4 - Events of the Syrian Uprising

Summary

This chapter narrates the events of the first year of the Syrian uprising, drawing on an original newspaper event catalog, activist-generated databases of non-state actor fatalities, and numerous town-specific reports composed by Syrian and other Arab research organizations. In doing so, it describes the outcome to be explained in subsequent chapters: variation in forms of challenge to the Syrian regime over time and space. The picture of contention in Syria presented in this chapter suggests that there was no lockstep progression from nonviolent, urban civic protests to ethnic insurgency in the countryside during the first year of the Syrian uprising. Rather, violence erupted almost immediately following the onset of challenge at some sites, while remaining absent at others throughout the entire first year of the uprising. Claims advanced by challengers varied similarly; they remained focused on civic demands throughout the first year in some areas, while quickly jumping from local to ethnic grievances in others.

Over the course of a single year, the Syrian uprising evolved from isolated demonstrations pressing demands for political reform in the capital to prolonged violent conflict fought largely along ethnic lines. But the processes pushing contention in the direction of ethnic war did not proceed in lockstep across all the country’s regions; widely differing types of contention occurred contemporaneously in different parts of the polity. Peaceful demonstrators pressed citizenship-focused claims in city centers while residents of small towns pressed local demands and engaged in sporadic violence with state authorities, and several cities were racked by neighborhood-against-neighborhood violence on ethnic lines. Other areas, including many Sunni Arab locales, remained free of demonstrations or were sites of pro-regime rallies throughout the first year of the uprising. While spatial diversity in forms of contention persisted throughout the uprising’s first year, some forms of challenge became far more common and others less so. In general, violence and claims involving ethnic symbols and demands came to dominate challenger–incumbent interactions over time, while nonviolent urban demonstrations pressing citizenship-focused claims on the central state and actions of local notables to advance locality-focused claims through informal channels each became less frequent.

This chapter sketches variation in challenger–incumbent interaction over time and space, then lays out in greater detail the forms of contention that constitute the outcome variation for subsequent chapters.

The National-Level Picture of Contention

The period under study in this book runs from the outbreak of the first demonstrations in February 2011 until March 2012, one full year after the initial popular challenge and repression in Darʿa. While challenge was diverse at any given point in time, a predominant form of challenger action is discernible in each of four general phases:¹

1. (1) overwhelmingly nonviolent demonstrations, many of which were aimed at reform and not the toppling of the regime (February–May 2011)

2. (2) intensified, mostly nonviolent protests calling for the fall of the regime, with isolated violent incidents consisting primarily of local residents carrying light weapons to demonstrations in order to deter regime attacks (June–September 2011)

3. (3) increasing defection from the army and movement of activists into the countryside, spurring formation of armed brigades under the Free Syrian Army umbrella (October–December 2011)

4. (4) full-scale civil war following widespread military defection and intensified violence (January–March 2012)

The Syrian uprising of 2011 began in the immediate wake of the Egyptian uprising that toppled President Hosni Mubarak. Initial demonstrations in February 2011 remained isolated, small affairs in the center of the capital, Damascus, until mid-March 2011, when a group of families, protesting the detention of their children in the small city of Darʿa, was fired upon by security forces. This incident intensified protest in Darʿa and spread it to locales throughout Syria. Within a matter of weeks, several cities witnessed near-daily demonstrations. By mid-April, momentum was building to create “Tahrir Squares” – occupations of a central public square on the model of the protests in Cairo’s Tahrir Square that toppled Mubarak – notably in the eastern suburbs of Damascus and in Homs, the country’s third-largest city. Security forces brutally prevented protesters from occupying the squares, using live fire on nonviolent crowds and killing tens or hundreds in each case. Though the most intense violence of the uprising’s first year did not occur until early 2012, regime violence was considerable even in the first weeks of contention. While peaceful demonstrations were dispersed without considerable violence in some places, like the center of Damascus and the Kurdish towns of the country’s northeast, the regime used considerable violence in many other areas. The rate of challenger and civilian deaths per day for the second half of April was the highest for all of 2011 (see Figure 4.1), and the death toll reached 500 during the week of April 18, 2011, during which the regime violently cleared the central squares in Damascus and Homs.²

Figure 4.1. Non-state fatalities over time

Source: Kass-Hout (2016).

Over the course of the summer (June–August 2011), multiple trajectories of contention became increasingly apparent. Nonviolent activism gained momentum, spurred on by the action of local coordinating committees, which were networks of activists on the ground disseminating information about where and how to protest. The first committees were set up in April in major cities, with committees springing up in smaller localities throughout the country by June (Yassin-Kassab and al-Shami 2016: 59, 78). Whereas nonviolent demonstrations were met with minimal repression in some areas of the country, other areas saw increasing violence. In spite of coordinating committee efforts to prevent the arming of challengers, some protesters began to carry rudimentary weapons to demonstrations in cities like Darʿa and Homs. Justified as a means of protecting protesters from state violence, this move provoked further, disproportionate response by the regime, which laid siege to Darʿa in late May. This siege prevented any movement in and out of the city and stopped food, water, and medicine from reaching residents; cut electrical and phone lines; and exposed individuals leaving their homes to sniper fire (Barout 2012: 240). Yet still other locales within Syria saw minimal levels of contention throughout the entire summer, including the centers of Damascus and Aleppo.

This mix of challenger tactics persisted through the end of 2011. Between October and December 2011, some locales saw no anti-regime activity, many hosted continued nonviolent protests, and some experienced continued regime siege tactics. But the balance among these activities tipped, in many cities, toward increasingly violent contention. Regime sieges and security forces’ raids of contentious neighborhoods in search of activists coordinating demonstrations pushed many civilian actors into the countryside. There they took up arms against the regime. Spreading news of regime violence impelled defections from the army, contributing to a self-reinforcing dynamic in which defectors clashed with state forces, encouraging further defections from the army. The end of September 2011 marks an escalation of military repression and the formation of local militias loosely organized under the Free Syrian Army umbrella (al-Haj Saleh 2017: 78). From September onward, many sparsely populated areas outside of cities began to fall under rebel control and operations to liberate peripheral areas of cities removed them from regime control for short periods of time.

Escalation toward even more violent confrontation continued in early 2012. A milestone was reached in January 2012, when Free Syrian Army units and local residents began fighting to hold territory in the town of al-Zabadani in the Damascus periphery, rather than “liberate” it briefly and cede it back in the face of regime incursions as they had done in previous months (Bishara 2013: 199). Similar operations – met by harsh regime reprisals – continued in other areas; the Homs neighborhood of Baba ʿAmr fell fully out of regime control in January 2012 and was not retaken by the regime until March 1, 2012.

Isolating the point at which the Syrian uprising tipped into civil war is largely a question of how one defines civil war. Under Stathis Kalyvas’ (2006: 17) definition, which holds that civil war obtains when armed combat takes place in a sovereign state that was under the control of a single authority at the beginning of hostilities, the regime-challenger confrontations occurring in the fall of 2011 would count as civil war.³ These clashes pitted the regime against the defectors and armed civilians organized loosely under the Free Syrian Army banner. They took place in peripheral areas of Damascus and Homs, as well as the countryside of Idlib, Dayr al-Zur, and Homs governorates. Yet if the number of casualties in a conflict is the primary indicator of a revolutionary situation becoming a civil war, Syria reached this point far sooner; the sieges of Darʿa and parts of Homs and Damascus, among other regime attacks in April and May 2011, raised the non-state death toll to 1,000 on May 21, 2011. The threshold of 2,000 deaths was crossed on July 31, 2011, and, by the end of 2011, the toll reached 5,828; it would nearly double, to 11,260, on the one-year anniversary of the beginning of events in Darʿa, March 15, 2012 (VDC 2016). This accounting of deaths in the Syrian uprising excludes regime forces and pro-regime paramilitaries; though the great majority of those killed in the first year of the Syrian uprising were nonviolent activists, uninvolved civilians, and anti-regime armed actors, this casualty figure remains an undercount of the total number of fatalities in the revolutionary episode.

Disaggregating Contention

More important than dating the beginning of civil war for the entire country is the extent to which patterns of contestation varied within the country at any given moment and how the claims made in a single site evolved over time. Understanding this variation requires subnational disaggregation and a more fine-grained account of the content of challenger–incumbent interaction than that afforded by fatality counts. To illustrate patterns of contention obscured at higher levels of aggregation, this section introduces original data on Syrian town characteristics and contentious events in the first year of the Syrian uprising.

Data Sources

The spatial unit of analysis is the Central Bureau of Statistics’ community-level census unit (n = 5,204), referred to here as a town (see Footnote 5 in Chapter 3 on census units). One dataset charts the majority ethnic identity of all towns in 2010. I collected this data jointly with Kheder Khaddour (Khaddour and Mazur 2018). We employed eight research assistants, who conducted 160 structured interviews with Syrians from all areas of the country to ascertain the ethnic composition of each town. Ethnicity coding was validated against maps compiled by local researchers (and shared with us privately), online subnational maps, and online lists of village ethnic identity posted by local community members. Because it is impossible to determine the exact percentages of two or more ethnic groups in a town, the empirical analysis uses only the majority identity of a town and whether or not a town’s population was ethnically homogeneous; 8 percent of all towns had residents of more than one ethnic identity.

A second dataset tracks contentious events. It contains 2,134 events and counts events at the town-day level, with multiday events coded as single events on their starting date. The dataset is based on multiple, diverse sources to minimize bias in event reporting. It draws on all relevant articles from the Associated Press, the daily digests of the Syrian Observatory for Human Rights (an opposition-leaning activist organization), and al-Thawra (the official political daily newspaper of the Syrian Baʿth Party). The event database records actions by challengers, the regime, and allies of the regime. Each of these actors can take one of several actions, coded on a scheme developed inductively from descriptions of events in secondary sources, the newspaper articles themselves, and my interviews with participants in and observers of the uprising (see the Appendix for a more detailed discussion of source selection and coding procedures).

The rest of this section describes the types of events recorded in the event database. Events coded as clash entail state security or military forces engaging in violent exchanges with armed challengers. This category includes battles along established front lines, regime forces’ skirmishes with defectors, and violent exchanges following a challenger ambush of regime forces.

Challenger actions short of the coordinated violence involved in clashes are classified as either nonviolent, in which a group gathers to make demands on the regime and no violent action is reported, or spontaneous violent, in which crowds initially amass to demonstrate nonviolently and shift toward the use of violence, such as throwing rocks or beating state allies.

Regime actions short of clashes are classified as follows:

1. (1) Crowd control describes actions directed at dispersing demonstrators without inflicting high levels of damage on protesters or monitoring them extensively. This category encompasses a wide range of actions by state agents, including some using far more force than those employed in industrialized democracies to police protests. Techniques range from nonviolently dispersing protests and arresting demonstrators to tear-gassing and beating demonstrators, to firing into the air (so long as doing so causes fewer than two deaths).

2. (2) Tactical control describes an organized form of violence and surveillance directed at a specific segment of the population of a city or town, but not the whole town or city. These tactics appear to be geared toward separating a contentious population from the rest of the town or city or punishing a specific subset of its residents. Examples include raiding a neighborhood to make arrests, encircling a neighborhood and cutting power and water for several days, opening fire indiscriminately on demonstrations, and using snipers to kill demonstrators, when doing so kills fewer than twenty people.

3. (3) Town destruction describes regime targeting of an entire town or major neighborhood of a large city indiscriminately. It inflicts heavy damage, defined as mass property destruction or the killing of twenty or more people. Actions in this category include the siege of entire cities, the shelling of a neighborhood, the burning of homes, and tactical control actions exceeding the twenty-death threshold.

Ally actions are those taken by actors without an official affiliation to the state in support of the incumbent. The actors include civilians who are not part of any organized group but turn out to demonstrations supporting the regime, and informally organized paramilitaries and thugs, commonly called shabbiha (al-Haj Saleh 2017: ch. 2; Khaddour 2015a). Allies can take two sorts of actions in the coding scheme: (1) violent action or (2) nonviolent action. When allies gathered to voice support for the regime – whether as a counterdemonstration or at an independently organized rally – ally actions count as nonviolent. Any sustained physical attack, from throwing stones at and using knives against anti-incumbent demonstrators to organized militias destroying villages assisting challengers, counts as ally violent action.

Temporal Trends in the Event Data

Several patterns of over-time variation emerge from these data. First, the frequency of nonviolent and spontaneously violent challenger actions declined as the uprising began to approximate a civil war in late 2011, but nonviolent challenge events continued throughout the period under study (see Figure 4.2). Even as conditions of full civil war engulfed large parts of the country, many challengers attempted to preserve the nonviolent, social movement character of the uprising. For example, on January 6, 2012, an estimated 50,000 demonstrators turned out for a peaceful protest in the small city of Duma, on the periphery of Damascus, as violent clashes took place between regime forces and rebels in surrounding areas (SOHR 2012a).

Figure 4.2. Biweekly count of challenger actions by type

Source: Mazur (2020).

Second, a small amount of violent challenger action was present from the very outset of the uprising, committed by local communities whose possession of weapons was tolerated by the regime as part of implicit, informal bargains with local notables. In late March 2011, for example, when regime forces fired on civilians, local youths in Darʿa returned fire using hunting rifles and pump-action shotguns kept by local residents before the uprising. Similar interactions took place in Tal Kalakh, a town in the Homs countryside close to the Lebanese border, where local residents were armed due to their involvement in smuggling networks. In May 2011, several residents of Tal Kalakh ambushed and killed an army colonel (Barout 2012: 241–56).

Third, the regime employed a diverse range of tactics against challengers at any given point in time (see Figure 4.3). Crowd control actions were used as a first response to challenge in many cities; on March 25, 2011, in the city of Hama, about 1,000 demonstrators moved from mosques toward the city’s central al-ʿAsi Square. Security forces met demonstrators before they could reach the square, dispersing them with traditional crowd control techniques, including water cannons, tear gas, and batons; no one was killed, and only a few suffered minor injuries. In other cases, regime forces employed tactical control actions to disperse crowds and hinder their organizers. On the same day, regime forces fired into the air and sporadically at demonstrators in the Damascus suburb of Darayya to disperse them, killing three (Karam and Mroue 2011a).

Figure 4.3. Biweekly count of regime actions by type

Source: Mazur (2020).

Finally, regime agents occasionally employed extremely violent tactics from the uprising’s first months. The first use of town destruction actions came on April 8, 2011, when regime forces fired into a funeral procession of thousands in Darʿa city. Several weeks later, the regime laid siege to the center of the city, cutting electricity and phone lines, raiding neighborhoods, and using snipers to kill residents leaving their homes (Associated Press 2011; Kennedy and Hadid 2011). Town destruction actions were initially used against urban demonstrators making national-level demands in Homs, the Damascus periphery, and Darʿa but would, by late 2011, come to be used primarily against violence defectors and armed activists in the countryside.

Variation by Town Size and Ethnicity

Though the forms of challenger–incumbent interaction were diverse, these interactions took place largely in Sunni-majority spaces and central public squares. Non-Sunnis played prominent leadership roles in the nonviolent opposition and average citizens of non-Sunni ethnic backgrounds attended demonstrations – in many cases as delegations from an identifiably non-Sunni urban neighborhood or village – but they had to move outside their local communities to do so. This pattern was not merely a product of mass actors’ orientations and motivations but the result of a concerted effort by the ethnic groups’ religious leaders and community notables to prevent group members from participating in challenge. The regime facilitated these patterns of social sanctioning by systematically avoiding the use of violence against non–Sunni Arab populations when they did participate in challenge.

Patterns of Challenge

When event data are joined with data on the majority ethnic identity of town residents, two patterns emerge: first, nearly all contention was undertaken in Sunni locales, and, second, many Sunni locales were sites of relatively little contention (Table 4.1). It bears repeating that this coding of identity captures only the majority ethnic identity of a town’s residents, not the identity of all residents, nor the identity of the individuals participating in a demonstration. While the great majority of participants in challenge were usually natives of the locale in which an event occurred, some challengers moved out of their neighborhoods and towns of origin, participating in demonstrations in large, diverse cities.

Table 4.1. Challenger actions by town characteristics (column percentages)

	Feb.–May ’11	June–Sept. ’11	Oct.–Dec. ’11	Jan.–Mar. ’12	all periods	number of towns
By majority ethnicity
Sunni Arab	90.2	92.1	91.7	93.5	91.8	3,019
ʿAlawi	0.7	0.4	0.8	2.6	0.9	1,104
Kurdish	7.6	5.8	3.0	0.6	5.2	654
Other minority	1.4	1.7	4.5	3.2	2.2	427
By population size
>1M	10.5	6.5	6.0	12.3	8.3	2
100k to 1M	27.9	30.5	33.8	23.9	29.2	12
50k to 100k	18.5	15.7	13.5	12.3	15.6	17
10k to 50k	39.5	40.9	33.8	35.5	38.8	168
1k to 10k	3.6	6.3	12.0	16.1	7.8	1,941
<1k	0.0	0.2	0.8	0.0	0.2	3,064
Event count	276	479	133	155	1,043

Sources: Central Bureau of Statistics of Syria (2004); Khaddour and Mazur (2018); and Mazur (2020).

Sunni Arab and Kurdish locales together account for 97 percent of all challenge events, and this trend is even starker when the events in non-Sunni majority towns are examined in detail. No homogeneously ʿAlawi locale was the site of challenger action, and 81 percent of the demonstrations occurring in non-Sunni locales were either spillover of demonstrations from neighboring Sunni towns or actions spearheaded by members of the town’s Sunni minority.⁴ Of the twenty-five demonstrations in towns that were both non-ʿAlawi and non-Sunni, nine took place in Christian-majority towns, seven in Ismaʿili-majority towns (including five in Salamiya), three in the Druze stronghold of al-Suwaydaʾ, four in ethnically mixed towns, and two in Turkman towns.⁵

Though the great majority of demonstrations took place in Sunni-majority locales, non-Sunnis were instrumental in organizing many demonstrations, and small groups of youths, in particular, moved out of their areas of residence to participate. Important activists and figures in the local coordinating committees organizing the early, nonviolent demonstrations against the regime were from non-Sunni backgrounds. ʿUmar Idilbi, a spokesperson for the Council of Local Coordinating Committees, estimated that half of the organizers of early demonstrations were from minority ethnic backgrounds; while this figure is unverifiable, it is clear from lists of protest coordinators gathered by activists that a considerable proportion of them were from non-Sunni backgrounds (Marush 2011; Wieland 2016: 232). These coordinators were involved in networks organized on lines of political belief and conscience, rather than ethnicity. For example, ʿAlawis involved in the early demonstrations in the city of Latakia were primarily members of the Communist Workers’ Party, who organized demonstrations with Sunni activists in central areas of the city (Shabo 2015: 11).

Outside the circle of seasoned political activists and well-networked members of local coordinating committees, many non-Sunni youths sought to organize local demonstrations in the same manner as members of Sunni local communities. Unlike their Sunni counterparts, however, they could not, generally speaking, hold demonstrations in their home neighborhoods or villages. The security presence and social opprobrium directed at any pro-uprising activity in the great majority of non-Sunni locales made holding demonstrations there impossible (Bishara 2013: 142–44).

Youths from ʿAlawi and Christian neighborhoods of major cities, such as Damascus and Homs, and surrounding homogeneously minority villages, came to demonstrations in city centers. They declared their presence as members of minority ethnic groups and expressed their solidarity with protests, often through statements on banners bearing the names of their home neighborhoods, communicating their ethnic identity indirectly but unmistakably (Bishara 2013: 144). In the ethnically ʿAlawi and Ismaʿili town of Masyaf, youths put on night demonstrations in June 2011 that were quickly suppressed by local community members and security forces; the subsequent security measures, which included arrests and checkpoints, were so severe that the youth activists were compelled to travel to nearby cities to continue demonstrating (Najjar 2017: 263; Orient News 2015: 22). Delegations of Masyaf youth attended demonstrations in numerous centers of protest, including neighboring Sunni towns and the cities of Duma, Aleppo, Hama, and Homs (ʿIssa 2013).

Just as non-Sunnis were not uniformly quiescent, Sunnis were not uniformly active in challenge. The most prominent examples of this phenomenon are the country’s two largest and wealthiest cities, Damascus and Aleppo, which are also historically Sunni. They account for far fewer of the demonstrations than their share of the population (see Table 4.1). Strong security presence played a role in blocking mass demonstrations in central squares but cannot entirely explain the low frequency of demonstrations in these cities (Wedeen 2013: 845). For example, a demonstration broke out in central Homs, which is subject to similar security presence and the site of the main national military academy, on March 25, 2011, and protests continued in and near the city center for weeks (Bishara 2013: 111–18).

Many of the residents of the central quarters of Aleppo and Damascus were direct beneficiaries of the regime, including senior government employees and businessmen. In Aleppo, in particular, poorer migrants were dissuaded from participation in demonstrations by both the direction of their wealthy employers and the tacit approval given to informal, illegal construction by state authorities as a concession (Barout 2012: 380). Small and medium-sized cities (with 10,000–100,000 residents) account for a disproportionate fraction of challenge events, particularly in the early months of the uprising; the combination of an educated, politically aware younger generation and high levels of exclusion from regime patronage networks in these cities made them particularly contentious (Bishara 2013: 157).

Multivariable regression analysis of the correlates of challenger action provides further insight into variation in participation in Sunni locales. It demonstrates that Sunni towns with higher levels of state employment or informal ties through customary tribal leaders were home to fewer challenge events during the first six months of the uprising, before civil war conditions prevailed (see Table A.1 for further details).

It would be unrealistic to expect that every Sunni town should be a site of contention – fear of repression, the absence of entrepreneurs to organize demonstrations, or simply small population size and isolation might account for a town’s quiescence (Pearlman 2016b: 24). Yet the frequency of ally actions in Sunni-majority locales suggests an active role for some Sunnis in supporting the regime during the uprising. This active support for the regime cannot be put down to fear or selective disincentives to participation. Of the 507 nonviolent actions taken by state allies, 66 percent took place in Sunni-majority towns, and 15 percent of all nonviolent ally actions occurred in homogeneously Sunni locales.

Patterns of Regime Action

Regime action evinces a clear ethnic patterning; the regime used violence almost exclusively in Sunni Arab–majority towns (see Table 4.2). When the regime interacted with challengers in Kurdish and non-Sunni locales, it assiduously avoided using violence.⁶ These broad patterns of regime conciliation with groups other than Sunni Arabs are borne out in accounts of local-level events. After a week of demonstrations in al-Suwaydaʾ, the largest Druze city in Syria, in late March 2011, the regime cleared its central areas of protests, but did so with techniques markedly different from those it used in neighboring Sunni Arab Darʿa. Whereas security forces opened fire on and beat demonstrators in Darʿa, official agents of the regime had a limited face-to-face role in dealing with challengers in al-Suwaydaʾ. Local regime supporters dispersed demonstrators, and security forces made no arrests, merely talking to demonstrators and sending them home (Ezzi 2015: 39–45). A similar dynamic was at play in Kurdish towns engaging in protest. Early demonstrations in Kurdish areas were met with light crowd control, including small numbers of arrests, tear gas, and security forces firing into the air to disperse protesters. Local notables and party members also helped to keep early demonstrations nonviolent, acting as intermediaries between the regime, Arab populations, and Kurdish populations (interviews with participants in demonstrations in Dêrika/al-Malikiya and al-Qamishli: Irbil, May 2, 2014; Irbil, April 23, 2014).

Table 4.2. Regime, clash, and ally actions by town characteristics (column percentages)

	crowd control	tactical control	town destroy	clash	ally	number of towns
By majority ethnicity
Sunni Arab	95.9	98.4	98.8	98.1	68.0	3019
ʿAlawi	0.0	0.7	1.2	0.4	16.2	1104
Kurdish	3.3	0.1	0.0	0.0	3.9	654
Other minority	0.8	0.8	0.0	1.5	11.9	427
By population size
>1M	19.5	4.7	0.0	2.3	24.8	2
100k to 1M	31.7	29.6	44.2	22.0	33.5	12
50k to 100k	18.7	12.5	8.1	10.6	10.3	17
10k to 50k	30.1	40.7	34.9	44.7	16.6	168
1k to 10k	0.0	11.6	12.8	17.4	13.1	1941
<1k	0.0	0.9	0.0	3.0	1.8	3064
Event count	123	760	86	264	513

Sources: Central Bureau of Statistics of Syria (2004); Khaddour and Mazur (2018); and Mazur (2020).

Regime conciliatory action toward challenge was not limited to non-Sunnis and Kurds. Areas with high concentrations of Sunni Arab regime clients received relatively light treatment in comparison to non-client Sunni Arab populations. Crowd control tactics were heavily concentrated in the two largest cities, Aleppo and Damascus, during the early months of the uprising; 51 percent of all crowd control actions before June 2011 happened in the two cities, which produced 18 percent of all challenge during this period. The proportion of crowd control events occurring in these cities declined as security forces made any form of challenge impossible in subsequent months, obviating the need for crowd control. A multivariable regression analysis of the correlates of regime violence evinces similar patterns, showing that regime forces systematically avoided using town destruction actions in towns with high levels of public employment and tribally organized populations (see Table A.3 for further details). The town destruction action was trained primarily on small cities and the countryside (i.e., with a population of less than 100,000). The second-largest size category of cities, with populations between 100,000 and 1 million, faced an outsize share of town destruction actions, but these events occurred only in Homs (32) and Hama (6), cities in which contention developed a violent dynamic separate from the other major cities. Darʿa was the only other sizable city to face this technique more than once – the rest of this violence was used against smaller locales.

Data Strengths and Limitations

Spatial units of analysis are an imperfect approximation of the identities and ties of individual actors engaging in challenge. Because revolutionary episodes are rare and unexpected occurrences, however, nationally representative survey data of this nature are very difficult to collect and, even when available, can be subject to social desirability bias in respondents’ self-description of participation (Beissinger 2013). While spatial units of analysis do not allow for definitive pronouncements about the proportion of members of one ethnic group participating in challenge compared to another, the sites at which challenge occurs provide an approximation of the characteristics of groups engaging in contestation. The town is the lowest level of aggregation available for data on Syria and constitutes an improvement over the regional- and national-level data often available to study violent conflict in autocratic polities. Data at the town level of aggregation also offer clues about the social dynamics pushing individuals to demonstrate at one site or another. The paucity of demonstrations in non-Sunni locales, combined with qualitative evidence that non-Sunnis who wanted to demonstrate did so in urban centers, suggests that non-Sunnis often faced enormous social pressure from their local communities against engaging in challenge.

Forms of Challenge

Having sketched temporal variation in challenge, the rest of this chapter describes the outcomes to be explained in the book: the scope of challenger identity claims and the extent to which violence dominates challenger–incumbent interaction. It develops in greater detail the five ideal types or forms of challenge proposed in Chapter 1 and tracks the extent to which challenger action approximates each of these types.⁷ The forms of challenge are an outcome variable; they describe challenger action and claims at a point in space and time but do not, alone, provide a causal account of why those actions and claims prevail in one place and not another or change over time – accounting for these patterns is the task of Chapters 5 through 8. Though challenger action is the primary outcome of interest, thinking in terms of the interaction with incumbents – rather than merely the statements, behavior, or demographic profile of the challenger group – is imperative because the identity of the challenger is constituted through an intersubjective process. The binary categories challengers employ to define the actors and stakes of contention and the categorization efforts made by their opponents influence one another, and both play crucial roles in how challenger action is received by the broader population. This reception matters greatly for characterizing challenger identity in the polity because it is what ultimately determines challengers’ ability to recruit others to their cause (McAdam et al. 2001: 56).

Table 4.3 summarizes the typology, and Figure 4.4 locates the cases described later in this section in terms of these dimensions. The remainder of this section describes the ideal types of challenge and offers examples of their clearest empirical manifestations.

Table 4.3. Forms of challenge in the Syrian uprising

	citizenship-focused	parochial	hybrid popular	ethnic insurgency	particularist
participant demographic characteristics	all ethnicities, led by educated youth and established opposition figures	natives of locality (always Sunni Arabs)	natives of locality (always Sunni Arabs)	Sunni Arabs, both urban educated and rural/poor, shifting to the latter over time	Kurds only
scope of claimed identity	entire polity	locality	range of local, ethnic, and entire polity	range of local, ethnic, and entire polity, but primarily focused on ethnic majority Sunni Arabs	Kurdish ethnic group
content of claim	new citizenship compact	redress specific grievance, respond to local regime action (overall political change secondary)	ranges from citizenship contract to establishment of Islamic law to redress of local grievances	ranges from citizenship contract to establishment of Islamic law and punishment of ʿAlawis	grant Kurds specific rights, local autonomy, or federalism
challenger violence	mimimal	spontaneous, scattered attacks	spontaneous, scattered attacks	continuous, coordinated destruction	minimal
coordination with other sites	polity-level networks plan many events, some localities organize independently but draw on shared concepts	minimal	polity-level networks coordinate with other sites but also much spontaneous activity	polity-level networks link many fighting groups, others locality-based and operate independently	polity-level networks coordinate, but only among sites dominated by the ethnic group
timing	primarily on Fridays	following a specific event	primarily on Fridays	continuous	not on Fridays
state response	crowd control, outreach to intermediaries, raids to find activists, violent dispersal	outreach to intermediaries, indiscriminate violence	outreach to intermediaries, raids to find activists, indiscriminate violence	outreach to intermediaries, raids to find activists, indiscriminate violence	toleration of demonstrations (and toleration of repression by ethnic group organizations)

Figure 4.4. Coding instances of challenge

Citizenship-focused contention makes national-level demands, in the name of all the residents of the country, primarily concerning the illiberal nature of the regime. These demands relate to how central state power is organized and exercised; if realized, this form of contention would affect the set of rights and obligations of all members of the population, in other words, what it means to be a citizen. Challengers in citizenship-focused events, such as demonstrations and sit-ins, present demands nonviolently and avoid direct violent confrontation with state authorities. Protesters disperse when state agents beat them or open fire on them, though repeated uses of state violence may push a small number of challengers at some sites to respond with violence – for example, by beginning to carry weapons to protests under the premise of protecting demonstrators.

Events approximating this form of challenge took place in cities’ central public spaces and Sunni neighborhoods but drew participants from multiple, diverse neighborhoods, across multiple ethnic groups. Challengers in this type of action coordinated with activists in other cities, organizing demonstrations on a common day, typically Fridays, and employed shared slogans and tactics. The primary coordinators across sites were young local activists linked to other sites through social media, though veteran political activists also played this role to some extent. Challengers presented themselves as a collective greater than any single ethnic group, invoking symbols to deny explicitly the ethnic or otherwise segmental nature of contention and link events in disparate localities (e.g., “One one one, the Syrian people are one,” “with blood and souls, we redeem you, oh Darʿa”).

In small, initial protests occurring in central Damascus in March 2011, for example, challengers made claims for the benefit of all members of the population, and challenger–incumbent interaction was mostly nonviolent. Activists staged sit-ins, and police dispersed them with little force, beating and arresting only a few participants. The organizers and attendees were primarily seasoned political activists and young people, including both generational residents of Damascus and migrants to the capital. This diversity of participant origins meant that participants came from all ethnic backgrounds (Bishara 2013: 421; Fadel 2019: 123). Moreover, ethnic identity was not relevant to challenger demands, which were for liberalization and the release of political prisoners; no explicit mention was made of ethnic identity and neither did the symbols and slogans employed invoke it, emphasizing instead themes of reform, dignity, and loyalty to the Syrian nation (al-Mustafa 2012: 43).

Parochial contention advances primarily local demands and invokes symbols related to local solidarities. Participants make specific demands related to local grievances, with broader change at the political center being a secondary demand. Events of this type emerge spontaneously in single localities, largely unconnected to that in other localities. These events begin with nonviolent gatherings or demonstrations and the vast majority of participants refrain from using any violence; what violence does erupt is uncoordinated, emerging in response to state violence in the course of a demonstration. Parochial contention often features more violent interactions with state agents than citizenship-focused contention because local communities engaging in this form of challenge often have implicit agreements with state agents allowing them to retain light arms. In addition, this form of contention emerges only in Sunni Arab locales; location-specific grievances of local communities of other ethnic identities are directed largely through informal channels of state access.

The contention breaking out in Darʿa a few weeks after the initial protests in Damascus most closely approximates the parochial type of contention. Though regime-challenger interaction in Darʿa in March 2011 featured violence, and the challenger group was entirely composed of local Sunni Arab residents, it cannot be described as “ethnic” because the primary identity highlighted by challengers’ claims was focused on the locality. Demands centered on an intensely local issue, the release of local boys detained by security forces, and the most prominent slogan of the initial protests in Darʿa explicitly invoked local solidarity; protesters chanted, “Where are you, oh people of fazʿa?” (waynkun, ya ahl al-fazʿa?). Fazʿa, which literally means “terror” or “panic,” has a specific meaning of “tribal or regional solidarity dynamics” when invoked in times of crisis, creating “a dynamic of collective defense” (Barout 2012: 184). Yet even this case falls short of the ideal type of parochial challenge; challengers also borrowed national political slogans used in previous weeks in Damascus, such as “God, Syria, freedom, and that’s it” (Barout 2012: 184).

Violence in the early Darʿa protests was intermittent and emerged mostly in response to regime violent action. The sit-in that began the episode of contention was dispersed on March 18, 2011, when regime agents opened fire, killing several challengers. Public gatherings continued in subsequent days with most challengers using no violence. Regime agents alternately policed the gatherings by erecting barricades and cordoning off demonstrators, negotiating with notables, and using force against demonstrators that ranged from beating challengers to arresting them to opening fire indiscriminately (Bishara 2013: 81–94). In response to the latter form of repression, some challengers began carrying weapons and opened fire on regime agents (Leenders 2012: 421).

Hybrid popular challenge presses demands for the fall of the regime, employing local, national, and ethnic symbols. Its participants include both poor, excluded local residents and civic activists of generally higher educational and income levels. This form of challenge combines internet-based organizing of citizenship-focused challenge with mobilization based upon local solidarities and dense neighborhood networks. Hybrid popular challenge is primarily nonviolent, consisting mostly of mass gatherings in central public spaces, but intermittently featuring isolated attacks on state agents in response to previous acts of state repression.

Two features of hybrid popular challenge are particularly notable. First, its participants constitute a broad coalition that includes civil society members and young activists coordinating demonstrations across sites and local residents advancing demands related to specific local grievances. The latter group’s demands typically include the release of long-detained local residents, the removal of a specific local official, and improvement of public services (Barout 2012: 210). Second, the ethnic and violent elements of hybrid popular challenge reflect patterns of pre-uprising governance. Local residents’ experience of state agents is dominated by arbitrary violence, decisions taken without recourse to formal rules or appeal processes, and exclusion from an economy largely controlled by regime allies. The intermingling of violence and an ethnic idiom with nonviolent protest and national slogans is a reflection of “the permeating and violent form of state power that forges its political identity” (Branch and Mampilly 2015: 21).⁸

This violence is distinct from the pure opportunism that often emerges under conditions of prolonged state breakdown. Violence for purely personal gain and looting became prevalent in Syria from late 2012 onward, when the state was absent for long stretches in many areas of the country. But such violence was highly limited during the first year of the uprising and conducted mostly by criminals released from prison by the regime (Bishara 2013: 343–45).

Ethnic claimsmaking that emerged in hybrid popular challenge also requires contextualization; it is at least as much a reflection of the structure of local community and state–society networks as it is a product of the dispositions of individuals invoking ethnic symbols. Sunni religious practice played an important role in the everyday social lives of many members of these dense local networks, making religious discourse and symbols available as an idiom in which to articulate the interests of the local community and opposition to the regime. The Islamist uprising of the late 1970s and early 1980s was central to the use of a religious idiom to define opposition to the regime; many of the communities engaging in hybrid popular challenge had hundreds of members arrested during this time – often on only the slightest suspicion of Muslim Brotherhood membership – and have received no information about them since (Lefèvre 2013: ch. 6). As a result, some members of the local communities engaging in hybrid popular challenge expressed their claims and opposition to the regime in a mix of local and explicitly Sunni religious terms; they chanted slogans against the ʿAlawi-dominated nature of the regime and asked for the release of, or information about, their relatives (Khaddour 2019: 8).

Demonstrations in the city of Homs in early April 2011 approximate hybrid popular challenge. These demonstrations were largely nonviolent and focused on national political demands, but also featured confrontations between challengers and counterdemonstrators, in which the former screamed anti-ʿAlawi insults in the faces of the latter, such as “we want to speak openly, we don’t want to see ʿAlawis here” (Barout 2012: 236). Similarly, in the city of Duma, on March 25, 2011, Islamist challengers chanted “No Iran and no Hizbullah, we want a government that respects God” and attempted to destroy government buildings. Other activists restrained them and their chants were drowned out by the great majority of challengers who were advancing nonethnic claims and stressing the nonviolent character of the uprising (Bishara 2013: 123–24). Demonstrations on the same day in the coastal city of Latakia also featured a range of slogans – some promoting citizenship-focused political goals and others having a Sunni tincture, signaling readiness for violent conflict. The latter category included “We can either live or die” (ya minʿish, ya minmut) and “No to Iran, no to Hizbullah” (Shabo 2015: 16). All these demonstrations arose primarily in response to news of demonstrations and state repression occurring in other locales, such as Darʿa and central Damascus.

Ethnic insurgency consists in armed challengers confronting state forces in battle, clashing with pro-regime paramilitaries, and engaging in clandestine attacks against state agents. A high proportion of challenger–incumbent interactions are violent and the scope of challenger identity claims is primarily ethnic but incorporates local and national elements. Participants in this form of challenge are overwhelmingly Sunni Arab but come from a wide set of geographic, educational, and occupational backgrounds, including activists who initiated the nonviolent urban demonstrations, military defectors, and poor and rural people pulled into the uprising by regime violence. The linkages underpinning this form of contestation are similarly diverse; locally organized brigades drew on dense networks based in individual localities, while defected military officers and activists who had been involved in nonviolent challenge attempted to impose a more formal command structure as the Free Syrian Army.

I term this type of challenge “ethnic insurgency” in spite of the diversity of claims and actors involved because nearly all participants were Sunni Arabs, and Sunni symbols and demands constituted its center of gravity. Whereas demands articulated in nonviolent protests were controlled mainly by urban, educated activists, organized violent challenge occurred mostly in small cities, towns, and villages. Consequently, its leaders expressed themselves in a religious idiom common to these areas; they mixed local solidarities with both a popular religious idiom and abstract demands for freedom – in effect, combining the majoritarian aspect of democratic rule with the fact that the country’s ethnic majority is Sunni. By the time challenge of this nature became prominent, in late 2011, the activists who had assiduously worked to prevent ethnic discourse from dominating urban protests had been largely pushed out of contention due to their arrest, killing, self-imposed exile, or being intimidated into quietism (al-Mustafa 2012: 101–05). Regime violence, which pulled many challengers in the aforementioned peripheral locales into the uprising, was frequently understood in ethnic terms and, consequently, elicited a response in similar terms. Explicitly ʿAlawi and anti-Sunni symbols employed by state agents in some cases facilitated this perception, as when elite army units defaced the ʿUmari Mosque in Darʿa while breaking up a sit-in in March 2011 (Bishara 2013: 90). In many other cases, though the regime did not explicitly employ such symbols, local residents conflated the regime and ʿAlawis in general; this perception was common in Sunni Arab local communities that had experienced fierce regime repression following the Islamist uprising of the 1980s and where powerful state agents and beneficiaries were ʿAlawi, including communities of the oil-rich northeast where remunerative oil service jobs were given mostly to ʿAlawis from the coast (ʿAin al-Madina 2015: 7; Barout 2012: 201).

Contention in Dayr al-Zur at the end of the period under study, in March 2012, approximates ethnic insurgency. By this phase of the uprising, protests in Dayr al-Zur city center had tapered off due to continual regime repression, making the primary challenger–incumbent interaction sporadic challenger attacks on government checkpoints and convoys. Local activists were organized into battalions based largely on extended family networks or ties of shared town provenance, and their names frequently had a Sunni inflection. One prominent fighting group, with members drawn primarily from the town of Muhasan, took the name of an early Muslim leader who is also closely associated with Sunni identity, ʿUthman ibn ʿAffan (Darwish 2016c: 23). This group engaged in the first truly urban warfare in Dayr al-Zur in March 2012, against an elite security force known to be heavily ʿAlawi. Fighting killed over twenty rebels and at least as many members of the security forces. In response, security forces tossed the bodies of rebels from the top floor of the apartment building in which the rebels had been hiding, left them in the street for residents to see, and denied their families the ability to reclaim the bodies for days. Nearly all the rebels killed came from Muhasan. Jamiʿ Jamiʿ, an important security official in the province, met with Muhasan local notables to try to repair relations with the local community and prevent further demonstrations and violence, but failed on both counts; a large funeral procession came from Muhasan into Dayr al-Zur city, and armed brigades continued to expand and engage in armed action against the regime. These brigades were organized on a local basis but employed Sunni symbols and, increasingly, discourse decrying ʿAlawi rule and demanding Islamic rule (ʿAllawi 2016; al-ʿAyad 2016).

Whereas a mix of town, kin, and ethnic identities undergirded claims made during armed conflict in Dayr al-Zur, the identity divisions animating armed conflict in Homs toward the end of 2011 were even more explicitly ethnic and aimed at both state authorities and local ʿAlawi communities. In Homs, by summer 2011, a combination of regime repression of largely Sunni protests and violent confrontations between residents of Sunni and ʿAlawi neighborhoods pushed contention toward violence along ethnic lines. Rebels began to gather in several neighborhoods, drawing in activists and fighters from a range of the city’s neighborhoods to engage in sustained clashes with regime agents (Barout 2012: 288). In January 2012, rebels fired rocket-propelled grenades into several ʿAlawi neighborhoods, killing several people and injuring scores, including women and children. This prompted ʿAlawi paramilitaries to invade Karm al-Zaytun, a nearby Sunni neighborhood, also killing women and children. Escalation occurred on the same pattern in March and culminated in ʿAlawi paramilitaries entering Karm al-Zaytun, slaughtering local residents, mostly with knives and swords, and leaving their mutilated corpses in the street. The regime refused to investigate any of the killers, and regime media described the perpetrators as Islamist extremists from a nearby neighborhood who had killed the (Sunni) local residents for their religious moderation and refusal to confront the regime (Bishara 2013: 335–36).

Finally, particularist contention describes social actors pressing claims that explicitly concern only a particular ethnic group and seeking accommodation for that group short of wholesale change in the political center. Challengers employ symbols of that group and engage in nonviolent demonstrations that are generally tolerated by regime agents, making challenger–incumbent interaction almost totally nonviolent.

Actions approximating this form of contention arose in Kurdish locales not long after ethnic insurgency became a prominent type of challenge in other areas of the country, in late 2011. Participants held rallies and made demands that included recognition of Kurdish cultural rights, extension of citizenship to Kurds whose families had been stripped of it in the 1960s, and, as the uprising entered the civil war phase, federalism or other forms of local autonomy for Kurdish areas. Security forces remained absent from most sites of demonstrations or exercised restraint, allowing demonstrations to go unhindered.

Demonstrations occurring in early 2012 in al-Qamishli, a city with a large Kurdish population, have many particularist elements. Al-Qamishli had been the site of citizenship-focused contention from the early months of the uprising, but by late 2011, contention in the city had become increasingly focused on Kurdish issues, with a formal coordinating body for Kurdish parties becoming involved in protests. Later in 2011, another Kurdish party, the Democratic Union Party (PYD), began holding separate demonstrations in the city, featuring only PYD and Kurdish national flags, with no slogans explicitly opposing the regime (interview with journalist from al-Qamishli, Sulimaniya, April 24, 2014). At the same time, the PYD took steps to suppress the Friday anti-regime demonstrations (International Crisis Group 2013: 10–13; Kurdwatch 2011b).

Variation in Forms of Contention across Space and Time

Any mapping of ideal types onto quantitative data requires indirect, approximate measurement. Even the clearest empirically observed examples of each type do not precisely conform to the ideal types laid out in this section, and quantitative event data generally do not provide information about identity claims and demands. Nonetheless, the original quantitative event data gathered for this book provide a rough, broad picture of the areas and periods where each type of interaction predominates, suggesting the sites to be examined in detail in subsequent chapters.

Contention resembling the parochial ideal type can be glimpsed in the occurrence of spontaneous challenger violence events during the uprising’s early months. Recall that parochial contention is characterized by the local nature of claims and symbols invoked and the intermittent use of violence in challenger–incumbent interactions. Spontaneous violent actions in the event data capture the sort of challenger violence often seen in parochial contention. Table 4.4 presents a list of all such events occurring before May 1, 2011. The secondary literature suggests that local demands featured prominently in many of these events. For example, much of the spontaneous violence in the Darʿa countryside was local residents attacking public buildings and state agents following repression in Darʿa city. On March 26, 2011, challengers in the town of Tafas, in the Darʿa countryside, attacked the local police station and Baʿth Party branch in response to attacks on demonstrators in Darʿa city (Karam and Mroue 2011a). Similarly, the early violence in Tal Kalakh was linked to detentions of local residents in response to a recent government crackdown on smuggling (Amnesty International 2011: 5; Barout 2012: 245).

Table 4.4. Count of spontaneous violent challenge events before May 1, 2011

governorate	capital	countryside	countryside town names (count)
Darʿa	6	7	al-Sanamayn (1), Nawa (2), Sayda (1), Shaykh Miskin (1), Tafas (2)
Homs	6	5	al-Rastan (1), Talbisa (1), Taldu (1), Tal Kalakh (2)
Idlib	1	1	Jisr al-Shughur (1)
Latakia	4	0
Tartus	0	3	Baniyas (3)

Source: Mazur (2020).

Challenge came to be less parochial and more closely approximated the hybrid popular challenge type as contention became common across many areas of the country in late April 2011 (Bishara 2013: 158). By this time, the scope of challenger identity claims was primarily at the national or ethnic level – challengers were well aware of the presence of similarly situated Syrians in other localities pressing similar claims – and the majority of spontaneous violence was the result of citizenship-focused, nonviolent contention facing incumbent attacks and turning into spontaneous violence. For example, after regime forces fired into the air to disperse protesters making national-level claims in Dayr al-Zur on May 27, 2011, civilians attacked and injured several police officers (SOHR 2012b).

A rough approximation of citizenship-focused contention’s prevalence is the predominance of nonviolent challenger action. Table 4.5 demonstrates the change over time on the national level; nonviolent contention predominated until autumn 2011, and the proportion of nonviolent challenge occurring in capital cities dropped over the summer, suggesting a spatial spread of this form of contention. Al-Mayadin, a small city in Dayr al-Zur countryside, exemplifies this dynamic. Several small protests occurred in early April, and demonstrations began to attract large numbers in late April and throughout the summer, before tapering off because the army entered al-Mayadin after confrontations with challengers in Dayr al-Zur city (ʿAbd al-Rahman 2016a); the event database records seven nonviolent challenge events in al-Mayadin between May and August, and none thereafter.

Table 4.5. Challenger actions over time, by type and level of violence

	Mar.–May ’11	June–Sept. ’11	Oct.–Dec. ’11	Jan.–Mar. ’12	all periods
total challenge events	310	528	252	289	1379
non-violent (%)	83	80	33	16	59
spontaneous violence (%)	17	19	22	38	23
coordinated violence (%)	0	1	45	46	18
share of non-violent events in a governorate capital (%)	44	38	50	43	41

Source: Mazur (2020).

The occurrence of challenge events on Fridays provides another indication of the prevalence of citizenship-focused contention, as protests coordinated across sites by local coordinating committees and other national-level activists occurred predominantly on Fridays. Whereas 57 percent of all challenge events before October 2011 took place on a Friday, only 18 percent of challenge between October 2011 and March 2012 took place on a Friday. Variation at the governorate level provides further clues as to the evolution of challenge. Thirty-nine percent of challenge in Darʿa between February and May 2011 took place on a Friday (compared to 62 percent nationally), but this figure rose to 72 percent between May and October 2011 (versus 54 percent nationally), suggesting that initial patterns followed a local logic, but later came to follow the logic operating in many other Sunni-majority locales, at that point focused primarily on national demands. In Dayr al-Zur, by contrast, 93 percent of demonstrations before May were on a Friday, but this number fell to 22 percent between October and December 2011, suggesting initial coordination with national-level activists and a later shift to local logics of armed confrontation.

The rise of challenger–incumbent interactions approximating ethnic insurgency began in the summer of 2011. Defectors in rural Idlib governorate announced the formation of the Free Syrian Army in June, and battalions began to form in other parts of the country soon thereafter (Bishara 2013: 273). Homs, for example, had been characterized by regime repression and confrontations at the borders of ʿAlawi and Sunni neighborhoods from the early weeks of the uprising, but armed challengers only began to organize beyond the immediate neighborhood level in July, with the formation of a battalion called Katibat Khalid bin al-Walid (al-Fares 2015: 27–30). In spite of the growing organization of armed challenger groups and rising intensity of violence in locales like Homs, the proportion of nonviolent events as a percentage of all challenger action hardly changed between the onset of challenge and October 2011 (83 percent from February through May compared to 81 percent from June through September), suggesting that the citizenship-focused type of interaction remained prominent.

From October 2011 onward, the percentage of events on Fridays declined markedly and the ratio of nonviolent to violent challenger actions tipped decisively toward the latter (44 percent nonviolent). This shift to increasing levels of violence was not uniform across governorates, however; al-Hasaka governorate largely remained outside of violent contention because particularist contention came to dominate in Kurdish locales (no clash or spontaneous violent events exist in the database). What little violence occurred in Kurdish locales came in the form of fights between PYD supporters and Kurdish protesters chanting national anti-regime slogans beginning in July 2011 in al-Qamishli and Dêrika/al-Malikiya. PYD and pro-revolution demonstrations became completely separate and operated in parallel across the major Kurdish cities beginning in November 2011 (Kurdwatch 2011b).

The increasingly violent interaction between challenger and incumbent in most governorates coincided with an increase in the ethnic content of claims being made. Identity claims and the nature of challenger demands generally require a more fine-grained view than what is possible through the available quantitative data, but a rough indicator of this change in content is the shift in posts to the Syrian Revolution Facebook group, a key site for coordinating early demonstrations and a forum for discussion in the uprising’s early phases (discussed in greater detail in Chapter 5). Hamza al-Mustafa (2012) conducted a content analysis of the comments posted on the group’s page. Randomly selecting from the site administrator’s posts and reading 1,000 group member comments per month, al-Mustafa found a clear shift toward ethnic discourse over the summer; slogans of sectarian hatred and attacks on the moderate opposition became more prominent than comments supporting civic, political goals (see Table 4.6).

Table 4.6. Type of comments on “Syria Revolution” Facebook page (column percentages)

	Mar.–May ’11	June–Aug. ’11	Sept.–Dec. ’11	early 2012
freedom/political	50	20	9	10
sectarian/hatred	20	45	44	10
sarcasm	30	-	-	-
against army	-	35	-	-
anti-moderate opposition	-	-	9	20
pro-opposition/FSA	-	-	24	40
encouraging intervention	-	-	14	20

Source: Content analysis by al-Mustafa (2012: ch. 2).

The content appearing on the Syrian Revolution Facebook page over the summer is further indicative of this shift. The majority of comments in al-Mustafa’s content analysis from this point forward expressed sectarian hatred (e.g., “God curse your soul Oh Hafiz,” “Iranians are Zoroastrians,” calling the regime shabbiha), insulted the army (e.g., the “Infidel Army” [al-jaysh al-kafir]), and encouraged further armed action against the regime. Beginning in autumn 2011, the discourse on the site increasingly incorporated calls for armed struggle and foreign intervention, with many comments explicitly asking for the “Libyan scenario” in which foreign powers intervened to support rebels. Additionally, fueled by violence in Homs, comments expressing ethnic venom continued to be a mainstay of discussion on the site, in spite of administrator efforts to police them.

Conclusion

The broad quantitative picture of contention in Syria presented in this chapter suggests that there was no lockstep progression from nonviolent, urban civic protests to ethnic insurgency in the countryside during the first year of the Syrian uprising. Challenger–incumbent interactions had an important element of violence from their inception in some parts of the country, and interactions remained nonviolent throughout the period under study in other locales. Still others saw no contention at all. Yet the overall level and nature of violence changed over the course of the Syrian uprising’s first year. Whereas early challenger violence in places like Homs and Darʿa was intermittent and largely a spontaneous reaction to regime violence against nonviolent demonstrations, violence from late 2011 onward took the form of irregular war and, by March 2012, came to be the dominant form of challenger–incumbent interaction.

To analyze the diversity of contention unfolding in different parts of the country, this chapter has inductively theorized the types of challenger–incumbent interaction found in the first year of the Syrian uprising. I propose that five ideal types of challenger–incumbent interaction provide a first approximation of the forms of contention found in the Syrian uprising: parochial contention focused on local grievances that intermittently tips into violence; citizenship-focused contention that is nonviolent and makes national-level political claims; hybrid popular challenge that makes a wide range of claims and includes civic activists and average local residents; particularist contention that is nonviolent and explicitly makes claims on behalf of Kurds only; and ethnic insurgency, which draws combatants almost exclusively from the Sunni Arab ethnic group and mixes local, religious, and citizenship-focused demands. Pairing these types with event data provides a picture of the evolution of challenger–incumbent interaction over time, from being mostly nonviolent and about sub- and supra-ethnic demands to being increasingly violent and featuring significant ethnic content (with national-level demands and nonviolent activism remaining present, if less prominent). Subsequent chapters investigate the mechanisms driving this evolution, elucidating the ways in which the initial forms of contention emerged, what pushed those forms of contention toward ethnic insurgency and particularist contention, and what state logics were at play in this process.