Patterns in Commenting on the YouTube Videos of Alexey Navalny

The final speaker in this I-POLHYS 2024 session, and indeed the symposium overall, is Aidar Zinnatullin, who shifts our focus to Russia. This will examine the period in Russia before the full-scale invasion of Ukraine (from 2015 to 2021), when it was already a depoliticised society under authoritarian leadership and political stability was the central mantra of Putin’s rule. The implied social contract here was to provide increased prosperity for the people as long as they did not become politically active.

The Russian opposition under Alexey Navalny managed to cut through this stasis by producing popular and engaging content and positioning Navalny as an alternative model of political leadership; they especially harnessed dissent towards the political elite for their corruption and kleptocracy. For audiences, connecting with Navalny’s message was often driven by periods of high public interest in him (connective effervescence), but such content might also attract incivility, toxic discussions, and hate speech. Such incivility can also foster community bonds between participants, and strengthen the solidarity between individuals on the same side of a political divide; however, it may also drive away other potential commenters.

Aidar studied this by gathering some 8.9 million comments from 1.8 million commenters on Navalny’s YouTube channel. Of these, some 5.3 million were top-level comments without a thread, and 580,000 were top-level comments that did start threads (and collectively attracted another 2 million comments). Commenters were also distinguished by whether they began commenting during times of especially high interest in Navalny (measured by the number of hits on Navalny’s Wikipedia page) or not.

The four most discussed videos generated some 23% of all comments; the top 10% of videos contributed some 50% of all comments – this indicates a high concentration of comments on a limited number of videos. In-thread comments are especially strongly concentrated on highly discussed videos. One-off commenters came in especially during periods of heightened attention, and there is a typical long-tail distribution of commenting activity across commenters. Commenters who appeared first during low levels of interest in Navalny were more likely to stick around.

Uncivil comments were more likely to start discussion than civil ones; in fact, according to standard automated toxicity assessments, there is a specific level of elevated but not extreme toxicity which produces the greatest number of comments. It would be interesting to extend this study to commenting on Russian platforms like VKontakte or Telegram, however.