

6.2.1. Factors Limiting Contact and Maintenance

The described unsupervised pattern of migration to the center of Tartus (see Section 3.3) led to the extension of the historic spatial segregation and the emergence of “demographic or linguistic islands” (Abdel-Jawad and Abu-Radwan 2013, p. 14) in the city center where the majority of the people residing in a locality share the same origin, linguistic variety, and religious/sectarian background. This extended segregation resulted in linguistic demarcation in Tartus Center, especially in the older part (Abu-Hajar 2018), where the [ʔ], [ʔ], and [q] dominate the localities of Khrab Islam (majority Sunnis), Khrab Masihiyah (majority Christians), and Al-Jameʿ (majority Alawites), respectively. Similarly, Al-Mina, Al-Malʿab Al-Baladi, and Al-Rabiyah, where [ʔ]-speaking Muslim Sunnis dominate, still maintain a higher level of little contact with others, and the same has happened in the localities of Al-Raml and Al-Ghamqa. Examples of such distribution are reflected in the data. For instance, speakers 92 and 83 are [ʔ] Sunni speakers from Khrab Islam. Moreover, speakers 87 and 91 are two [ʔ] Sunni speakers from Al-Mina. However, participants 14 and 15 are categorical [q] Alawite speakers from Al-Jameʿ locality. Furthermore, speakers 18, 20, and 29 are Alawites from Al-Raml, and the three are categorical [q] speakers. Such extended segregation within the city center prolonged the maintenance of the communities’ original linguistic features.

The religious endogamy prevalent in Syria is not different from that of Tartus. Inter-communal marriages have been rare throughout the various stages of development, and such traditions play a role in maintaining the differences between the different groups, including the linguistic ones. None of our participants, for example, is married or is the child of such mixed marriages.

This situation of little contact was further consolidated by continuing the older links with the migrants’ original places and, consequently, their dialects and the [q] variant. Most Alawites originate from the nearby rural and mountainous regions and often keep links by regularly visiting their own rural houses, lands, and relatives (Nakkash 2013, p. 1). Furthermore, the rural range in Tartus is considerably close to the Center, with a good network of roads connecting them. This geographic and linguistic proximity is an essential factor in maintaining the original features of such migrants (Dodsworth 2017, pp. 333–34). Added to this is the possible role of government and private jobs in Tartus Center, which are based largely on employees from the rural regions who commute daily to the city for work.

These factors played a significant role in extending speakers’ maintenance of [q], especially the older age groups. In Mohamad (2022a), age emerged as statistically significant among the same urban Alawite sample in Tartus Center. While the older generation (between 40 and 49 and 50 and above) showed a higher tendency toward maintaining their original rural [q] feature, the younger age groups (19 and below, between 20 and 29, and between 30 and 39) showed a tendency toward adopting the [ʔ]. These generational differences can indicate the subsequent and gradual change in the city center among the rural Alawite population, which is also reflected in this study’s apparent variation between the [q] and [ʔ] among the generational groups.

In this respect, it is vital to speak of other factors that countered the earlier continuity of segregation along the religious/sectarian lines, leading to this intra- and inter-linguistic variation in our data. As the discussion shows below, these factors encourage contact between groups within Tartus Center and with other urban centers such as Lattakia, Hims, and Damascus.

6.2.2. Factors Increasing Contact and Variation

The emergence of mixed localities in Tartus Center varied across the decades, with Al-Berraniyah being one of the early localities that developed with the expansion of the old city in the last century (see Section 3.3). Muslim Sunnis constitute the majority within this locality, with a few Alawite families and fewer Christians. Compared to the localities where one sect or religion is dominant, speakers within these mixed localities are likely to exhibit more variation between the [q] and [ʔ] with a possibility of a change toward the [ʔ] among the [q] migrant speakers. In our data, speakers 82 and 89 are Muslim Sunnis from Al-Berraniyah, showing slight variation toward the [q], for example. Participants 1, 37, 46, 51, 53 and 60 are from the recently emerging mixed localities, and all show some variation.

Several factors may decide the dynamics of such variation and the direction and pace of change in such localities and others. Speaker 55 is a 27-year-old Muslim Alawite female from Al-Berraniyah. She is categorized as [ʔ] speaker with 21% instances of [q] and 79% of [ʔ]. Her comments highlight some of the essential factors common in the urbanization process, which are realized in a different pace among the various groups, with a few tending to adopt the urban linguistic features quicker than others. These varying tendencies lead to the emergence and co-existence of multiple patterns in the same household, such as female–male patterns, parents’ patterns, and generational patterns (Abdel-Jawad and Abu-Radwan 2013, p. 9). Speaker 55 is the daughter of two parents from the rural region. Her father moved to this neighborhood in 1959 after his father bought land there and started building a house in phases. She highlights the generational and gendered differences within the family. While she considers her parents as categorical [q] speakers, she highlights the gendered variation among the rest. On the one hand, she and her two sisters show a higher tendency toward the [ʔ] with a few instances of [q], and her two brothers use the [ʔ] occasionally but show a tendency toward the [q] variant. Migrants, especially the younger generation, develop these varying tendencies at the earlier stages of migration, with boys showing an inclination toward maintaining the local [q], unlike girls, who show a higher tendency toward the urban and supralocal [ʔ] (Mohamad 2023).

Schools within such mixed localities are among the earlier public sites for contact among speakers of various dialectal backgrounds (Smith and Durham 2019). The role of education as a space for contact continues to influence speakers at a later stage when they go to universities as well. For speaker 55, contact with other urban groups happened when studying at Tishreen University in Lattakia, which exhibits a more significant degree of urbanization and variation. Before the establishment of Tartus University in 2015, a few colleges were present, which were part of Tishreen University in Lattakia. During that period and before, a huge number of students had to continue their education at universities in cities such as Lattakia, Hims, and Damascus. In this respect, education emerged as a key factor for introducing innovative features, such as [ʔ], due to its role in increasing learners’ mobility and social contact, especially at the earlier stages of urbanization (Al-Wer 2002). Students like these commute daily, weekly, or monthly to these cities. Such movements fall within what is referred to as “circulation” (Zelinsky, 1971, cited in Kerswill 2006, p. 2275), and the agents in this are called “language missionaries” who form “a potential bridgehead for the introduction of innovations or for dialect leveling” (Trudgill 1986).

Moreover, the increased availability and expansion of media have been seen as essential factors in pushing toward such erosion through dialect leveling (Holes 1983, pp. 413, 435). In the Syrian context, such influence has been in the direction of the Damascene dialect, where the urban [ʔ] is common. Similar to many other capital dialects, it became the national standard, and its influence is not only national but also among many Arabic-speaking regions through media (Hachimi 2013). Within Syria, such virtual exposure to the dialect and its [ʔ] feature can indirectly affect speakers’ attitudes toward their dialects or dialectal features, which can eventually influence people’s linguistic choices and practices (Kristiansen 2014) and, in our case, toward the [ʔ] variant. Thus, the capital’s dialect and its [ʔ] constitute the target of speakers wishing to move away from their original features.

Furthermore, the 2011 Syrian Crisis has also contributed to increasing variation in the context of Tartus Center. As no conflict happened in the city during the war, it served as a safe haven for the locals and returnees who had migrated to other cities for work. It also became a peaceful destination for thousands of internally displaced individuals of various religions/sects and groups from different regions of Syria4 (Balanche 2018, pp. 21, 37). As a result, the population within the city increased, creating more contact opportunities between speakers from various cities. These dynamics increased the diversity within the City Center and introduced the local people to more urban features such as [ʔ].

6.3. The Qaf Indices: Bringing Things Together

It is common for dialects to develop specific associations not inherent to them (Al-Wer et al. 2015; Germanos and Miller 2014). These associations are mainly shaped by the history and present of their speakers. In such contexts, a dialect and its distinctive variants can become sources and sites for further social differentiation (Suleiman 1999, p. 10; 2004, p. 16; Schmid 2001, p. 4; Darquennes 2015, p. 12).

Thus, before the 1970s, the [q] developed specific associations that were instigated by the social status and geographic distribution of its speakers, rendering the [q] as a rural and non-prestigious variant compared to the urban and prestigious [ʔ]. The economic and political status that Alawites had compared to other groups must have added another layer of social meaning, depriving the [q] of symbolic power, unlike the [ʔ], which accumulated such power. Furthermore, the dominance of [q] among Muslim Alawites in the region has been a significant reason behind its historical association with them, unlike the [ʔ] shared by different religious groups in Tartus and other cities in Syria and the Arab World. The comments referred to above (see Section 6.1) are the only written examples I could find on the evaluation of Qaf variants in the center of Tartus and the rural regions: while rural [q] speakers ridicule those speaking with [ʔ], the latter ridicule the rural [q] speakers as well.

These dynamics largely prevailed until the 1970s transformation. Similar to the Jordanian case (Suleiman 2004; Al-Wer and Herin 2011), an association between using [q] and symbolic power later developed. Such association was primarily due to the arrival of Asad to the presidency in the 1970s and the said disproportionate representation of Alawites in the central military units and security bureaus (Balanche 2015, p. 86). These changes happened without altering the already existing social associations in the country. On the national level, the [q] remains associated with rurality and non-prestige, unlike the [ʔ], which is still regarded as an urban and prestigious variant (Habib 2005, 2010). However, with migration and the subsequent demographic change to the center of Tartus, the [q] enjoyed a more positive evaluation, as Alawites are the majority there now. Even those who acquired the [ʔ] in childhood or adopted it later on do not largely socially discriminate against [q] speakers, as one or more of the family members, including parents, can be a [q] speaker. However, such stigmatization based on [q]’s usage has not stopped (Mohamad 2022b). The sectarian associations continued from the previous period, and the gendered differentiation developed as males and females started to vary between [q] and [ʔ].

Thus, using the [q] during this period could indicate an array of indices, including regional, sectarian, social, and gendered ones. This paper argues that in the absence of any prevailing tension, switching toward [ʔ] for a [q] speaker was mainly for social and gendered associations. However, with the beginning of the Syrian Crisis in 2011 and the gradual increase in tension that took a sectarian dimension in a few cities such as Hims, Lattakia, and Damascus, the political division interacted with the regional and sectarian distribution, dividing the country and turning Tartus into a government city (Khaddour 2015, p. 38). This transformation made the Alawite dialect and its distinctive [q] a stereotype of pro-government people and mainly Alawites, rendering its usage powerless in non-government-controlled cities.

These changing dynamics had linguistic repercussions that worked differently in government and non-government control regions. Here, the paper argues that the sectarian and power indices did not influence the long-term linguistic tendencies and the target of change in Tartus Center, as it has been marked by peaceful existence throughout the Syrian Crisis. This has been confirmed by the absence of any reports of leveling by [ʔ] speakers toward the [q], or vice versa, for these reasons. Instead, this was the case in other cities where tension prevailed. In such contexts, switching to [ʔ] by a [q] speaker emerged from the will to avoid any possible social, regional, sectarian, or political identification. In Excerpt 1, the speaker is a 38-year-old Alawite female who resides in Al-Ghadeer locality. She is a categorical [q] speaker, and she spoke about a recent incident that happened with her during a visit to Damascus, where her male cousin asked her to “soften the [q] that is as big as the sofa.” She stated that she refused to do so as she does not belong to the other side (i.e., [ʔ]-speaking side); she expressed her pride in the dialect and the [q] that she used.

Excerpt 1 Speaker 10: Age (38), Alawite, female, urban, [q].

I am going to tell you about my workplace; the percentage that I see, the people who speak this, three-quarters of them talk with the [Ɂ].

[…]

I could not take this side [I could not speak with [Ɂ]]. I maintained what I am and not what I am exposed to; I am like this. My reality is this. This is the reality that represents me. I did not feel… I went to Damascus where I visited Al-Hamidiyeh, and my cousin told me: ‘Please, soften the [q] that is as big as the sofa… enough, do not speak.’ Why would I do this, my brother? [said even to someone who is a cousin]: This is my dialect, and I am proud of it.

[…]

We do not have the culture of being what you are.

In Mohamad (2022b), many comments were reported regarding such associations and the need to switch away from the [q] in specific contexts or cities to avoid danger. A female Muslim who is a [q] speaker from the rural regions describes this as follows:

Excerpt 2 Questionnaire respondent: Age (51), Muslim, female, rural, [q] speaker.

Switching between [q] and [ʔ] can happen to avoid being associated with a particular social background… it can have political associations… In specific contexts, you may need to change your dialect and avoid using [q], especially while passing on a checkpoint of ISIS or armed militias, to prevent harm or death.

Thus, the indices of [q] and [ʔ] have significantly affected peoples’ tendencies and distribution across the various developmental stages. The degree of such influence and the indices involved in the process varied according to the speakers’ region, political background, social status, and gender role. They further interacted with the religious/sectarian indices the [q] acquired contrary to the religiously neutral [ʔ].

7. Conclusions

It has become clear that the transformation and the later changes in Syria have had linguistic repercussions on the distribution and status of Qaf variants in Tartus Center. Two groups of factors have been competing: one pushing toward maintenance of the local [q] variant among Alawites and the other pushing toward leveling in favor of the [ʔ] variant.

Thus, despite the dominance of [q] mainly among Muslim Alawites, the data reveal variation and ongoing change among them and the dominance of [ʔ] among others. In Tartus Center, the direction of such leveling is not toward the dialectal speech of Muslim Sunnis or Christians, despite being dominated by [ʔ]. Instead, it is toward the supralocal dialect of Damascus or an approximation of it, commonly referred to as “the white dialect” or “Al-Lahje Al-BayDa.” This “white” dialect and its [ʔ] are socially prestigious and are resources of differentiation in the gendered differentiation happening in Tartus, mainly among migrants. The two play a role in shaping the direction of change in Tartus. Both also represent a religiously, regionally, and politically neutral and unmarked resource for [q] speakers wishing to dissociate themselves from the regionally, sectarianly, and politically marked [q] variant outside of Tartus and mainly at danger zones during the crisis.