We presented a paper called, “Reconstructions of the Covid-19 Protests in Poland: Using Text, Pictures, and Video Data to Derive Visions of Social Change.”

See also: NCN Grant on Pandemic Protests in Poland: Presentation Abstracts in 2023

This presentation and the paper we submitted for the conference was funded by the National Science Centre (NCN), Poland (2021/43/B/HS6/01155) grant entitled “The Construction of Post-Pandemic Society: Covid-19 Street Protest in Poland.”

The presentation was by Anna Radiukiewicz, Institute of Philosophy and Sociology, Polish Academy of Sciences, Joshua Dubrow, Institute of Philosophy and Sociology, Polish Academy of Sciences, and Alan Żukowski, University of Wrocław & Institute of Philosophy and Sociology, Polish Academy of Sciences

We wrote this paper for discussion within the conference, “Analyzing protest in the digital age: Challenges and opportunities in combining text and video sources,” WZB Berlin Social Science Center, December 7-8, 2023. We used AI-assistants in the writing of this paper.

The purpose of this paper is to present how we extract and use audio-visual sources (A/V) to reconstruct protest events for our research grant project. From the point of view of feasibility and ethics we discuss: (1) the potential added value of the “reconstructive” function of A/V to better understand the protest; (2) the bias of media, organizers, and researchers; the ethics of collecting and archiving these data. We illustrate our approach with an example of how we used text and A/V to reconstruct a COVID-19 protest in Poland.

Covid-19 Protests in Poland: NCN Grant Project

As the COVID-19 pandemic challenged and pressured democratic institutions and social relations, it generated protests worldwide. Protests, in our view, are attempted social (re)constructions of reality. From this perspective, we view street protests as visions of a future, post-pandemic society.

Our research question is: In Poland, what are, and what drives, protesters’ visions of post-pandemic society?

We define visions as imagined futures and perceive them as empirically observable paths that are the perceptual basis of the coming social change. We reconstruct protesters’ visions of post-pandemic society from our knowledge of what happened during the protest.

To address our research question, we need to reconstruct what happened — context, actions, interactions, demands, and spaces — during the protests. We examine only COVID-19 protests, i.e. those protests whose goals are explicitly and primarily about the pandemic. In this effort, we are building a qualitative dataset that consists of the near-universe of pandemic-specific protest events in Poland. Ours is a text-based, qualitative expansion of ACLED’s (Armed Conflict Location & Event) special protest event dataset called COVID-19 Disorder Tracker. From ACLED, we have a list of protests and some leads as to when, where, what happened, and who was there, and media sources of information. We seek to build a simple, effective, and low-cost way to reconstruct what happened during the protest in order to obtain the protesters’ visions of society.

To reconstruct protests, we focus on:

● Demands and reasons (What and Why)

● Participants (Who)

● Performance: actions and interactions (How)

● Location/place (Where)

In our methodology, we adopt a multimodal approach, i.e. the use of multiple data sources to comprehend phenomena. Multimodality, in scientific parlance, denotes the amalgamation of diverse types of data sources to gain a more comprehensive and nuanced understanding of the event. In the realm of social sciences, this approach resonates with the notion of “triangulation,” which encompasses the use of more than one method or data source. The strength of a multimodal approach lies in its capacity to furnish a more precise and comprehensive understanding of events, thereby potentially mitigating biases or constraints arising from reliance on singular data sources. By harnessing a spectrum of data streams, we aim to construct an accurate narrative, thereby enhancing the robustness and depth of analysis.

Our textual data come from mainstream news articles and statements from the social movement organizations that organized the protest, when available online. The text provides insights into the narratives, emotions, and specific details related to the protest. Many mainstream media outlets provide pictures of some kind, and photographs of people, banners, and placards are a visual snapshot of protest moments. Videos capture some details and dynamics of the event – the movement of crowds, interactions between protestors and authorities, the progression of the protest, and any incidents that might occur. Videos also capture sequences that might not be evident in still images. Video is usually accompanied by audio. The audio track can provide additional layers of information, such as speeches, chants, songs, announcements, and ambient noises. This can be particularly important in understanding the sentiments and motivations of the protestors, as well as the reactions from other stakeholders. 

We recognize and address potential biases from media sources, organizers, and our own research perspectives, aiming for a balanced and rigorous analysis. Biases from the media, protest organizers, and ourselves are a significant concern. Journalistic practices and the prevailing media system impact the media’s portrayal of protests and protesters, which can lead to negative or sensational depictions of protest events. This bias in media reporting, combined with the potential biases of the organizers who may selectively highlight certain aspects of the protests, and the researchers’ own perspectives, presents a significant challenge in constructing an unbiased, holistic view of the protests. We endeavor to be critical of all sources of information to counter these biases and achieve a balanced analysis of the protests.

Analyzing audiovisual (A/V) materials poses inherent challenges due to time-intensive processes and uncertain reliability. Relying solely on A/V content may engender a biased perspective of protests, as textual resources offer comprehensive insights not attainable through A/V sources alone. Nevertheless, A/V materials remain valuable, offering unique perspectives and augmenting our understanding of protest events. While A/V content enriches our comprehension of protest occurrences, its limitations underscore the indispensable role of textual analysis in providing nuanced and multifaceted perspectives.

The paper also discusses the ethical implications of collecting and archiving data from protests, highlighting the tension between preserving historical events and respecting individual privacy and consent. We underscore the necessity of balancing the protestors’ “right to be forgotten” with the public interest and the advancement of knowledge, especially considering legal protections like GDPR. It also explores the feasibility of data collection within ethical, time, and resource constraints, advocating for anonymization to mitigate privacy concerns while enabling scientific research and historical documentation.

We are preparing an article based on the presentation to be submitted to an academic journal.