Sheyril Agarwal and Joyojeet Pal
to cite (Agarwal, S., and Pal, J. (2023) Coordinated Targeting of Creative Work on Social Media in India: The Politics of #BoycottPathaan. Online at:https://joyojeet.people.si.umich.edu/pathaan)
A week into Pathaan’s successful OTT release, which followed its blockbuster performance at the box office, it would appear that attempts to #BoycottPathaan failed to have any box office impact. The persistent calls to Boycott Bollywood arguably had as their lead target the uncrowned ‘King of Bollywood’ Shahrukh Khan. For months preceding the release of Khan’s Pathaan, there had been calls to boycott the film on social media, and the week that the film released, there were incidents of attempted protest at movie theaters. While these appeared to be organized by a range of fringe groups as well as loosely organized individuals, there have been frequent insinuations that this action is driven by supporters of the BJP government.
To examine these, we did the following:
- We collected messages that included the string “#BoycottBollywood” or “#BoycottPathaan”, analyzed their use over time
- We created an archive of “Boycott Bollywood Users” by seeking out the most frequent users and profiling their behavior
We present two thematic buckets of findings. First, we describe the discursive style of the Boycott Bollywood trend online by identifying the most common arguments put forth to attack the film industry, and the network characteristics of such attack. Second, we present findings about the accounts that have high activity related to Boycott Bollywood, and find that their political alignments are consistent with each other.
We find that the Boycott Bollywood agenda is closely tied to the idea of hitting a news cycle, and by extension, the attention economy. In other words, the accounts that engage in significant messaging around this sense that there is newsworthiness, which is both helpful in terms of getting the message out, but is also valuable to the accounts that output this material, since it gets them more attention. Ironically, this could also arguably benefit the movies being boycotted, since the trolling adds buzz to the film’s release, especially when there are teasers or trailers for the film being boycotted.
Figure 1: Daily timeline of #BoycottBollywood tweets and retweets (Link to interactive plot: https://chart-studio.plotly.com/~sheyril/85)
We mapped the tweeting activity for the shaded region in the previous blog. When we plot the same graph for a longer time period, we observe the same pattern of coordinated release attacks as before. We annotate the peaks in fig.1 and explain the events related to them.
Looking just at BoycottPathaan, we see that early Twitter activity related to the film began in August 2022, but the major activity did not take off till December 2022.
Figure 2: Daily frequency of #BoycottPathaan Tweets and Retweets (Link to interactive plot: https://chart-studio.plotly.com/~sheyril/83)
While we see a dip in activity during the later part of September and month of October, the movement was revived with a sudden surge in tweeting occurring on 2nd November owing to the release of Pathaan’s teaser. Another peak on 24th November was a result of a quote tweet by Indian actress Richa Chadha commenting on a statement by an Indian military officer, claiming they would take territory from Pakistan by force, to which she sent a snarky response about a border skirmish with China, in which China took border territory from India by force.
The most significant peak related to Pathaan happened in mid-December, when the film’s song “Besharam Rang” was released. The boycott activity here focused on the film’s female lead, Deepika Padukone’s appearance. The messages related to these tie in closely with other common attacks on Bollywood that tend to be around the appearance or characterization of female actors in the films.
Figure 3: Tweet examples for the campaign against Pathaan, highlighting two major themes – negative commentary on star Deepika Padukone using clips from the song Besharam Rang, and misinformation attributed to Gems of Bollywood that Shahrukh khan is promoting and funding Pakistan
The second high activity period, however, shows a closer integration of the Boycott Bollywood movement with national politics. The driver here was the film Ram Setu, an Akshay Kumar action film, called a nod to the Hindu right by multiple film critics. While the film got both support and trolling, the attention came in part from Kumar’s meeting with UP Chief Minister Yogi Adityanath, and the appeal from Hindi actor Suneil Shetty to Adityanath requesting the Boycott Bollywood activity be reined in. This action in and of itself was meaningful because it suggested the actor felt the politician could influence those running the troll farms. Besides Shetty, there was also a statement from the Federation of Western India Cine Employees (FWICE) urging the government to help against the ongoing Boycott Bollywood trend.
The messaging triggered by Ram Setu as well as Shetty’s engagement with Adityanath highlighted that the Boycott Bollywood movement may not be amenable to easy management even with political intervention. First, the attacks on Ram Setu took place irrespective of the Hindutva undertones of the film, since Akshay Kumar was being targeted by the Sushant Singh Rajput trolls irrespective of the content of his cinema. Second, the plea by Suneil Shetty ended up giving more fuel to the Boycott group, which took to arguing for why the movement must go on.
Figure 4: Tweets examples highlighting Bollywood’s response to the #BoycottBollywood movement and appeals to CM Adityanath
A key feature of the BoyottBollywood network is that the content is driven by a small number of accounts, then circulated throughout a network of accounts that mainly regurgitate content from these alpha accounts. We visualized (see fig. 5) the key users, whose content was central to the spreading of BoycottBollywood, and found that the overwhelming majority of accounts that were constantly tweeting about the subject were accounts with limited online following or impact, with the exception of three accounts – @Dharmes56192706 @ErayCr and @KreatelyMedia. Of these, @Dharmes56192706 belongs to an individual who has not identified themselves, offers reciprocal “following” and almost all of their content is far right leaning. However, @ErayCr and @KreatelyMedia have both been seen as key players in other trolling incidents involving right wing causes.
Figure 5: Tweeting activity of key accounts in #BoycottBollywood and their online reach in terms of followers (Link to interactive chart: https://chart-studio.plotly.com/~sheyril/94)
The work of collusive messaging to make “BoycottBollywood” trend on a regular basis is carried out through a network of users who indulge in “copypasta” activity, where the same block of text is used repeatedly to suggest that a certain idea has greater public appreciation than it actually does. We classify a group of tweets as “copypasta” if the exact tweet content is copied and pasted into separate messages 5 or more times. In our dataset, we come across 1,379 such tweet collections, with 18 users that get copypasta-ed more than 10 times. Evidence of collusion is seen in that of the 1379 cases of copypasta, 400 tweets were copypasta-ed in the same day they were originally posted, and 266 tweets saw coordinated copypasta activity within the same hour that the original was tweeted.
The second network strategy we see is the use of high-influence accounts as well as high-frequency output accounts to put out content. To examine these, we use the social network formed by using users as nodes and the retweet relationship to form edges between them. To capture relationships between two users and highly influential nodes, we apply two thresholding methods. First, we only consider users that have been retweeted more than 50 times in the sample. Second, we only consider edges where a user has retweeted another user more than two times. Using eigenvector centrality, we were able to identify top 30 users that play a crucial role in the flow of information in the #BoycottBollywood community.
Of the thus identified 30 core users, we find that only two users seem to have been established with the sole purpose of propagating the #BoycottBollywood movement, with practically no content outside of this topic. From the remainder, we find two patterns. The first is the existence of “SSRians” or the community whose social media activity revolves primarily around the film star Sushant Singh Rajput and theories around his passing. The second is right-wing influencers, who tweet on a variety of issues of interest to those on the right of the political spectrum in India. While these were the core community that drove the Boycott Bollywood movement online, a second layer of influential users emerge when we threshold the tweet frequency at over 70 and with a retweet count (>=20). Merging this set of users with the 30 users, we were left with a set of 282 users.
We pulled the tweets for these 282 users from Aug 1, 2022 to Jan 30, 2023, the week after the release of Pathaan. A visualization of the cluster of accounts whose activity is largely related to the film Pathaan shows that these are much more likely to be recently created accounts.
Figure 6: Accounts spreading #boycottpathaan content are more likely to be created fairly recently, suggesting older accounts get suspended, and a likelihood of bot activity
Thereafter, we studied the activity of the core set of 282 users to see what else they do online besides engaging on topics related to Bollywood. We visualized the accounts that were most retweeted by this core set.
Figure 7: Times core #BoycottBollywood users retweeted top users from their following network where node sizes indicate number of followers (Link to interactive plot: https://plotly.com/~sheyril/96/)
Figure 7 is an important indicator of the spread of authority in the BoycottBollywood network. A high authority score (shown on the x axis) indicates that the user has a significant following from the BoycottBollywood community while the size of the bubble represents the size of the user’s overall followers. Here, we see the accounts most followed and retweeted by 282 accounts at the core of BoycottBollywood activism. We see @narendramodi has the highest “in-degree” value, which means that his account is followed by the highest number from among these accounts, and thus has a very high authority score among these users. While several of the high-authority accounts belong to political accounts (@myogiadityanath, @amitshah, @sambitswaraj, @bjp4india, @KapilMishra_IND) or commentators (@ARanganathan72, @vivekagnihotri) associated with the BJP, among the top 10 accounts is that of @elonmusk.
We see that the three top accounts most retweeted are @GemsOfBollywood (which is by far the most influential in terms of the total presence it has in the output of these 282 core Boycott Bollywood users) @opindia_in, and @KreatelyMedia, all of which are generally associated with pro-BJP or far right content.
Next, to understand the individual targets of the BoycottBollywood messaging, we visualized (see fig. 8) the most frequent hashtags without the words “boycott” and “bollywood”, and the visualization clearly shows that Sushant Singh Rajput remains the driving force behind this community. We see the key individual targets include Shahrukh Khan, Deepika Padukone, Shalman Khan Kareena Kapoor Khan, Richa Chadha, Karan Johar, and Ranbir Kapoor, all of who are regularly targeted, but also Akshay Kumar, who has generally been seen as a pro-Hindu actor. Besides Pathaan, films most attacked include Laal Singh Chaddha and Bhrahmastra, however, the interesting thing is that we see a fair share of marketing activity – particularly of individuals associated with the Big Boss series, a reality television show that often uses online influencing to get viewers interested.
Figure 8: Word Cloud of most frequently used hashtags (excluding “boycott” and “bollywood”)
Specifically, one of the questions long lingering was whether the attempts to Boycott Pathaan were even more specifically galvanized around Shah Rukh Khan’s Muslim identity. To examine this, we examined the sub-themes in messaging about Boycott Bollywood that went viral.
To understand what elements of BoycottBollywood content go viral online, we annotated the 200 most retweeted messages in the network alongside a random sample of 200 original tweets from the network. We manually coded two collections of tweets: 200 most retweeted tweets and a random sample of 200 original tweets. To detect the different kinds of agendas, we make use of 7 categories. Since tweets can have overlapping themes, these categories are non-exclusive. The annotated set of 400 tweets is available here: coded_400
Our findings show that subcategories that have thematic intersections with a second area of interest are much more likely to make a tweet about Boycott Pathaan go viral. Intersectional propaganda, defined as having a second line of conceptual framing that aligns with a propagated partisan position in the public discourse. Thus, a notion of Muslims as incompatible with the idea of patriotic sensibility for India, or westernized women as incompatible with the normative understanding of Indian society are examples of such partisan positions, and both would be considered propaganda, as they are typically propagated through networks through methods of coordinated messaging.
We find that propagandist content is much more “effective” in terms of numerical outreach. We find that in the highly retweeted set, 82.5% tweets identified as propaganda as compared to the random sample in which 46.5% of tweets qualified for the same, and we find this result to be significant (T-test, P ≤ 0.05). So our first conclusion here is that at the top end of the spectrum, i.e. the most retweeted messages about Pathaan, some co-located element of intersectional propaganda is more likely to be present.
We further break the highly retweeted messages into categories and subcategories of the accompanying theme in the tweet, the results are in table 1 below:
First, we see that Religion and Nationalism are the two overwhelmingly large categories of intersection. The basic argument for the Boycott Pathaan messaging is either to attack one of the core characters for their religion (ie Shahrukh Khan as a Muslim) or to present the overall idea of patronizing the film as anti-Hindu. On Nationalism, the majority of articulations tend to be around abstract notions of the film industry as being filled with anti-nationals, loosely defined as associated with organized crime, or undermining the pride of India by criticizing it.
An important construct within nationalism is the notion of culture, and by extension, a definition of what is legitimate and illegitimate culture. We find that language is an important element of culture – thus the term “Urduwood” is frequently used, suggesting that Urdu speakers are by definition culturally incompatible with India. Interestingly, in the last several years, the use of Urdu script to introduce a Hindi film has declined dramatically, suggesting that filmmakers want to avoid ruffling feathers on this issue.
The second related element of cultural exclusion is the use of misogyny. The use of misogyny is in and of itself used as a means of suggesting cultural incompatibility, especially with an abstract notion of Indian tradition. The last intersecting theme is that of SSR. Although the SSR community itself is central to understanding the drivers of trending activity around Boycott Bollywood and Pathaan, it does not have the broad-based appeal of engagements with nationalism or religion. The SSR community is largely seen as a fringe, conspiracy theory community, especially since the mainstream media has stopped engaging the case. Consequently much of the Boycott Pathaan retweeting that co-occurs with SSR is almost entirely engaged by communities that also otherwise propagate SSR content.
We find that toxic content generates more reaction from users. We verify this by comparing engagement metrics like likes, retweets, quotes and replies between toxic and non-toxic tweets for the combined sample of 400 tweets. To do so, we ran a T test to check whether there was a significant difference in engagement when the tweet contained inflammatory content. We see a significant increase across all metrics in cases where we detected some kind of toxic content.
Additionally, on average, toxic tweets receive twice as many likes and retweets as compared to non-toxic tweets. Despite toxicity in their content, tweets rarely get ratio-ed with a mere 2.5% toxic tweets receiving more replies than retweets.
Finally, a commonly used argument against Bollywood is its history of nepotism, which indeed was a central part of the SSR community’s grouse against the industry. We see in fact that the driver of the BoycottBollywood agenda is not in fact nepotism, which is very miniscule as a share of messaging, as compared to religion. Fig. 9 gives some insight into how some major themes of discourse have changed based on how often they appear in a week. While discourse around the film industry and their depictions of Hinduism has been prominent since the beginning, a shift in focus away from nepotism can be seen in the visualization. Interestingly, the period in which this shift takes place coincides with the rise in Pathaan related tweeting.
Figure 9: Weekly frequency of Tweets based on different narratives in the Boycott Bollywood movement
Attacks on Deepika Padukone were a key part of the social media engagement, although we found that the negative messaging extends beyond the Pathaan star to misogyny more broadly. While most of the misogynistic messaging aimed at her happened in December 2022, there are attacks on Kareena Kapoor and other women who have married Muslim men.
To better highlight the nature of misogyny in Pathaan related posts, we isolate and analyze 12,636 tweets from the entire dataset that mention Deepika Padukone. By clustering the commonly found words, we built thematic blocks that best describe the framing, and found two most common notions – body shaming and moral policing.
We use the following keywords to capture these two themes:
- Body/Slut Shaming: bikini, body, nudity, clothes, softp0rn, porn, cheap, shame, vulgar
- Moral Policing: toxic, values, culture, india, hindu
We found that “Moral Policing” was the largest cluster, with 1,306 tweets focusing on how various components of the film and its songs have a bad effect Hindu culture and Indian values. Another major theme of discourse in this category was “love jihad”, with tweets linking it to the religious identities of Pathaan’s lead actors. Following this, Body/Slut Shaming was present in 1,152 tweets, with most tweets making personalized attacks on the actress. These types of tweets also ranked quite high in toxicity owing to the kind of language used to frame these attacks.
Figure 10: Sample of tweets related to Boycott Pathaan that show misogyny directed towards Deepika Padukone
We also see tweets that are misogynistic but not specifically directed towards Deepika. A major theme here is the mention of “Love Jihad”, which in this case is also a reference to the Muslim male star Shahrukh being paired with (the Hindu) Deepika. These add a direct dimension of anti-Muslim prejudice intersecting with the misogynistic messaging.
Figure 11: Examples of misogynistic content with underlying anti-muslin themes.
Several of the hashtags used by the movement are fundamentally misogynistic in their framing. These include #KHAAN_GANG_TBKC and #KHAAN_GANG_TMKC, wherein “TBKC” and “TMKC” are acronyms for Hindi language abuses aimed at women. Another misogynistic hashtag is “#रेंडीबाज_बॉलीवुड” which is often accompanied by images that objectify women.
In conclusion, #BoycottPathaan may have scared a small number of people from showing up at movie theatres for the fear of incidents, but failed to get any negative outcome for the blockbuster collections of the action film itself. Ironically, the film is crafted as standard hyper-nationalistic outing, with a vilified Pakistan, selfless members of the military fighting to the death for the nation – quite unlike the romantic fare Shahrukh Khan typically stars in. However, the collections aside, Boycott Bollywood undermines Shah Rukh Khan or Deepika Padukone as heroes for a young generation.
In this sense, what we see with the #BoycottBollywood trending is less about the individual targets of the film industry and more about undermining elements of what the industry stands for. The attacks on Shahrukh Khan and other Muslim lead performers offer a means of invisibilizing Muslims from spaces of authority in the public sphere. They simultaneously offer a way to ensure the narratives emerging from the content in the film industry align with the hegemonic interpretation of nationhood.