Feminist Encounters: A Journal of Critical Studies in Culture and Politics, 7(1), 08

This article examines Twitter publics to map how the ‘ dadis of Shaheen Bagh’ (grandmothers of Shaheen Bagh) emerged as political subjects through transnational media space even though they themselves did not directly access social media. A team of feminist media researchers examine how social media networks were mobilised strategically to gain international visibility and traction. Through a feminist close reading of Twitter data and a select few in-depth and unstructured interviews with various associated actors, this mediated visibility of seemingly subaltern women is mapped. This article draws on transnational/postcolonial feminist frameworks to examine digital public presences. This study uses a multi-methods approach that includes qualitative interviews with activists (local and transnational) and related actors as well as a situated feminist data analytics and critical digital humanities approach to examining big social data online. In examining the mediated production of this visibility, however, the study does not wish to imply that the women whose presence is amplified in international media are not actual protestors or to deny, discount or appropriate their agency or labour as activists.


INTRODUCTION
The three quotes above frame our investigation. In this article, we examine Twitter publics in an attempt to map how the Muslim 'dadis of Shaheen Bagh' (grandmothers of Shaheen Bagh), such as Bilkis Dadi 2 , emerged as political subjects through transnational media space even though they themselves did not directly access social media. Our interest here as feminist media researchers is to shed light on and raise questions about how women of the community are highlighted as active protestors through the formation of social media-based affective networks (Papacharissi, 2015). In examining this phenomenon, we look at how the visibility of women as protestors expands the utility and scope of social movements in varied ways. We note how the presence of women at protest sites has directly or indirectly been repurposed via Twitter while making 'identity-based cultural and political demands' (Jackson, Bailey and Foucault-Welles 2020, p. xxv), with hashtags becoming the lingua franca of this phenomenon.
As Srinivasan (2013) who researched the 'Arab Spring' as an ethnographer on-site observes, responses to social media used in protest movements tend towards either 'distortion' or 'deification ' (p. 52) in characterising the use of new communication technologies in protest spaces. Zeynep Tufekci, another researcher of the Arab Spring, has argued that we need to study the how of networks to move past an inappropriate debate of binaries that ignores the power of social context around technology use and its appropriation (Tufekci, 2011). In this piece, we attempt to move beyond this binary, studying both the mediatisation of the Shaheen Bagh protests and the symbiosis between online and offline feminist protestors through analysing hashtags. Social media use in activism was comparatively new in 2011 when Tufekci wrote, but now in 2022, we have seen the impact of the Black Lives Matter movement and can hardly discount the importance of social media use in activism. However, if we think of social media as similar to political pamphlets, newsletters, and other forms of media used by activists, we begin to see that these are just tools but with their own unique yet limited affordances to help build networks with great speed.
We are a team of researchers some of whom are part of a research team at Bowling Green State University where we have been working to examine the use of computational tools and data analytics software through a critical humanities lens. The team was started for graduate students working with Radhika in 2019. Co-authors Emily Edwards, Debipreeta Rahut and Ololade Faniyi are members of this team. Other co-authors of this article joined us in this investigation as we started to do interviews for the current paper. Dyuti Jha, Aiman Khan and Jhalak Jain were in Delhi at the time of the Shaheen Bagh protests and were able to provide us with onsite information and contacts for us to interview. Saadia Farooq joined us while we were finishing up some drafts of writing for the overall project looking at the data from Shaheen Bagh because of her interest in Muslim feminism in South Asia. While Jhalak Jain, Aiman Khan and Dyuti Jha provided us with onsite contextual information, Dyuti Jha and Saadia Farooq advised us on the different histories of Muslim feminisms in India and Pakistan, thus allowing our team to work on this project without making broad generalized assumptions about Muslim feminism. The information we got from those we interviewed on site, as the rest of us did a close examination of the data from Twitter and Instagram, allowed us to understand the geographical specificity of Muslim feminism in India vs other regions in the world. We have listed co-authors to reflect their contribution to the writing of this article. The first four co-authors (Radhika, Emily, Debipreeta and Ololade) are responsible for majority of the writing in this article.
Overall as a team, we have been watching how Twitter hashtags have promoted the role of women protestors on-site in the anti-CAA/NRC protests. The anti-CAA/NRC protests refer to the national and transnational series of protests in response to the Citizenship Amendment Act (CAA) enacted by the Government of India in December 2019. Along with the CAA, the National Register of Citizens (NRC), an official record of legal citizens in India, was also proposed. The list would include undocumented migrants of Hindu, Sikh, Jain, Parsi, Buddhist, and Christian religious backgrounds from Pakistan, Afghanistan, and Bangladesh who entered India before 2014. However, Muslim undocumented migrants were conspicuously missing from the NRC list. Protests soon followed, decrying blatant religious discrimination. A ban, as well as the scrapping of the NRC, was demanded (Vishwanath and M. Sheriff, 2019).
In this article, we retrospectively examined the social media terrain of the protests by looking closely at smaller sections of larger datasets through the use of data analytics tools. In the process, we see how geographically and politically dispersed offline users are linked under particular hashtags and how the network is shaped through algorithmic logics that users have no absolute control over. Although many social media users and digital activists attempt to 'make' a hashtag trend, the mechanics of the algorithm are opaque and often uncontrollable (Edwards, et al 2021). In looking at the data we had collected around the events in December 2019 and then in January 2020, we found tweets that cross-referenced the #shaheenbaghprotests #dadisofshaheenbagh #womenofshaheenbagh © 2023 by Author/s 3 / 15 and #shaheenbaghdadis. All hashtags referred to protests led by Muslim grandmothers of Shaheen Bagh, a community in South Delhi, who were portrayed as keepers and protectors of the home that was under threat with the passing of the CAA/NRC protests which gained momentum during the period between December 2019 and April 2020. These hashtags were reflective of the strong offline visible presence of women from the community who were on the streets protesting. Overall, the dadis (paternal grandmothers) of Shaheen Bagh seemed to draw more significant attention internationally than anti-CAA/NRC protests featuring Muslim men or female Muslim students (Kadiwal, 2021).
In the current article, therefore, we map how this mediated visibility of subaltern women protestors on-site is produced through digital publics. We do this through close feminist readings of Twitter data and a select few indepth, semi-structured interviews with various actors. Our multi-methods approach presents a situated feminist data analytics and a critical digital humanities methodological lens to examining big social data online (D'Ignazio 2020; Jackson, Bailey and Foucault-Welles 2020;Klein 2014;Rettberg 2020;Risam 2018). In terms of theory, we draw on transnational/postcolonial feminist frameworks to examine these digital public presences. We chose Twitter as our overall focus for collecting data because it is commonly used in India and is also mentioned by interviewees. Scholars have observed how Twitter has the potential to serve as a counter-public sphere (Rambukanna, 2015;Cheema, 2020).
This article also demonstrates how critical feminist and qualitative/interpretive researchers might be able to retool computational data analysis software. For instance, we show how social network analysing and visualising tools such as Gephi and Netlytic can be used against the grain of quantitative epistemologies that might flatten the analysis and create broad generalisations that erase the specificities of particular intersections of social media use. Thus, we take a 'different approach to studying voluminous digital corpora that combines the scale of computational research with the depth of [critical, feminist] qualitative analysis' (Shahin, 2016: 29). We must also note that our collaborative authorship is consistent with our methodological approach to the parsing out of the data through critical feminist close reading. We attempted to include multiple voices in the collection and analysis of the evidence. Co-authors include not only academics from the Global North but also researchers and participants on-site and who were engaged in reporting on events locally.
In Part I of this article, we will lay out the context of the anti-CAA/NRC protests and follow this by detailing our methods and methodological approach. We explain how we handled the two kinds of data we collected; qualitative, in-depth, semi-structured interviews and through data scraped from Twitter via an iterative process that allowed us to view the data as situated (Rettberg, 2020). In Part II, we discuss two main features of the movement's strategy to amplify the women of Shaheen Bagh that became evident through an examination of Twitter datasets collected from January to mid-March 2020. Thus, while the creation of visibility for the subaltern and older women protestors (the dadis) of the community contributed significantly to the overarching strategy to visualise an ethos of non-violence and the local performance of an ethic of care in the Shaheen Bagh protest space (Bhatia and Gajjala, 2020), the strategies and tactics for amplification of the protests internationally also relied on inviting transnational activist coalitions, international non-profit groups, Human Rights organisations as well as on the spontaneous efforts and participation of local and global celebrities. In this way, we see the criticality of both online and offline networks in producing the Shaheen Bagh protests as both a local manifestation of a new iteration of Indian Muslim female-led activism and a global, digital feminist protest.

PART I
The anti-CAA/NRC protests happened in several places in India between December 2019 and March 2020. The reasons for the protests and the demographics of who was protesting were sometimes different, particularly in the case of Assam. Thus, for instance, while in the case of Delhi, the anti-CAA protests were concerned with the marginalisation of Muslim citizens, at the same time, in Assam, there was intense tension between indigenous Assamese, undocumented Bangladeshi migrants and Bengali Muslims. Therefore, while our paper focuses on the Delhi protests, we must acknowledge that the anti-CAA/NRC protests have complex regional nuances. With this in mind, there has already been an excessive focus on academic writing and news reporting internationally on the Shaheen Bagh site. Why one more? We argue that this very global hypervisibility provides the rationale for our article. As with the 'Arab Spring,' the questions regarding social media-based hypervisibility centre around examining the strategic deployment of various types of media by local activists and digital tool-wielding witness allies. However, we do not argue that the protest itself was possible because of Twitter or that somehow the digital activists were more important than the local physically on-site and at-risk activists and community participants (Aday, et al., 2012;Srinivasan, 2013).
Past work on social media protests dates back to work on Occupy Wall Street's use of social media in the Global North (Juris, 2012) and Egyptian and Tunisian protest movements' use of social media in the Global South (MacKinnon, 2012; Papacharissi, 2015; Tufekci and Wilson, 2012; Srinivasan, 2013). More recently, in the past five years or so, we have seen research on the Black Lives Matter movement (Florini, 2019;Jackson, Bailey, and Foucault-Welles, 2020), on #metoo as it happened in India (Guha, 2021), the Delhi 2012 Rape protests (Dey, 2018), and so much more. All these scholarly writings have a nuanced popular perception of social media use by activists and sympathisers as either slacktivism or a democratising tool that somehow facilitates revolutions (Rotman, et al., 2011). While most past research on such social media publics amplifying offline movements is either based on offline ethnographic work or close textual readings and interviews, research by scholars such as Shahin, Nakahara and Sanchez (2021), Jackson, Bailey and Foucault-Welles (2020), Papacharissi (2015), Bruns and Burgess (2015), as well as various other contributors to the edited collection on 'Hashtag Publics' edited by Rambukkana (2015) engage more fully with the digital landscape. We also attempt to use computational tools to engage big data publics to understand not only the use of social media in activism but also the discursive terrain that such activist use produces. Activist groups are organised, facilitated, and assisted by offline community experts, and skilled technology workers are often voluntarily part of such teams. The mass of users who jump into the action spontaneously -the slacktivists and performative activists -who can consider themselves contributing to the movement may amplify the cause. This article contributes to and extends theoretical and methodological discussions initiated in the previous works through a specifically feminist epistemological framing of methodology (Alcoff and Potter, 1993), engaging with feminist and postcolonial interventions in the use of computational tools for critical research (D'Ignazio and Klein, 2020;Jackson, Bailey and Foucault-Welles, 2020;Rettberg, 2020;Risam, 2018).
There has been previous research on social media use, particularly Twitter, as with Papacharissi (2015) and Jackson et al.'s respective works (2020), and considerable work has been published on the topic of the Shaheen Bagh protests. Scholars have pointed to the visibility of subaltern women at the protest site and care as a protest strategy (Bhatia and Gajjala, 2020;Kadiwal, 2021;Mustafa, 2020;Salam and Ausaf, 2020). Some research has also begun to look at data scraped from Twitter publics while arguing for a situated data analytics approach to studying such protest publics (Edwards, et al., 2021). We apply this latter framework to our data to provide further evidence for how Twitter publics worked in sync with the offline protests. Thus, our focus here as media researchers is to examine how social media networks were mobilised strategically by offline local groups of protestors and how they were used to gain international visibility and traction on behalf of the protest movement, with the seemingly unlikely dadis, the subjects, at the forefront of the movement. Our analysis connects with Suddhabrata Deb Roy's (2021) observation that the anti-CAA movement is perhaps the only movement in modern Indian history that has witnessed such large-scale mobilisation of people from all walks of life. This raises some very pertinent questions regarding class, caste, and gender as it relates to this immense mobilisation which will surely be the subjects of many studies to come. Spaces like Shaheen Bagh are 'resistances in themselves' when we note the subaltern working-class character of the women who led the protest and thus merited additional study to illuminate how this protest, in particular, not only featured the leadership of previously marginalised female political subjects, the dadis, but also incorporated a variety of feminist actors online and offline, in India and beyond (Roy, 2021).

Context
What has now become known as the 'Shaheen Bagh movement' began as a response to the passing of the Citizenship Amendment Act, National Register of Citizens and National Population Register, or CAA/NRC/NPR which happened on 11 Dec 2019. As Aiman Khan and Madhur Bharatiya (2020) point out, the CAA is exclusionary, particularly when paired with the National Population Register (NPR) and the National Register of Citizens (NRC). It targets the Muslim population as it 'grants Indian citizenship to people of all religions (except Muslims), who have allegedly faced religious persecution in their home countries, namely Pakistan, Bangladesh, and Afghanistan' (Khan and Bharatiya, 2020). The Shaheen Bagh protests (15 December 2019 to 24 March 2020) featured a significant number of Muslim women. Some of the women interviewed on-site by a member of our larger research team emphasised that the women of the community led the organising. One of them stated, '[i]f we do not speak, who will speak for us because the majority of people are already against us and this is the time when we need to do something on our own.' The presence of a high number of middle-aged and older women was also noted. Their presence apparently created an aura of safety for younger women from the neighbourhood and women visitors from nearby colleges and universities to express their solidarity and support. The protest became further popularised through social media. Photos and videos were shared from the physical location of protests to social media with hashtags such as #womenofshaheenbagh, #ShaheenbaghProtests, #NotoCAANRCNPR and more. Sites of protests included Shaheen Bagh in Delhi, Park Circus in Kolkata, and Mumbai Bagh in Mumbai, along with several other small protest gatherings in locations such as Lucknow, Allahabad, Bangalore, as well as some locations in the state of Assam. While not all these protests shared the same concerns about Muslim marginalisation, several of the sit-in protests that happened simultaneously in other parts of Delhi, such as Gandhi Park in Hauz Rani and Maujpur, shared similar concerns. Notably, the Shaheen Bagh protests took place in the same neighbourhood as Jamia Millia Islamia (JMI) University, which witnessed police brutality on 15 December, 5 / 15 2019, as retaliation for JMI students having joined the anti-CAA/NRC protest in huge numbers. As per an eyewitness interviewed, Delhi police brutalised students without provocation and stormed into the library and hostels. They used tear gas, stun guns, canes, and bullets on students resulting in several serious injuries.
The women of Shaheen Bagh became the face of the protest movement. In light of how Muslim women are viewed globally as being allegedly oppressed by their men (Ahmed, 1992;Stabile and Kumar, 2005), the mediated visibility of the women of Shaheen Bagh as protestors created an international impact. Western media outlets and researchers began to make claims that made it seem that this expression of agency by the subaltern women of the community worked to change the story and to create an ethic of care (Edwards, et al., 2021;Bhatia and Gajjala, 2020,). The progressive framing of the dadis' protests as an 'affective strategy' thus emphasised non-violence and tableaus of domestic activities such as making tea for protestors at the site, thereby pushing back at narratives of the protests as driven by forms of anti-state violence. While these observations are not incorrect since #changethestory was intentional and the creation of a non-violent caring aura at the site was strategic, such claims need to be understood in relation to regional histories of Muslim women exerting their agency socially and politically in India.

Muslim Women's Movement in India
The political exclusion of Muslim women must be viewed through a wider context of exclusion faced by Muslims in India; despite this, Muslim women across various regions in India have been involved in protest movements visibly for a long while. Here, we only briefly map their organising in the interest of touching on the relevant points in these histories starting with the Shah Bano case. In the 1980s, a 62-year-old Muslim woman Shah Bano filed a petition demanding maintenance from her divorced husband. This case became a moment of challenge in the history of Muslim women's struggle for equality. This case resulted in the Muslim Women Bill being adopted in 1986. The bill prioritised Muslim Personal Law over criminal law on the issue of Muslim women's right to maintenance. Thus, the secular state ironically reinforced patriarchal religious control over women.
Since the 1990s, however, Muslim women's networks in India became visibly diversified. Key players in movements include the Muslim Women's Rights Network (MWRN) and the Bharatiya Muslim Mahila Andolan (BMMA). These networks created shifts in how Muslim women's issues were taken up. For instance, although the BMMA advocacy deals with the issue of matrimonial rights and the reform of Muslim Personal Laws, it also tackles issues such as the socio-economic and political marginalisation of the community (Kirmani, 2011). Both MWRN and BMMA move beyond common issues such as veiling, polygamy and triple talaq 3 (divorce). Women's rights are also highlighted by drawing attention to problems related to violence against the community and how these impact the women in the community.
While MWRN is committed to a women's rights approach, the BMMA experiments with Islamic feminist approaches by asserting that it is not Islam alone that has oppressed Muslim women, but that centuries of patriarchal interpretation of Islamic texts are to blame. Thus, the BMMA argues that standing up for women's rights does not require abandoning one's religious identity. Women are encouraged to engage with Islamic texts while claiming rights related to inheritance, marriage, and political empowerment (Kirmani, 2011). In the Indian context, 'Islamic feminism' thus poses a dilemma at the theoretical-conceptual level. On the one hand, some women dislike being associated with Islamic feminism. On the other hand, in actual practice, Muslim feminist organising has been heavily influenced by both Islamic texts as well as texts of modern feminist thought.
However, despite these shifts in Indian Muslim feminist thought over the past few decades, Indian national and international media still focus on reporting only the feminist engagements that question Islamic religious authorities. The Shaheen Bagh protests thus signalled a watershed moment in traditions of Muslim feminist activism in India in terms of both the national scale of the protests and the hypervisibility of working-class Muslim women rather than the leadership of formalised secular feminist non-profits and Islamic feminist organisations.

Methods
Our methods are interdisciplinary, following scholars such as Jackson, et al. (2020: xxxv), who found it helpful to transfer 'online counter-publics into networked data' to understand how activists mobilise digital publics. Close feminist readings of the data collected emphasise data visualisations and networks as providing partial pictures of more complex phenomena. We focus on particular hashtag datasets from January to March 2020 accessed through the Twitter Application Programming Interface (API). These datasets, around #Shaheenbaghprotests #womenofshaheenbagh and #sheinspiresus, were chosen because of their visibility in anti-CAA/NRC protest tweets. They were scraped via Netlytic and the open-source software Gephi. These two software tools afforded us social network analyser and visualiser tools to capture and transform data collected on the #Shaheenbaghprotests into concise networks showing the 'nodes' (the people involved) and the 'edges' (how they connect). On Netlytic, we set the system to collect #Shaheenbaghprotests starting from 15 January 2020 for 62 days, and then we set it up to collect #sheinspiresus and #womenofshaheenbagh in March 2020 for 62 days. With Gephi -one of us ran it on her Mac for approximately two hours every day for the first three months of 2020. Team members interviewed on-site activists, witnesses, and diasporic South Asian activists. During our interviews with local and transnational/diasporic activists and with witnesses on site, we verified some patterns we saw emerge through mapping Twitter publics. The interviews and surrounding news reports and commentary essays collected during the protests and subsequently until summer 2022 also helped us make sense of the data collected.
In order to get a full-scale reading of what was happening, we needed to look not only at who was saying something and to whom but what they were saying and how they were connecting with the larger conversation. Netlytic and Gephi allowed us to search and filter specific hashtags and phrases to find what we were looking for. We were able to do two kinds of analyses of the tweets -content and network analysis. Our network analysis with these tools showed us who was talking to whom and how communication and ideas spread across the Internet and worldwide. Tracing the nodes (the users in the dataset) and edges (how they are connected) also allowed us to attach actual people to the data and connect the digital space with the groundwork of activists. We argue that groundwork does not have to be limited to on-the-ground activists and protesters, as our work revealed that the digital space is a socio-cultural environment with its own nuances and power dynamics. Several transnational groups we found did not have a physical grounding with the Shaheen Bagh protestors but did have a digital connection that brought attention to the women on the ground. Therefore, we maintain that studying just one or the other does not tell the full story. Comparatively, while in a 'big data' approach to information, the identification of these clusters and patterns would be the end of the analysis (Kitchin, 2014: 8), in applying a critical, feminist framework to this data collection and analysis, visualisation became a starting point to engage in further data collection and to dig deep into the meaning of the nodes and clusters.
In addition, we also did over 25 unstructured interviews with various people through snowball sampling. Respondents included transnational activists, local activists, social media influencers from India, and transnational influencers. Ten were Muslim women, including three from the community of Shaheen Bagh. Two Muslim women were college students, and one was a high school student about to enter college. Two other Muslim women were diasporic South Asians, and two were professionals with postgraduate degrees. Another ten interviewees were diasporic South Asian non-Muslims, some of whom formed transnational activist coalitions working to support the cause, and three of this pool of interviewees self-disclosed their Dalit-Bahujan location. The remaining interviewees were male. As such, these interviewees composed a diverse pool in terms of regional location and conditions of class, caste, religion, and gender, including both on-site protestors and digital activists. What follows are snapshots from the digitally mediated environment of the protests as we explore intersections and relations of various groups and actors. Three of our co-authors were present in Delhi at the time of the protests and started as interviewees. Consistent with our research team's feminist approach to collaboration, we invited interviewees who wished to join us as co-authors. However, we are also conscious of the need to protect the privacy and identities of interviewees. Anyone who chose to co-author did so voluntarily.

Snapshots from the Twitter publics
The vignettes described in this section provide us with snapshots of what is brought into existence through a mix of strategic and spontaneous use of social media around the protest site at Shaheen Bagh. These 'snapshots' are of digital actors/nodes/space/clusters visualised through arrangements of scraped networked tweets. These snapshots allow a glimpse into a process by which the grandmothers of Shaheen Bagh emerged as political actors in the Twitter sphere, even though none of these older women was actually using social media. As an interviewee pointed out, these older women do not have the technological access or social media literacy to be present as Twitter or Instagram users. However, the on-site strategies of activists, including welcoming younger generations of visitors who possessed this social media literacy, did contribute to the production of online visibility. The dadis of Shaheen Bagh thus emerged as political subjects in transnational space through social media, even without being Twitter users.

The 'Dadis of Shaheen Bagh' as Political Subjects
Our analysis of the Twitter data was circular and iterative. We ran several visualisations and textual cloud features, retrieved reports and graphs from the Netlytic site and collected and imported data to Gephi. Then we looked closely at the interview transcripts to identify connections between what the interviewees told us and what we could see in the scraped Twitter data. We also kept an eye on current news around protests against the Indian authorities and Prime Minister Modi both on social media and in traditional news outlets. Therefore, for instance, 7 / 15 when we saw that Bilkis was listed on the TIME 2020 list of 100 most influential people, we returned to the datasets and searched 'dadi' (paternal grandmother). The search revealed a focus on 'dadis of Shaheen Bagh' in the datasets scraped using hashtags 'Shaheenbaghprotest,' 'womenofshaheenbagh' and 'sheinspiresus.' However, in the dataset for #sheinspiresus, we saw that there were many 'dadis' being named and celebrated by even the majority BJP pro-Modi (Prime Minister Narendra Modi's Bharatiya Janata Party loyalists) Twitter accounts-in what might be read as a concerted attempt no doubt to sabotage the visibility of the activist dadis of Shaheen Bagh (see Figure 8). Yet, what this revealed to us was that the term dadi as a political subject had gained brand value in the social media space.
Snapshots (in Figures 1 and 2) we share below are careful arrangements of nodes in the larger datasets that name the presence of the dadis. We can see that the presence of 'dadi' in Twitter space comes about because of the actual physical presence of the dadis on-site in Delhi. The brand value then is derived from the actual protest site where the physical activist presence of the older subaltern woman is a fact. We identified this by filtering the term dadi and enlarging the particular nodes to make them visible. We then dragged them to encircle the larger hairball mesh containing the larger mass of tweets to highlight and represent the presence of the dadis as political subjects in Twitter publics. 4 The reports and visualisations were developed by Radhika Gajjala using software such as Gephi and Netlytic.

Celebrities
One of the interviewees informed us that the sit-in protest of Shaheen Bagh was initially started by fifty women who knew each other. Then the movement picked up. The interviewee further noted that the strategy of digital connectivity started in specific locations, for example, Jamia Milia Islamia. This interviewee also noted that the local print or broadcast news in India did not cover the protest until it became visible via social media. The protests became prominent internationally when celebrities tweeted about them. We see then that this sort of visibility created an 'opportunity space' for independent actors to become active and also to form international coalitions. It further opened up opportunities for transnational activists to respond and connect. Through this transnational visibility, the scale of the movement shifted and led to connections with 'common networks among actors from different countries with similar claims' (Tarrow, 2012: 32).
Some protestors gained more visibility internationally than others because of their access to social media, while others-celebrities like Swara Bhaskar-came to the physical protest site to express solidarity and attracted large crowds. Periodically, through what Tarrow (2012) calls 'non-relational diffusion' and 'mediated diffusion,' these celebrities became more of the focus of the media attention than the protest cause itself. Yet these moments of 'non-relational diffusion' also disseminated information far and wide. Celebrities, then, acted as 'brokers' who may not necessarily participate but served as connectors for otherwise unconnected sites. Some people partaking in the protest, like the celebrity Swara, were visible because of their presence on the site and their 'accessibility' to social media. Interestingly though, as an interviewee confirmed, offline protest sites in Shaheen Bagh were highly fractured spaces -not as cohesive as the online visibility of a few groups might imply.
In Figure 3, we see that celebrities such as Bollywood actors and high-profile activists are tagged by a pro-Modi account in a negative Tweet. The tweet says, 'shame on you,' but it uses the #womenofshaheenbagh and tags celebrity accounts with a large following. This account was later suspended, so it is hard to say if it was just a bot or a real person handling the account. However, since the algorithm does not really care about the meaning of the words, this sort of post would also significantly amplify the #womenofshaheenbagh and contribute to its 'trending' on the Twitterverse. Swara Bhaskar gained much visibility via social media because of her public anti-CAA statements in mainstream press spaces. Even though she herself was not active on social media in the time period of the dataset we examined (although it is possible that in later archives of Twitter, she became active), what is known as the 'in-degree' centrality for her Twitter user ID (node) was high. In-degree centrality is a simple measure to map network connections and refers to the number of edges directed towards a central node (user in this case) being examined. Thus, in this case, the number of inbound ties directed to Swara established her Twitter handle as significant in that dataset at that particular time. In such a case, her virtual presence is signalled through the fact that her Twitter handle is tagged and addressed regardless of whether she herself was actively sending out tweets or not.
Yet, in this particular scenario, the algorithmic measure of her importance is also directly connected to realworld events. Swara Bhaskar explicitly expressed her support during the anti-CAA/NRC protests and shared her views with the media. After an incident in January 2020 at Jawaharlal Nehru University (JNU) in which masked people attacked and injured students and teachers (Yasir and Perrigo, 2020), Swara, a former student of JNU, tweeted her anger and urged people to come and show solidarity. As shown in Figure 4, @reallyswara is centred and highlighted as the second-highest ranking cluster in the data around #shaheenbaghprotests in January 2020. Netlytic reports also revealed that @reallyswara had a high in-degree centrality (Figure 6), with visualisations in Figures 5 and 6 showing the significance of this node. We highlight this cluster not just because of the statistical significance of the node in the larger dataset containing 99,950 messages and 52,577 unique posters, as evident from the report from Netlytic (Figure 6), but also because of our contextual knowledge that this Twitter handle belonged to Swara Bhaskar.
The visualisation above depicts all the tweets tagging @reallyswara from 6 January 2020 to 20 January 2020. A smaller cluster was extracted from the larger dataset to show how accounts that tagged @reallyswara pulled her into visibility. It is also interesting to note that some of the accounts that tagged @reallyswara became non-existent after sending tweets to popular figures like Swara Bhaskar. It is key to highlight here that while these tweets tag Swara Bhaskar, there is no tweet in this set that comes from her Twitter account. This activity around Swara Bhaskar happened because she took a public stance against PM Modi in support of the women of Shaheen Bagh. . The large node represents @reallyswara and the lines represent the edges. The smaller notes are the users who were tagging @reallyswara. This pushed her visibility to a higher point in the network so that she featured as one of the top ten (© Gajjala 2022) Most of the tweets tagging her are negative tweets. Nevertheless, this gave her visibility as one of the top ten 'indegree' Twitter accounts, as we saw in the report on the dataset around #Shaheenbaghprotests collected starting 17 January 2020, via Netlytic (see Figure 6).
In data space, even negative publicity produces visibility and creates a presence of particular Twitter users as they are tagged while contributing to the trending of hashtags. Thus, we see that celebrity presence in the discourse around the protests, whether or not Swara was physically at the site or tweeted about the protests, amplifies both the hashtag and the celebrity herself. As Jackson, Bailey, and Foucault-Welles (2020) point out, celebrity participants help spread the message but not necessarily always in the direct ways we imagine. Further, it is also possible that celebrity participation might inadvertently create problems. While there were examples of other celebrities tweeting, we chose to describe Swara Bhaskar's cluster from the larger data visualised through Gephi because this illustrates how online networks behave as particular hashtags get amplified.

Feminist Encounters of the Dadis and Younger Muslim Women Activists
While there was also much internationally visible activity on Twitter around other hashtags such as #delhiprogrom, #delhiburning, and #antiCAA, some of which fed the global stereotype of violent Muslim protests, the tweets amplifying the presence of the dadis (grandmothers) of Shaheen Bagh took centre stage eventually. Thus, the strategy of amplifying the offline presence of the older women protestors through social media shifted the focus of the reporting about the protests, highlighting their demands and their agency. The story that eventually remains in the international memory is the story of non-violent protest led by subaltern Muslim women of the community like Bilkis, who practiced an ethic of care and respect. Twitter visibility emphasised nonviolence and the ethic of care on-site. As Ziya Us Salam and Uzma Ausaf (2020: 14) have noted, 'the women refused to move as if their life depended on it; and it probably did. They protested peacefully, persistently and persuasively, asking the government to repeal the CAA'.
India has had a complex relationship with feminism with its culturally diverse population and complicated colonial history. It has thus resulted in both vilification and veneration, often simultaneously, of feminist movements and activists. Feminist movements are further nuanced by issues of class, religion, caste, age, sexuality and more, as shown in the Islamic feminist struggles discussed earlier in this article. It cannot be said that India lacks any form of feminist struggle, especially as the Shaheen Bagh protests exemplify a new feminist wave within the Indian subcontinent. However, we must point out here that this feminist wave did not necessarily enact the models of Islamic feminism discussed earlier in this article, despite the Muslim religious identity of the women onsite.
Contemporary Indian feminist protests are frequently defined by the interrelation of digital and physical activism (Khan, 2020). Adrija Dey (2016) has argued that Indian feminist activism today is more broadly and increasingly defined by its global, digital dimension, despite the persistence of a digital divide, drawing particular attention to how student movements in India, the United States, and in the United Kingdom have mobilised to protest and show solidarity. Past research has identified among Indian digital feminist movements where "subaltern images" are deployed to spread a sense of affect and authenticity (Gajjala, 2017). The strategy of mobilising the women of Shaheen Bagh by the digirati on Twitter became clear to us when we visualised the position of various network actors. Our analysis revealed a concentrated effort among anti-CAA/NRC Twitter users to emphasise a narrative of political protest defined by strategic representations of made-to-feel-at-home-ness and an ethic of care. The sit-in location thus became a significant site for dadis and younger activists as they bonded not on the basis of victimisation but with their energies directed to a political commitment. As the successful strategy to make the protest visible through the hypervisibilising of the 'feisty' and 'gutsy' dadis of Shaheen Bagh via Twitter became apparent to even the pro-Prime Minister Modi and pro-CAA/NRC actors, these groups tried to draw attention away from the dadis on International Women's Day by highlighting other Indian female public figures with #sheinspiresus (Edwards, et al., 2021). While this was not an explicit blacklisting of the dadis, it emphasises concerted attempts to make the women's protests and their cause invisible and banal through seemingly harmless yet purposeful tactics.
In a larger dynamic visualisation on Gephi, we saw that the visibility of the hashtags #womenofshaheenbagh and #sheinspiresus spiked as they competed with each other with significant spikes on 4, 8 and 9 March 2020. Through interviews with transnational activists, we learned that this might have been the result of a concerted effort to take over and sabotage the hashtag introduced by Prime Minister Modi's account. The #sheinspiresus garnered over 20,000 tweets between 8 and 9 March 2020, while #womenofshaheenbagh had a spike of 10,000 tweets on 4 March of the same year (see Figure 8).
In our investigation, we see not only the struggle for minoritised identities asking for inclusion of their identity as part of the Nation, but also see women's struggles with layered patriarchies in local, national, and international contexts. Through a committed resistance to these layered patriarchies, supposedly 'subaltern' women and unlikely activists form counterpublics and become part of big social data archives. Here we see a shift from earlier hashtag movements such as those around the Delhi 2012 Rape and those around the #metoo since 2012, in which the primarily visible protest agents were metropolitan college-educated women (Bansode 2020;Dey, 2018;Guha 2021;Gajjala 2019).
How then might we read this 'big data' visibility of subaltern groups of global south women as agents of change? For instance, in the case of Shaheen Bagh, the community members of Shaheen Bagh led the charge on the ground and through their own Twitter account, @shaheenbaghoff1. However, the affective deployment of contentfigures, videos, text -and the strategic populating of particular hashtags by a larger network of digital participants and activist coalitions, local and diasporic, generated affective intensities, which in turn produced publics. These digital publics focused and reoriented the transnational gaze on the seemingly benign bodies of older women rather than on the rage of the younger feminists and the men of the community. Thus, a concerted strategy was involved in making the dadis visible globally through digital publics.
When we asked an interviewee about this international media visibility of dadis and Time Magazine's influential feature of Bilkis, she noted that this was strategically done to highlight Indian culture, specifically that of care, love, and respect, instead of rage. She said one could counter rage, but it is harder to counter a peaceful protest. This was clearly a strategy invoking a Gandhian non-violent resistance to reassert the community's identity as Indian. Groups like 'United Against Hate' also came forth to push for 'compassion' and the 'peaceful' nature of the protest, and Bilkis, an elderly woman, became the face of the movement. However, we argue that rage manifests in varied ways and that the dadis sitting in for three months in determined protests against the CAA and NRC exemplifies the driving force of rage and resistance despite the attempts at the reductive framings. The dadis made clear their identity-based political demands with non-verbal disruptive tactics that were just as eloquent as the tactics of younger women and men activists protesting the CAA. Their quietness was not weakness; the dadis refused to move from the roadways as if their lives depended on it, their driving force a deeply affective rage as their home was under threat with the passing of the CAA and NRC in India.

Internationalisation and Transnational/Diasporic Coalitions
Transnational digital activists also played a role in amplifying the narratives emerging from the local activist sites in India. As Giuliana Sorce and Delia Dumitrica have observed in their study examining the Fridays for Future's (FFF) digital protest communication on Facebook, there is an awareness on the part of current youth activists "of the necessity to act in a 'vertical' manner that goes beyond national, geopolitical milieus and towards a transnational activist sensibility" (Sorce and Dumitrica, 2021: 2). Activists and observers of the anti-CAA/NRC protests in India that we interviewed also clearly articulated their awareness of this fact and talked to us about the conscious, strategic manner in which student activists and community organisers mobilised social media while also connecting with each other through back-channels via WhatsApp, Twitter, Facebook, and Instagram with activists internationally.
In looking at the data, we examined the participation of international organisations, diasporic populations, and transnational activist groups in the data. Various actors shape this internationalisation, including NGOs, labour movements, and transnational coalitions. International actors include 'double-edged institutions' like the World Bank. Thus, the conditions of possibility for progressive transnational social movements are based on contradictory and ambivalent politics. Tarrow's (2012) mapping of internationalisation as a precondition of transnational and global activism is relevant here. In this context, it is worth noting that the activity we recorded through the databases was scraped on and a few days after International Women's Day, 2020. The two databases, #womenofshaheenbagh and #sheinspiresus, also included #internationalwomensday. Most of the activity around these two hashtags happened from 4 March to 10 March 2020. Twitter users implicitly adopted the strategy of internationalisation on both sides of the debate. On the one hand, the pro-Modi groups concentrated on making the #sheinspiresus trend more significantly on the Twitter global stage than the Shaheen Bagh protests. On the other hand, the anti-CAA activists made a concerted effort not only to keep #womenofshaheenbagh visible but also to hijack PM Modi's followers' #sheinspiresus, following the announcement by Modi on Twitter that he would hand over his Twitter handle to seven inspiring women on International Women's Day.
In terms of media engagement by diasporic populations in their home-nation politics, such as in the case of Tunisia, Marwan Kraidy (2017) notes a shift in the internationalisation of media that formed a precondition for what was characterised as the Arab Spring. Kraidy calls these digital spaces and the connections articulated 'crucial training grounds for the revolution' (Kraidy, 2017: 33). Activists changed not only their strategies but also their platforms, flocking to social media such as Facebook because of the open ethos of these platforms. We saw similar kinds of formations emerge in relation to the anti-CAA/NRC protests as well.
Continuing our circular and recursive process of examining the data, we looked for diasporic group handles that came up either in our interviews or in various news reports. We also looked to see if any such handles appeared in the 'top ten' graphs we visualised via Netlytic. See Figure 7 for one such report from which we were able to identify Rana Ayyub (ranaayyub), the author of the article in Time Magazine about Bilkis and Swara Bhaskar (reallyswara), both of whom are located in India. However, R. J. Sayema (sayema), a 'Secularist Feminist Rockstar' of 'Radio Mirchi' 5 fame, was also one of the top ten. While Bollywood's popularity amongst South Asian diasporas is well known, 'Radio Mirchi' also has an international/diasporic following as evidenced by news reports documenting abuse targeting Bhasker, visible women personalities were trolled and harassed by pro-CAA users illustrating larger trends of trolling within the Indian digital political sphere on Twitter (Mishra, et al., 2021). However, the attention they received -whether positive or negative -served to make even traditional media pay attention to the cause. In Figure 7, we see that the top ten Twitter handles in the #womenofshaheenbagh dataset include transnational diasporic activist groups, local groups, and high-profile women activists. It is no surprise that the top handle is @narendramodi since many anti-CAA/NRC protestors and pro-Modi protestors likely tagged his handle, implicating the Prime Minister in this dataset meant to amplify the activist dadis.
In yet another example of how we went about looking for the role of transnational digital activists, when we saw reports concerning the banner 'Resign Modi' that appeared on the London Westminster Bridge on India's Independence Day anniversary, 15 August 2021, we scrambled back to our datasets to search for the 'South Asia Solidarity' group that was allegedly responsible for the banner and the press release that followed. We found that this group had a significant presence (N= 124) in the #womenofsshaheenbagh dataset, even though this handle did not make it to the top ten. Compared with the N=320 of the @australiansi handle count in the dataset, the number of Tweets referencing them was far less, yet given our overall examination of evidence beyond Twitter, we recognise that this is a potentially influential group.
However, the involvement of diasporic communities was not restricted to joining the protests via Facebook, Twitter and Instagram. There was street activism as well in international sites. For instance, on 25 January 2020, South Asian activists in London convened outside the official residence of the British Prime Minister. The group 5 https://onlineradiofm.in/stations/mirchi

Pro-CAA/NRC
Finally, the presence of pro-CAA/NRC content needs to be mentioned lest we give the impression that all media outlets and Bollywood actors and actresses were sympathetic to the protests. Pro-CAA and pro-Modi activity on Twitter was also heavy during this time and within these two datasets. The pro-Modi Twitter handles included high-profile actresses and other visible figures too. Although the anti-CAA/NRC protests were able to reach international audiences because of the transnational handles amplifying a significant majority of Indian diasporic populations, despite the visibility of progressive diasporic groups on Twitter, many remain in support of the right-wing Hindu politics currently prevalent in India. 6 Thus, when we examined the dataset #sheinspiresus as a whole for instance, the most prominent Twitter handles are @narendramodi @Kanganateam @bjpforaurangabad and so on.

CONCLUSION
In the current study, we looked at datasets from Twitter using Gephi and Netlytic software. We then proceeded to make sense of them through a close feminist reading of texts, visualisations, and reports acquired using data analytics tools for computational analysis. We also made connections with the interviews we had conducted with activists and witnesses on-site in Delhi and with transnational digital activists from the South Asian diaspora. We were able to see that the significant transnational visibility produced through social media activity for the subaltern 6 Witness for instance, the 2022 Hindu-Muslim flare-ups in August and September in the city of Leicester, UK.