‘Willing’ ethnic-nationalists, diffusion, and resentment in India: A micro-foundational account

Abstract Using evidence regarding the consolidation of Hindu nationalism in India we put forward new ethnographic data about the variety of popular support for a Hindutva project and a new framework that proposes an interactive theory of social identity. This framework helps us understand how Hindu nationalism becomes embedded in society. We assert that Hindu nationalism in India could be fruitfully analysed by focusing on the processes through which ideas of exclusive nationalism spread among ordinary middle-class people and are expressed in micro-level psychological changes at the individual level. The consolidation of Hindu nationalism in India is being authored not only by parties or the state, but also by societal actors, specifically, ordinary middle-class Indians. Hindu nationalism has been spreading in micro-public spheres in a time of apparent peace and between elections, and with the participation of willing supporters. Building on our fieldwork and research in psychology and history, our conversations have also helped us to identify profiles of different types of nationalists, which we categorize as willing ethnic-nationalists, hardliners, bystanders, and moderates. Further, we suggest the need to focus on inter-linked micro-level mechanisms such as diffusion and emulation of Hindu-centric beliefs and ideas, mobilization by hardliners and organizations, and impunity resulting from protection by state agencies, which helps to create willing ethnic-nationalists and sustains Hindu nationalism. Evidence regarding social interactions from a variety of survey organizations concurs with our findings and our ethnographic material allows us to delve deeper into varieties of Hindu nationalist support across diverse ordinary people.

the BJP are Hindus (Sardesai and Attri 2019;Heath 2020). 7 Simultaneously, a dramatic rise in 'social hostilities' has also been observed, with India occupying a score of 9.7 on a scale out of 10 on the Social Hostilities Index calculated by the Pew Research Center (Majumdar and Villa 2021;Majumdar 2018). 8 Public opinion surveys by different organizations-CSDS-Lokniti, CSDS-APU, and the Pew Research Center-conducted in 2015 all found that support for Hindu nationalism is much broader than initially supposed and that a high degree of social hostility and distance, especially between Hindus and Muslims, has become a feature of Indian life (CSDS-Lokniti 2015;CSDS-APU 2017Pew Research Center 2021).
While a variety of public opinion data point in the direction of a consolidating Hindutva identity, scholars have not yet explored the variety of belief systems underlying these attitudes and the processes through which social attitudes congeal and coalesce. Despite an emerging recognition that Hindu nationalism is seeding into society, the micro-foundations of these changes remain a puzzle to which we need answers: how did some family members, friends, neighbours, and co-workers become willing participants in a Hindu-centric political discourse and sensibility? How do ordinary members of society frame, interpret, and transform their identities in a majoritarian direction? Attempting to answer these questions about micro-level attitudes-how people think-and societal changes also helps us to focus not only on the important question of why parties try to create a Hindu bloc but also, as Kalin and Sambanis ask: '…why the manipulation of such emotions resonates with an audience in the first place' (Kalin and Sambanis 2018, 241), or how 'individuals understand and practice national belonging' (Mylonas and Tudor 2021, 119). 9 We assert that Hindu nationalism in India could be fruitfully analysed by focusing on the processes through which ideas of exclusive nationalism spread among the ordinary middle classes and are expressed in micro-level changes at the individual level. We use detailed interviews with middle-class voters to unpack some of those processes and offer a typology of varieties of support for Hindu nationalism that helps us to go deeper into the demand side of Hindu nationalism. The consolidation of Hindu nationalism in India is being authored not only by parties or the state, but also by societal actors, specifically, ordinary middle-class Indians. 10 (Jaffrey 2021, 232). 9 Varshney (2003, 86) also discusses the need to analyse the 'value-rational micro-foundation' of national identity formation and attend to both the supply and demand side of ethnic mobilization. 10 The motivations of ordinary men and women have been studied in genocide studies with a focus on perpetrators (Browning 1992;Lyons and Strauss 2006;Strauss 2017;Mann 2000). One of the few studies on India that recognizes the role of ordinary people is a book on the 2002 Gujarat pogrom. The author states: 'For me, the single most disturbing experience during the violence in Gujarat was not the complicity of politicians and orchestrations of large parts of the state machinery but the psychological Gleichschaltung (coordination) of "ordinary" Gujaratis with whom I was acquainted' (Ghassem-Fachandi 2012, 10). Our study suggests the need to extend that shift of perspective to studies of right-wing nationalism, a much more commonplace phenomenon that goes beyond riots or violence. Our study also moves from a focus on hardliners to also offer profiles of willing supporters and bystanders, profiles that we believe are important to observe in relation to a conservative turn in nationalism. theorizing on Hindu nationalism, this is not only a top-down project, and ordinary middle-class Indians are not passive objects of this emergence. Hindu nationalism has been spreading in micro-public spheres in a time of apparent peace, between elections, and with the participation of willing supporters, bystanders, and hardliners, a typology that we offer. Anti-Muslim sentiments and a resurgence of Hindu primacy persist even in the absence of riots and pogroms. This tripartite typology is partially drawn from our fieldwork where we noticed a variety of beliefs, as well as theories and research in history, sociology, and psychology.
This article, thus, describes a research agenda for how scholars and journalists can begin to conceptualize the demand side of Hindu nationalism by focusing on individual level attitudes, the diffusedly structured support for Hindutva, and the two-way dynamics of identity politics rather than assume an exclusive top-down effect. In our view, Hindu nationalism is a dynamic and circular process involving both bottom-up and top-down mechanisms. India's 'ethnic democracy' (Jaffrelot 2017) or majoritarianism is the joint product of state-party actions but is also socially produced through multiple micro-level changes in Indian society and urban demographics, which have psychological effects of ethnic attachment, resentment, 11 and othering. 12 Our conversations and fieldwork have also helped us to identify different individual profiles, which we categorize as willing ethnic-nationalists, hardliners, bystanders, and moderates. While the hardliners constitute the core support base of the BJP, the expansion of the BJP's validation among the ordinary middle classes also points towards a growing support from what we call 'willing' ethnic-nationalists and the silencing of the moderates. Willing ethnic-nationalists may not be party workers or rioters but are ordinary members supporting a Hindu-centric view and their numbers may have grown exponentially in the past decade or so. 13 We call them 'willing ethnic-nationalists' to denote the need to look at them, not as passive recipients of a Hindu specific version of nationalism, but rather as active participants of a majoritarian and exclusive vision of nationalism. Bystanders are ordinary people who may be more neutral or passive recipients of a Hindu-centric world view, but one of the striking developments has been the transformation of many bystanders into willing supporters. So, some of the willing supporters started as bystanders in that they were more agnostic about a majoritarian belief system but they became more committed supporters over time (Staub 1989a). 14 Moderate voices are not absent but are rendered impotent in the face of such a resonating consensus, as people opposed to this Hindu-centric, anti-Muslim 11 Cramer (2016, 9) defines the politics of resentment as when '…resentment toward fellow citizens is front and center. People understand their circumstances as the fault of guilty and less deserving social groups, not as the product of broad social, economic and political forces.' 12 Nandy (2009), Appadurai (2006, and Kakar (1995) also urge a focus on the psycho-social aspects of Hindu nationalism.
13 They now constitute the core supporters of the BJP, which has expanded beyond upper castes. Some may argue that the BJP's recent victory in Uttar Pradesh (UP) elections (March 2022) shows the role of this stable core support base; the BJP could not be dislodged from power in UP despite dissatisfaction with the incumbent party and government.
14 Genocide studies have focused on the tripartite distinction between perpetuators, victims, and bystanders, which has some overlap with our categorization (Hillberg 1992). We draw the idea of bystanders from this literature (Staub 1989b) but also extend it to a completely different context of commonplace Hindu nationalism and add some new profiles, such as willing ethnic-nationalists and hardliners.
world view become reluctant prisoners of a social sensibility around them. Attention to micro-level elements of Hindu identity must, thus, attend to important variations in the belief systems and internal dispositions of people within a community. We also relate the micro-foundations of individual-level feelings and beliefs to mechanisms adopted by parties, organizations, and states, leading to a two-way understanding of social identity.
This article urges us to go beyond the organizational forms of Hindu nationalismthe 'elite planners' (Chappel and Stammers 2017) and 'producers' (Brass 2004)-to conduct more research into why ordinary people, who may or may not belong to parties and organizations, support the world view of exclusive or ethnic nationalism without being active perpetuators of riots or violence (Ghassem-Fachandi 2012, Berenschot 2011, Brass 2003, Valiani 2011, Das 2007. 15 This argument leads us to go beyond an exclusive focus on the electoral realm or party politics to look at the societal and psychological facets that are sustaining Hindu nationalism in India as well as the horizontal processes of diffusion and mobilization through which willing ethnic-nationalists are created and hardened. So, in addition to identifying varieties and profiles of individual supporters, we suggest the need to focus on inter-linked micro-level mechanisms such as diffusion and emulation of Hindu-centric beliefs and ideas, mobilization by hardliners and organizations, and impunity accorded by state agencies. 16 We argue that studying Hindu nationalism by disaggregating the mechanisms of diffusion and mobilization is valuable in highlighting the processes underlying public attitudes gleaned through public opinion surveys and how parties and organizations inculcate social change. These ideas together lead us to create a framework that focuses both on the incentives and political strategies of parties as well as the varieties of bottom-up support and popular consciousness. This framework also makes a contribution by connecting top and bottom levels of analysis and offering a social theory of politics. While we lack the evidence to determine the exact timing and origins of these sensibilities, these arguments are confirmed by a unique survey of popular attitudes across 23 states conducted by CSDS-Lokniti and Azim Premji University (APU) titled 'Society and Politics between Elections', the writings based on those surveys (Swaminathan and Palshikar 2021;Varshney et al. 2021;Kumar and Gupta 2020;Swaminathan and Shastri 2020;Venugopalan 2020), and the Pew survey 'Religion in India: Tolerance and Segregation' (Pew Research Center 2021).
Figuring out the dispositions of people and their deeper attitudes and psychological aspects is very difficult to do in large-scale surveys but was possible in our interviews. We find that ordinary people from the Indian middle classes and professionals have become the agents of and participants in Hindu nationalism. 17 Many ordinary middleclass Indians have come to think that 'Muslims' are a problem not because they were manipulated to do so but because they have developed passionate beliefs regarding Hindus, Muslims, and terrorism. Many people also believe that Hindus are besieged and the social context around them is 'hindu-phobic', even as these perceptions are not based on facts. 18 Not all is negative about ordinary Indians, who may believe in religious or even civic nationalism and so we also have to respect and recover the sense of civic and inclusive nationalism that is pervasive among ordinary middle-class Indians (Yadav 2020;Singh 2020). But, the conviction and commitment of ordinary willing ethnic-nationalists have acquired an agency of their own and need independent analysis that goes beyond the electoral strategies of parties. This focus on a people-centred view of nationalism and a two-way theory of social identity allow us to go beyond the idea that individuals are manipulated by elites and parties or have been 'led astray by power hungry politicians' (Berenschot 2009, 415): rather, committed supporters have mushroomed across middle-class society through a process of diffusion, mobilization, and impunity. We are faced with a social contagion that spreads through insular but powerful social networks in our societies and all political parties-including opposition and regional parties-have to adjust to this new landscape. Our goal is to distil these insights into a more empirically cogent model of Hindu nationalism while also contributing a new conceptual framework for future research into varieties of bottom-up popular support for a Hindutva project.
A wide range of writings are beginning to hint at new ways of thinking about Hindu nationalism, labelled as BJP's ideological hegemony and saffronization (Kumar and Gupta 2020;Palshikar 2019; Chhiber and Verma 2019), hard and soft Hindutva (Anderson 2015), 'institutionalized everyday communalism' led by parties and local organizations (Pai and Kumar 2018;Narayan 2014), 19 the working of latent Hindu nationalism (Zavos 2010), in new public spaces and regions (Anderson and Jaffrelot 2018;Anderson and Longkumer 2018, 372;Longkumer 2021), 'vernacular Hindutva' (Reddy 2011(Reddy , 2018Hansen 1996), 'banalization of Hindutva' (Jaffrelot 2013, 82), and 'neo-Hindutva' (Reddy 2011;Anderson 2015;Palshikar 2015b). 20 Scholars have urged a focus on the appearance of Hindu-Muslim conflict as 'mundane everyday ethnic conflict' (Mehta 2010, 15), in 'quieter times' (Brass 2003, 9), and on 'quotidian violence going beyond large scale violence' (Basu 2015, 8). 21 Despite attempts by scholars to identify these new terms and frames, we have limited information and concepts about how people think and become willing supporters of a Hindu-centric project. Only a few of these studies explore the views and sensibilities of ordinary men and women, and most of them focus on parties and organizations even as they speak of new forms of politics. These ideas, thus, need a more grounded theory of how everyday communalism becomes rooted among ordinary people going beyond parties, institutions, and the supply side of politics. We also need more analysis of the linkages that connect local and national issues as well as the processes and mechanisms that link the macro phenomena with micro processes. 22 So, this recent scholarship can better be understood within a new interactive framework that focuses on a people-centred basis of India's Hindu nationalism, and directs attention at how macro and micro mechanisms interact (see section below: 'A two-way theory of social identity').

Methods and timing
How did we gather our information? We relied on social attitudes surveys conducted by CSDS-APU and social media posts (see Appendix 2 for images). We also conducted interviews among professional and educated middle-class citizens across two Indian cities-Bangalore and Delhi. A total of around 40 open-ended interviews were conducted between 2017 and 2020 (see Appendix 1 for the interview method). Delhi and Bangalore were chosen because they are two large cities spanning the North-South divide and had identifiable middle-class neighbourhoods. We recognize that other cities may throw up different findings which deserve further research. We went into communities of middle-class voters to understand their 'reasons' and the way they thought about politics. 23 While this phenomenon of exclusive Hindu nationalism among ordinary men and women may go beyond the middle classes, for the purposes 20 'Neo-Hindutva' focuses on diasporic Hindutva, defining it as: 'idiosyncratic expressions of Hindu nationalism which operate outside of the institutional and ideological framework of the Sangh Parivar' (Anderson 2015, 47). These writings build upon a large literature on Hindu nationalism including: Hansen (1999), Brass (1996, Andersen and Damle (1987;, Jaffrelot (1996), Zavos (2000), Zavos et al. (2012), Varshney (2002), Basu (2015), Bhargava (2002), and Wilkinson (2004). A journal issue examines Hindu nationalism within courts, social media, and the Northeast (Anderson and Longkumer 2018). 21 Similar arguments are made by Pai and Kumar (2018) and Das (2007). 22 Basu (2015, 8) observes that Hindu nationalists have been uniquely successful at 'bridging the localnational divide'. Tambiah (1996) also urges attention to be paid to the dynamic between 'nationalization and localization'. Mehta (2013, 116) suggests a focus on the 'mechanisms' by which a social consensus is created and 'sustained'. Despite these fragmentary suggestions, we need more research into how topdown incentives connect with demand side factors, which would allow us to flesh out how party strategies or movement politics respond to, and shape, bottom-up feelings, preferences, and resentments. 23 Philosophy makes a distinction between 'reasons', understood as meanings that we give our actions, versus 'causes', which explain why things happen (Macklin 1972). of a first-cut analysis it may be important to assess its contours in one section of society (the middle classes)-one that plays a crucial role in shaping discourse and populates institutions of the state such as the civil service, judiciary, and the police. The CSDS-APU surveys allow us to assess if similar attitudes can be found across India and across classes (see the section 'Additional survey evidence' below).
Interviews were held in Bangalore by one of the co-authors and then in Delhi by both co-authors, and with the support of a research assistant: (a) in Summer 2018 and September-October 2018, one of the co-authors interviewed well-to-do, educated professionals, mostly Kannadiga, in the urban locality of Jayanagar, Banashankari, and Indira Nagar in Bangalore; (b) interviews and focused group discussions with technology sector youth were held in June 2019 in Bangalore; and (c) in Delhi, interviews were conducted during the eruption of the JNU controversy in 2017, the first wave of the coronavirus pandemic between March and June 2020, and after the rioting that took place in Delhi in early Spring 2020. In addition, ethnographic research was undertaken at other times. 24 While we could not conduct interviews across southern states (Kerala and Tamil Nadu might throw up important variations) or across rural areas 25 and all sections of society, the pervasiveness of our findings among middle-class Indians in northern, western, and some parts of southern India (Karnataka) is confirmed by national social attitudes surveys conducted by CSDS and APU, and implies that a deeper analysis of ordinary people as agents of Hindu nationalism is warranted.

The question of timing
Arguably, a majoritarian sensibility may have started to coalesce before 2014, but we cannot be sure of its origins. Palshikar (2004, 5430) argued, as early as 2004, that even when the BJP was defeated, Hindutva forces had 'secured' the Hindu mind and majoritarians are to be found across parties such as the Congress and the BJP. In a similar vein, Nandy noted in 2008 after a Gujarat election: 'The [middle] class has found in militant nationalism a new self-respect and a new virtual identity as a martial community…' (Nandy 2008). Despite these intriguing observations about the origins of a Hinducentric project, social science lacks the time-sensitive survey data required to assess when these ideas and sensibilities began to emerge. Our interviews took place after 2017. The surveys conducted by CSDS-APU only start from 2017 and are not timeseries. Panel data, which analyse the same set of people over time, are not available. 26 The question that Palshikar relies on to assess a majoritarian sensibility is not conclusive evidence of an ethno-nationalist world view. 27 So, we are unable to assess through interviews or survey data when willing ethnic-nationalists begin to emerge. The only qualified claim we can make is that if Palshikar's assessment of the BJP cornering the 24 Professor Manisha Priyam conducted interviews in Bangalore while both authors conducted interviews in Delhi. 25 There is some evidence that rural areas are also beginning to see a similar bottom-up resurgence (Kumar and Gupta 2020). 26 Feedback from a reviewer. 27 The CSDS-NES survey asks: 'Do you agree that in a democracy the will of the majority should prevail?' Based on the results, Palshikar (2015a, 731)  broader Hindu vote in 2004 is correct, then that could explain the bottom-up support we found, and give further impetus to analysing ordinary Indians' beliefs and ideas. 28 In essence, if majoritarian sensibilities had shifted the 'middle ground of politics' (Palshikar 2004, 5427) as early as the 2000s, popular consciousness has been driving the BJP's victory to some degree or increasing its vote share even when it does not win. 29 Yet, irrespective of when support for Hindutva spread, or even if the BJP's national victory in 2014 and 2019 has galvanized Hindu popular consciousness, scholarship on Hindutva must expand beyond its exclusive focus on parties and organizations to move to an analysis of ordinary people.

A two-way theory of social identity
A process of redefining India's national identity is under way. What is an identity? Shayo (2009, 147) notes: an individual identifies with group J '…if (1) he or she cares about the status of group J; and (2) he or she wants to resemble the members of group J'. Building on this definition of identity we make two conceptual advances in this article: first, we suggest that identity should be conceptualized as an overlap between identity, as belonging to a group, and perceptions. 30 For example, identity refers to feelings of belonging and may arise from your participation in a larger community of believers (of being a Hindu, for example (Chhibber 2014)). Perceptions refer to how people perceive the issues and conflicts of the day and may change over time. A person may belong to a specific ethnic or religious group but their perceptions about the salience of that identity or notions about who gets what, and how, as well as ideas about the 'out-group', may change over time. Perceptions of threat, even if false, may also make the identity more salient or exclusive, consolidating 'we' and 'they' feelings and identities. 31 Thus, perceptions play a major role in making an identity salient, relevant, and more or less intense-a process of dynamic reinvention, a shift in identities or group formation (Chandra 2012;Appadurai 2006;Shani 2007). 32 Together, identity and perceptions create an 'us versus them' mentality and a psychological state that views other communities as outsiders or a threat. 33 Some evidence suggests 28 Palshikar (2004, 5427) also argued: '…the BJP does not require a communal campaign to win Hindu votes any more'. 29 Despite the Aam Admi Party's (AAP) 2020 victory in Delhi, the BJP's vote share in Delhi has stayed high (38.5 per cent in 2020). Similarly, while the BJP lost West Bengal in 2021, its vote share in Bengal was still 38.3 per cent. In general elections, the BJP's national vote share has increased from 18.6 per cent in 2009 to 37.4 per cent in 2019. 30 Cramer (2016) defined it as 'rural consciousness'. Also see Palshikar (2004, 5429) for a discussion of 'distorted perceptions'. 31 In a recently concluded Dharma Sansad (religious assembly) in northern India, a threat perception that Hindus were under attack circulated widely. The two-day deliberations were held on the theme 'Islamic Bharat Mein Sanatan Dharm ka Bhavishya' which can be translated as 'What will happen to Hindus in a possibly Muslim dominated India?'. 32 Tambiah (1996, 21) also suggests that ethnic conflicts rely on both identity creation and its 'reification' but also on the making and remaking of people's identities through 'flexible' and 'volatile' processes. 33 In a similar vein, Appadurai (2006, 51) argues that predatory identities 'whose social construction and mobilization require the extinction of the other …' may emerge 'by mobilizing an understanding of itself as a threatened majority'. that it is not that middle-class people in India have become more religious (Chhibber 2014) but that their perceptions regarding other communities and their own sense of being a besieged majority has changed. As an interviewee said: '… then came this particular [corona] virus and so suddenly, a community became the target… It was an anti-Muslim perception.' 34 Figure 1 outlines a dynamic model of social identity which discomposes identity into two components, but also includes the mechanisms of diffusion, mobilization, and impunity. This may translate perceptions into a 'collective "we" identity', and thereby hardening against the 'other'.
Second, our theories of identity-in-formation in India rely on elite-centric and strategic models, focusing on party strategy and electoral politics, or on accounts that centre on the 'institutionalized riot systems' (Brass 2004) and social capital (Varshney 2002). Basu's work (2015) linked riots and violent events to a variation in the partymovement nexus across Indian states. Theories of both electoral results, and of riots and pogroms, focus on the rational-political-benefits and incentives of leaders and parties, and do not address the attitudes and perceptions that go beyond violence. A few scholars have conducted bottom-up analysis of the participants in and perpetuators of the Gujarat 2002 riots, examining the notion of sacrifice and ethnic cleansing (Ghassem-Fachandi 2012) and micro-connections with local patronage politics (Berenschot 2011;Shani 2007). Building on these insights, we argue that movement and party politics need to be understood in terms of how they have changed the 34 Interview No. 3, Delhi, July 2020. psychology of ordinary people on the street (and at home) through diffusion, by increasing salience but also by creating more societal space for bottom-up Hindutva across India. These ordinary individuals consume, and then, demand a more aggressive Hindu-centric discourse and policies.
In order to understand this phenomenon, we need a two-way theory of identity formation that thinks about how parties behave but also how ideas of social movements and parties circulate and disseminate at the societal level and ricochet back. The model in Figure 2 highlights that even when the government takes pre-emptive action, endorsement by local level leaders and finding resonance among ordinary people make the campaign stronger and more durable, and reinforces perceptions in society. This may, even, create a new language and norms that reinforce social hostilities (the phrases 'Jai shri Ram' [Glory to Lord Ram] or 'Bharat Mata Ki Jai [Long Live Mother India]'). As an important complement, our ethnographic research and literature from psychology help us to create a profile of different types of supporters. Some of them-hardliners and willing nationalists-become intermediaries between parties and grassroots voters. This intermediation role allows the two-way dynamic to become enduring and sustain itself. In a similar vein Tambiah (1996, 257) identifies two types of processes-'focalization and transvaluation'-as well as 'nationalization and parochialization' to identify how small incidents become bigger and centre-periphery processes take place, thereby creating a chain-like, cumulative effect. 35 Importantly, while we recognize that impunity and mobilization harden the bystanders and willing nationalists, these processes are taking root in a fertile ground of 'resentful hearts' (Mehta 2019). These twin ideas-accounting for the role of perceptions and identity-and conceptualizing Hindu nationalism in terms of an interactive dynamic of elite-driven and bottom-up processes helps us to account for the specific form of ethno-nationalism witnessed in India today and explains the sensibilities of the median voter revealed through public opinion surveys in a new way.

Profiles of support for a Hindu-centric world view: hardliners, willing ethnic-nationalists, bystanders, and the silencing of moderate liberals
An implication of the dominant models of violence and electoral strategy is that ordinary people are pliable and can be easily manipulated by leaders and parties who flood them with negative messages. While leaders are assumed to be utility maximizers and, therefore, rational and materialistic, voters and citizens are seen as innocent cognitive vassals to receive those ideas. Our conversations have shown us that ordinary voices may affect parties' strategies from the ground up. We use our fieldwork and readings from political psychology to create individual-level profiles of different kinds of support for Hindutva: bystanders, willing ethnic-nationalists, moderates, and hardliners, which form a continuum of belief systems, ideas, and speech acts.
Apart from hardliners, who may belong to the RSS or the Vishva Hindu Parishad (VHP) or local organizations such as the Bajrang Dal and Hindu Yuva Vahini, there is a group of people who can be characterized as '…"semi-active participants," if not in the perpetration of genocide, at least in the system…' (Staub 1989a, 42). We label these people as 'willing ethnic-nationalists'; their numbers may have increased dramatically in the last few years, although we cannot be sure when this started. A bystander is an individual who is 'neither a perpetuator or victim' (Staub 1989a, 42) nor strongly committed to a Hindu-centric view. Hardliners show their commitment to the Hinducentric view by participating in organizations related to the Hindutva project, while moderates are against the Hinduization of Indian society.
Dinesh Paril, the district head of the Bajrang Dal group in Maharashtra stated: 'Everyone in this world is born Hindu. They are turned into Muslims when they are circumcised and Christians when they are baptized' (Siddiqui et al. 2017). Surendra Jain, joint secretary for Viswa Hindu Parishad, said: 'If the sentiments of the majority community are respected, there would be no such incidents. Can we demand pork in any Gulf country?' (Siddiqui et al. 2017). These are the hardliners who pursue an active agenda to defend the Hindu faith in a public and political way, and hold anti-Muslim views. Dinesh Arya, state head of the Gau Raksha Dal, acknowledged his group is breaking the law: 'Seizing cattle is not legal and we know that well. We are not authorized to do this, it's the police department's work,' he said. But he claims a higher calling: 'Our religion has given us the right to stop our mother being butchered', referring to 'gau mata' or 'mother cow'. 'We have forcefully taken that right' (Siddiqui et al. 2017). Many speakers at Dharma Sansad in Hardwar calling for the 'killing of Muslims' and, seeking inspiration from Myanmar's genocide against Muslims, are classic examples of such hardliners. They also celebrate their closeness with BJP leaders such as Yogi Adityanath, among others, although the RSS has distanced itself from the Dharma Sansad (The Wire 2021).
These hardliners refuse to listen to official statements, even by Prime Minister Modi himself, revealing that the boundaries between the party, the RSS, and society have become porous, and that hardliners are beginning to push the BJP and RSS from the bottom up in new ways. 'The cow protection movement totally belonged to the BJP before 2014,' said a local Hindu group's leader, Pawan Pandit, a part-time software engineer. 'Now, groups like ours [Bhartiya Gau Raksha Dal] have the momentum.' Many self-styled local groups and activists have emerged, seeking to shore up their 'hardliner' credibility. Balen Baishya, head of the Hindu Youth-Students Council of Assam, said that he believed that the local party leadership was not made up of 'hardcore believers and decided to take matters in their own hands' (Siddiqui et al. 2017).
These hardliners form the core support base and get a lot of media coverage, but underneath this group is a larger group of willing ethnic-nationalists, who, while not formal members of organizations, strongly empathize with their views. We met many such individuals during our fieldwork. Sumesh [name changed] is a professional in his fifties and grew up as part of the Nehruvian world view. Over time, especially in the 2000s, he became disillusioned by the terrorist attacks by Pakistani groups, and now fervently believes that 'Hindus need to show Muslims a lesson.' Similarly, a large group of professionals sympathize with Hindu-centric views, and repost many 'fake' stories about Muslim brutality to their friends on social media. In conversations they defend Hindu activists even when the latter skirt around the law or their actions lead to violence such as the death of a Muslim trader who was transporting cows for dairy. Some of the willing supporters are housewives, students, and ordinary middle-class Indians. Many of them are not explicitly religious (that is, they do not pray regularly or observe rituals, and so on) but they believe that 'Hindus deserve preferential treatment over Muslims.' 36 They see their task as rousing other Hindus against Muslims. Similar beliefs can also be found in Udupa's fieldwork with technology-savvy, ordinary people such as Vasisht, who trolls her moderate friend for taking a 'liberal-secular position'. Her interviews were conducted around 2013-2014 (Udupa 2018, 456-61) and, in our view, confirm that we need a new category to differentiate such people from the hardliners.
The journalist Sheela Bhatt's interview with a fan of Modi in 2014 is notable. Pramod Singh said: 'To preserve one's identity one has to do something. I am not from the RSS. I have never been to their shakha. I don't think this country should be a Hindu rashtra, but in this country at least we should be respected' (Bhatt 2014). A US-trained former mechanical engineer defended the use of violence by Hindus: '…the outfit's [Bajrang Dal's] use of violence is justified because no one listens to Hindus otherwise. The aggression of Muslims makes them heard but the meekness of Hindus leads people to ignore them. Thus, it is not wrong for Hindus to pay back the abuse they have endured for centuries' (Mehta 2017, 12). Chhibber and Osterman (2014, 2) would call such supporters 'vote mobilizers': 'individuals whose support for a particular party goes beyond simple voting and instead involves monetary donations, door-to-door canvassing, leaflet/poster distribution etc.'. Brass's (2003, 32) concept of 'fire tender' and a 'conversion specialist', 'whose task it is to decide when a trivial, everyday incident will be exaggerated and placed into the communal system of talk, the communal discourse, and allowed to escalate into communal violence' is also relevant here.
Radikha [name changed] is a professional who was and remains quite non-political. She never used to vote in elections and rarely followed political trends and events. Her interaction with her family and friends is usually over everyday issues. She does not like rioting and violence but has come to believe that 'Muslims look different and behave differently.' She is a classic bystander who at certain times could become a willing ethnic-nationalist but is more wary of violence. She also does not condone vigilante attacks but feels that Hindus need to be defended and protected. In a similar vein Dipanker Gupta argues: '…for a successful ethnic operation, innocent insiders must be convincingly cast against seditious "outsiders"… They did not accept the politics of separatism, but admired the "heroics" of those who did' (Gupta 2011, 29). Importantly, people who strongly oppose such behaviours and attitudes-moderates-are mostly silent and act like bystanders: unable to collectively challenge the status quo perceptions and anti-Muslim vitriol. The moderates note that the feelings of animosity towards and distance from Muslims have become so pervasive that it is impossible for them to fight on all fronts-in your neighbourhood, workplace, or within your own family.
Psychological and international relations research into the motivations for terrorism and racial violence has identified the importance of the bystander phenomenon. Staub observes: 'The psychological effects of passivity are similar to but less intense than the effects of perpetration. To reduce their own distress, passive members of the perpetrator group tend to distance themselves from victims, in part by accepting justifications offered by perpetrators, and by blaming and devaluing victims' (Staub 2006, 872). So, bystanders may become willing supporters under the right social-political conditions (Staub 1993). Another way to think about bystanders is that, in the context of tension between communities, bystanders interpret every event in 'communal terms' (Brass 2003, 18).
Many bystanders and willing ethnic-nationalists emulate and translate the dominant ideas pursued by hardliners, highlighting the processes of diffusion we discuss below. They spread stories, songs, and pictures that talk of the 'Muslim' threat. These willing ethnic-nationalists may be educated, unemployed, or employed youth and professionals who seek to defend the Hindu faith and proselytize their ideas to their friends and co-workers. The assumption that ethnic-nationalists are only unemployed youth is incorrect; such views are found among professionals, the college educated, and well-off people in large measure, as our interviews with professionals confirmed, and as survey data revealed (Venugopalan 2020). In the early decades of India's independence, moderates shaped dominant discourse and ideas; now it is the willing ethnic-nationalists and hardliners who do so.
It would be remiss to fail to acknowledge the group of moderates who are opposed to this larger world view, although their numbers are relatively quite small. We found many important voices of resistance and opposition, who oppose or are uncomfortable with the polarization that is sweeping Indian middle-class society. Many of them are linked to Indian central or regional universities or are associated with civil liberties movements, and are writers, journalists, students, and educators. Yet, what was also palpable was the silence and fear among moderate liberals. Some noted that they have become careful about explicitly criticizing the social consensus. An interviewee expressed this angst in the following way: First, they said they are like ISIS, they are terrorists and everybody knows Tablighi Jamaat, whatever they may be, they are not terrorists… But there is absolutely no reason to label everyone. You are able to target an entire community largely because they get isolated. And they are visible and easy to target. 37 This four-part profile of willing ethnic-nationalists, bystanders, hardliners, and moderates raises the important, but difficult, issue of participation and complicity in the emerging discourse and sensibility. While hardliners and willing supporters play an active role in the enduring power of exclusive nationalism, are moderates as complicit as bystanders in their effective inability to challenge this system and beliefs? We argue that the beliefs, attitudes, and speech acts of all four groups combine together to create a larger social consensus or the popular common sense of Hindu nationalism, albeit in slightly distinct ways. While bystanders and moderates seem to be the passive receivers of the larger consensus around them, their helplessness does have political consequences in reinforcing the power of Hindu nationalism's discourse and ideals.
Yet, even as the volume of willing ethnic-nationalists accelerates the resonant discourse of Hindu nationalism, the bystanders and moderates' relationship to the larger system is vicarious, mediated, and more indirect. For example, moderates might not vote for the BJP, while bystanders might. Moderates tend to use strategies theorized by James Scott as 'weapons of the weak' (Scott 2008), although many of them also write publicly against exclusive nationalism and make their voices heard in whatever way possible. Simultaneously, it is difficult to offer wholescale or collective resistance to the dominant social consensus. The notion of complicity relies on the presupposition of individual control and responsibility that affect different profiles in varied ways. Resisting the dominant system in quieter ways, as evident in many moderates' actions, is not surrender nor complicity. Importantly, though, through the actions and views of a variety of individuals, the power of a large Hindu-centric social sensibility has acquired a social and political force that affects the national-popular common sense of all actors as well as what is possible to talk about in social and public interactions.

Additional survey evidence, 2015-2021
Our interview findings can be supplemented by a survey of public opinion and social attitudes conducted by CSDS-Lokniti and APU (Swaminathan and Palshikar 2021;CSDS-APU 2017 and recent writings based on that survey (Varshney et al. 2021;Swaminathan and Shastri 2020;Kumar and Gupta 2020). In addition, the just concluded Pew survey on religions in India is also relevant (Pew Research Center 2021). The CSDS-APU survey was conducted across 23 states in three separate rounds, and sampled 48,542 respondents. It shows that ordinary people's attitudes mirror the stark Hindu-centric feelings we identified in our interviews. The attitudes of conservatism are the greatest among the middle class, as we found, and have also extended to the rural areas (Kumar and Gupta 2020, 209), perhaps to the middle class there. Varshney et al. (2021, 23) also found that the 'middle-class and upper-class' variable correlates strongly with Hindu nationalism, which, they argue, is 'unexpected and needs more analysis' (Varshney et al. 2021, 24). The recent Pew survey conducted between November 2019 and March 2020 across 29,999 people finds that 64 per cent of Hindus say that it is very important to belong to a Hindu religion in order to be truly Indian (Pew Research Center 2021), which is, arguably, an ethno-nationalist claim.
Below, we present data on three prominent questions deployed in the CSDS-APU survey: whether people (1) want the government to punish beef eaters, (2) punish those who refuse to say Bharat Mata Ke Jai, and (3) support punishment for conversion, and on questions that could help assess anti-Muslim attitudes. In 2017, surveys of people from Odisha, Karnataka, Haryana, and Gujarat found that close to 75 per cent believe in a strongly majoritarian (or illiberal) view of national identity, while 6 per cent believe in a strong liberal nationalism and 17 per cent take a weak liberal nationalist position (CSDS-APU 2017. 38 It notes: 'A striking finding is that the highest proportion of respondents holding a majoritarian nationalist position are among those college educated and above. Unlike other societies, there appears to be no correspondence between higher levels of education and a liberal view of nationalism' (CSDS-APU 2017. Rather the reverse. In 2018 the findings from surveys from Andhra Pradesh, Bihar, Madhya Pradesh, Maharashtra, Rajasthan, Jharkhand, Chhattisgarh, and Telangana found that on average 68 per cent of respondents believe that 'those that do not stand for the national anthem should be punished'. As many as 80 per cent of Hindus believe the same, which means that this sentiment is quite strong among Hindus. In addition, 75 per cent of all Hindus and 80 per cent of the urban population believe that the government should punish those who eat beef or cow meat across these states, and 69 per cent of respondents are supportive of punishment for conversion across all states. In 2019, 72 per cent of Hindus felt that someone who eats beef cannot be a Hindu; while this number is lower in the South, it still comprises half of the Hindu population (50 per cent). In the same year, the survey across Assam, Delhi, Jammu and Kashmir, Kerala, Mizoram, Nagaland, Punjab, Tamil Nadu, Tripura Uttarakhand, Uttar Pradesh, and West Bengal found that 28 per cent of all college-educated people support a conservative nationalism position, which is defined by a combination of views on punishment for beef eating, conversions, and punishment for people not standing for the national anthem or for those who do not say Bharat Mata Ki Jai. What this report defines as 'majoritarian nationalism' appears to hold strongly across a wide variety of states and especially among Hindus and collegeeducated Hindus. People from the states of Kerala, Jammu and Kashmir, Mizoram, and Nagaland adopted more liberal nationalist or centrist positions. The recent Pew survey also found a strong correlation between various aspects of a Hindu-nationalist position: 'For example, Hindus who take a strong position against eating beef are more likely than others to say they would not accept followers of other religions as their neighbors (forty-nine per cent vs. thirty per cent) and to say it is very important to be Hindu to be truly Indian (68 per cent vs. 51 per cent)' (Pew Research Center 2021).
What about attitudes regarding Muslims? A 2015 survey of around 5,081 people across 21 states found that 40 per cent of Hindus have either little or no trust in people belonging to the Muslim religion and lack of trust persists among the college educated or middle classes (CSDS-Lokniti 2015). A 2017 survey found that only 13 per cent of Hindus think that Muslims are highly patriotic, while 77 per cent of Muslims think of themselves as 'highly patriotic' (Express Web desk 2017). These findings continue for the 2019 survey where more than 30 per cent of Hindus think of Muslims as unpatriotic and 13 per cent think of them as 'highly unpatriotic' (CSDS-APU 2017. This means that close to 50 per cent of Hindus think that Muslims are unpatriotica striking figure. Interestingly, this view is one-sided: in large measure Muslims do not consider Hindus to be unpatriotic. Similarly, 13 per cent of all respondents believe Muslims to be peaceful, which is a low figure, while 31 per cent believe that Muslims are violent. Perceptions about Muslims' 'unpatriotic and violent behavior' as compared to other religions score at the highest level, revealing an Islam-specific religious polarization. Importantly, being 'college educated' reduces the perception that Muslims are a peaceful community (CSDS-APU 2017, which is in line with our interview results. Venugopalan (2020, 159-160), using the CSDS-APU survey, also found that education, especially college education, increases prejudice towards Muslims (emphasis added), noting: 'For instance, college-educated Hindus-around 11 per cent of the Indian population-perceive Muslims more negatively than their non-literate counterparts' (Venugopalan 2020, 163). She also observed that the intensity of those feelings is relatively strong, with most, even when they are friends with Muslims, weakly or strongly prejudicial and 'the predominant notions in the society of suspicion and negativity towards the Muslim minority are entrenched' (Venugopalan 2020, 163).

Empirical analysis of contours of Hindu nationalism: campaigns and mechanisms
How are willing ethnic-nationalists and hardliners created and hardened? Tambiah (1996) defined the processes of identity formation as a product of opposing forces as local events become aggregated into national themes and issues, and national mobilization reverberates into local arenas (Tambiah 1996, 257-258). Building on his insight, but also going beyond it, we argue that in addition to identifying the processes through which identity boundaries are reified, we also need to identify the mechanisms that create different kinds of supporters of Hindu nationalism as well as the intermediaries and linkages between top-down drivers and bottom-up demand. We highlight the tripartite mechanisms of diffusion and emulation, mobilization, and impunity by focusing on campaigns such as the response to coronavirus and legal-social campaigns such as the beef ban and 'love jihad'. 39

Diffusion and emulation
Diffusion has been defined as 'the spread of something across space' (Elkins and Simmons 2005, 36) or as a process in which conflict and ideas in one area alter the likelihood of similar actions occurring elsewhere (Kuran and Sunstein 1999) or as a process of 'uncoordinated interdependence' (Elkins and Simmons 2005, 35) wherein people follow ideas and actions found in other contexts and domains through a method of uncoordinated spread. Revati Laul (2018, 26) called it a 'tidal wave that built on itself ' to refer to the feelings of a perpetuator of the 2002 Gujarat riots. Diffusion can happen through a variety of routes: 'learning, imitation, bandwagoning, emulation, and mimicry' (Elkins and Simmons 2005, 35). Some of it is a greater circulation and dissemination of pro-Hindu ideas that begin to shape the ideas of individuals by updating their beliefs about different communities as well as ideas about the costs of discriminatory actions. It could also work as a demonstration effect where ideas keep being repeated but then generate a set of new norms. It could also work through emulation and translation when certain beliefs and perceptions begin to resonate with people as their statements are reinforced by others. 40 We found that diffusion has begun to affect every member of the middle-class household: housewives, educated college-going adults, retirees, even people who were, until recently, non-political or apolitical. The process of diffusion has also created micro-ideologies and slogans (for example, Jai Shri Ram), some of which have been generated from below and then adopted by hardliner activists, BJP/RSS party workers, and officials of the state. Diffusion also works to connect local and national processes of polarization through social media (Sircar 2020), even music and memes. Tambiah (1996, 257) observes: 'microevents at the local level, through chainlike linkages, accelerate and cumulatively build up into an avalanche…'.
A widespread worry about Muslims began affecting everyday life and ordinary people's relationships with their fellow citizens during the first wave of coronavirus in 2020 (Asif 2020). It started with a super-spreader event that had been organized by an orthodox sect of Islam-Tablighi Jamaat-at the Nizamuddin Mosque in Delhi (Dutt 2020). Yet, similar events by Hindus and Sikhs did not create the same cycle of hostility. An interviewee told us: 'I mean definitely, Muslims have become the collateral damage, you know, most people cannot differentiate, they feel "oh masjid, Muslims, blame karo, blame Muslims".' 41 Around 23 April 2020 a man refused to accept delivery of groceries from a Muslim delivery man. 42 Ration dealers denied food to Muslim families in Bihar. They said: 'Muslims don't get any rations. Muslims spread coronavirus.' 43 Natasha Badhwar, a journalist, reported that when an NGO 40 Diffusion through emulation was evident when in February 2016, a PhD student at India's premier Nehru University was arrested. At issue was the shouting of pro-Pakistan slogans at an event on JNU campus, which he denied making. Initially, as information about police action and doctored videos put the government and party spokespersons on the backfoot, they were hesitant in their defence of the arrest. Soon after, however, middle class people and professionals such as bankers, lawyers, and even housewives articulated their criticism of 'slogans in JNU', after which the BJP began to criticize JNU speech as 'antinational'. Individuals from the middle classes also began to emulate, repeat, and thereby diffuse these ideas across the micro-public spheres. The progression showed a two-way interactive process with the BJP activists becoming more aggressive after the middle-class support for the arrests became evident (participant observations by Aseema Sinha). 41 Interview No. 10, July 2020. 42 https://twitter.com/ndtv/status/1253376561140494337?lang=en; https://www.ndtv.com/india-ne ws/near-mumbai-man-refuses-to-accept-grocery-from-muslim-delivery-person-2217212, [ was distributing food kits to migrants, they were asked if the person delivering the food kits was Hindu or Muslim, and when told that they were a Muslim they refused to accept the rations (Badhwar 2020). A new term-'Corona-jihad'-and images of Muslims entering a religious-type structure and coming out with a coronavirus cell on their heads began circulating in social media as well as false claims about Muslims spitting on food. 44 These stories were circulated and forwarded very widely. Such conversations and fears were expressed in many homes and on the streets. Some comments called for the government to issue 'sight and shoot' orders against Muslims, a reference to an Indian police practice called 'shoot-at-sight' orders that are usually employed in a riot-like situation. 45 From 28 March to 3 April, the tweet with the hashtag #coronoajihad appeared 300,000 times and could have been seen by 165 million people on Twitter (Perrigo 2020). One such tweet was retweeted 4,200 times with 503 replies. This anti-Muslim 'virus' reached the Middle East where some Indian expatriates faced punishment over their 'vile Islamophobic remarks' (Staff Report 2020). Some of these comments went beyond coronavirus. In March 2020, 'Indian chef Trilok Singh who worked at a restaurant in Dubai was fired for making an online threat to rape a Delhi-based law student Swati Khanna over her views on the controversial Citizen Amendment Act' (Staff Report 2020). The diffusion of anti-Muslim sentiments is, at times, organized by people employed by the BJP through their IT-cells (Singh and Venkatanarayanan 2021; Chaturvedi 2016) but decentralized bottom-up views of middle-class actors are also common, and once activated, the BJP tries to use and manipulate them. The role of the media, with some channels escalating anti-Muslim rhetoric (Verma and Razdan 2020), is important, but they go down this route because their ratings and viewership have increased. There is a demand for this kind of polarization among the watchers of Indian media channels, who seek a reinforcement of their feelings and anxieties. BJP spokespersons are aware that the Indian reading and watching public 'approves' of these ideas, and this explains their ratcheting up of anti-Muslim rhetoric at key moments. In essence, a new language of Hindu nationalism has pervaded households through diffusion and emulation. NDTV did some research into the use of hateful and divisive language by high-ranking politicians (Jaiswal et al. 2018). Some of these phrases are: 'Jai Shri Ram', 'Modi Raj vs. the Mughal Raj', 'Mohamadad', 'corona-jihad', and so on, which may have emerged in local areas but are then used repeatedly by hardliners and willing ethnic-nationalists. The CSDS-APU surveys confirm widespread support for these ideas and the circulation of a new Hindu-centric set of phrases. Overall, we argue that the BJP and many affiliated organizations seek to nurture, consolidate, and channel underlying antagonisms, but that societal resentments have acquired an independent force and feed back into BJP's strategies in ways we do not yet understand. There is no doubt that mobilization and escalation by interested actors and the state (impunity) are part of this interactive two-way dynamic; we analyse that next.

Mobilization and role of impunity: beef ban and 'love jihad'
We do not posit an exclusive role for bottom-up mechanisms of diffusion, and acknowledge that these horizontal processes are enhanced by a resurgent BJP, and RSS, which seek to spread Hinduism across society and gain politically and culturally from these larger trends. Mobilization is an act of bringing forth focused coordination through campaigns; organization is intrinsic to the concept. Building upon recent writing on 'everyday communalism' (Pai and Kumar 2018), we suggest that many local Hindu organizations, such as the Hindu Yuva Vahini, compete for the definition of the Hindu nation, in the process ratcheting up Hindu-centric voices and ideas at the grassroots level and through 'every day vigilantism' (Jaffrey 2021). The RSS has always focused on cultural mobilizational campaigns targeted at many 'social' issues, in keeping with its cultural face, while the BJP is focused on the political side. Historically, campaigns were about cow protection and the Ram Mandir. More recently, they focus on what has been termed 'love jihad', 'Ghar Wapasi', and cow protection. 'Love jihad', a so-called 'strategy deployed by Muslim men to marry Hindu women', the issue of conversion in Ghar Wapasi, and the larger issue of banning beef were the focus of mobilization and escalation by a variety of local Hindu organizations more recently. The fire of campaigns has been kept alive both by top-down strategic choices but also bottom-up campaigns from local organizations that seek to use the election of an RSS Pracharak to increase their power and influence. Gupta (2018) documents how a campaign on 'love jihad' started in the villages and small towns of western Uttar Pradesh by organizations such as Dharma Jagaran Manch, the Viswa Hindu Parishad, and the Bajrang Dal. The RSS began publicizing these by devoting prominent space to it in its publications (in Organiser, for example) and its officials began speaking about it. However, these mobilization tactics are not new; the RSS and local Hindu organizations have been focusing on these campaigns and issues ever since the early twentieth century (Gupta 2018, 88); what has changed is their reception among Indian society. A 2018 social attitudes survey found that a majority of Hindus believe that beef should be banned (CSDS-APU 2017, an example of such acceptance within society. Importantly, these local mobilization campaigns prepare the minds of ordinary people for pro-Hindu and anti-Muslim rhetoric, creating a fertile connection between top-down and bottom-up processes, using and deploying existing resentments and connecting local issues with national ones (Ghassem-Fachandi 2012, 266;Tambiah 1996). Even though it is possible that this opposition to eating beef existed earlier, the RSS-sponsored campaigns have increased their salience at the local level and diffused these ideas across society, creating a new standard (see Figure 2). This explanation points towards an interactive dynamic between local grassroots movements, organizations, and elite party strategy as well as the connection between the RSS's cultural face and the BJP's political aims that our theory of social identity foregrounds.
Many new Hindu organizations and locally rooted hardliners have mushroomed, and seek to establish their credibility and influence to mobilize people. Their goal is to gain the attention of the RSS and BJP so that their members can catch the party's eye and possibly secure tickets to fight elections. 46 So, there is a bottom-up logic to mobilization too. The friendly competition of local Hindu organizations with the BJP 46 Berenschot (2009, 417) outlines individual-level motivations of a local politician in Ahmedabad (Gujarat), who led attacks on Muslim homes during the 2002 riots, and through his actions attracted popular support and won a seat in the assembly after such actions. Also see Kanungo (2012) on mobilization. and RSS expands the reach of organizational and exclusive Hindu nationalism. There is some disquiet within the RSS at the behaviour of these local vigilante groups and some attempt to officially distance itself (Anderson and Jaffrelot 2018), but other RSS leaders are too tempted to develop a coherent policy against such actions. A senior RSS official noted anonymously: 'Hindus never had the courage to stand up for their religion and now they are standing up' (Siddiqui et al. 2017). Thus, mobilization also works through horizontal processes and competition among a variety of new Hindu organizations rather than exclusively through a united front of BJP-RSS strategizing to achieve their electoral and cultural goals.

Impunity or the role of the state
Mobilization works hand in hand with support from the state but across different levels of the state, including local police departments, regional states, and a wide variety of state agencies. 47 Notably, both the urban and rural middle-classes populate these institutions at all levels. So, in keeping with our current models of religious violence and politics, we outline the role of impunity and collusion (Kim 2017;Chatterjee 2017), which highlights the reaction of state actors who look the other way and initiate pre-emptive legislative action such as the Citizenship Amendment Act 2019 or a beef ban. For example, the BJP's manifesto for the 2022 state-level Uttar Pradesh elections promises punishment and fine of Rs 1 million for 'love jihad', an example of direct support and impunity provided by the ruling party in Uttar Pradesh (UP BJP 2022). Wilkinson's (2002) research focused on how the state leadership acts to prevent riots only when Muslims constitute a significant portion of a state's population, and there is high competition between parties. When aggrieved Muslims go to the police, they are met with ignorance or even harassment. In Maharashtra the regional government scrapped positive reservations for Muslims, and in 2015 the sale and possession of beef was made punishable by a fine and a prison sentence of up to five years, and religious conversion was made difficult (Jaffrelot 2017). Many other states followed with introducing such laws, even appointing 'animal welfare officers' in every district combined with new cow shelters (in Uttar Pradesh, for example). One such cow vigilante asserted: 'The police have to listen to us because the BJP is in power' (Siddiqui et al. 2017); this is a classic demonstration of the mechanism of impunity. Going beyond laws and state policies, official statements by BJP MPs, cabinet members, and even the prime minister during election campaigning has created a chilling effect on public discourse and made it possible for ordinary people to be emboldened and confirmed in their anti-Muslim sentiments. Such acts of impunity also have a micro-effect in that they encourage, even urge, many hardliners to make provocative statements publicly and boldly, and by doing so, normalize and sustain such rhetoric and actions among bystanders and willing ethnic-nationalists. These acts of omission and commission by state actors have contributed to tacit-even explicit-support for underlying resentments, which feeds back into society. 47

Conclusion
Political scientists and journalists have written about the top-down mechanisms of polarization focusing on electoral and communal strategies: 'At the macro-level, the self-reinforcing dynamic of political polarization… is well understood' (Schedler 2020). Essentially, this powerful consensus views the turn towards Hindutva as a strategic and ideological choice by leaders and parties of the Hindu right to win more votes. While we acknowledge the indispensable role of top-down polarization strategies adopted by the BJP and its affiliates-what we call mobilization and impunity-current Hindutva politics is also born of consent and complicity by the larger population, especially the middle classes. While all nationalisms are rooted in community-centred sensibility, the project of Hindu nationalism's roots in civil society and the changing common sense of the people needs more research and analysis, to which we offered a nascent conceptualization and new data, based on interviews. Understanding these roots must turn to micro dynamics and diffuse, bottom-up, and sideways processes, in addition to the BJP's attempt to polarize voters. This calls attention to the micro-foundations of the underlying collective psychology of nationalism, resentment, and anger that is sweeping ordinary people along, especially those belonging to the middle classes. Theories of macro-institutions and system-wide strategies must combine with micro-level analysis of ordinary people, and how strategic choices transform but also respond to everyday street-level and home-based Hindutva. When societal majoritarianism becomes routinized and diffused, it begins to cumulatively affect party strategies and create new norms of Hindu-centric behaviour and discourse, leading to the mainstreaming of anti-Muslim discourse. The cumulative effect of this two-way process is to ensure broader acceptance of a majoritarian vision-what has been referred to as 'ideological hegemony'. Our findings suggest that we need a new explanation to help understand the consolidation of Hindu nationalism in India, one that combines attention to the social psychology of a variety of ordinary men and women, and an interactive and mutually reinforcing relationship between Hindu nationalist parties and affiliated organizations that manipulate but also respond to a changing Indian sensibility, especially among India's professional and middle classes. Organizational and party-centric Hindutva is being enhanced by non-institutional-everyday and societal Hindutva-that needs our attention.
While we offered a typology of a variety of belief systems among Hindu nationalists-willing ethnic-nationalists, bystanders, moderates, and hardliners-our focus is on what they think, believe, and their speech acts. While implicit support for violence was striking, at least among willing ethnic-nationalists, and actions must subsume speech acts as well as direct action, we are unable to verify empirically if supporters of a Hindu-centric world view would participate in riots or explicit violence. But, there is no doubt that the discursive boundaries of what Indians can say has shifted, and it has become more acceptable to express anti-Muslim ideas and views directly and publicly. It is also important to acknowledge that these are not fixed or essentialist categories but a continuum of traits and ideas about belief, which can also change. The fact that bystanders can turn into willing supporters at certain moments foregrounds that our theory accounts for the possible dynamic nature of this typology and the constructivist nature of identity. Sometimes those classified as willing ethnic supporters may become bystanders or even moderates when faced with changing situations. 48 Importantly, the purpose of our two-way theory is to suggest that people are not organized into these categories in a rigid or static manner, but may be part of a large, diffuse community of believers with a Hindu-centric sensibility that will make people more willingly 'adapt' (Luft 2020) to an ethnic-nationalist set of ideas, which they can also reverse, though with difficulty. Future research must also analyse if and how these everyday micro-level anxieties are deployed during riots or election campaigns, which are the more intense forms of political activity-but their everyday presence during times of apparent peace is palpable.
Our article offers a new research agenda for future studies of Hindu and rightwing nationalism rather than assume the last word on the role of bottom-up Hindu nationalism, as well as ways of combining the top-down and bottom-up dynamics and mechanisms of Hindu nationalism. In concluding, we point towards future directions for research into how psychological polarization and rising resentments among India's middle and professional classes are conditioned by the specific context of crisis events (such as terrorist acts) combined with fragmented civil societies and polarized information environments which engender reinforcement of biases and resentments. Clearly, social media such as WhatsApp and Facebook plays a major role. Settle's (2018) book argued that Facebook can create feelings of psychological polarization, wherein even non-political individuals begin to view out-groups as outsiders. Indian cities are showing signs of such psychological polarization. In the words of an interviewee: What happens is you're inclined to think in a certain way and if you get WhatsApp messages that corroborate such things, then you go all out and believe them without thinking, could this be all right, is there another side to it. You don't. Because that is your half-formed impression in any case. And if there are messages to the contrary, then you dismiss them by saying that this must be the anti-government lobby, this is the anti-so and so lobby, that is why they're saying it, so you dismiss them. 49 Furthermore, we must do more research at the micro level-into neighbourhoodsgoing deeper than civil societies or states (Berenschot 2009). Resident welfare associations (RWAs) across various colonies in Delhi and Bangalore have become the arenas for ideas of Hindu primacy and antagonism towards Muslims to circulate (Das et al. 2021). The overall effect is rising resentment and intolerance combined with the diffusion into the social public sphere of Hindu ideas, symbols, and festivals. India is beginning to reveal the dark side of social solidarities, which needs to be accounted for in the various celebratory treatises on civil society, democracy, and public action (Mehta 2017;Berman 1997;Mann 2004). Psychological polarization brought about by many horizontal and bottom-up processes is, then, mobilized further by party strategy, organizational Hinduism as well as home-level and street-level Hindutva politics, which in turn affects electoral outcomes. Electoral politics, however, are a mere tip of the iceberg; beneath the wins of the BJP in 2014 and 2019 lies a massive iceberg of a variety of public attitudes, ideas, and resentments.
While most of the current scholarship has focused on studying Hindu nationalism in terms of electoral results, party strategies, or riots, this article offered a two way and reciprocal theory of social identity that builds a micro foundational theory of individual beliefs and motivations. Such a theory builds a more robust demand-side theoretical foundation for our understanding of nationalism and right-of-centre politics the world over. We argued that ordinary Indians are India's willing ethnic-nationalists and even vigilante activists. In addition, bystanders are more complicit in a new world view, even though they may not actively perpetuate intolerance. Leaders of the BJP and members of the various Hindu organizations enhance this phenomenon and make use of it-through the mechanism of mobilization and impunity-but the BJP and RSS both represent and are responding to a fast-growing ethno-nationalism at the societal level, and are trying to use it to their advantage. They are as much the outcome of wider processes in Indian society as its agents. Top-down drivers such as party strategy, beliefs of leaders, and electoral calculations combine with micro-level changes in people's sensibilities and everyday resentments to produce a deep, wide, and assertive Hindu-centric wave that is here to stay. hours and at times were followed by repeat interviews. First, we asked our respondents questions about their background and their family stories and histories. We asked them about their fathers and mothers to get a sense of their political and social leanings. We were also interested in how long they have held their views, although this was difficult to assess. We asked about their religious practices and also about their friends and kinship networks. Then, we turned to their views and perceptions about current politics in the relevant city or state. We used current controversies-for example, about Tipu Sultan in Karnataka-to get at their political sensibilities. We did ask them about who they voted for but our purpose was to find out why they voted for the party they did. We tried to be very careful and emphatic that we were seeking to understand their world views and their ideas about Hindus and Muslims. We also asked interviewees if they were active in their resident welfare associations. The structure of questions varied across campaigns and cities. So, for example, in Delhi we asked about the JNU controversy in 2017, the coronavirus and its spread across communities in 2020, and about terrorism. In Karnataka, we asked about the elections, especially when interviews were done after the assembly elections. In 2020, the Delhi interviews were conducted over the phone, but all other interviews were done face to face. Thus, the structure of questions changed across cities and time periods.