In this paper, we study tweets with the hashtag #metooindia, towards understanding the #metoo phenomenon as it has unfolded/unfolds in India. This study by Nayantara Ranganathan (researcher and lawyer working on the politics and culture of technologies) with support from data-analyst Piyush Aggarwal, was commissioned by All Women Count-Take Back The Tech! (AWC-TBTT!) project at the Association for Progressive Communications Women’s Rights Programme (APC WRP). Through quantitative methods, we examine the possibilities, limits and contradictions of studying a movement through a dataset of tweets centred around a phrase.

The hashtag #metooindia and the public and private conversations that it sparked assembled widespread righteous anger around the prevalence of rape culture and misogyny in India. People’s disillusionment with the “due process” of law opened up a collective space for personal testimony and validation. But such a collective eruption quickly called forth many other reckonings: of complicity, of punitive justice, of a lack of significant participation beyond upper caste Hindu women’s lives and professional worlds and so on.

As much of the phenomenon unravelled on social media, the movement is also inextricably tied up with the nature and function of specific corporate American companies like Twitter, Instagram, Facebook etc. This study proceeds with the understanding that the design and affordances in the current platforms are intentional and not inevitable, and have shaped the public around #metoo, while acknowledging that users nevertheless often navigate platforms subversively and with personal agency; the intentions of users cannot be overdetermined by platforms. The study also tries to locate #metooindia within a time and place, and not simply as a satellite of an abstract global #metoo movement.

This study proceeds with the understanding that the design and affordances in the current platforms are intentional and not inevitable.

#metooindia: a short history

The phrase and hashtag #metoo called attention to the structural problem of sexual harassment and abuse. The phrase itself has its roots at least as far back as 2006, with African-American activist Tarana Burke using it towards breaking the silence and isolation around racialised sexual harassment and abuse. More than a decade later in 2017, Alyssa Milano, a Hollywood actress, urged people to use “me too” to call attention to the prevalence of harassment and abuse in the film industry. Since then, the hashtag #metoo has gone viral, and grown and spread to challenge the limits of legal and procedural frameworks of redress for sexual harassment and abuse across the world. The extension of the generalised #metoo campaign to local-historical contexts has also been visible in how people have adopted country associations (e.g. #metooindia, #metooethiopia, #metookenya etc.) to frame their activism.

While fights against sexual harassment and abuse have their own histories of activism across the world, #metoo seemed to find resonance, convened old and new voices, and brought dark secrets of entire industries to light. Through testimony and personal narrative, the hashtag seemed to spread across the world and emerge within different cultural and geographic contexts. Through the momentum of the hashtag, people sought to disrupt a culture of tolerance around harassment and abuse. Its successes and failures have been fiercely discussed in public and private spaces, and to some extent, have challenged the understanding of the neatly divided realms of public and private.

Initially, the movement centred around experiences of harassment within workplaces and industries: film, sport, journalism and so on. Over time, through its wider appropriation, the hashtag also came to be used to testify and narrate everyday experiences beyond formal institutions and fields. As elsewhere, #metooindia was captured by the voices of upper-class, upper-caste working women while propagating easy binaries of oppressor-oppressed. However, already existing campaigns and movements against sexual harassment sought to use the momentum to talk about the failures of formal mechanisms to address routine harassment in, for example, garment factories and within domestic work.

As elsewhere, #metooindia was captured by the voices of upper-class, upper-caste working women while propagating easy binaries of oppressor-oppressed.

The questions raised by #metoo as a feminist strategy to address sexual harassment are several: What happens after the testimonies and naming? What kinds of solidarities has #metoo been able to forge and what faultlines have become clearer? As many postcolonial feminists have argued, “women” is not a homogenous category, and often does not capture intersectional experiences along the lines of caste, sexuality, class or disability. Other criticisms of the movement point to its centring of identity politics and heterosexuality in a totalising manner,1 its “individualist methodology” that does not resonate with structural forms of inequality2 and so on.

Dataset

Our starting point with the study was the hashtag #metooindia on Twitter.

At the time of data collection (June 2020), we decided to collect all tweets with the hashtag, dating back to the beginning of 2017. The dataset was collected on 22 June 2020. The tweets were collected by trawling Twitter’s advanced search.3We discarded the possibility of collecting data from the Twitter API. This would only allow collection of tweets of the seven days before the time of collection, using their free plan. Twitter’s advanced search allows Twitter users (account needed) to collect strings for specific periods.

This collection yielded 21GB of tweet data. For every tweet containing the hashtag #metooindia, we collected all the fields of information available about the tweets.

Analysis

Our analysis begins with the data made available by Twitter: content of tweets, number of retweets, likes and replies to the tweets and some information about the tweeting accounts like username and number of followers. Even as these fields of information may not always be the most important facets of conversations, we undertook a basic exercise in quantitatively interpreting text and language, and drawing conclusions from their volume.

Dominant narratives and personalities

The hashtag #metooindia is a site of convergence for people narrating their testimonies, people threatened by stories, news reportage, memes, academic publications and opinions. Co-located hashtags, or those most commonly appearing alongside #metooindia, reveal popular interpretations of the strategy and its arrival in the Indian public sphere. People tweet about #metooindia on a spectrum of allegiance: in positive terms (#metoomovement), more ambivalently (#metoocontroversy, #metoodebate) and often negatively (#metooliars). Other hashtags capture some of the themes underpinning the moment: a shift or change in business-as-usual (#TimesUp).

Names of perpetrators and survivors populate the dataset. Easily quantifiable artifacts like hashtags show that tweets with perpetrators and survivor names are anchors in reportage and discourse. However, unsurprisingly, these markers are dominated by “high profile” cases of politicians, entertainment personalities, etc. #MJAkbar, former member of parliament and cabinet minister, #AlokNath, an actor famous for playing a righteous and religious patriarch, #TanushreeDutta, a Bollywood actress, #SeemaSapra, a lawyer are all top hashtags appearing alongside #metooindia. Apart from the influential list of names, tweet data confirms some of the tendencies noted by critics about #metoo: that it has come to focus attention on individuals, even while uniting many voices attesting to the experience of harassment.

Much like Hollywood’s prominence within the American #metoo discourse, testimonies around Bollywood dominate conversation in volume. #Bollywood appears not only as one of the top hashtags, but most top tweets are also around Bollywood personalities. One possible reason is the high uptake by entertainment and gossip news sites of stories involving celebrities from the industry. Other industries like #journalism, disciplines like science and technology (#metooSTEM), and workplaces (#SupremeInjustice referring to the Supreme Court) are also visible in the tweets. There are likely many tweets around harassment in other industries and fields, but they are not rendered legible through hashtags data.

Tweets that have received over 500 retweets: mostly Bollywood related testimonies(link is external). Hover over the tiles to see the full text of the tweets, and other details around it. The bigger the tile is, the greater the number of times it has been retweeted.

Co-located hashtags.(link is external) Hover over the tiles to see the most popular co-located hashtags. Bigger tiles indicate a higher number of times the hashtag has appeared alongside #metooindia.

Men’s rights activists, misinformation brigade

Accounts identifying authors as men’s rights activists (MRA) advocate for eliminating laws that supposedly demonise men, with hashtags #LegalTerrorism, #MenToo, #MeTooLiars, #FakeCases etc. Several of these tweets list helpline numbers for men to contact when they have been falsely accused. These tweets include memes perpetuating unsurprising tropes about feminism: that women trick men into sexual advances, that feminism advocates for mistreatment of men and so on. Very often, entire tweets are simply hashtags appearing together with #metooindia suggesting victimisation of men by the movement (#malesuicide, #husbandsuicide, #metooliars). A large number of these tweets are discussing merits of cases where they involve Bollywood or other film industry personalities.

Accusations have spanned the entire political spectrum, but communal or polarising gains have also been attempted in the interpretation. Testimonies about the sexual violence by Bishop Franco Mulakkal in Kerala have been co-opted by members of the political right, to discredit Christian communities. Hashtags like #francomulakkal and #bishopfranco appear with #urbannaxals, feeding the narrative that discredits minorities and dissenters under the label of “urban naxals”.

For instance, political inclinations of accused persons (#AnuragKashyap) who are read as centre or left-leaning are invoked to question why Kashyap was not held to account in the same way as persons who support the ruling party BJP. Tweets with the hashtag #NanaPatekar claimed that the testimony against him by Tanushree Dutta was because of his support of BJP politician Nitin Gadkari. These tactics show an intentional co-option of the #metooindia hashtag by agents of misinformation to channel the discomfort around testimonies towards strengthening certain political perceptions.

These tactics show an intentional co-option of the #metooindia hashtag by agents of misinformation to channel the discomfort around testimonies towards strengthening certain political perceptions.

Controversial methods and feminist movements

Certain methods that were associated with #metoo in India had already been harshly denounced in 2017. When LoSHA, or List of Sexual Harassers in Academia, was created by Raya Sarkar, their methods were questioned. LoSHA was a list that collected accounts of harassment by influential men in academia. Accused of vigilantism, LoSHA was met with a backlash from several people, most notably by people identifying as women and feminists, for naming sexual harassers “with no context or explanation.”4 Others note the significance of LoSHA for its tactics of subversion, for the questions the episode raised about homogeneity and stasis of identity politics, and the transnational nature of academia and the digital public sphere.5 “Such networks have existed for as long as sexual predators have, but this instance was quickly understood to be replacing judicial mechanisms with vigilantism,” writes Morais dos Santos Bruss.6 “The anxieties about such a ‘viral’ object verbalised by upper-caste (savarna) Indian feminists inadvertently reveal and repeat historical anxieties about caste and a non-savarna subaltern national authenticity that queered the politics of identity in the post-colony.” Very few tweets draw intentional connections between #LoSHA and #metooindia.7

In 2017, no tweets using #metooindia were accompanied by #LoSHA. The higher volume of tweets in 2018 containing references to #LoSHA mainly draws attention to articles written about the movement. Some note that the caste and gender (queer) identification of Raya Sarkar led to the difference in reception between #LoSHA and later #metooindia, and also that it “laid the groundwork”8 for the uptake in #metoo in India in 2018.One of the main lines of attack against LoSHA was the strategy of anonymity/pseudonymity that has now become common with #metoo mobilisations across the world. In #metooindia tweets, women/people offering testimonies were often anonymous, but so were people attacking#metooindia. Accounts where the identity is either withheld or unclear exist across the board. One account undefined seems to aggregate anonymous testimonies of sexual harassment and violence, while remaining an anonymous account without any explicit allegiances or authorship claims. An Instagram account undefined sought to anonymously relay testimonies about harassment in the Indian art scene, faced a criminal complaint for defamation by the artist Subodh Gupta. The anonymity of the account as well as the survivors were both matters of contention before the court.

“Should the survivor be allowed to remain anonymous and have her story reported or should news reporting be restrained until she reveals her identity?” asks a report by the Internet Democracy Project.9 The failures of institutional and formal processes have triggered cynicism around the law and a need for other avenues for processing events, and bearing witness. Anxieties around anonymous accusations raise the question whether these discussions are only thought of as legitimate when the accuser’s identity is known and verified, regardless of whether this poses a risk to their well-being, job security and even physical safety.

Law and due process

In India, progressive developments around sexual harassment and abuse have come about through the subcontinent’s own struggles and landmarks. “Almost every legislation for women in India was passed after the death of a woman, a widely reported rape case or a mass agitation that shook the country” writes Ashwaq Masoodi reporting on significant events that changed the course of the law.10 The course of these developments has, for example, shifted what constitutes rape in the eyes of the law (Tukharam case), what material evidence is needed to corroborate the survivor’s story and where the burden of proof lies, what responsibilities lie with employers to avoid sexual harassment at the workplace (Bhanwari Devi case) and so on. More recently, the Nirbhaya case of 2012 has prompted calls for lowering the age for prosecution of accused, increasing punishment for rape etc. However these developments in the law have not been accompanied by implementation11 or by any significant shifts in attitudes, and so raising the question of the limits of the law within Indian societies that are checkered with gender-based, caste-based discrimination. This disenchantment with due process of the law forms the backdrop of the #metooindia phenomenon.

Some of the top accounts tagged alongside tweets around #metooindia are of government authorities. Even as the movement is seen to be extra-judicial and about the court of public opinion, the tweets commonly address persons in power or institutions, or follow judicial enquiry processes that have commenced. Top accounts tagged are the Prime Minister’s Office, The Home Minister’s Office, Delhi and Gurgaon police. This includes US authorities like the Federal Bureau of Investigation, President of the United States, The Justice Department, The White House and even Donald Trump.

Cases where legal proceedings have begun are also the ones that remain the longest in the public eye. Perhaps unsurprisingly, the volume of #metooindia tweets gains momentum around court hearings relating to the testimonies. Spikes in the usage of the hashtag are visible around high profile lawsuits like the defamation cases filed by ex-parliamentarian M.J. Akbar and artist Subodh Gupta. To that extent, both mainstream media newsworthiness and court cases stir discourse on Twitter around the #metooindia hashtag.

In a society and a justice system where some women are seen as more respectable and worthy of redress than others, caste and heteronormative privilege matter.

It also helps when the testimony is made by someone who fits the normative idea of a respectable woman. In a society and a justice system where some women are seen as more respectable and worthy of redress than others, caste and heteronormative privilege matter.12

Busiest accounts through time.(link is external) The visualisation contains busiest accounts over time. This means that through the years, accounts that have posted the most number of times appear in the visualisation. Each dot represents a user account.

Limits of the analysis

It is important to hold onto the complex questions that #metoo has raised, the methods that the movement has made available even when controversial. However, attempting to study the movement through some of the standard quantitative methods leads us to the very places that are most dangerous to set in stone.

Social media search and retrieval tools

When data is collected from the search and retrieval tools of social media platforms, it is important to keep in mind the limited extent of public control over this information. Social media posts disappear from view over time. They are ephemeral and their archival is at the mercy and discretion of the hosting platforms. People use reporting mechanisms available on the platform to force tweets off the platform influencing what remains in public memory and in data.

The understanding we can glean is very much shaped by the parameters of data available for every tweet using the hashtag. The parameters of data available, as well as standard formats of data and therefore of analysis do not always help answer the most pertinent questions.

Quantitative analysis of data speaks to dominant narratives: popular narratives, popular personalities. “Collecting and cleaning the data” has meant privileging English language or dominant language discourses over the vernacular, non-English language narration of experiences. For example, we did not systematically look for #metoo written in other Indian languages (like Kannada language #ಮೀ_ಟೂ) at the time of collection or processing of the data, even though the phrase has travelled to signify this particular movement in the vernacular. Beyond that, significant gaps in the data exist for understanding sexual harassment: like the extent of abuse and harassment in the informal sector.13

Quantitative analysis of data speaks to dominant narratives: popular narratives, popular personalities.

Hashtags as starting point

Limitations of using hashtags to isolate data have been noted by researchers. Alice Marwick says: “Some hashtags do function as spaces of expression with recurring actors (Bruns & Burgess, 2011), but in other hashtags the participants do not interact with each other. Moreover, hashtags can be used for a wide variety of purposes besides identification. And the majority of Twitter users do not use hashtags, as they only appear in between 5–11% of tweets (boyd, Golder, & Lotan, 2010; Suh, Hong, Pirolli, & 20 Chi, 2010). While this can be a convenient method, it is also an inadequate one.”14

The #metoo movement in India, and the people who have been able to use digital social networks more broadly for building solidarities are women who exist within particular employment, caste and class contexts. Industries which have been infamous for the systematic prevalence of sexual harassment and abuse, like garments factories or informal workers within domestic contexts, are of course largely absent in the dataset. “Sometimes it feels like #metoo was only for those women who know lots of people on Facebook,” says a garment worker from Maddur, in a news article noting unchanged conditions of the industry.15 While garment worker groups, domestic workers etc. had already been mobilising against sexual violence at the workplace, they did use the momentum of #metoo to organise meetings to boost their demands, however their voices make their way into Twitter discourse through reportage by others.

“Sometimes it feels like #metoo was only for those women who know lots of people on Facebook,” says a garment worker from Maddur, in a news article noting unchanged conditions of the industry.

Speculating on identity

The spirit behind #metooindia, like #metoo, has been to privilege and centre the voices and experiences of women on their own terms and despite countervailing circumstances of, such as absence of judicial record of misconduct, passage of time since offence, consent given within oppressive and unequal power dynamics etc. However, a quantitative analysis of the tweets is necessarily an exercise in drawing meaning from testimonies among other content, so feminist and ethical concern and question that remains is how do we treat any understanding derived from this.

Can you fix someone else’s identity while studying the tweets? Notwithstanding sophistication of methods, what does it mean to infer gender/intention etc. from language analysis? These questions are very much related to the questions #metoo raised in the indian context: how do we unravel and start thinking about who can lay claim to the term “feminists”, even as it is also a string of letters searchable in a language dataset? Can that be something we define through the data? And before we exit the idea, what would it look like to try?