How States Remember their Past: Conflicting Representations of Political Violence in Italy, 1969-2020

How do liberal democracies deal with their controversial pasts? How do institutions remember political violence that threatened the democratic order? This dissertation engages with these questions by examining the production side of collective memories of political violence in Italy (1969-2020), focusing on the political class and mainstream media. The central concern of this work is understanding the meaning-making process of institutional memories of far-right bombings (“stragi”) and the mechanisms that drive changes (or lack thereof) in these collective memories over time. To address this concern, this study analyzes the official memories, in the form of narratives promoted by the political class and newspapers, of several far-right bombings that occurred in Italy between 1969 and 1984. It provides a historical and sociological account of these events, arguing that their memories and legacy are still well alive in Italian society, but that they are conflicting and confusing. Narratives vary across generations and are characterized by judicial, political, and moral controversies. My main thesis is what I define as the “de-memorialization” of controversial events. I use this label to indicate situational contexts where the past is neither deliberately suppressed nor fully remembered by society’s legitimate authorities, but rather is left continuously open to multiple interpretations, resulting in confused memories. In the case of Italy, I argue that “de-memorialization” has been displayed by creating a void, defined as the absence of a State narrative; a void that was filled with noise, in the form of continuous, contentious public interpretations about this period of violence that kept it present in national discourse without coming to closure. A range of narratives, including radical and conspiratorial ones, have occupied the discursive space. These “unregulated” memories have facilitated and amplified the already existing confusion around several events of violence, especially among the youngest generations, pushing toward a collective forgetting by rendering every bombing that entered the institutional discourse as equally complex and unsolvable, but also equally irrelevant. Beyond Italy, I argue that this concept helps explain cases of contested past events within democracies where multiple interests overlap and where actors appear stuck in perpetual fights over time. This dissertation contributes to current literature on the State and collective memory by proposing a framework that goes beyond the dual opposition between the State as a producer of hegemonic narratives about the past and local communities as sites of resistance and counter-memories. Departing from widely held scholarly and public understandings, I conceptualize the Italian State as a relational and multivocal institution, which, as I argue, failed to produce a dominant narrative on key events of violence. This research provides substantial methodological advance to the field of memory studies by combining computational and qualitative approaches to the analysis of large amounts of textual data. Data include three primary sources: newspaper articles from the three leading Italian newspapers (about 44,900 articles after data cleaning), reports from the meetings of special parliamentary commission (256 meetings and 19 reports, about 10,000 pages), and 18 in-depth interviews with politicians, journalists, magistrates, and members of victims’ rights associations. Methods comprise unsupervised and semi-supervised learning for text classification, dictionary approaches, and qualitative analysis. Keywords: collective memory, narratives, heteroglossia, events, conspiracy theory, computational social science, natural language processing, NLP, LDA, political violence, democracy, far-right, political extremism, Italy, years of lead, stragi, Piazza FontanaRelease after 08/03/202