Paperwork, patronage, and citizenship: the materiality of everyday interactions with bureaucracy in Tamil Nadu, India

This article explores the material practices through which lower-caste and poor villagers engage with bureaucracy in contemporary India. We take documents and paperwork – such as ration cards and community certiﬁcates – as a ‘lens’ through which to explore how paper materiality is infused with the politics of power, patronage, and identity. The article brings ethnography from rural Tamil Nadu, South India, in conversation with two bodies of literature: one on the materiality of bureaucracy and one on the nature of political mediation in contemporary India. We demonstrate how everyday engagements with paperwork as well as processes of applying, form ﬁlling, and securing recommendations are constitutive of social and political relationships and, ultimately, of citizenship itself. Political mediation around paperwork and bureaucracy generates a hierarchy of citizens rather than equal citizenship for all, yet ordinary villagers transpire as anything but passive. Drawing on patronage networks, engaging in affective performances, and navigating a politics of identity, they actively negotiate access to the state in an attempt to claim their rights as citizens.

that shape ordinary people's engagements with bureaucracy and its representatives (Fuller & Harriss 2000). What do the poor and lower castes make of the paperwork, documents, and certificates that mediate their access to state bureaucracies and actors? How do they mobilize networks, affect, and identities in the process of making claims and pursuing meaningful citizenship (Bakewell 2007;Shinde 2016;Srivastava 2012)?
In showing how people draw on caste relations, patronage, political parties, affective performances, and (religious) identities, the article sheds light on the ways in which ordinary people navigate bureaucratic and political entanglements, and the sorts of networks and identities they selectively cultivate and activate. Seeking to avoid reifying 'the state' , we explore the multi-scalar processes, actors, and materialities that make up the bureaucratic encounters experienced by ordinary Indians.
Drawing on a study of two villages in Tamil Nadu, we bring ethnography in conversation with two bodies of literature: one on the materiality of bureaucracy, and one on the nature of political mediation in contemporary India. We do so by presenting evidence of people's everyday material experiences of bureaucractic processes in order to enhance our understanding of a number of aspects of state-society interaction. First, we shed light on paperwork as a tool through which a range of political mediators facilitate or obstruct people's access to state resources, and include people in or exclude them from meaningful citizenship (Martin 2014). Second, we reveal how political recognition and citizenship are actively pursued and shaped by marginalized communities themselves. Or, as Doron put it, '[M]arginality no longer implies "victimhood", nor does it suggest duty and obligation towards the state, rather it denotes defiance and distinctionproducing a heightened sense of entitlement derived from being relatively deprived ' (2010: 772). We document some of the ways marginalized people navigate bureaucratic paperwork in rural Tamil Nadu. This is about more than just agency. We are concerned with the skilful manipulation of contingent networks, communicative modalities, and personal identities in the pursuit of specific material and symbolic outcomes. We pay particular attention to the 'role of ascriptive identities such as gender and/or caste in shaping the channels through which people attempt to access government and services' (Kruks-Wisner 2011: 1151. Third, we contribute to an understanding of citizenship as 'emergent' (Das 2011), rather than as 'static' and 'just there' , and as a process that requires continual negotiation and engagement. A focus on the material dimensions of people's engagements with various state actors and mediators reveals how citizenship is an ongoing and necessarily incomplete process of applying, form filling, and claim making through which different types of citizens are created. Building on Sriraman's broad conceptualization of citizenship 'as sharply contested and socially precluded claims to political equality' (2018: i), our evidence suggests that rather than extending citizenship to all and producing equal and full rights-bearing citizens, political mediation around paperwork and bureaucracy generates a hierarchy of citizens in which the poor, lower castes, and women, more than anyone else, need to resort to a series of interventions -including drawing on patronage networks, engaging in affective performances, and navigating a politics of identity -in an attempt to obtain what they are legally entitled to. As Heller and Evans argue, ' [C]itizenship is always multi-scalar. Citizens are made not only at the national level through constitutions and elections, but also in their dayto-day engagements with the local state ' (2010: 435). Indeed, access to basic citizenship rights and entitlements is extensively mediated through political society, leaving those at the bottom of the social and economic hierarchy also politically most vulnerable and Journal of the Royal Anthropological Institute (N.S.) 26, 495-514 C 2020 The Authors. Journal of the Royal Anthropological Institute published by John Wiley & Sons Ltd on behalf of Royal Anthropological Institute exposed. Our evidence thus contributes to debates about the nature of political society in India today (Chatterjee 2004) and to an understanding of the nature of political agency in those everyday interactions where organized, civic movements are largely absent.
But what leads to the production of compliant 'patients' rather than equal citizens in places such as India, Argentina, and other postcolonial states? Should official paperwork not reflect the indiscriminate nature of the bureaucracy? While a Weberian reading of bureaucratic artefacts -such as ID documents -might indeed consider them key tools in modern bureaucracies' pursuit of impartiality, transparency, and legibility, ethnographic research has challenged such a view by revealing the porousness of state-society boundaries, the embeddedness of bureaucracies within society, and the influence of societal make-up on the functioning of the state (Fuller & Harriss 2000;Hull 2012). Viewed this way, bureaucratic documents and paperwork not only carry society's imprint, but also (re)produce inequalities of caste, class, religion, and gender, amongst others. Or, as Hull put it, documents are 'not simply instruments of bureaucratic organizations ' (2012: 253), but constitutive of the way bureaucracies function and the ideologies, practices, rules, and subjectivities they engender.
There is no doubt that some bureaucracies are more effective in delivering services in a more impartial and egalitarian manner than others, producing more equal citizens in the process. So we need to ask why in certain places, like Tamil Nadu, bureaucratic processes continue to fall short of producing the Weberian neutrality and accountability that we might find in other places. Heller (2000), for example, has shown that Kerala has succeeded fairly well in facilitating access to welfare services in a broadly inclusive manner. This, he shows, was largely due to the long-term presence of broad-based leftist forms of social mobilization and an active civic associational life, which curbed extra-democratic influences and paved the way for more far-reaching redistributive reforms (2000: 499-511). Over the last two decades or so, Tamil Nadu, too, has witnessed a shift from clientelist to programmatic party politics -as described by Wyatt (2013) -in which good governance and welfare provision have now become increasingly expected among the electorate (Carswell & De Neve 2014a). Nevertheless, this shift has not led to the eradication of clientelist politics altogether. Rather, in Tamil Nadu, as in much of India, upper-caste control of state institutions remains the norm, participation in civil associations is limited among the lower castes and classes, and economic dependency on higher castes reproduces clientelist relations as key to the bureaucratic engagement of subordinate groups. Therefore, Heller's observation that people's 'ability to exercise citizenship rights is circumscribed by the persistence of traditional forms of social control ' (2000: 492) clearly applies to Tamil Nadu. Socioeconomic structures of control and domination -largely shaped by caste and class inequalities -continue to undermine the accessibility, impartiality, and accountability that bureaucratic processes may well deliver elsewhere.
In our study villages, classed, gendered, and caste-based discriminations are the most obvious inequalities being reproduced through paper-based bureaucratic encounters. It is precisely through the politics of signatures and recommendations that 'patients of the state' are brought into being and that documents play a central role in the making of differentiated citizenship. Much of what we describe below applies to villagers across class and caste divides, and will no doubt be recognized by citizens across the Indian subcontinent. Everyone will be familiar with the frustrations of having to obtain the right document, being made to wait for a recommendation, or having to get the necessary signature. However, it is particularly the marginalization and exclusion of the poor, lower castes, and women from full rights-bearing citizenship that is actively entrenched through the material processes described here. They are the ones on whom the burden We therefore focus on the social and political life of documents, and the patronage networks in which they get embroiled, while also recognizing the highly 'affective' processes they engender. The materiality of state-society interactions is infused with the politics of affect (Hull 2012). Fear, intimidation, humiliation, despair, despondency, as well as anger and outrage, permeate and shape interactions around paperwork, and play a key role in the reproduction of subjugation and compliance. We follow Navaro-Yashin's call to study the 'affects retained, carried and effected by documents, as they are produced, exchanged, transformed, and transacted among their users' . She argues that documents 'are perceived or experienced as affectively charged phenomena when produced and transacted in specific contexts of social relations' (Navaro-Yashin 2007: 79). Building on this, we show how paperwork's affective impacts are further enhanced by the social networks and interactions through which it comes to be obtained and circulated. We show that documents generate strong emotions as people apply for cards, collect signatures, and make claims to state resources. It is precisely because the very possession of official documents is so important to the rural poor that the processes of acquiring them are so strongly charged with affect. Applying and claim making are infused with high levels of uncertainty and insecurity, and are experienced through feelings of anxiety and humiliation, which in turn contribute to the constitution of ordinary people as (compliant) clients of the state (Auyero 2011;. And yet paperwork also provides spaces for claim making by ordinary villagers. In contemporary Tamil Nadu, people are continually engaged in navigating identity documents in an attempt to be 'seen' by and encounter state representatives in ways that are advantageous to them (Street 2012). As such, we question a view of ordinary people as mere 'victims' or 'passive' patients of 'the Indian state' . Instead, spaces for representation are opened up by the fragmented nature of the local bureaucracy and the layered structuring of political society. Indeed, as Fuller and Harriss remind us, rather than resisting the state, ordinary people engage pragmatically with state actors and institutions, and seek to 'use the "system" as best they can ' (2000: 25). We examine such spaces of contestation by exploring how people use 'affect' around paperwork and manipulate paper-based identities to access state resources.
In terms of affect, our analysis draws on writings by Appadurai (1990) and Cody (2009), who analyse affective practices of begging, crying, and submissiveness as communicative modalities through which political representation is exercised. In an insightful analysis of a petition to the Collector by low-caste, partly literate villagers in Tamil Nadu, Cody describes how the petitioners used a language of supplication and deference -and sought to perform subjugation in front of the Collector -in order to obtain a specific effect: that is, to have their requests heard and acted upon (2009: 361-9). Such public affective performances, Cody argues, draw on well-established modalities of communication in Tamil Nadu through which kinglike patrons and political superiors are compelled to listen to those who submit to them in this way. Through begging, praise and flattery, a 'model of pleading' is mobilized to activate bureaucratic processes and make those in power respond (Cody 2009: 363-4). Similarly, Appadurai observes that beggars in India typically bless and praise their targets in order to make them generous. This amounts to what he calls 'coercive subordination' , in which subordination is performed in order to force the dominant to show kindness and compassion and to respond to the pleas of supplicants (1990: 101; see also Cody 2009: 363). In our case Ltd on behalf of Royal Anthropological Institute study of Joy below, we draw on this concept to make sense of her pleading with the village president for his signature, which, we will show, constitutes a form of political representation through coercive subordination.

Patronage, citizenship, and democracy
A focus on the materiality of bureaucratic interactions can also advance debates about the role of patronage in the making of citizenship and democracy in contemporary India.
Having moved beyond merely vilifying political brokers as obstructers of democracy and development, research has begun to recognize the role played by various political actors in the production of democracy in India (Piliavsky 2014), with Chandra (2004) calling India a 'patronage democracy' . But it has been questioned whether such political patrons enable or hinder oppressed groups' political representation, citizenship, and access to state welfare and development agendas. This remains an important analytical question, and we set out some recent positions in the debate before indicating how a focus on the materiality of bureaucratic mediation can shed new light on the workings of political patronage in contemporary rural Tamil Nadu. While we fully recognize that the terms 'patronage' and 'patronage politics' have long been subjected to normative evaluations as premodern, non-democratic, and a general shorthand for bad and corrupt politics, we build on recent critical but sympathetic analyses of patronage that see it not as 'a feudal residue' , but as 'a current political form vital in its own right' (Piliavsky 2014: 4). It is therefore our explicit aim to eschew normative interpretations and instead evaluate patronage as central to how political representation, citizenship, and democracy function in contemporary Tamil Nadu. Chatterjee (2004) sees postcolonial democracy in India as played out across two separate spheres: civil society as the arena where the elite make lawful claims as citizens and draw on the state's formal institutions; and political society as the arena where broader populations engage in political manoeuvring at the margins of or even outside legality -often driven by survival needs and dependent on patron-client relationships. However, even poor and marginalized communities do not only act through political society, but also mobilize collectively through civil society. Scholars have described the 'civility' of subaltern actors, or the ability of marginalized groups to draw on an array of rights-based movements, community-based organizations, and civic associations to make the state accountable and claim rights (Corbridge, Harriss & Jeffrey 2013;Harriss 2006;Pattenden 2016). However, as Harriss has pointed out, civil society is 'distinctly stratified' , and even in urban India the working poor remain 'largely excluded from active participation in civil society organizations ' (2006: 445). As a result, in the absence of civic associations or when marginalized within them, the rural poor often have little else to draw on than their everyday manoeuvring within political society to claim rights and gain recognition as citizens.
Spearheaded by Chatterjee's argument (1998;) that political society is key to the poor's access to public resources and basic freedoms, some call for a recognition that political actors are indispensable to bringing the people to the state and the state to the people. As most ordinary people cannot access the state through civil society channels, they take recourse to actors in political society to make state institutions and actors more responsive to their needs. Others are more critical of this political mediation and show how it is shaped by localized power relations, can divert resources away from the poor, and makes entitlements dependent on political loyalty or votes. Indeed, Witsoe reminds us that any 'interaction with state institutions is mediated . . . by the structuring of local power' . According to him, the 'political mediation of the Indian state' is always and everywhere 'about struggles over domination and subordination within local sites' (Witsoe 2012: 48; see also Witsoe 2011). Veron, Corbridge, Williams, and Srivastava (2003: 4) similarly question the ability of political society to offer a viable route to access the state and represent the interests of the poor, as argued by Chatterjee. In a critical evaluation of the role of political society actors in enabling access to the Employment Assurance Scheme in eastern India, they make the point that 'poverty and powerlessness are the fate of those who lack political connections and who cannot successfully break the law when required to do so' (Veron et al. 2003: 5). Unless the poor can access political society in a meaningful way, there is little guarantee that political mediators will prioritize their needs or substantially improve their fate.
An explicit critique of political mediation is made by Martin in an ethnographic study of political patronage in Pakistani Punjab, which demonstrates that political society is actually 'integral to processes that dispossess people of their rights and to the reproduction of elite power' (2014: 419). Rather than helping the poor, political mediators and brokers of the dominant Jat caste possessed the power to 'subvert policy implementation, the rule of law and bureaucratic procedures' so that they served their own interests (2014: 421). The result was clear: 'instead of access to health care, communal spaces, a decent infrastructure and educational system and justice, poor villagers just got a cash payment in exchange for their vote' (2014: 427). Political brokerage did not help people access state resources; rather, Martin concludes, 'ad hoc clientelistic exchanges that characterize political society reinforce existing power structures and therefore undermine rather than expand popular freedoms' (2014: 432).
Others again have explored the gender, caste, and class dynamics that undermine political society's ability to ensure broad-based representation of ordinary people's interests (Pattenden 2016;Picherit 2015;Still 2011;2015). Caste-based modes of subaltern engagement and collective (political) organization through which disenfranchised groups seek representation and claim citizenship have been well documented. Studies of the Dalit Panthers of Tamil and the Yadavs of Uttar Pradesh illustrate how caste is central to the way in which mobilization and representation are pursued across both civil and political society (Gorringe 2016;Michelutti 2004;. Moreover, lower-caste mobilization has recently been weakened by new forms of patronage that emerge around labour circulation, access to state resources, and capital investments in booming sectors such as construction. Picherit (2015;2018) has shown how earlier forms of collective organization by Dalits in Andhra Pradesh have given way to more individualized patronclient networks between Dalit migrant workers, small-time Dalit labour brokers, and higher-caste contractors and political patrons (see also Pattenden 2012). What such renewed dependency on labour brokers, political patrons, and bosses of sorts reveals is that lower-caste collective action is being undermined and replaced with individualized patron-client networks that mediate access to state resources.
And yet, despite this 'dark side of political society' (Martin 2014), the poor are anything but despondent. There is plenty of evidence that ordinary Indians persist in making claims to state resources and in struggling for recognition as 'fully fledged members of society and rightful citizens of the state' (Doron 2010: 767). Central to political recognition and meaningful citizenship is access to basic public goods, such as health and education, as well as to a fair share of rights-based welfare programmes (Carswell & De Neve 2013;2014b;Heyer 2012). Much of this struggle for citizenship and welfare is an ongoing process, experienced by ordinary citizens through routine Households in rural Tamil Nadu hold an array of cards and paperwork. At the time of our field research, the ration card was a key document, held by every household, and listing all household members. It was used to access the Public Distribution System (PDS) through ration shops, as well as other goods that the state government periodically distributes. The Aadhaar card is a new identity card held by individuals. Launched in 2010, it holds biometric and residential data, stored in a central database. Initially supposed to be voluntary, the Aadhaar card has, for all intents and purposes, become compulsory. Today, Aadhaar is linked with the transfer of welfare payments such as MGNREGA (Mahatma Gandhi National Rural Employment Guarantee Act) wages, pensions, and LPG (liquefied petroleum gas) subsidies. The MGNREGA 'job card' is a booklet given to anyone who registers for work under the National Rural Employment Guarantee Scheme. Rolled out across India since 2006, this scheme gives all rural households the right to 100 days of employment on public works (Carswell & De Neve 2014b;Khera 2011). Each year a new job card is issued to any individual seeking MGNREGA work. The voting card is given to all individuals over the age of 18 and is essential for being able to vote at elections, but is not discussed in detail here (see Carswell & De Neve 2014a).
In addition, there are a wide range of other cards and certificates held by most households, such as residence and community certificates. All these documents have become particularly significant given the state's ever-larger role in ordinary people's lives as provider of social protection. Over the last decade, the Government of India has rolled out multiple social protection policies to alleviate poverty and provide basic social welfare (Carswell & De Neve 2014b;Drèze & Khera 2017;Khera 2011), but -crucially -access to these remains mediated by paperwork and people. While the possession of the right documents might not in itself ensure access to state resources, ownership of appropriate documents and identity proof is always a first and necessary requirement in citizens' engagement with bureaucratic processes (Carswell et al. 2019 In what follows, we focus on how ordinary -particularly lower-caste, poor, and female -villagers experience their encounters with this paper bureaucracy and the sorts of mediation and patronage politics this involves. Through extended interviews with more than fifty individuals across both villages, we collected narratives of bureaucratic interactions, of people's perceptions of political mediators, and of the affective performances and identity politics they activate to access state resources and gain recognition as citizens. While we interviewed some local-level state representatives and mediators, such as panchayat presidents, ward councillors, and political leaders, our main focus was to understand how ordinary villagers themselves recount their experiences. Rather than treating these narratives at face value, we see them as situated accounts that need to be interpreted from within the specific social contexts and interactions in which they occur, including the affective performances, idioms of communication, and identity politics that they draw upon.

'You need a recommendation': the politics of accessing cards and resources
In order to get almost anything done in relation to the state in India -such as buying rations, changing schools, or applying for a pension -people need a signature or a recommendation. To steer their way through the labyrinth of forms, offices, and approvals, applicants need to mobilize any networks and patrons available to them. Villagers are very aware of the range of patron-client relationships that can expedite their individual application processes and determine outcomes. While they are often left exasperated and confused by the lack of transparency and by what they experience as the arbitrariness and unfairness of bureaucratic processes (Gupta 2012), many are skilled at leveraging whatever networks they have to pursue their entitlements.
Before we draw on the stories of applicants themselves, we first hear from Allapuram's influential former panchayat president and current district councillor, Mr Kulandaisamy, how the processes of recommendations and approvals work: For all these [recently launched] schemes, a signature is needed from the president and the district councillor. We district councillors can recommend people to the president, but he has to sign. The district chairman or MLA can also recommend people for signatures but, again, the president has to sign. If the president refuses to sign, I can interfere and request him to sign a particular application. If we [the president and district councillor] are from the same party, then people can easily get a signature and we won't check on the president. But if we are from different parties, say DMK and ADMK, 3 then I will keep an eye on the president to see that he signs off for the people we send to him.
Kulandaisamy is a Mathari Dalit, ex-panchayat president (elected under a reserved seat for the SC/ST [Scheduled Castes/Scheduled Tribes] community), and current district councillor, which gives him significant status and respect in the area. He is very active in the ADMK party and has close connections with senior (non-Dalit) ADMK activists and politicians, vying for an MLA (Member of the Legislative Assembly) ticket himself. He recognized that while he possesses notable negotiating skills as an individual, his position (or 'posting'), too, gives him power to get things done. He explained: 'An individual doesn't have any respect, it is your posting . . . I will keep the power from the posting as a pickle [i.e. I will use a little at a time], but I will use my skills as an individual and add a little bit of pickle to move things around' . As Hull wrote of the Islamabad bureaucracy, '[C]ontrol over the speed and path of files is a means by which officials exercise power over people and things and is metonymic of power and influence . . . A powerful person can move a "stuck up" file. Those without influence have to "put wheels on it" 'posting' and his individual 'skills' , but his prerogative to recommend applicants and sign documents is key. Kulandaisamy is a huge asset to the Matharis of Allapuram, and as panchayat president he managed to direct resources to his own community. As district councillor, he has a wider reach across the district and is able to keep a check on the current panchayat president, who belongs to the dominant Gounder caste. More than in other places, Dalits here have benefited from well-implemented MGNREGA works, extensively funded self-help groups, and other state resources such as scholarships and housing schemes. People thus need to be 'recommended' by different officials and secure the relevant signatures to get their eligibility approved or application moving. In this process, they are largely at the mercy of personal networks and patronage relations, in which some villagers are better placed than others. But, in addition, individual applicants can become prey to party, caste, and personal politics that shape the relations between different actors at various levels -such as between the local panchayat president and district-level councillors or MLAs. The panchayat president emerged as a key figure in people's narratives of bureaucratic encounters as most cards, certificates, and schemes require his signature or recommendation, and without this support paperwork can easily 'get stuck' . Getting on the 'wrong side' of the president -whether owing to caste affiliation, party politics, local electoral politics, or personal relations -can be a costly affair and cause exclusion from key state resources, even with the right documents in hand.

The know-how of bureaucratic engagement
It is not only that the president must be on your side, but before you even get to him, you need to know whom to contact in times of need and how to do so. The following cases show the lack of information available to many poor, low-caste villagers about bureaucratic processes and networks, and illustrate how a lack of bureaucratic knowledge makes ordinary people dependent on the goodwill of local mediators within political society.
Let us look at the ration card, probably the most important document for poor households, as it enables access to the all-important rations available through the PDS. Maheshwari is a Hindu Mathari woman from Allapuram whose family's ration card was accidently left behind in the ration shop by her husband. Out of spite, it seems, someone known to the family stole their card. As a result, Maheshwari struggled for a whole year to get a replacement. She first approached the staff at the ration shop, who were most unhelpful and did not tell her how to apply for a replacement card. For months she was unsure what to do until one day she told Kulandaisamy, a fellow Mathari, that she was without a ration card. He immediately contacted Mohan, the ward councillor (and, crucially, a fellow ADMK activist), to ask him why he had failed to help her. Mohan in turn asked Maheshwari why she had not informed him in the first place. This was because Maheshwari had believed that issuing ration cards was the responsibility of the PDS staff and was unaware that Mohan could help. When recounting the whole episode to us, Maheshwari was still unclear whose signatures were needed or who finally did the paperwork. But after Kulandaisamy's intervention, Mohan completed the necessary forms and saw the process through. Five months later, Maheshwari received her new card.
Maheshwari's case reveals not only her lack of knowledge of bureaucratic processes, but also her reliance on key local actors whose support is crucial to get things moving. This dependence is not just one of ordinary citizens having to rely on the co-operation of elected government officials; it is shaped by the politics of gender and caste too. Gender shapes bureaucratic engagements at different levels. While Maheshwari's husband lost the card, it was she who ended up with the responsibility for applying for a new one, a responsibility that emanates from her role as wife, mother, and provider of food. Having been burdened with this 'work' , she then had to appeal to male actors outside the family for help with the paperwork. In the case of Kulandaisamy, their shared caste identity facilitated the communication, but her interactions with Mohan the ward councillor (a higher-caste Gounder) were shaped by the inequalities of both gender and caste. Having to interact with higher-caste men itself undermines Dalit women's confidence to appeal for bureaucratic assistance.

Navigating the personal and the party political
The centrality of the panchayat president, and the ways in which the personal and the party political intersect, are crucial in bureaucratic processes. Whilst Maheshwari had Kulandaisamy's support, members of the other Dalit group in the village found themselves on the wrong side of the party political lines that shaped the patterns of bureaucratic patronage in the village. The Adi Dravida Christians in the village were known to be DMK supporters, with several being extensively involved in the DMK party. A significant fault line in Allapuram was between the ADMK ward member Mohan and the DMK-supporting Adi Dravida Christian community. This is illustrated by the experience of Selvi, a Christian Adi Dravida woman in Allapuram. Following a work accident eight years earlier, her husband had been unable to work and had applied for a disability card to receive free public transport and a monthly allowance. To get this card, however, they needed the panchayat president's signature, but he refused to sign. Selvi attributed his unwillingness to sign to different party affiliation: 'If we voted for the ADMK, we would long have had this card now!' explains Selvi. 'All us Christians here vote DMK and my husband has a formal DMK post . . . My husband also tried to get a pension through Mohan, the ward member . . . [but] he is ADMK, so my husband got frustrated with the whole process and left it' .
Selvi was not the only one struggling to obtain support from Mohan. Rukmani is a rather vocal Christian Adi Dravida woman, married with three children, whose husband has long campaigned for the DMK. Rukmani recounted how she had given Rs 2,000 to Mohan in order to obtain a water connection to their house. She complained, But he just spent the money and no water pipe was provided. Like that he gets money! But when he failed to get us a connection after taking my money, I got really angry and shouted at him: 'What are you? Are you a Paraiyar [name of her own Dalit caste] or what?' As Mohan is a higher-caste Gounder, referring to him as a Dalit amounted to a huge insult, suggesting he had lost his dignity and good name as a Gounder and formal officer, and that he had fallen to the level of behaviour typically associated with lower castes. But Rukmani remained powerless in the face of a higher-caste ward member who belonged to the opposite political party from her husband and yet on whose support she relied to access a basic public resource. Party affiliation and caste identity combined to act against her interest and to keep her deprived of access to water, the most basic public good. Here we are reminded of Cody' Ltd on behalf of Royal Anthropological Institute the means to be heard and seen, even where they take the risk of raising their voice, as in Rukmani's case.

Navigating local panchayat politics
A panchayat president is elected every four years, and who occupies this position has huge significance for villagers. In a new settlement on the edge of Mannapalayam, our second study village, two widowed Gounder women complained bitterly about how badly they were treated by the new president, and their narratives reveal the interconnectedness of personal and local electoral politics. Samathal is a widow with two unmarried adult sons who moved to a resettlement area in the village, where she built a one-room brick house on a few cents of land allocated to her by the then president. However, ten years on, she still had not received the patta (land document) that proves her ownership of the land. When we asked her why she had been unable to obtain the document from the president, Samathal replied: Oh, that one! The previous president was okay, but this one is useless. The previous one always listened and helped as much as he could. For example, when we were given the land and I needed some money to build the house, he would help. But this one, if we now ask for a water connection, he immediately says, 'I can't do it, the government has to do it' . But when I tell him that only he can do it and that we voted for him, and when I ask him please to do something for us, he asks, 'Who voted for me? The people from my place voted for me, the people from Kapuram, so go away!' The first thing he always asks is, 'Who are you, where are you from?' I spoke to him very frankly but he still wouldn't help.
Rajamma, another widowed and impoverished Gounder woman, joined us and told a similar story. Following the death of her husband, her brothers-in-law disputed her claim to a share of the family land. She managed to obtain the patta for her share, but her brother-in-law then stole her voter ID card. Since then, Rajamma had struggled to get a replacement voter card, and when she approached the president about it, he shouted at her: 'Are you going to keep losing it? Do you think we are a charity or what?' The current president -a Gounder from a separate hamlet, Kapuram, within the panchayat -was widely seen to be more sympathetic to the people of his own hamlet. Rajamma reflected, 'You see, even our own people [referring to her Gounder caste] are now turning against us!' and explained how the president only looks after 'his own people': that is, the people of Kapuram. Shared caste membership does not in itself guarantee that one's pleas will be heard by panchayat officials or that one's subordination will succeed in eliciting their support (Cody 2009). Local electoral politics play a key role too. Here, panchayat politics, shaped by locality-based voting patterns, ended up trumping caste. Many informants confirmed that there was a clear rift in how the president's patronage worked, with the president being seen as favouring people from his own hamlet whom he perceived to have voted for him and whom he therefore recognized as the rightful citizens of the panchayat -citizens with a right to make claims and request his support.
A group of women at the MGNREGA worksite echoed the difficulties others had with the current president, explaining: We've asked him [for things] and he says, 'Who are you? You didn't vote for me, I didn't win the election because of your vote, so I won't do it for you' . . . Even if you go to him for a signature, he asks Rs 2,000. When my daughter went and asked for his signature, he said 'I don't know who you are' . So if you go without any connection, he asks for money, so what can we do?
These statements reveal that material processes of applying and appealing for support act as sites where subordination and dependency are dynamically produced through  (Navaro-Yashin 2007). Many informants, like Samathal, were deeply offended and personally upset by being asked the question 'Who are you?' It is no coincidence that this was brought up by many, as the question is highly loaded and suggests much more than merely 'not knowing' a person, their name, or place of residence. Rather, it contains a sense of refusal to acknowledge them as legitimate members of the panchayat who have the right to approach their elected president for support. It expresses not only an unwillingness to help, but also a much deeper denial of their rights to basic resources and of their very existence as right-bearing citizens. This is why the questions 'Who are you?' and 'Where are you from?' are such powerful discursive tools in the hands of the president. They not only undermine villagers' confidence to request support but they also challenge the very legitimacy of their appeal and presence in the president's office. Moreover, replies such as 'Are you going to keep losing it?' and 'Do you think we are a charity or what?' publicly humiliate the applicants and project them as non-deserving poor who have only their own negligence to blame. Such affective responses are extremely effective in reproducing the subordination of applicants and curtailing their ability to act as rightful, claim-making citizens.
Similar to Auyero's accounts of waiting and queuing, affective rebuttals and public humiliations act as governing techniques through which docile and compliant welfare clients are created (Auyero 2011: 24-5). Navaro-Yashin therefore calls documents 'affectively loaded phenomena' , in that 'when placed in specific social relations with persons, documents have the potentiality to discharge affective energies which are felt or experienced by persons ' (2007: 81). Here too, located within the gendered and caste-based hierarchies between elected officials and ordinary villagers, documentary processes release powerful affective expressions of intimidation and humiliation that limit women's ability to make claims and often -but not always, as we will discuss below -turn them into reticent clients of local state actors.

Affective expressions as idioms of communication
The affective expressions of subservience by the poor have a further complexity to them, as illustrated by the story of Joy in Mannapalayam. Joy was a recently divorced young mother of two whose bitter dispute with her husband turned into a dispute over documents. Elsewhere we have described how her attempts to get a duplicate birth certificate turned her into a 'patient of the state' (Carswell et al. 2019). Here, we discuss the centrality of affective performances of subordination to her ability to navigate the bureaucratic maze and exert political voice. While pregnant, Joy approached the panchayat president about a small grant that was available for pregnant women. The money had to be paid into her bank account, but her husband had kept her bank book since their separation. As a young woman with little confidence or education, she sought help but was passed between different officials: the president's office, the women's police station, the bank, and then again to the president. She finally got help from a local teacher to fill in the required form, and then asked her ward councillor to phone the president in an effort to get him to sign the form. But even then, 'When I went to him, he still wouldn't sign. Then I went for a second time and I cried, and then he said, "Don't cry, I'll sign it"' .
Joy's case reveals a number of issues. First, being separated from her husband made her particularly vulnerable, not only because of the missing documents (birth certificate and bank passbook), but also because she now had to deal with all the Ltd on behalf of Royal Anthropological Institute paperwork herself. Second, Joy had very few social connections to leverage in her struggle with the bureaucracy and completely depended on her ability to gain the goodwill of a local teacher and ward councillor. Not having a good connection with the panchayat president, who lives in a different hamlet from her own, further weakened her already vulnerable position. Finally, Joy's story illustrates how, despite her vulnerabilities, she effectively resorted to the well-recognized cultural idiom of pleading to obtain the president's signature. She managed to get his signature by openly subjugating herself through begging and crying in his office. Here, Joy activated a form of 'coercive subordination' (Appadurai 1990: 101): by publicly enacting a submissive performance in the president's office, she enforced his reciprocity and secured his signature. Where patronage and networks fall short of delivering desired outcomes, affective performances can be mobilized to get things moving. By drawing on a cultural idiom of communication that uses praise, flattery, and submissiveness to elicit sympathy, affective performances of public subordination become a means of political representation and claim making rather than being an expression of passivity or compliance (Cody 2009). It was by begging and crying in his office that Joy ultimately obtained the president's all-important signature.

The politics of identity
In addition to mobilizing a range of connections and mediators, and drawing on affective performances to make claims, villagers play on their paper-based identities in ways that suit their interests. While scholarship has paid some attention to the role of identity in bureaucratic documentation, much of this draws on identity documents and their history in the West (Torpey 2000), on transnational communities ( Here, we describe how lower-caste and poor villagers negotiate their identities around documents and bureaucratic processes. The manipulation of identity has been illustrated by Shinde, with reference to the de-notified (ex-criminal) Kunchikorves of Mumbai, who actively 'manufactured' their own caste certificates in order to benefit from state reservation policies (2016: 96).
Here, the bureaucracy's willingness to accept unconventional proof of identity reveals that state actors are 'not only discriminatory or controlling, but also integrative and mediating' (Shinde 2016: 107). Shinde's study shows how ordinary people exploit the disjointed nature of local state bureaucracies, and turn this fragmentation to their advantage. Srivastava, discussing fake IDs and other counterfeit documents in a Delhi slum, similarly describes the multiple strategies used by the urban poor for gaining security of livelihood and residence in the city, including 'simulation and dissimulation, feigning and duplicity, and passing and pretending ' (2012: 79). Securing the right paperwork and identity proofs to enable a life in the city, Srivastava points out, 'is not only surrounded by anxieties and apprehensions' , but also by more creative 'cultures of making' the required documents (2012: 80). Indeed, the acquisition and circulation of documents open up spaces for negotiation, appropriation, and even production, in which people manipulate both identities and documents of identification. In the context of rural Zambia, Bakewell explored different manifestations of national identity: distinguishing between a 'handheld' identity, defined by papers and changeable, and a 'heartfelt' identity, being what a person feels and usually more lasting ( An individual's handheld nationality might differ from their heartfelt nationality, and Bakewell observed that 'self-identification is clearly socially constructed and must be expected to vary with the context ' (2007: 17). It is precisely in the 'gaps' between social and paper identities, or heartfelt and handheld identities, that a space for political agency emerges, and it is in this space that ordinary people navigate official documents and their meaning.
The 'flexible' use of identities can best be illustrated with reference to the community certificate (jati chaandrathal), a document that confirms an individual's caste/community membership. Supposed to be held by all individuals, the certificate is crucial at certain points in life, such as when children join or change schools. For Dalits in our study villages, the community certificate carries a range of benefits and is therefore highly valued. Reservations, emanating from quota-based affirmative action, give members of SC communities access to a percentage of government jobs as well as places in government-funded educational institutions. Some Dalits from our villages benefited from such reservations, for example by obtaining jobs at the Electricity Board, but one of the biggest advantages today is seen to be the educational scholarships that are given to members of SC communities. Crucially, however, anyone registered as Christian -irrespective of their caste -is automatically classified as BC (Backward Caste) on their community certificate, and as a result does not qualify for the benefits and policies reserved for members of SC communities. For example, Visanti, a young educated Mathari woman, told us that the only benefit of belonging to her community was that she had received several scholarships. Poornima, a woman from the Christian Adi Dravida community, similarly emphasized the advantages of having 'SC Adi Dravida' rather than 'BC Christian' on her community certificate: 'It is better . . . because you get scholarships and all that' .
The Christian Adi Dravidas in Allapuram were acutely aware of their entitlements as Dalits and how best to 'adjust' their paper identity in order to access state resources.
When talking about what was recorded on their community certificates, Rukmani explained that only her eldest daughter had 'Christian Adi Dravida' as community on her certificate, while all other household members had 'Hindu Adi Dravida' , and hence SC, on their certificates. Rukmani explained this to the second author as follows: People clearly strategize around their cards and adapt their paper-based identities in order to access benefits and resources. Rukmani's family retained their Hindu names and Hindu Adi Dravida community certificates as these allow them to be registered as SC and benefit from social policies meant for SC communities. On the other hand, Another woman from the same Christian Adi Dravida community confirmed that she, too, had 'SC Adi Dravida' on her certificate. Poornima explained that she and her husband saw themselves as Christian in every way: 'From the beginning I am Christian, and my husband is Christian . . . I have always been Christian' . But she was happy to admit that, knowing the advantages that being classified as SC brings, they had 'provided the information for the VAO' (as to what caste they were), admitting that this was a deliberate decision. The reason for it was very clear: an awareness of the benefits that different paper identities entail. This is not to suggest that villagers merely act as hard-nosed pragmatics, singlemindedly focused on maximizing state benefits. Many Christians preferred a paper identity in line with their social identity, even if this meant forgoing certain benefits, as exemplified by Ravi. As a Christian Adi Dravida, Ravi had worked for the Post Office for forty-two years, and was hoping to be promoted before retirement. However, he had been told that to be promoted, he would have to become Hindu first and be registered as 'SC Adi Dravida' on his community certificate. Ravi objected to this: 'I am attached to my religion, and even though I may have to miss my promotion, I will remain Christian. I won't convert. All our cards say "Christian" and we are BC' . Similarly, Rukmani's consideration to change her children's names on their certificates to their Christian names points to a strong desire to have a 'handheld' identity in line with their 'heartfelt' identity. It is clear, however, that compromises are made in order not to miss out on what constitute vital sources of support, and even survival, for many. Villagers' acts are perhaps better conceived of as practical tactics that enable them to make the most of a material bureaucracy that is hard to navigate, not least when one's caste identity, party affiliation, or local politics fail to produce the patronage needed to make headway. Cards and paperwork are inscribed with the cultural politics of identity, while simultaneously acting as the medium of agency through which identities can be recrafted and resources can be accessed.

Conclusions
Dalits and other poor villagers in rural Tamil Nadu engage with bureaucratic processes through a wealth of paperwork and documents required to evidence identity and access welfare. In order to make headway in the maze of state bureaucracy, an array of documents, certificates, signatures, and recommendations are needed. Access to these is mediated through multiple 'political society' actors, including village office clerks, ward and district councillors, party politicians, and panchayat presidents. In line with the critiques by Martin (2014), Veron et al. (2003), Witsoe (2011;, and others, we question political society's capacity to expand freedoms, produce full-fledged rightsbearing citizens, and facilitate the access of poor and lower-caste people to state welfare schemes. The role of political mediation is ambivalent, to say the least. On the one hand, there is little to dispute that ordinary people's access to pensions, rations, and scholarships depends on political mediation (Chambers forthcoming;Piliavsky 2014;Witsoe 2011;, and without the intervention of various political actors the rural poor would struggle to access a range of citizen rights and welfare benefits.
On the other hand, political mediation, rather than guaranteeing equal citizenship to all, also reproduces dependencies and inequalities, along the lines of caste, locality, and party affiliation, among others. These dependencies are reproduced at the state-society interface, where villagers across class and caste need to mobilize political affiliation, caste networks, or local intermediaries in their bureaucratic engagements. For poor and lower-caste villagers, such mobilizations become particularly vital because they rely more than anyone else on the meagre pensions, MGNREGA wages, or educational scholarships that the state provides. In term of mediation, therefore, the rural poor remain at a considerable disadvantage compared to better-off and better-connected higher-caste villagers, and women are considerably more vulnerable than men.
Rather than extending citizenship to all and producing full rights-bearing citizens, political mediation generates a hierarchy of citizens in which the better-networked villagers have access to different -and usually superior -means of political representation than those at the bottom of rural society. The latter in particular need to resort to a series of interventions -including engaging in affective performances and navigating a politics of identity -in an attempt to obtain their legal rights and entitlements. Access to basic citizenship rights as well as invaluable state resources is thus extensively mediated through political society actors, leaving those already economically and socially disadvantaged also politically most disenfranchised. Particularly crucial, we have argued, is the role of the local panchayat president as a key facilitator or obstructer, given that almost all documents need his signature, approval, or recommendation in order to be taken forward. How one relates to the president -be it through caste, gender, class, locality, electoral politics, or political affiliation -is thus vital in one's material encounters with the bureaucracy.
Unsurprisingly, therefore, the politics and processes of acquiring cards and certificates are a highly 'affective' matter, fraught with emotions, fears, and anxieties (Hull 2012). The uncertainties, dependencies, and public humiliations produced by bureaucratic encounters leave many villagers feeling despondent and disempowered in their everyday engagements with state actors and help to reproduce inequality and subjugation. And yet people avail themselves of various tactics to gain visibility, enact political voice, and obtain material resources. One of these tactics involves drawing on familiar communicative modalities such as pleading, begging, crying, and submitting in front of powerful patrons in order to gain their sympathy, approval, or signature. As Appadurai (1990) and Cody (2009) have argued, these communicative modalities are a means through which political representation and citizenship claims are made. While such tactics should be recognized as a mode of political agency, they are also wrought by the politics of gender, caste, class, and religion, and therefore continue to produce citizens with unequal means of representation.
Another tactic involves the manipulation of paper identities and bureaucratic classifications (Shinde 2016;Sriraman 2018;Srivastava 2012). Villagers seek to adjust their paper identities to gain visibility and claim rights wherever possible. They hide their Christian identities in order to access state resources, or foreground them in order to benefit from church scholarships. While this may produce a rupture between their 'heartfelt' and 'handheld' identities, as described by Bakewell (2007), it enables them to play on bureaucratic forms of identity classification, seek visibility from state actors, and construct themselves as deserving and rights-bearing citizens. As Street pointed out in a study of hospital patients in Papua New Guinea, as much as being identified and classified by the state, people themselves can use technologies of legibility and classification to compel the state to see them and to enact a state relation (2012: 2). In searching for recognition as rights-bearing citizens, the rural poor are thus rarely just passive 'patients of the state' , even though in Tamil Nadu -as elsewhere -the tools of political representation remain highly unevenly distributed across rural society.