To substantiate my theory, I conducted two fieldworks in Algiers (Algeria) and over six months of archival research in the National Archives at College Park (Maryland, United States) with the support of funds from the Social Sciences and Humanities Research Council of Canada (Joseph Armand Bombardier Doctoral Scholarship), the Center for International Studies at the Université de Montréal, the Schull Yang International Experience Award, the George Washington University’s Project on Middle East Political Science Travel Research and Engagement Grant, the BEAR Network Mobility Award and several McGill University internal scholarships. ​​​

CATALOGUE OF PROTEST NARRATIVES

My empirical work relies on the social narrative analysis of 2937 protests held in Algeria during the presidency of Abdelaziz Bouteflika (1999-2019). A catalogue of protest narratives was built by intersecting protest event analysis methodology with what critical theorist and historians call an "anarchival approach" (Brozgal 2020; Jarvis 2021; Hartman 2019). The latter was developed by scholars studying violence targeting marginalized communities in Algeria and elsewhere to generate a diverse corpus of data by first building an archive (reflecting mainstream accounts) and an "anti-archive" based on competing voices on the same events or phenomena. The goal is to put in dialogue competing accounts and memories of the same event, in view of generating theories that systematically account for the effects of violence on knowledge production. 

​

 To conduct the narrative analysis of recurring protests, I first used the equivalent of an archive listing and coding protest events in Algeria, developed by the Armed Conflict Location & Event Data (ACLED). Employing first a thin approach to narrative analysis emphasizing the protest stories and the characters involved (Shenhav 2015, 83), I recoded every protest event (the unit is event-location-day), by retracing the source used by ACLED researchers, and triangulating it with alternative local sources. The re-coding aimed at rebuilding accounts of protests (stories or plots) by emphasizing and systematically capturing some relational components (actors involved, targets, reported motivations or grievances, actions undertaken/repertoires employed, link to other protest events/developments), and is inspired by the relational and semantic grammar structures used in protest event analysis to reflect better protest narratives ( Franzozi 1998, 2004; Tilly 1978, 1995 ). Coding entries reflect local and plural representations rather than “hard facts.” For instance, under the category of “actors”, content is understood as reflecting less demographic characteristics and more representations of protesters and their reported mobilizing identities. 

​

 I treat pluralism and divergence in protest accounts not as a methodological challenge but as a welcomed resource, a heuristic device. Acknowledging that there may be diverse and potentially conflicting accounts of a single protest event, the catalogue allows for multiple entries for each category. Silences, euphemisms, rumors, analogies, and omissions were also noted and constituted an integral part of the analyzed corpus. The narrative analysis of protest events accounts aimed at teasing out the various ways recurring forms of protest interactions were represented and identifying stories generated from this contestation. 

​