Anthropological perspectives on documents. Ethnographic dialogues on the trail of police papers

ARTICLES

Anthropological perspectives on documents. Ethnographic dialogues on the trail of police papers

Laura Lowenkron; Letícia Ferreira

ABSTRACT

This article discusses implications and potentialities of anthropologically thinking and researching with documents, arguing that they consist of ethnographic artifacts, which can be particularly fruitful in certain research contexts. In order to do that, it resumes some movements of distancing and rapprochement between anthropology and documents and places two recent ethnographic experiences in dialogue. One a research through police inquiries of Brazilian Federal Police concerning "human trafficking for sexual exploitation" crimes, and the other focused on administrative procedures related to "missing persons" cases investigated by Rio de Janeiro Civil Police. In addition to revealing the heuristic potential of "following the paper" and calling attention to the micro politics of the interactions among those that document and those who are documented, the paper aims to contribute to larger discussions about the challenges that dealing with documents in field work situations poses to anthropology and its self-representations.

Keywords: documents, ethnographic artifacts, Federal Police, Civil Police, human trafficking, missing persons

RESUMO

O artigo discute implicações e potencialidades analíticas de se pensar e pesquisar antropologicamente com documentos, argumentando que estes constituem artefatos etnográficos especialmente rentáveis em certos contextos de pesquisa. Para tanto, recupera alguns movimentos de distanciamento e aproximação entre a antropologia e os documentos e coloca em diálogo duas experiências etnográficas recentes: uma pesquisa em meio a inquéritos da Polícia Federal brasileira relativos a crimes de "tráfico de pessoas para fim de exploração sexual", e outra voltada para procedimentos administrativos em torno de casos de "desaparecimento de pessoa" investigados pela Polícia Civil do Rio de Janeiro. Além de revelar o potencial heurístico da escolha por "seguir o papel" e chamar atenção para a micropolítica das interações entre documentadores e documentados, o texto busca contribuir para reflexões mais amplas acerca dos desafios que a lida com documentos em situações de trabalho de campo aporta à antropologia e às suas autorrepresentações.

Palavras-chave: documentos, artefatos etnográficos, Polícia Federal, Polícia Civil, tráfico de pessoas, desaparecimento de pessoas

Recent anthropological research dealing with documents has contributed to fundamental debates on classic conceptions of both ethnographic practice and ethnographic authority. This article seeks to discuss some theoretical and methodological dilemmas, implications and potentialities of anthropological thinking and researching with documents. In order to do that, it resumes some movements of distancing and rapprochement between anthropology and documents, and places two ethnographic experiences in dialogue. It departs from the assumption that one of the tasks that anthropology systematically faces is to rethink both the construction of its objects, and the selection of the artifacts from which ethnographies are built.

The "post-modern" view of ethnography as text has led anthropologists to turn their attention to other textual genres that had either been set aside or regarded as less important once anthropology moved away from the "armchair" and gained "scientific" authority and legitimacy through malinowskian fieldwork. On the one hand, "archives" came to be viewed as privileged spaces for understanding relationships between colonial governments and peoples traditionally studied by anthropologists (Stoler, 2002, 2009; Souza Lima, 1995). On the other, different modes of document production present in the so-called Western societies came to be analyzed as artifacts and/or knowledge practices that are crucial to understanding ethnographic universes that anthropologists have long been sharing with other disciplines, such as history, sociology and political science (cf. Riles, 2001; Latour and Woolgar, 1997).

Since documents are paradigmatic artifacts of modern knowledge practices (Riles, 2006), which define both ethnography and "native" knowledge practices in some contexts, a large portion of anthropologists' activities during ethnographic fieldwork consists of reading documents produced by their interlocutors (about whom they also take notes). This has happened, as discussed throughout this article, in our own ethnographies (Ferreira 2011; Lowenkron 2012 e 2014). As Vianna (2014) suggests, working with documents, rather than raising questions about their content and reliability, involves questioning oneself about the very act of documenting. Our recent research experiences have contributed to a realization that ethnographic document analysis in the social space within which it is produced, circulated and archived offers a privileged look on the act of documenting. That is, a look on how it is accomplished, what it means and what effects it produces in different contexts.

Reading inquiries and other police documents was one of the central ethnographic activities of the research carried out by one of the authors of this article on "human trafficking for sexual exploitation"¹ (Lowenkron, 2014). Since a crucial part of police work is documenting investigations, "the interest our research subjects have shown in documents had become contagious" (Riles, 2006, p. 8). During fieldwork, the researcher not only looked up inquiries in Federal Police headquarters and produced ethnographic accounts about them, but was also able to personally follow the production of some of its documents by its agents, police commissioners and registrars during criminal investigations. Reading and directly observing the production of these artifacts highlighted how those who are regulated and constituted by these papers are not inert with regard to state documentation practices.

Contact with police reports and other documents pertaining to "missing persons" cases that were produced, circulated and/or archived by a sector of the Rio de Janeiro Civil Police was a central element in making fieldwork possible in the research carried out by this article's other author (Ferreira, 2011; 2013). On one level, the documents worked as an inescapable material support for affirming authorities and giving authorizations without which fieldwork would not have been possible. This reinforces the prevailing notarizing logic of Brazilian police units (Kant de Lima, 1995; Miranda et al, 2010). However, on a second level, the documents became the most crucial research object, both due to how they are viewed by the inspectors, investigators and commissioners who produce them and due to the crucial role they play in police management of "missing persons" cases in Rio de Janeiro.

Based on these ethnographic experiences, which show how documents, seen within their production, circulation and archiving contexts, are especially profitable ethnographic artifacts in certain research contexts, we seek to contribute to broader discussions not only on the dynamics, effects and powers mobilized by documentation practices, but also on the challenges that dealing with documents in ethnographic fieldwork brings to anthropology and its self-representations. In order to do so, we analyze how these challenges presented themselves in our researches, proposing a dialogue with some discussions that have weaved through contemporary anthropological debates of the theme.

Distancing and rapprochement between anthropology and documents

In the mid-20^th century, Evans-Pritchard (1950) was already drawing attention to the a-critical use of document sources as one of the problems of the not only a-historical, but anti-historical approach adopted by structural-functionalist anthropology, which established "participant observation" as its privileged method. If his warning "had little resonance at the time, it certainly has more today" (Stoler, 2002, p. 90). Over the past few years, several anthropologists have highlighted the need to go beyond documents' informational and instrumental dimension, analyzing them as producers of knowledge, relations, effects and affective responses (Stoler, 2002 and 2009; Cunha, 2004; Navaro-Yashin, 2007; Gupta, 2012; Ferreira, 2013) or even as artifacts whose material, aesthetic and formal properties, as well as social life (or socio-technical processes) can no longer be easily ignored (Latour and Woolgar, 1997; Riles, 2006, Reed, 2006; Hull, 2012b).

Despite their recent rehabilitation as ethnographic artifacts, anthropologists' relation to documents is still shot through with tensions, doubts and anxieties, no longer because it threatens the specificity of anthropology's knowledge practices (if we believe that these dilemmas have, or should have been, overcome), but because it produces new analytical challenges. Since ethnographic knowledge was historically represented as "diverse from, and even opposed to, that which results from documentary research" (Cunha, 2005, p.36), the task of ethnographing documents may still seem strange and obscure to most ethnographers, in contrast to the allegedly self-evidence of the expression "participant observation".

The epistemological devaluing of these objects, which delayed their ethnographic scrutiny, may be attributed to different reasons. On the one hand, ethnographing documents does not guarantee the same strategies of authority derived from the certainties (if not of objectivity, at least of controlled partiality) produced by the experiences of direct contact and dialogue with the persons who inhabit ethnographic narratives. On the other, it requires that the researcher establishes a dialogue with, and consider the agency of, those who are not immediately identified as subjects in the research: the persons being documented and the very papers or other material supports of graphic, textual, sound and/or audiovisual records. Furthermore, there are those who consider that bureaucratic writing was long neglected because it is a very similar knowledge practice to that of anthropologists (Riles, 2006b), because it is tedious and lacking the symbolic richness or density of meaning which make a geertzian ethnographic description possible (Graber, 2012), or because it is easy to see documents as simply giving immediate access to what they document, denying the role of mediation (Hull, 2012b).

The most classic way of dealing with documents is precisely to look through them, and not so much at them (Hull, 2012a, p.253), in order to produce ethnographic narratives about scenes, discourses and events that were not directly observed by the researchers. That is, that were not registered in their field notes firsthand, but rather came to them previously documented by others (Vianna, 2014). Starting with the "historical turn" of the 1980s, the idea of an ethnography of/in archives (Castro and Cunha, 2005; Cunha, 2004 e 2005; Castro, 2008; Zeitlyn, 2012), in its multiple modes and metaphorical extensions, became increasingly less exotic and more frequent.

Since then, new methodological strategies have emerged for dealing with documents. As Zeitlyn (2012) discusses, one of these strategies is reading archives against the grain², that is, subverting the modes of understanding that were imagined and intended by the administrative rationalities that produced and maintained them. This epistemological posture led to the development of critical and imaginative approaches to traditional sources. Its main goals were to excavate subaltern and silenced voices, to rescue actions (or agencies), perceptions and small gestures of resistance from those situated at the margins of official records and, thus, to promote the "insurrection of subjugated knowledges" (Foucault, 1980, p. 81).

More recently, Stoler (2002, 2009) questioned how most colonial archive scholars could so quickly turn to readings against the grain without first moving along the grain. According to her, "assuming we know those scripts (...) diminishes our analytic possibilities" (Id, 2002, p. 100). By analyzing the archive as a cultural artifact, she seeks to understand the perspectives and concerns of its producers and administrators, giving particular attention to conventions that shape what can and cannot be recorded: repetitions, acts of forgetfulness, different modes of silencing and the credibility hierarchies that delimit qualified and disqualified knowledges.

By moving anthropological self-representations away from ethnography as "participant observation" or as writing, toward ethnography as modes of reading, these lessons from ethnographies of/in archives allow us to broaden the analytical possibilities of "current files"³ read during ethnographic field-work in contemporary bureaucratic organizations. As we will see, the combination of these two reading strategies (against and along the grain) in the ethnography of "human trafficking" inquiries seeks to render explicit the power relations that permeate the production of these documents and are fixated in these papers. At the same time, it seeks to turn documentary ethnography into a way to subvert the finality and logic of this administrative police procedure. In order to do so, we must remember that, as Hull (2012b) suggests, analytically restoring documents' visibility means treating them as mediators, that is, things that "transform, translate, dislocate, distort and modify the meaning or elements that they supposedly carry" (Latour, 2005, p.39).

Restoring documents' visibility and mediation role, especially of those documents produced in bureaucratic organizations, has led recent ethnographies to explore the material dimension of this paperwork. These ethnographies, many of which are heirs to works that promoted important changes in the way that anthropology deals with objects (cf. Appadurai, 1986; Miller, 2013), pursue not only modes of reading, but also carrying, handling and trading documents to their ultimate analytical consequences. They draw attention to what makes papers similar, in terms of their effects, to other equally ordinary artifacts from these institutional spaces. From this perspective, "for many genres of documents, it is often less important what they stand for than, like tables and desks, how they arrange people around themselves" (Hull, 2012b, p.134). It is therefore up to the ethnographer to face them as more than recording instruments used by bureaucrats and to apprehend in what way they constitute, hierarchize, separate and connect people.

A paradigmatic example of this movement is Hull's ethnography in Pakistani bureaucratic offices, which explores the centrality of documents and documentation forms in Islamabad's governance (2012b). Working from the premise that the material qualities of governmental papers are not only elements mobilized in the production of meaning, but are also driving forces behind other types of processes, his work analyzes, among other issues, the relationships between carrying, signing and producing copies of documents and the diffusion of individual responsibilities in favor of a certain type of "collective agency". Navaro-Yashin (2007), in turn, emphasizes the mutually constitutive relationship between producing, exchanging and transforming documents and the affects retained, transported and experimented by those who handle them. In Brazil, the pioneer and crucial works by Peirano (1986, 2006a, 2006b, 2009) have long drawn attention to the fact that the possession of documents has the capacity, in certain situations, to performatively make the citizen. For identification documents, according to the author, this capacity is largely derived from the redundancy between material elements found in them, such as the documented persons' signature, 3X4 photo and fingerprint (cf. Peirano, 2006a, p. 36).

The growing attention to the materiality of documents in ethnographic works has thus shed light on the fact that papers and forms, as well as photos, signatures, seals, stamps and other graphic artifacts⁴ regularly used in bureaucratic organizations, play a role both in controlling and coordinating procedures, agents and administrative actions and in constructing subjectivities, affects, persons and relationships that extrapolate organizational universes. Thus, ethnographies that pay attention to these artifacts, and not only to what can be seen through them, have made clear that document production, circulation and archival must not be viewed as isomorphic to the organizational structures in which they take place. They have also illuminated the socialities and entanglements that these processes produce, as well as the borders they are able, at times, to cross and, at others, to reaffirm.

Following the trail established by these works and exploring the materiality of documents in a research on police reports of "missing persons"⁵ we are able to ascertain responsibilities, obligations and affects that are created by the production, circulation and archival of police reports and other papers brought together as inquiries within a sector of the Rio de Janeiro Civil Police. In this context, boundaries between what would be police responsibilities and what are considered attributions of "families of missing persons" are both produced and traversed by the papers handled by police officers. Furthermore, in this context, documents covered by certain signatures and stamps function as conditions of possibility for establishing certain socialities, among which the one established between the researcher and her interlocutors.

We will now present the reasons for, and analytical profitability of, our choice to "follow the paper" (Hull, 2012b, p.22) inscribing them within the broader context of these theoretical-methodological discussions. This choice is presented for the two studies already discussed: the study in a Brazilian Federal Police precinct that sought to understand administrative practices and rationalities responsible for constructing and managing the crime of "human trafficking for sexual exploitation"; and the study carried out in a sector of the Rio de Janeiro Civil Police that sought to analyze how "missing person" cases are managed in Brazilian police units. Both ethnographies pay particular attention to the micro-politics of interactions between documenters and the documented subjects.

The art of "reducing to writing": along and against the grain of a police inquiry

As some authors have pointed out, we cannot mistake the inquiry for the police investigation (Misse, 2011; Vidal, 2013). The inquiry's main goal is to ascertain that a particular criminal action took place and to establish its authorship. It is a "piece composed of technical reports, accounts recorded by a notary public and a legally-oriented report, signed by a police commissioner with a degree in Law" (Misse, 2011, p.19). It is also, however, an instrument through which the police investigative work is formalized and officialized⁶. As legal jargon puts it, "what isn't in the documents isn't in the world", so that an undocumented investigation is as if it had not existed.

If the inquiry is the artifact through which the Federal Police's investigative and administrative actions become legible and real before their own artificers, ethnographing these practices requires that one follows the paper. The initial idea was to approach these documents as a means to understanding the administrative rationalities responsible for constructing and managing the "human trafficking" crime and its respective social characters ("traffickers" and "trafficked persons"). This meant reading them along the grain. However, it gradually became clear that the process of producing, dislocating and fixing official and unofficial conceptions of "human trafficking" in police inquiries did not only depend on the performances of their multiple documenters, but also of the people being documented. In order to understand how the alleged "victims" of this crime construct "spaces of agency" (Piscitelli, 2013) within certainly asymmetrical interactions with state apparatuses, one had to simultaneously read the inquiries against the grain.

In this sense, the ethnography of "human trafficking" inquiries establishes a dialogue with other works that, instead of a "victimizing" perspective, suggest giving voice to the experiences of people who participate in these mobilities, making it possible to confront their narratives with hegemonic accounts about them (Agustín, 2005; Piscitelli, 2008). Based on this confrontation, research has shown that these migrants' trajectories generally do not fit the myths and stereotypes disseminated by the media and anti-trafficking campaigns (Piscitelli, 2008; Teixeira; 2008; Blanchette e Silva, 2010).

This suggestion is in line with the police perception of the phenomenon, so much so that they most often characterized "human trafficking" less in terms of its legal definition than based on a contrast with what they assumed "outsider" interlocutors (including the researcher) expected of this crime.

Almost all federal police officers heard during the research began conversation by alerting the researcher that there were few inquiries on this crime. They seemed not to assign much relevance to the subject, partly because they considered the investigations fruitless. "Sometimes the investigation turns up nothing because, simply, there is nothing. Just because an inquiry was started does not mean that there is a crime", one commissioner points out. "There's plenty of revenge-hotline", according to the agents, who claim that people use the Public Security Office's "Crime-hotline" in order to harm someone because of some conflict or personal enmity. However, police officers were unanimous in considering that the biggest difficulty in "human trafficking" cases is that "victims" don't see themselves as such, which reveals that authorities assign them the co-responsibility for elucidating (or not) the crime. As one of the agents put it: "usually the victims deny everything. They don't deny it only when they are very badly harmed. Then they come clean. But usually they deny it, because they do well out of it."

One commissioner claims not to like to work with this type of crime because no one wants to cooperate with the investigation. He says the "victims" are a sort of "accomplice to the crime", that is, they collude with and protect the criminal. He states that the "victim" is almost always aware of the risks, chooses to engage in sex work and wants to go abroad where she has the opportunity to earn more. Thus, they seek to differentiate the reality of "human trafficking" found in police investigations from the idealized representations according to which "trafficked persons" would be deceived as to the purpose of their dislocation and/or forced into prostitution, emphasizing their consent and agency in order to deconstruct the idea of violence, even when they acknowledge the "typical conduct", that is, defined and criminalized in criminal law.

However, instead of discussing the agency of alleged "victims" in their migratory and mobility projects, as emphasized by some academic research on the subject and as noted by the police officers, we want to propose a different analytical focus and consider their agency in the face of state administrative apparatuses. This implies, on the one hand, taking into account the foucauldian suggestion that subjects are constituted ("subjectified") amidst a multiplicity of correlations of force, through dynamics of subjection and resistance, both of which are equally important in thinking about the notion of agency (Mahmood, 2006). On the other hand, one must remember that, in state bureaucracies, documents are privileged means of diffusing responsibility and of producing collective agency (Hull, 2012b), or, in this case, an agency shared between documenting agents and documented persons.

Without losing sight of the asymmetrical nature of this relationship in analyzing individual performances, the possibility of observing interactions between individuals and the "State" (as an administrative apparatus) means that the Federal Police, as a field for research, is a productive space in which to further the reflection on the potentialities of the concept of agency. Agency here is never understood as free will and not necessarily viewed as resistance, but rather "in an anthropological sense, as the socially and culturally mediated capacity for action" (Piscitelli, 2013, p. 22). What is noticeable is that alleged "trafficked persons", based on the relative authority they are given as privileged "witnesses" in these police inquiries, manipulate information according to their own interests. Thereby, they usually resist the criminalizing/victimizing process, since they rarely report on, or contribute to establishing the veracity of the offense of which they are, in theory, "victims".

Based on the premise that writing is one of the most important state activities and government technologies (Gupta, 2012; Das, 2007), the ethnography of police documentation practices aims to observe how these persons' statements are "reduced to writing"⁷ in police inquiries, as well as to render clear the regulating structures that govern these discursive formations and deformations. The so-called "Term of statements" is the main type of bureaucratic writing that guarantees that discourses of alleged victims (converted into "witnesses") "subsist and, at the same time, regularly change" (Foucault, 2014, p.159) throughout processes of document production and superposition.

Statements "reduced to a term", which are enunciated amidst the embarrassment of police "interviews", are registered in a formally rigorous document, marked by repeated narrative conventions and a legal form of its own. Converting oral speech into writing, in this case, means reducing and substituting answers given by questioned persons with the interrogator's indirect discourse. The questions are silenced and only signaled through the generic formulation "when inquired about the facts, ANSWERED: THAT...", situated at the beginning of the document, after the date, place, deponent's complete "qualification"⁸ and the name of the police officer who presided over the interview. At the end, the term must be read and signed by all (deponent, police commissioner and registrar), attesting its trustworthiness and making it official.

Reading "Terms of statements" and their ramifications along the grain of "human trafficking" inquiries showed how the margin of agency of witnesses and suspects is only reduced in documentation procedures when their versions are confronted with the "social force" of other documents (Ferreira, 2013), such as transcripts of phone taps. Observing a police interview in person, in turn, enabled the researcher to follow the police art of "reducing to term"⁹ in action and, thus, to understand and subvert the discursive formations and deformations that this type of bureaucratic writing produces.

Furthermore, observing police interviews in persons also enabled the researcher to comprehend the micro-politics of interactions between documenting agents and documented persons. To present these micro-politics, we will now describe the production and the uses of some documents, part of one of the few inquiries included in the research that did establish that the "human trafficking" crime took place and who its authors were¹⁰. This inquiry was initiated based on a "letter rogatory" that formalized an international cooperation request, noting the participation of a Brazilian citizen in a criminal group that brought women from Brazil to Portugal to engage in sex work. Most members of the group had already been arrested and convicted in Europe, not for "human trafficking", but for the crimes of "pimping" and "aiding illegal immigration" – which reveals that individuals who were defined as "victims" in Brazil were viewed and treated as "undocumented immigrants" in Portugal. However, this divergence of classifications is silenced over the course of the police inquiry, without damaging diplomatic relationships that were established and renewed by the international circulation of official documents.

The only accused who lived in Brazil was "indicted"¹¹ in the Federal Police based on copies of documents extracted from the Portuguese inquiry and from a "Term of statements" produced by the Federal Prosecutor who received the request for international legal cooperation. Initially interrogated at the Federal Prosecutor's Office, the woman claimed to be a domestic worker and that she received financial assistance from a daughter who lived abroad, who was married to a Portuguese man and whose pay came from the beverages consumed by customers she accompanied at the nightclub where she worked. The document further notes that the woman claimed her daughter did not engage in sex work, but was unable to say whether or not there were any sex workers at the nightclub. She also assured the officer that she had never discussed any subject involving sending Brazilian women abroad for purposes of sex work over the phone with her son-in-law. The term closes with a description of the scene in which the suspect is emotionally destabilized by records of a phone call, followed by the bureaucratic demonstration that the Federal Prosecutor followed legal formalities:

After being confronted with the phone transcripts present in the inquiry, she became reticent and, instructed by the Federal Prosecutor that she was not obligated to answer and that she could ask for a lawyer's guidance, she preferred to say that she did not have that conversation and to remain silent, stating that she would no longer answer any questions. Thus, it is determined that the interview be terminated and she is instructed to seek the Federal Public Defender's Office.

During the ethnographic fieldwork, the researcher also directly accompanied the process of producing the "Term of statements" of a "victim"/"witness" in the same police inquiry. Before beginning the "interview", the commissioner quickly analyzes the inquiry and marks the pages he considers most important to help him during questioning. The witness also appears carrying documents. Besides the copy of the summons, she brought along a folder with the lease of an apartment located in the city where she lived and the diploma of a beautician course she attended in Portugal. Since these papers did not contribute to establishing that the crime did (or did not) take place, only the personal identity documents required for making her presence in police headquarters official were examined and registered at the beginning of the "Term of statements".

The commissioner was convinced about the purpose of her trip based on the Portuguese police inquiry. Thus, he merely sought to confirm and document how and by whom the young woman had been invited to travel, as well as to obtain her confession about engaging in sex work, a necessary condition for establishing that the crime took place. The witness, in turn, sought to protect her own reputation. If, by selecting what will and will not be recorded, the police officers are responsible for "simultaneously rendering certain accounts visible and invisible – or audible and inaudible" (Vianna, 2014, p. 54), this does not mean that the documented person passively submits to the official documentation act. On the contrary, the deponent seeks to negotiate the choice of words and shape, up to a point, the architecture of the narrative that will be fixed on paper. Thus, the police inquiry can be read, against the grain, as a vehicle for her perspectives, experiences and aspirations, despite the evident deformations produced along the mediation process.

The young woman, who seven years before had twice traveled to Portugal to work at the same commercial establishment, recounted that, on the first time, her sister-in-law had told her that a friend had made a great deal of money at a "singles' bar", receiving 50% of the value of drinks offered to, and consumed by, customers. They decided to leave together for Portugal because it was "just drinking". They were then introduced to an older woman (towards whom she evidently still felt anger) who was responsible for attracting women in Brazil and mediating the entire immigration process (obtaining a passport, buying plane tickets and contacting the club abroad). The commissioner interrupts her, stating that it is of no use if it is not on paper. He then reads aloud everything that was recorded and asks that the young woman confirm the story and that she continue more slowly.

The commissioner asks if there were "striptease" shows or rooms for sexual encounters at the bar, making explicit a suspicion that could have been previously ruled out had he read the inquiry documents more attentively. She denies, "no way, none of that! It was a bar", stating that customers went there only to drink and talk. "But the Portuguese are really dumb if they go to a nightclub and pay more just to talk!", the commissioner says, ironically. She corrects him, "it's not a nightclub, it's a bar", and then counters that in Brazil things might be different, but Portuguese men went there after companionship, because they were "needy", so much so that they would then want to have lunch, dinner, even marry the girls. She recounts that most of the Brazilian women married Portuguese men, including her sister-in-law, who was " happily married", and that she also lived for four years with a man she met at the "singles' bar" in Portugal. The dialog was translated by the police authority and thus "reduced to term" by the registrar:

THAT the work at the "singles' bar" consisted of encouraging customers to consume alcoholic beverages and that they would get 50% of the value consumed, while the other 50% remained with the nightclub; THAT at the place there were only music and sale of alcoholic beverages, with no striptease shows or places for sexual encounters, being an actual bar. (researcher's emphasis)

Despite the somewhat anxious and hurried tone, the woman seemed firm and sure of what she wanted to tell or not tell the police officers, trying to present the relationships' dynamics in her own terms. The first time, she traveled by her own initiative and paid her own tickets. That way, if she did not like it, she could leave. The second time, she at first did not wish to go, but the recruiter insisted. She then accepted a loan in order to buy the plane tickets, but settled her debt in one month because she was smart and did not drink, but rather pretended to, avoiding losing control so as not to be "swindled" during payment, which happened to the women who got drunk. The first part of the information was registered, but the more detailed explanations regarding the possibility of being swindled and her strategies of resistance were ignored and silenced. This does not mean that the commissioner wanted to erase the agency of so-called "trafficked women". On the contrary, he was emphatic in underlying her consent, even if by law that is irrelevant in configuring the "human trafficking" crime. Contrasting the concrete case with the idealized version of the crime, the officer notes:

THAT during the entire period she worked at the nightclub she did so of her own volition, not having been held anywhere against her will; THAT her passport was not held by any person, because it was in her possession the entire time she was in Portugal.

Retaining passports for the purpose of collecting debts, despite also not being a legal requirement in formally characterizing the "human trafficking" crime, is one of the most common images in political and media representations of the phenomenon, symbolizing the curtailment of freedom. It is worth noting, however, that for immigrants in an irregular situation, possessing a passport with an expired or wrong visa (tourist instead of work visa) does not prevent them from being viewed and treated as "undocumented". Stripped, after a few months, of its supposed power of performatively making the citizen (Peirano, 2006a), this document is converted, beyond national borders, into an "ephemeral object" (Navaro-Yashin, 2007), whose strength of materiality subsists only as a fetish of the Nation-state.

The young woman questioned at the Federal Police recounts that singles' bar managers charged more or less double the value of the plane tickets, claiming this was done to cover other costs, such as free food and housing for the first weeks. The commissioner then interrupts her: "did they refrain from charging you if you arrived on time at the nightclub?". She answers she does not remember exactly. The commissioner tells the registrar that she vaguely remembers that she had to arrive by 6pm. She says she does not remember that. "That's why I said 'vaguely remembers'", the police officer offers as justification. He then reads a portion of the transcripts of a phone call that confirmed this information:

"... if I'm on time, like, if I arrive at six, I don't have to pay rent, housing, or food, dinner". "I have to arrive at six, half past six at the nightclub, so I don't have to pay housing and food, get it?" (a recently-arrived employee explains to her Brazilian interlocutor how the work regime at the Portuguese club functions).

Still not satisfied, the commissioner makes one last attempt to extract from the alleged "victim" a confession capable of establishing that the crime took place, saying all the clients questioned during the Portuguese investigations, including her ex-boyfriend, claimed the girls at the "club" engaged in sex work. The witness defends herself from the accusations: "If they did it, it was when they went out with the customers, when they went out to lunch or have dinner with them, or accompanied them to a mall and then, I don't know, maybe they went to a hotel... But I never did it and never knew of it". Insisting on her argument, she says: "do you understand? The woman had to go out with the customer until she won him over and married him, that's how I met my boyfriend and got together with him." Complementing what the young woman said with other information recorded in the inquiry documents (in other witnesses' "Terms of statements", in phone transcripts and Portuguese police and court reports), her explanations were thusly "translated to term":

THAT she is aware of girls who worked at the Club who engaged in sex work, but this was done during the day and outside the Club; THAT many of them dated and even married men they met at the place; THAT the DEPONENT herself dated a Portuguese man, for almost 4 years, whose qualifying data can be found in pages 580/581 of the inquiry; THAT she even lived with this Portuguese citizen, in Portugal.

If, on the one hand, the "deponent's" answer denotes an effort to erase the more commercial dimension of sexual exchanges (prostitution), which was reintroduced by the commissioner through the decodification of her discourse fixed on paper, on the other, it reveals that relationships between singles' bar employees and customers did not always fit the "prostitution" box, according to their own perspectives. The relationships established between men and women in these spaces were diverse and constituted a continuum of exchanges of sex, affection and money (Piscitelli et al, 2011) that ranged from commercial exchanges around the consumption of alcohol inside the club, included "going out" (which could involve actual prostitution or less formal and commercial meetings, such as an invitation to dinner), and went up to relationships and marriages.

Before signing the document, the deponent reads everything attentively. She corrects a misspelled word and the value charged for the plane tickets. She confirms everything is "alright". The registrar prints out the "Term" once more and hands it to her so that she signs it. She re-reads the whole content and complains about the word "nightclub". The registrar argues that, in Brazil, "nightclub" is a place that sells alcoholic beverages... She says "okay" and, then, finally signs and receives a copy of the "Term of statements". Before she leaves, she asks if she will be called again. The registrar says it is unlikely, "at least, not by the police".

Going back to Ginsburg's (1991) analogy between inquisitors and anthropologists, reading these documents along and against the grain enables us not only to critically view this form of state documentation and administration, but also to rethink similar artifacts in anthropological knowledge practices, since the same analytical operations could be carried out toward the anthropologist, who silently witnessed and recorded this interaction. Furthermore, the analysis sought to draw attention to the fact that neither the "inquisitorial" character of interviews in police headquarters, nor the re-shaping produced in the mediation process that converts witness speech into "Terms of statements" turn documented persons into inert objects of administrative documentation practices. It is not for nothing that these procedures rarely lead witnesses to report the crime of which they are supposedly "victims".

To conclude this account we must highlight that, unlike the case chosen for analysis, most "human trafficking" inquiries initiated in this Federal Police precinct led not so much to the construction than the deconstruction of the materiality of the crime (evidence). Thus, instead of leading to a criminal report and a judicial process, they generally wind up being archived by the Federal Prosecutor's Office. However, given that statistics are one of the main state narrative and administrative genres (Gupta, 2012), by moving from the documents' qualitative dimension to their quantitative effects, we can observe that their supposedly inexpressive result does not mean they "go nowhere", as the police officers used to say. In official reports about the phenomenon, the number of inquiries is more important than their content, serving as a support for policy formulation. Thus, the heterogeneity and "the messy details of cases are replaced by the precise factuality of numbers" (ibid, p. 156), which are the main criterion in measuring the productivity and effectiveness of this and other criminal policies.

In this sense, the police inquiries' institutional ramifications, if evaluated beyond their criminal aspects, make clear how bureaucratic government is exercised not only through the articulation between knowledge and power (Foucault, 1980), but also through the everyday production of ignorance and even stupidity, in Graeber's (2012) terms. The political strength of ignorance can be perceived in the way in which so-called "quantitative data" is used at the highest levels of the state administrative hierarchy, which are responsible for formulating national reports and maintaining a dialogue with supranational government agencies. In Brazil, the low number of "human trafficking" inquiries and their low "productivity" in terms of criminal convictions are converted into uncertain statistics and policies of confrontation that are reproduced and legitimize themselves less through what is actually documented by state organs (such as the police) acting in the operational front lines to combat this crime than by the proclaimed need to improve the regulation of something that is imagined not to be duly documented.

Inside and outside of drawers: relationships, feelings and commitments on paper

If we cannot mistake the police inquiry for the actual investigation, the same cannot be said about documents produced in "missing persons" cases reported to the Rio de Janeiro Civil Police. In the state's precincts, every "missing person" case is the object of a "Verification of the Merits of the Information" (Verificação da Procedência da Informação – VPI), which consists of an administrative set of documents endowed with a different status from the inquiry. Though it designates documents, the VPI was originally conceived and regulated by the Brazilian Criminal Code as a preliminary stage of all police investigation, which must precede the inquiry and determine whether or not it should be initiated. However, in the precincts' day-to-day routine, this expression came to name a set of papers, establishing, in practice, a coincidence between investigative procedures and documents (cf. Eilbaum, 2012).¹²

VPIs are not initiated only for "missing persons" cases, but for a variety of occurrences reported to the precincts. Nevertheless, the coincidence between procedures and documents implied in its name is emphasized here in order to introduce the reasons for, and effects of, the choice for dealing with documents in a research specifically concerned with "missing persons" cases.

The research's main objective was to understand, through ethnographic fieldwork, how cases classified as "missing persons" are administered in Brazilian police units, apprehending both practices and representations about this type of occurrence that are present in precincts. Initially, there was no predefined choice for the use of documents as artifacts based on which the ethnography would be constructed. Since the negotiations that made the research possible, however, it became imperative to deal with papers, given that, in that context, processes and documents maintain close and definitive relationships. Though extremely revealing, the implied coincidence found in the VPI initials is only one of these relationships.

"Missing persons" is, among so many others, a category used in Brazilian Civil Police precincts to name a specific type of occurrence. Cases of adults who leave their homes, jobs and routines without informing their whereabouts; public calamities, natural disasters and inclement weather episodes whose victims are not located; children and adolescents who run away from home, shelters or youth detention centers; elderly or mentally ill individuals who get lost in their own cities. Amongst many other, dramas, plots and characters like these constitute "missing persons" occurrences registered and investigated in VPIs. "Missing persons" cases are part of the universe of so-called "atypical facts" registered in precincts: events that are not established in Brazilian criminal law and, therefore, are not crimes (cf. Paes, 2008, p. 173).

Ethnographic reflections on Brazilian public security and criminal justice have long shown that police officers hierarchize occurrences with which they deal, adding to formal categories such as "occurrence", "missing persons" and "atypical fact" a series of informal classifications, labels and stereotypes (cf. Kant de Lima, 1995; Kant de Lima, Pires and Eilbaum, 2008; Miranda et al, 2010). One of these informal classifications is the one separating the so-called "feijoada"¹³, a set of occurrences considered to be of little relevance, such as threats, domestic disputes and thefts for which there is no evidence, from the so-called "actual crimes", that is, homicides, kidnappings and robberies, for example¹⁴. The research's initial objectives included the interest in understanding the place "missing persons" cases occupied in police officer's informal classifications.

In order to approach this issue ethnographically, the researcher sought access to a sector of the Rio de Janeiro Civil Police that specializes in "missing persons" cases: the Sector for Discovering Whereabouts (Setor de Descoberta de Paradeiros – SDP), part of the state capital's current Homicide Unit. During her first visit to the SDP, the researcher was greeted by the sector's then-interim chief, inspector Fernando, and quickly instructed to formally request an authorization to carry out the study. The authorization should be addressed to the office of the Rio de Janeiro Civil Police Chief. The researcher followed the instructions exactly and quickly, but the document only arrived at the office after passing through different sectors of the Civil Police and receiving more than a dozen stamps and small notes on its margins until, finally, being inscribed with the short sentence that allowed the work to be carried out: "There are no objections to the researcher's academic aspirations."

The need for not only a piece of paper, but one that circulated through several sectors of an organization at a pace independent of the anthropologist's wishes put her in contact with what is understood in Brazil to be the oppressive materialization of the ideals of formalism, impersonality and anonymity prevalent in bureaucracies: the inescapable need for records, letterheads and signatures in order to set a process in motion or to obtain all manner of rights (cf. Reis, 1998 e DaMatta, 2002). Furthermore, the need for explicit authorization also made clear a key aspect of the way in which bureaucratic organizations work: the imposition both of uncertainty and of waiting as an efficacious resource of power (cf. Hoag, 2011, p.86).

Additionally, reading the short sentence that, at last, authorized the research presented the anthropologist with "an ethnographic fact even before reaching the 'field'" (Peirano, 2009, p.54). Reading documents became not only a routine activity, but also an involving datum of the then-beginning fieldwork. Though the request had been deposited in an SDP drawer, from which it was only removed for inspection once, other papers, organized as VPIs, were routinely removed from drawers and systematically read, transcribed and handled by the researcher. These acts, though very similar to what the police officers themselves did everyday, were seen by inspector Fernando and his colleagues as almost opposite tasks to what could be seen there, should the anthropologist have decided to observe the sector's day-to-day routine.

At the initial negotiation regarding the fieldwork routine, held between the researcher and inspector Fernando, the latter presented "observation" and "dealing with documents" as opposite options in terms of the effects they could produce on the SDP day-to-day routine. Immediately after the researcher stated that she intended to "observe the sector's day-to-day routine and talk with police officers", the inspector warned her: "that word observe will cause you trouble". According to Fernando, talking to police officers and observing their work would lead her to a crossroads: either they would tell her what they do and ought to do in an idealized way and she would write something "pretty, but unrealistic", or she would write about what actually happens in a police unit, producing "realistic" work, but drawing the officers' animosity. He suggested, as a (in his words) "great alternative" to circumvent the crossroads, that she turn her attention to the papers received, produced and stored in the SDP's drawers and lockers, what he and his colleagues referred to as "our archives". Research with documents, said Fernando, would be "easier".

By referring to documentary research as a "great alternative", "easier" because it was dissociated from what was "real" in the day-to-day SDP activities, inspector Fernando expressed a graphic ideology (cf. Hull, 2012b) common among bureaucrats and other documenters. These agents "much like scientists, claim to represent, engage with, or constitute realities 'in the world' independent from the processes that produce documents." (Ib, 2012b, p.5). For Fernando, after all, while the researcher was dealing with documents produced and handled by himself and his colleagues, she would not be observing what they "really" do. Furthermore, if this graphic ideology is present among all kinds of documenters, for SDP police officers specifically it is reinforced by two characteristics of the papers with which they deal. First, their status as administrative procedures and the lack of importance that follows from it. Second, the incongruence between material elements and narrative conventions present in these papers and the case plots registered therein.

The specific VPIs that circulate within the SDP are initiated in Rio de Janeiro city police precincts. The precincts are first reported of "missing persons" cases by people classified by police officers, often regardless of factual connections or blood ties, as "families of missing persons". Once a "family of missing person" reports a case in a precinct, a police officer must open a VPI. Some time after this initial step, VPIs must be forwarded to the SDP for specialized investigations¹⁵. Unlike police inquiries, VPIs are administrative procedures initiated for non-criminal occurrences. The central issue, however, is not their distinct nature but rather the different status attributed by police officers to VPIs and inquiries. Police officers, reproducing the hierarchy implicit in the distinction between "real crimes" and "feijoada", understand that inquiries are properly police documents and, thus, endowed with evident relevance, while VPIs are, in their words, "only administrative procedures", unimportant, incapable of revealing the "reality" of police work and, for this very reason, liable to being touched, read and transcribed even by a researcher¹⁶.

Nonetheless, within the heterogeneous universe of VPIs, police officers understand that those initiated for "missing persons" cases are even less important than the rest. Using this other informal classification, that at times is added to the separation between "real crimes" and "feijoada", SDP police officers often state that "missing persons" cases are "family affairs", in opposition to "police matters". While homicides, kidnappings and robberies, among others crimes, are "police matters" demanding street operations, team investigations and other activities part of "police work", "missing persons" cases are, to sector agents, "family affairs" that should not be left up to police units. Police officers say that, when faced with these cases, they have little to do but "to fill out pieces of paper" (cf. Ferreira, 2013) and to tend to the so-called "families of missing persons", carrying out a type of service they classify as "social work"¹⁷.

Inspector Fernando usually bases the use of informal classifications such as "family affairs" and "social work" on one of the main state narrative and administrative genres: official statistics, in this case produced by the Rio de Janeiro Public Security Institute (Instituto de Segurança Pública – ISP). For some years, ISP has claimed, based on quantitative data that the majority of "missing persons" cases reported to the state's precincts end with the return of the missing persons to their homes or at least with the discovery of their whereabouts¹⁸. For Fernando, that data prove what he sees everyday: that most occurrences are family conflicts leading to someone running away from home, to couples separating or to one of the persons involved in the conflict explicitly deciding to leave home. These conflicts would erroneously arrive at police units, leaving officers like himself and his colleagues to "merely fill out pieces of paper" – not any paper, but especially unimportant documents – and await their resolution. Nonetheless, for Fernando, these numbers also lead to SDP's work being "an illusion": while he and his colleagues "merely fill out pieces of paper", most cases are solved and counted as such, which enables their superiors in the state hierarchy to use numbers suggesting the SDP is an efficient service for locating missing persons. What the numbers do not show, however, is that many cases are solved because the missing person voluntarily returns home, and not as a result of successful investigations.

A clear incongruence between the material elements of the VPI documents and the cases plots also accompany this "illusion" produced by quantitative data and the lack of importance attributed to the papers. This incongruence adds to the VPIs' inferiority the idea that they are ineffective documents in terms of investigations, reinforcing the "illusory" character of SDP's work. They both lead to the production of vague and imprecise records of the cases and the choice, by many police officers, to leave fields blank or to fill them out with expressions such as "ignored", "unknown", "does not know", among others.

Reports of occurrences, official correspondence, and "terms of statements", among other documents included in VPIs, contain inscriptions and insignias of the Rio de Janeiro Civil Police and/or the precincts and sectors in which they are produced, as well as categories, fields for certain kinds of information and narrative conventions that are exhaustively repeated. Categories such as "author", "victim", "witness", "location of occurrence", "date and time of occurrence" and "goods involved", for example, appear in the police reports, guiding the work of every police officer who fills out this documents when faced with a case – whether or not it is a "missing persons" case. In "terms of statements", official correspondence, dispatches and subpoenas, as well as in some reports, some of these categories must be filled out, juxtaposed with fields where information about the "victim", such as "name", "parents' name", "occupation", "race" and "age", must be registered. If these categories appear repeatedly and even redundantly in the documents, at the same time there is no space for photos of those involved to be attached, nor for physical descriptions beyond the "race" attributed to them.

The inescapable presence of these categories and fields in the documents, as well as the lack of space for photos and physical descriptions of those involved, are material elements of VPIs that are incompatible with "missing persons" plots. "Missing persons" plots are not crimes and, therefore, it is not possible to speak of "authors", "victims" and "witnesses", at least not with the meaning assigned to these categories in criminal occurrences. Furthermore, a persons' "disappearance" cannot be narratively treated as an "occurrence" that took place at a certain "place", "date and time" and in front of certain people. What can be said about a "disappearance", instead, is only the place where the "missing person" was last seen by others who came to view them as "missing" based on a moment or process that may have no connection with events situated in time and space.

For this reason, photos of "missing" persons, graphic artifacts that are usually taken by their "families" to the precincts where they report the cases, are frequently glued, in a manner at once improvised and already standardized, at the edges of papers included in the VPIs. In most cases, this is done at the upper right corner of police reports, where there are no insignias or inscriptions that would be obscured by the images. Physical descriptions of the "missing persons", including information about the clothes they were wearing when they were last seen, striking facial features and marks or tattoos they may have, are registered in fields meant not for information about persons, but rather for descriptions of the so-called "dynamics of the occurrence".

As Paes (2008) exemplarily describes, the process of filling out documents in precincts is precisely characterized by improvisations, alterations and subversions of what categories and fields such as "dynamics of the occurrence" seem to presuppose, such as those employed in "missing persons" cases in order to include photos and physical descriptions of the "missing" in the VPIs. These improvisations and subversions often become fixed patterns of recording, as revealed, for example, by the practice of filling out the "dynamics of the occurrence" field not with information about the occurrence, but rather with accounts of activities carried out by the police officers since becoming aware of it.

This practice, just as other patterns of recording used in precincts, are interpreted by Paes (2008) and Miranda et al (2010) as effects of the not only inquisitorial, but also notary system that orients police actions in Brazil. As a result of this system, "the logic of the document that must be registered and filed precedes the logic of the record as a resource for the investigation" (Paes, 2008, p.175). In records about "missing persons", this logic of documents that must be filled out, even despite incongruences between their material elements and the cases' plots, ends up reinforcing the lack of importance attributed VPIs. Moreover, it also ends up endowing "missing persons" cases with an enigmatic character that results not from the cases themselves, but from the imprecision provoked by those incongruences¹⁹. Furthermore, it also reinforces the idea that "missing persons" cases are not "police matters", since even the papers routinely filled out, which are the most trivial part of a precinct's routine, are shown to be inadequate in the face of their characteristics, plots and characters.

Despite being considered unimportant, and despite the incongruence between their material elements and the plots of occurrences therein registered, documents on "missing persons" cases play a crucial role in the way police agents and units such as SDP manage these cases. By systematically reading the documents, the researcher was able to understand that much of the content recorded in these papers points not to what happened during the "disappearance", but rather to a process of delegating responsibilities put in motion by the production and circulation of documents itself. If, for officers like inspector Fernando, "missing persons" cases are "family affairs", in "merely filling out pieces of paper" about these cases those agents delegate responsibilities for managing and even solving them to the so-called "families of missing persons".

Based on what they record, how they record it and the fact that what they record is set in police documents – despite their own conceptions regarding the "social work" nature of the activities demanded by "missing persons" cases –, police officers attribute tasks such as looking for those "missing", controlling their movements and even stopping them from disappearing again to the very persons who went to precincts to request that they, the officers, would perform these tasks. As the three cases we will now present show, in filling out documents, the officers return to the "families" the responsibility for managing the cases, through advice and suggestions given both verbally and through the documents' texts themselves. These suggestions and advice are responded by the "families" with commitments, feelings and obligations that are also registered in VPIs and, thus, deposited in the SDP's so-called "archives", to be known by anyone who dares open their drawers.

Let us then discuss the cases involving Arlete, Cinira and Melissa, three young women who were reported as "missing" and whose cases, all solved, are currently in the sector's drawers. The excerpts of the cases reveal guidance, advice and suggestions given by police officers to the "families" of the three "missing" women, while also revealing commitments, feelings and obligations provoked both in these "families" and in the "missing" women themselves by the guidance, advice and suggestions fixed on paper. Let's begin with Arlete.

On January 7, 2008, Regina, a domestic worker, left early for work, as usual. She returned at the usual time and found her 16-year-old daughter, Arlete, watching TV. The two had dinner together and then Regina went to bed, "leaving her daughter awake, and her daughter said she was going to her friend Ana's house and would sleep there". The following day, she worked as usual, but, upon returning home, did not find Arlete.

In the next four days, Regina followed her routine, hoping to find Arlete every night. This did not happen. On January 12, she went to a precinct and obtained a record of her daughter's "disappearance". Almost a month later, she was once again at the precinct, this time to inform of the opposite: Arlete had returned home. Though already solved, in June 2008 the case was sent to the SDP. Before archiving it, the police officer who received it saw fit to call Regina and end the case with the following log:

Over the telephone the reporter declared that her daughter returned home, after spending some days in the company of classmates; she also said that, since then, she has sought to maintain a greater control over the minor and even punished her with a beating when she came home. The mother was instructed to provide the minor with an I.D.

Something similar happened with Melissa, a 14-year-old student who was used to spending a day or two away from home without informing anyone. In April 2008, she spent three days without contacting any family members and her mother decided to seek the police. At the time, a "missing person" report was made for Melissa. A week later, her mother returned to the unit and reported that she was already home. Based on her account, the officer who took her report registered:

That according to the information given by the reporter, her daughter Melissa was found in Quintino neighborhood at a friend's house and was taken home so that certain measures could be taken regarding her behavior.

The last case is somewhat richer in detail. Cinira was a 19-year-old college student who left the house where she lived with her mother and sister and returned ten days later. As soon as she returned, Cinira herself went to the police precinct where her "missing" had been first reported. The records of the "missing" person's visit to the police precinct are similar to those of Arlete and Melissa's cases: they fix on paper commitments with controlling behaviors. In Cinira's case, however, they are commitments to self-control and to what would be a good conduct toward herself, and not others, as in Arlete's and Melissa's cases.

What Cinira said to the police officer was thus recorded:

that in fact no crime related to her disappearance had occurred; that she left the house where she lives with her mother at approximately 10 pm, not giving anyone notice, not getting in touch in any way and reappearing 10 (ten) days later, saying she was upset with life, tired and at her limit; that the declarer regrets not having contacted her family, that when she goes out and takes somewhat longer to return in the future, she will call someone in her family to let them know where and with whom she is; the declarer said (...) that during the time when she was away she did not feel like talking to anyone, an attitude she regrets a great deal.

The "beating" in Arlete's case, the guilt, the regret and the commitments to control her own desires, feelings and conduct in Cinira's, and the obligation to punish her daughter for her behavior, in Melissa's mother's case are effects of advice, suggestions and guidance provided by police officers in their interactions with "families of missing persons" who seek precincts to report "missing persons" cases. An example of this kind of guidance is the suggestion to provide Arlete with an I.D., not only stated by the officer who took her mother's report, but also registered in the case's VPI.

It is precisely through the articulation between advice, suggestions and guidance such as this one and the commitments, feelings and obligations they provoke in the so-called "families" that police officers manage "missing persons" cases reported to Rio de Janeiro precincts. If, from these agents' point of view, "missing persons" cases are "family affairs", it is precisely the "families" that end up managing them, through "beatings", guilt and commitments fixed on documents considered to be the least relevant in police units. Furthermore, what is particularly interesting is that this mode of management becomes effective because it finds its way on paper, starkly revealing that official documents "bear the double sign of the state's distance and its penetration into the life of the everyday" (Das and Poole, 2004, p.15).

On the trail of police papers: ethnographic dialogues and final considerations

In this article, we sought less to compare our "fieldwork" data than to render clear the profitability of documents in anthropological research. We did so by proposing a dialogue between recent theoretical-methodological discussions on the theme and two ethnographic accounts of studies carried out on the trail of police papers. Based on the option to deal with documents, our ethnographies addressed the notary character that is widely criticized in the literature about Brazilian police units (Kant de Lima, 1995; Miranda et al, 2010; Misse, 2011, Vidal, 2013) in its positivities. That is, instead of associating documentation practices with an alleged inefficiency that would prevent police investigations from being properly carried out, we sought to seriously consider the idea that writing is one of the most important activities and government technologies of state routines (Gupta, 2012; Das, 2007) and to explore its analytical potentiality. Thus, we argue that in order to understand police administration of "human trafficking for sexual exploitation" and "missing persons" cases we had to follow the paper and produce reflections that questioned the tediousness, lack of symbolic richness and density of meaning attributed to documents (cf. Graeber, 2012).

In following the paperwork routinely produced and handled by federal and state police officers, we could observe that they assign less importance to the documentation work they perform, whether by pointing out the inefficacy (or unimportance) of inquiries that "turn up nothing" or differentiating the activity of "merely filling out pieces of paper" from the investigative practices they should ideally privilege. Thus, they frequently delimit borders and hierarchies very similarly to how anthropologists traditionally conceived of and defined ethnographic practices: opposing dealing with documents with "real" police work, characterized by street investigations and geared toward "actual crimes".

As we know, the legitimacy and authority of anthropology, anchored in the mythic model of the malinowskian fieldwork, were founded on an opposition between "armchair anthropology", based on documentary research and considered distant and uninteresting, and an anthropology emanating from fieldwork, understood to be capable of apprehending the "reality" of native life. Considering this analogy's potentialities, we note that if, on the one hand, our ethnographic accounts showed the central role of documentary artifacts in the everyday routine of police institutions, even when their importance is downplayed by their artificers, on the other, they revealed the little-debated distance between certain idealized self-representations about anthropological research and the activities we effectively carry out while "doing" anthropology (cf. Pacheco de Oliveira, 2009).

By following the trail of police papers and moving away from certain stereotypes about anthropology's knowledge practices and about the arti facts we ought to observe, handle and describe when carrying out fieldwork, both accounts made clear and were based on different analytical strategies. The first focused on the ethnographic description of the art of "reducing to writing" in the Federal Police. The second focused on exploring the materiality and the enactments produced by police reports and other documentary artifacts related to "missing persons" cases in the Rio de Janeiro Civil Police. Both works paid special attention to the micro-politics of interactions between documenting agents and documented persons, also enabling reflections about the interaction between documenting agents and one of the researchers.

By regarding documents as ethnographic artifacts, we were able to note that the resolution (or lack thereof ) of the cases investigated in the institutions where we carried out our researches depends not only on the action (or inaction) of their artificers, but also on different modes of delegating responsibilities to the documented persons. This is due to the fact that police officers believe they do not have much to do when faced with these cases other than producing papers, acknowledging the agency of persons technically defined as "trafficked" or "missing" over managing their own lives and bodies. Both analyses also revealed that this delegation of responsibilities guarantees to documented persons the possibility of actively intervening in the bureaucratic fate of administrative procedures and/or in the inscriptions of part of their lives in state documents.

The diffusion of responsibility, which produces a kind of collective agency, as noted by Hull (2012b), or shared agency, as we suggest, occurs through paper – something that was prominent throughout the accounts. The main difference is that, in "human trafficking" cases, the documented persons' agency takes on the form of resistance, while in "missing persons" cases, the agency takes place through a logic of subjection. We highlight that both processes are crucial to understanding the notion of agency, as Mahmood (2006) suggests. In our studies, we were only able to understand and analyze these different modes of agency, that are given form and materiality on paper and whose effects can be perceived through the bureaucratic fates of these administrative procedures, once we decided to analytically pursue the mediations and enactments which follow the processes of producing, circulating and archiving police documents.

Received 12^th August, 2014, approved 20^th September 2013

• AGUSTÍN, Laura. 2005. "Migrants in the mistress's house: other voices in the 'trafficking' debate". Social Politics, 12(1): 96-117
• APPADURAI, Arjun. 1986. The social life of things Cambridge: Cambridge University Press.
• ARAÚJO, Fábio Alves. 2008. "Falta alguém na minha casa". In: R. Kant de Lima (org.), Antropologia e direitos humanos 5 Brasília: Booklink/ ABA/ Fundação Ford. pp. 166-225.
• ARAÚJO, Fabio Alves. 2012. Das consequências da "arte" macabra de fazer desaparecer corpos: sofrimento, violência e política entre familiares de vítimas de desaparecimento forçado Tese de doutorado, Programa de Pós Graduação em Sociologia e Antropologia da Universidade Federal do Rio de Janeiro, IFCS/UFRJ.
• BLANCHETTE, Thaddeus Gregory ; SILVA, Ana Paula da. 2010. "Mulheres vulneráveis e meninas más: uma análise antropológica de narrativas hegemônicas sobre o tráfico de pessoas no Brasil". In: Ademir Pacelli Ferreira et al. (orgs.), A experiência migrante: entre deslocamentos e reconstruções Rio de Janeiro: Garamond. pp. 325-359
• BRASIL. Decreto-Lei 2.848, de 7 de dezembro de 1940 – Código Penal (CP). 1940.
• BRASIL. Decreto-Lei 3.699, de 3 de outubro de 1941 – Código de Processo Penal (CPP). 1941.
• CASTRO, Celso. 2008. Pesquisando em arquivos Rio de Janeiro: Jorge Zahar.
• CASTRO, Celso.; CUNHA, Olívia Maria Gomes da. 2005. "Quando o campo é o arquivo". Estudos Históricos, 36: 3-5.
• CLIFFORD, James; MARCUS, George (eds.). 1986. Writing culture: the poetics and politics of ethnography Berkeley & Los Angeles: University of California Press.
• COMAROFF, Jean; COMAROFF, John. 1991. Of revelation and revolution Volume 1: Christianity, colonialism and consciousness in South Africa. Chicago: University Chicago Press.
• CUNHA, Olívia Maria Gomes da. 2004. "Tempo imperfeito: uma etnografia do arquivo". Mana. Estudos de Antropologia Social, 10(2): 287-322.
• CUNHA, Olívia Maria Gomes da. 2005. "Do ponto de vista de quem? Diálogos, olhares e etnografias dos/nos arquivos". Estudos Históricos, 26: 7-32.
• DAMATTA, Roberto. 2002. "A mão visível do Estado: notas sobre o significado cultural dos documentos na sociedade brasileira". Anuário Antropológico, 99: 37-67.
• DAS, Veena; Poole, Deborah. 2004. "State and its margins: comparative ethnographies". In: Veena Das e Deborah Poole (eds.), Anthropology in the margins of the State Santa Fe: SAR Press. pp. 3-33.
• DAS, Veena; Poole, Deborah. 2007. "The signature of the State: the paradox of illegibility". In: Veena Das. Life and words: violence and the descent into the ordinary Berkeley: University of California Press. pp.162-183.
• EILBAUM, Lucía. 2012. "O bairro fala": conflitos, moralidades e justiça no conurbano bonaerense São Paulo: Hucitec/ANPOCS.
• EVANS-PRITCHARD, E. E. 1950. "Social anthropology: past and present, The Marett Lecture, 1950". Man, 50: 118-124.
• FERREIRA, Letícia Carvalho de Mesquita. 2011. Uma etnografia para muitas ausências: o desaparecimento de pessoas como ocorrência policial e problema social Doctoral dissertation, PPGAS, Museu Nacional, UFRJ, Rio de Janeiro.
• FERREIRA, Letícia Carvalho de Mesquita. 2013. "'Apenas preencher papel': reflexões sobre registros policiais de desaparecimento de pessoa e outros documentos". Mana: Estudos de Antropologia Social, 19(1): 39-68.
• FOUCAULT, Michel. 2014. A arqueologia do saber Rio de Janeiro: Forense Universitária. 8a ed.
• FOUCAULT, Michel. 1980. Power/knowledge: selected interviews and other writings, 19721977 Brighton, UK: Harvester.
• GINZBURG, Carlo. 1991. "O inquisidor como antropólogo: uma analogia e as suas implicações". In: A micro-história e outros ensaios Rio de Janeiro: Bertrand. pp. 203-214.
• GRAEBER, David. 2012. "Dead zones of the imagination. On violence, bureaucracy, and interpretive labor". HAU: Journal of Ethnographic Theory, 2(2): 105128.
• GUPTA, Akhil. 2012. Red tape: bureaucracy, structural violence and poverty in India Durham and London: Duke Univerty Press.
• HARTMAN, Saidiya V. 1997. Scenes of subjection: terror, slavery and self-making in nineteenth-century America New York, Oxford: Oxford University Press.
• HOAG, Collin. 2011. "Assembling partial perspectives: thoughts on the anthropology of bureaucracy". PoLAR: Political and Legal Anthropology Review, 34(1): 81-94.
• HULL, Matthew. 2012a. "Documents and bureaucracy. Annual Review of Anthropology, 41: 251-267.
• HULL, Matthew. 2012b. "Government of paper: the materiality of bureaucracy in urban Pakistan" Berkeley, Los Angeles, London: University of California Press.
• ISP INSTITUTO DE SEGURANÇA PÚBLICA. Pesquisa desaparecidos Available at: http://www.isp.rj.gov.br. Access on: 29 December 2009.
• JARDIM, Tarciso Dal Maso. 2011. Brasil condenado a legislar pela Corte Interamericana de Direitos Humanos: da obrigação de tipificar o crime de desaparecimento forçado de pessoas Brasília: Centro de Estudos da Consultoria do Senado/Textos para discussão, n. 83.
• LATOUR, Bruno. 2005. Reassembling the social: an introduction to actor-network-theory Oxford: Oxford University Press.
• LATOUR, Bruno; WOOLGAR, Steve. 1997. A vida de laboratório: a produção dos fatos científicos (Trad. Angela R. Vianna). Rio de Janeiro: Relume Dumará
• LOWENKRON, Laura. 2012. O monstro contemporâneo: a construção social da pedofilia em múltiplos planos Doctoral dissertation. PPGAS, Museu Nacional, UFRJ, Rio de Janeiro.
• LOWENKRON, Laura. 2014. Gênero, fronteiras e Estado: a construção do tráfico de pessoas em investigações e inquéritos policiais Relatório de pesquisa de pós-doutoramento apresentado à Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP), (processo nş 2012/11629-4), março de 2014.
• MAHMOOD, Saba. 2006. "Teoria feminista, agência e sujeito liberatório: alguma reflexões sobre o revivalismo islâmico no Egipto". Etnográfica, X(1): 121-158.
• MILLER, Daniel. 2013. Trecos, troços e coisas: estudos antropológicos sobre a cultura material Rio de Janeiro: Zahar.
• MIRANDA, Ana Paula Mendes de et al. 2010. "A reinvenção da 'cartorialização': análise do trabalho policial em registros de ocorrência e inquéritos policiais em 'Delegacias Legais' referentes a homicídios dolosos na cidade do Rio de Janeiro. Segurança, Justiça e Cidadania, 4: 119-152.
• MISSE, Michel. 2011. "O papel do inquérito policial no processo de incriminação no Brasil: algumas reflexões a partir de uma pesquisa". Revista Sociedade e Estado, 26(1): 15-27.
• NAVARO-YASHIN, Yael. 2007. "Make-believe papers, legal forms and the counterfeit: affective interactions between documents and people in Britain and Cyprus". Anthropological Theory, 7: 79-98.
• PACHECO DE OLIVEIRA, João. 2009. "Pluralizando tradições etnográficas: sobre um certo mal-estar na antropologia". Cadernos do Leme, 1(1): 2-27.
• PAES, Vivian Ferreira. 2008. "'Quem domina a regra do jogo': sobre a reforma da polícia e os registros policiais". In: Michel Misse (org.), Acusados & acusadores: estudos sobre ofensas, acusações e incriminações Rio de Janeiro: FAPERJ/Editora Revan. pp. 165-186.
• PEIRANO, Mariza. 1986. "Sem lenço, sem documento". Sociedade e Estado, 1: 49-63.
• PEIRANO, Mariza. 2006a. "De que serve um documento?" In: M. Palmeira; César Barreira (orgs.), Política no Brasil: visões de antropólogos Rio de Janeiro: Relume Dumará, NuAP/UFRJ. pp. 25-49.
• PEIRANO, Mariza. 2006b. "A lógica múltipla dos documentos". In: A teoria vivida e outros ensaios de antropologia Rio de Janeiro: Jorge Zahar Editor. pp. 135-153.
• PEIRANO, Mariza.2009. "O paradoxo dos documentos de identidade". Horizontes Antropológicos, 32: 53-80.
• PISCITELLI, Adriana. 2013. Trânsitos: brasileiras nos mercados transnacionais do sexo Rio de Janeiro: Eduerj.
• PISCITELLI, Adriana. 2008. "Entre as 'máfias' e a 'ajuda': a construção de conhecimento sobre tráfico de pessoas". Cadernos Pagu, 31: 29-63.
• PISCITELLI, Adriana; ASSIS, Gláucia Oliveira de; OLIVAR, José Miguel  Nieto (orgs.). 2011. Gênero, sexo, amor e dinheiro: mobilidades transnacionais envolvendo o Brasil Campinas: Unicamp/ PAGU.
• REED, Adam. 2006. "Documents unfolding". In: A. Riles (ed.), Documents: artifacts of modern knowledge Ann Arbor: University of Michigan Press. pp. 158-177.
• REIS, Elisa Pereira. 1998. "Opressão burocrática: o ponto de vista do cidadão". In: Processos e escolhas Rio de Janeiro: Contra Capa Livraria. pp. 239-269.
• RILES, Annelise. 2006. "Introduction: In Response". In: A. Riles (ed.), Documents: artifacts of modern knowledge Ann Arbor: University of Michigan Press. pp. 1-37.
• RILES, Annelise. 2001. The network inside out Ann Arbor: The Michigan University Press.
• RIO DE JANEIRO. 1991. Secretaria de Estado de Segurança Pública. Resolução SEPC 513 de 16 de dezembro de 1991. Rio de Janeiro: Secretaria de Estado de Segurança Pública.
• SOUZA LIMA, Antonio Carlos de. 1995. Um grande cerco de paz: poder tutelar, indianidade e formação do Estado no Brasil Petrópolis: Vozes.
• STOLER, Ann Laura. 2009. Along the archival grain: epistemic anxieties and colonial common sense Princeton: Princeton University Press.
• STOLER, Ann Laura. 2002. "Colonial archives and the art of governance". Archival Science, 2: 87109.
• TEIXEIRA, Flavia do Bonsucesso. 2008. "L'Italia dei Diveti: entre o sonho de ser europeia e o babado da prostituição". Cadernos Pagu, 31: 275-308.
• VIANNA, Adriana. 2014. "Etnografando documentos: uma antropóloga em meio a processos judiciais". In: Sérgio R. R. Castilho; Antonio Carlos de Souza Lima; Carla C. Teixeira (orgs.), Antropologia das práticas de poder: reflexões etnográficas entre burocratas, elites e corporações Rio de Janeiro: Contra Capa. pp. 43-70.
• VIDAL, Paula Chagas Lessa. 2013. "Os donos do carimbo": investigação policial como procedimento escrito Rio de Janeiro: Lumen Juris.
• ZEITLYN, David. 2012. "Anthropology in and of the archives: possible futures and contigent pasts. Archives as anthropological surrogates". Annual Review of Anthropology, 41: 461-80.

1

This crime is defined in article 231 of the Brazilian criminal code: "to promote or facilitate the entry, into national territory, of someone who will engage in prostitution or other form of sexual exploitation within it, or the exit of someone who will engage in these activities abroad. Sentence – incarceration for three (3) to eight (8) years. § 1 The same sentence applies to anyone who obtains the services of, entices or buys a trafficked person, as well as, being aware of this condition, transports, transfers or harbors them. § 2 The sentence is increased by half if: I – the victim is under eighteen (18) years of age; II – if the victim, due to illness or mental disability, lacks the necessary judgment to engage in the act; III – if the perpetrator is the victim's parent, step-parent, sibling, step-child, spouse, partner, tutor or curator, preceptor or employer or if he assumed, by law or other form, an obligation to care, protect or watch over the victim; or IV – there is use of violence, grave threat or fraud. § 3 If the crime is committed with the purpose of gaining economic advantage, a fine is also applied" (BRASIL, 1940).

2

See, for example, Comaroff and Comaroff (1991) and Hartman (1997).

3

"That is, not archives, but active file and other documents that are currently in use" (Hull, 2012b, p. 28).

4

The expression "graphic artifact" was coined by Hull (2012b) in order to designate the diverse documents and types of records produced in bureaucratic offices, such as maps, copies, cards, signatures and lists, indissociably drawing attention both to their material dimension and their role as crucial mediators in producing the persons to which they refer.

5

As a title given to police reports, "missing person" is not to be confused with the international criminal type "enforced disappearance", known, in the common sense language, as "political disappearance". "Missing person", as discussed later in this article, is an administrative category attached in Brazilian police precincts to what has been called, in some occasions, "civil disappearance" (cf. Oliveira). "Enforced disappearance" is a crime against humanity, defined by the International Criminal Court in 1998 and regulated by the so-called Rome Statute. "Enforced disappearance" means "the arrest, detention or abduction of persons by a state or political organization, or with the authorization, support or acquiescence of a state or political organization, followed by a refusal to acknowledge the deprivation of freedom or to give information about these persons' fate or whereabouts, with the intention of keeping them outside the reach of legal protection for a prolonged period." (Jardim 2011:14). For an anthropological analysis of the current incidence of "enforced disappearances" in Brazil, which also explores the possibility that this crime remains concealed by the administrative category "missing person", see Araújo (2008, 2012).

6

This formalization of investigations on paper is what guarantees that the prosecutor, judge and the target of the investigation are able to follow the legality of actions carried out in police headquarters (Vidal, 2013, p. 36).

7

The expression is used in article 9 of the Criminal Process Code (Brazil, 1941): "all pieces of police inquiry shall be, in a single process, reduced to writing or typed and, in this case, signed by the authority".

8

"Qualification" is the formal manner of identifying individuals in legal acts and documents, with at least some of the following information: full name, sex, date of birth, nationality, place of birth, marital status, parents' names, educational level, occupation, number of documents, phone number and address.

9

The formal procedure through which the interview is "translated" and registered in a "term of statement".

10

The research included 11 of the 14 "human trafficking for sexual exploitation" inquiries ongoing in the Rio de Janeiro Federal Police precinct chosen for the study. At the time, that precinct had the largest number of "human trafficking" cases of all Federal Police regional offices, which are situated in each Brazilian state. Only one inquiry (the one analyzed) led to an indictment. In addition to this inquiry, the veracity and authorship of the crime were established in only one other case that had been investigated by the Civil Police and whose criminals had already been prosecuted and convicted for state crimes, such as "sexual exploitation of children and adolescents" and "pimping". Since this investigation also gathered evidence of "international human trafficking", a copy of the inquiry was sent to the Federal Court and the Federal Police.

11

The "indictment" is the act of formalizing suspicion within the legal inquiry, usually at the end of investigations. For a more detailed explanation, see Vidal (2013).

12

Quoting Eilbaum (2012), "the VPI initials comes from article 5, paragraph 3 of the Brazilian Criminal Process Code. It refers to the fact that, when a criminal infraction is reported to the police, it must, after verifying the information, initiate a police inquiry. Based on the act of verifying information, the police created a bureaucratic administrative piece of their own, called VPI." (2012, p.414)

13

Reference to a traditional Brazilian dish made from beans and pork. (Translator's note)

14

The consolidated interpretation of these hierarchizations, developed in reference anthropological studies on public security (cf. Kant de Lima, 1995; Kant de Lima, Pires and Eilbam, 2008; Miranda et al, 2010) is that police authority is constituted, in day-to-day routine, through the ways in which agents not only hierarchize but also criminalize (or not) events with which they deal. As Kant de Lima, Pires and Eilbaum (2008) put it, in police precincts "cases of theft for which there is no evidence, threats, domestic disputes, among others, are filtered, set aside, avoided. That is because "real" cases, cases that "resonate", that are relevant, are homicides, kidnappings, cargo robberies, drug trafficking, among others." (p. 166)

15

As Ferreira details in the dissertation (2011) that resulted from the research, the SDP is responsible for investigating "missing persons" cases registered in a specific set of regular precincts (from the 1st to the 44th police precincts) of the Rio de Janeiro State capital. The precincts which first register these cases have fifteen days to carry out an investigation and close them. After that deadline, if the investigations have not solved the case, the SDP then takes responsibility for all "missing persons" cases registered in those precincts (Rio de Janeiro, 1991).

16

If, on the one hand, the status assigned to VPIs is not only distinct, but also inferior to that of police inquiries, Kant de Lima's (1995, p.68-69) ethnography reveals that these administrative procedures offer a greater margin of autonomy to police officers than inquiries, since they do not require an authorization by a judge or prosecutor, but only a commissioner's decision, in order to be archived.

17

The pioneer work by Oliveira (2007) also dealt with police officers' questions and reflections in terms of their attributions and responsibilities when faced with "missing persons" cases.

18

In December 2009, the ISP, a state agency dedicated to research and to training public servants in the public security field, divulged a pioneer study on "missing persons" cases. Titled "Research on the Missing", the study, which has been periodically updated since its release, sought to map out cases in the state in order to identify, on the one hand, the profile of "missing persons" and, on the other, the circumstances and possible causes of their disappearance. Among the results, there is one information frequently evoked at the SDP: both in the original study's sample and in its follow-ups, over 70% of cases ended with the "missing persons" returning to their homes or at least with their whereabouts being identified.

19

In the dissertation that resulted from the research (Ferreira, 2011, p.94-112), there is a detailed reflection on this enigmatic character and its relationship with the documents' material elements. For reasons of economy, the issue will not be explored here.

Publication Dates

• Publication in this collection
  28 Jan 2015
• Date of issue
  Dec 2014

History

• Accepted
  20 Sept 2013
• Received
  12 Aug 2014