INTRODUCTION

The courtroom was full. I was seated in a corner on one of the six benches assigned to the public, and it took me some time to realize that other attendees had taken very specific places depending on their relationships to the protagonists of the case: those who knew the defendant sat on the three benches directly behind the desk of the defense attorney, whereas those who knew the victim sat on the three benches behind the desk of the prosecutor. Earlier in the afternoon, the friends and family members of both victim and defendant had been arriving and congregating in the hallway of the court building. They were chitchatting about the alleged assault, phoning other people, and speculating about the potential outcome of the hearing.

It was the last case—and the most serious one—being handled by the court that afternoon. The defendant was accused of assaulting his live-in partner and had been found by police in possession of a firearm without the appropriate permits—allegedly, of course, because at this stage in the judicial process, the evidence is preliminary, and a trial needs to take place before a defendant can be considered guilty. That the woman who sat next to the prosecutor and who was the victim in the case had indeed been assaulted seemed obvious, however. She had bruises on her face and arms and had walked into the courtroom with some difficulty.

After the gendarmes brought the defendant in handcuffs to the courtroom, the prosecutor spoke first, presenting the particulars of the case, reading aloud excerpts from the police report and elaborating on them. The previous night, the defendant—angry because the victim had arrived home later than she had said she would—beat her up severely. The victim then took a picture of her injuries and sent it to her mother, who in turn called the police. Several policemen went to the victim's home, found a firearm hidden in a closet and arrested the defendant. The prosecutor asked for pre-trial detention [prisión preventiva] while more evidence was being gathered in preparation for the trial. In response, the defense attorney argued that the defendant did not know about the firearm; that it belonged to the victim's husband, who was in prison at the time; that the charge of illegal possession of weapons [porte ilegal de armas] was not relevant because the home, and therefore, the weapon, belonged to the victim; and that, with the information available at that point, his client could only be accused of assault, a charge for which house arrest would be sufficient. Ultimately, the judge, after asking the prosecutor to repeat the contents of two reports in the file—one issued by a physician describing the severity of the injuries and another assessing the domestic violence risk as “high”Footnote ¹—ruled that the defendant would remain in jail. We heard a muffled curse—“bitch” [perra]—muttered by one of the relatives of the defendant, and, heavily surveilled, he left the courtroom.

In my fieldwork, I had approached these judicial hearings skeptical of their ability to deliver the criminal justice system's promises of fairness. I had read about their tendency to punish people with certain characteristics—members of poor and marginalized populations, as was probably the case of the defendant here. And yet, I left that courtroom absolutely convinced that this man had assaulted that woman. The fact that the beating had happened only the night before, the tense conversations and intense looks exchanged by people attending, and the visible bruises of the woman reinforced this perception. Yet how could I be so certain that the man was guilty? Why did this aggression seem so obvious to me? Was it similarly obvious to the prosecutor, the defense attorney and the judge in the case? To the police officers who went to this couple's home and found her injured? In fact, this case was one of many that are brought to these courts every day, cases in which defendants are found in flagrante delicto by police, clearly committing the criminal offense, and are therefore arrested. Through these actions, the criminal justice system is attributing a degree of indisputability to these criminal offenses, its enforcement and judicial arms convinced that these criminal offenses did indeed occur and were committed by the defendants, who are then brought, in handcuffs, to these courtrooms.

Yet, this matter-of-fact acceptance of the facticity of the defendant's guilty conduct is precisely the subject of debate in the criminal justice system. In an adversarial legal system such as that in Chile, legal procedures seek to devise a series of mechanisms that structure the confrontation carried out between the two opposing sides. No matter how determinative of guilt it may be, no video, picture, or testimony alone can render the legal procedures unnecessary. “It is impossible, in any case,” wrote Reference LatourLatour (2010[2002], p. 151), “to define the expression ‘to say the law’ if we eliminate from it the hesitations, the winding path, the meanders of reflexivity: the reason why we represent justice as blind, and holding scales in her hands, is precisely because she hesitates.” Where are these hesitations in the case of flagrant criminal offenses? How is “evidence”, understood not as items but in the sense of being “evident”, confronted, challenged, acknowledged, or rejected during these specific legal procedures? How is this particular legal notion, the flagrante delicto, applied, and what role does it play in judicial “truth-making routines” (Reference Maguire, Rao, Maguire, Rao and ZurawskiMaguire & Rao, 2018)?

In this article, I explore the way in which flagrant criminal offenses are constructed. Drawing on ethnographic data gathered in lower criminal courts and one unit of the Public Prosecutor's Office in Santiago, as well as on the scholarly literature on legal technicalities (Reference RilesRiles, 2005, Reference Riles2011), I show that the flagrant character of criminal offenses, as these offenses move from their encounter with police to the courtroom, is constructed through specific ways of connecting facts and law, truth and procedures. Paying special attention to documents, procedures, and ostensibly minor aspects of bureaucratic work, I describe how the flagrant character of a crime conveys certain epistemological assumptions about how to determine what happened and what constitutes the criminal offense. More specifically, it indicates what cannot, for the moment, be known and that, therefore, can be ignored throughout the judicial process. In other words, in these cases, everything is organized so that actors in the criminal justice system are able to avoid asking themselves the same question that bothered me so much: “How can I be so sure?”

DETENTIONS IN FLAGRANTE DELICTO IN CHILE

Unlike most Anglo-Saxon criminal justice systems, in which detentionFootnote ² is a police power based in common law, civil law traditions explicitly declare when someone can be legally detained. The Chilean Criminal Procedure Code describes two such situations: when a judge has issued an arrest warrant [orden de detención] or when the person is caught in flagrante delicto. This latter case is described in Article 130 of the CodeFootnote ³:

The messiness in the way that this article is written is due to the many amendments made to it since 2000, when the Criminal Procedure Code was adopted.Footnote ⁴ Since then, the Chilean Congress has increasingly extended the scope of the article to cover more potential situations. For example, the original article did not specify the length of time that could elapse between commission of the offense and detention other than to refer to the “just committed” criminal offense; this time period was later defined as no more than 12 h. And, originally, only victims could identify the alleged culprit, but this provision was later expanded to include witnesses, as well.Footnote ⁵ Although amendments made to the text of the article have been criticized by some jurists, who have argued that its current version violates the spirit of criminal procedure reform (Reference Meneses Pacheco and PúblicaMeneses Pacheco, 2010; Reference Vitar Cáceres and MaureiraVitar Cáceres, 2011), most controversies surrounding the legal regulation of crime in Chile do not focus on amendments to articles regulating detentions in flagrante delicto, but rather on those regulating the police's rights to ask for IDs and to strip-search people (Reference DuceDuce, 2016).

Article 130 itself, therefore, raises little public controversy, which is consistent with its role in criminal procedures in Chile: the flagrancy reflects the way in which the offender is detained, rather than the offense itself. Under the article, this flagrancy determines how the alleged offender is detained and should not, in principle, have a significant impact on the adjudicative outcome of the case. Its provisions are intended to always be followed along with those of the criminal code that describe the actual offenses. Because Article 130 does not refer to any specific criminal offense, any crime from murder to issuing verbal threats, including sexual harassment, robbery, and driving while intoxicated, can lead to detention if the offender is caught in flagrante delicto.

This kind of marginal role of Article 130 contrasts with the widespread public attention given to what has been defined as flagrant criminal offenses in discussions on delinquency in Chile. The apparently insignificant character of this article contrasts, as well, with the way in which the criminal justice system treats cases that start with the detention of someone in flagrante delicto. Despite little systematic evidence thereof (Reference Fondevila and Quintana-NavarreteFondevila & Quintana-Navarrete, 2020), flagrant crimes are seen as “good” cases for prosecutors, who can build their cases on suspects, witnesses and evidence that have already been identified,Footnote ⁶ making it more likely to obtain a guilty verdict in a trial (Reference Ríos LeivaRíos Leiva, 2012). Studies in Argentina (Reference KostenweinKostenwein, 2018) and Chile (Reference Castillo Val, de los Ángeles Tapia Mansilla and SalvoCastillo Val et al., 2011; Reference Fandiño, Rua, Moreno and FiblaFandiño et al., 2017; Reference PásaraPásara, 2009) have shown that the perception of flagrant crimes as “good” cases works as a self-fulfilling prophecy, thereby casting doubt on the effectiveness of the criminal justice system in handling criminal offenses that are not flagrant.

After the police detain someone in flagrante delicto, they must inform a prosecutor within 12 h of the arrest; at that point, the prosecutor must decide whether the person should be released or not. The case will continue its judicial treatment regardless of the decision made about the defendant's detention at this point; however, when the prosecutor decides that the person will remain detained, he or she must be brought to court to take part in a detention review hearing [audiencia de control de detención] before another 12 h have elapsed.Footnote ⁷ The legal objective of these hearings is to determine whether the defendant's detention was conducted by police in accordance with their prerogatives and, when applicable, the articles regulating detentions in flagrante delicto. In practice, at these hearings, fewer than 1% of detentions are declared illegal (Reference Fandiño, Rua, Moreno and FiblaFandiño et al., 2017), and the application of Article 130 is seldom questioned (Reference Rebolledo, Moraga, Careau and AndradeRebolledo et al., 2008). Rather, these hearings serve as the defendant's arraignment [formalización], during which the prosecutor explains the charges and proposes a legal procedure to apply to the case, depending on the offense, the defendant's criminal record, and the prosecutor's judgment. In some less serious cases, the prosecutor can offer a kind of probation [suspensión condicional] or request an expedited procedure [procedimiento abreviado or procedimiento simplificado] to close the case quickly. Or, as it happens in most serious cases, the prosecutor will ask for precautionary measures [medidas cautelares] while more evidence is gathered, before an eventual trial.

None of these procedures and hearings are limited only to defendants detained in flagrante delicto; nonetheless, in practice, most of those detained for committing flagrant offenses are processed through the judicial system using these procedures and hearings. The defendants are caught and arrested, and if the prosecutor decides so, they are arraigned in a court hearing within 24 h of their detention. In practice, therefore, flagrant criminal offenses—defined as those in which the defendants have been detained by virtue of Article 130—constitute a peculiar kind of offenseFootnote ⁸ in the Chilean criminal justice system, one that structures a specific way of producing legal facts.

ANALYTICAL FRAMEWORK: FLAGRANT CRIMINAL OFFENSES AS “TECHNICALITIES”

How does law get to “know” things? The question has recently prompted much scholarly work, mainly inspired by Latour's invitation to study the “making of law”, inquiring into law's mechanisms of enunciation, qualification, and imputation, its “obsessive effort to make enunciation Reference Latourassignable” (Latour 2010[2002]: 274). Latour's focus is not so much on individuals' and institutions' moralities, subjectivities, and discretion or on policies and the state's projects, but rather on the legal tools themselves—rules, standards, documents, and files—that make this continuous imputation possible. The literature inspired by Latour's work has tried to answer the question of how law produces facts, how it participates in the “making” of realities, in “the fabrication of persons and things” (Reference Pottage, Pottage and MundyPottage, 2004). As a kind of “non-epistemological knowledge” (Reference PottagePottage, 2014), law is conceived as a type of performativity that operates through associations (Reference BlomleyBlomley, 2014).

One of the tools through which law produces facts, or rather shapes them, is what Annelise Riles calls “technicalities.” Precisely because they appear as playing a marginal, “technical” role in socio-legal configurations—“the mundane technocratic dimensions of law, precisely those dimensions that fail to engage humanists' theoretical, critical, or reformist passions, that are the most interesting artifacts of lawyerly work” (Reference RilesRiles, 2005, p. 1029)—technicalities have not attracted the interests of socio-legal researchers as much as other phenomena. However, as recent scholarship on the performativity of law and on its “machinery” (Reference ValverdeValverde, 2009) has shown, law's technical tools—just like those used by scientists (Reference Coopmans, Vertesi, Lynch and WoolgarCoopmans et al., 2014)—are folded into epistemological claims about how to get to know the world. To focus on technicalities is an invitation to explore closely those tools that seem just a means to produce knowledge, but that end up doing more than that.

One of these tools is legal fictions. In her fieldwork exploring the work of lawyers in charge of drafting and filling out the multiple forms and agreements required for the creation of contracts in large financial institutions in Japan, Reference RilesRiles (2011) focused on the role and use of collateral in international financial markets. She analyzed the specificities of collateral as a legal tool that coordinates parties' actions but, at the same time, seems to play a marginal role in the overall transaction, appearing to be simply a means to an end. Yet, as Riles explains, international markets function on the assumption that the other party will honor its obligations, an assumption which is ultimately uncertain, “a command, or a mutual agreement, simply to act ‘as if’” (Reference RilesRiles, 2011, p. 173). By treating collateral as a legal fiction, as “a guarantee of something that by definition cannot be guaranteed” (Reference RilesRiles, 2010, p. 804), Riles is not so much referring to legal theoryFootnote ⁹ as she is to practices in which collateral is a tool for action. Other examples of socio-legal studies in which researchers have followed similar analytical paths include the work of Barbara Reference YngvessonYngvesson (2007, Reference Yngvesson2010) on adoption as upheld by legal fictions on kinship and nationality and the study by Reference Pottage and ShermanPottage and Sherman (2010) on patent law and its definition of “invention” as a result of the fictionalization of ideas' scarcity. By looking at the seemingly mundane aspects of technical legal devices, researchers can learn about the ways in which legal tools articulate assumptions of truth and falsehood, playing a series of epistemological “tricks” on reality to resolve situated dilemmas.

Documents are another legal tool that have been overlooked as merely technical aspects of legal work. Perhaps because of their previously neglected materiality (Reference HullHull, 2012; Reference KafkaKafka, 2009; Reference Latour, Lynch and WoolgarLatour, 1990), they have been rediscovered as paradigmatic tools through which law produces facts (Reference RilesRiles, 2006). Among many other studies, Reference BarreraBarrera (2008), Reference HullHull (2012), and Reference WellerWeller (2018), have used ethnographic approaches to explore the indexical role of documents in addition to their semiotic role, highlighting both the mundaneness and complexity of bureaucratic artifacts, such as forms, certificates, and files. These file-making operations characteristically bring about a way of working case by case (Reference WellerWeller, 2018), “vertically” (Reference VismannVismann, 2008), progressively improving their performed knowledge of what is defined as “the case.”

According to this perspective, court files have a specific way of producing facts. In the logic of record-taking and its claim to accuracy (Reference WheelerWheeler, 1969), they operate as if they were transparent in regard to what they represent: the case. Through techniques of authorization, authentication and a self-referential logic of communication (Reference Van Oorschot and SchinkelVan Oorschot & Schinkel, 2015), as well as inscription and certification (Reference SureshSuresh, 2019), documents serve as a legal tool that produces facts precisely by claiming that they indeed represent the reality they are supposed to represent. But more than that, they play a role that could be analyzed independently of their content, as their mere existance—their materiality, their asthetics, their signatures (Reference HetheringtonHetherington, 2008; Reference HullHull, 2012; Reference RilesRiles, 2006)—is enough for them to contribute to the production of facts. In the following sections, I turn to the way in which flagrant criminal offenses are built on specific documentary practices that ensue from the application of Article 130. As I will describe, in the assembling, treatment, and use of the court files for flagrant criminal offenses, clerks, prosecutors, judges, and defense attorneys focus on the possibility of referring to them more than on the their actual content.

METHODOLOGICAL FRAMEWORK: TRACKING FLAGRANT CRIMINAL OFFENSES IN SANTIAGO

From a strictly legal perspective, flagrant crimes do not exist in Chile. As shown with the analysis of Article 130 in the Criminal Procedure Code, what is flagrant is how the defendant was caught—detained in flagrante delicto—and not the offense itself. And yet, flagrant criminal offenses seem to be everywhere. How the lower criminal courts and the Public Prosecutor's Office are organized in Santiago revolve around the many detentions that are carried out daily by police on the basis of this article. This was indeed one of the first findings of my ethnographic research in Santiago: most cases that are treated by the Chilean criminal courts involve the application of an article of the Criminal Procedure Code that seems both crucial and marginal for the criminal justice system.

Over three stays in Santiago for a total of 18 months between 2018 and 2020, I carried out 300 hours of observations in lower criminal courts [juzgados de garantía], specifically for the one type of judicial hearing I describe here, detention review hearings [audiencias de control de detención], which are open to the public. Since 15 of these courts are located in the same building in Santiago, and I did not aim to use quantitative sampling in my research, I combined my observations from these different courts. While the kind of cases presented in the different courts varies mainly according to the different parts of the city they cover—for example, more shoplifting cases in parts where there are more boutiques—and the different styles of judges, the way in which cases are handled from a legal standpoint is the same, as the same laws are applied.

Subsequently, I carried out 150 h of observations in one of the four units of the Public Prosecutor's Office that act as a link with police in cases of offenders detained in flagrante delicto in the city of Santiago. I obtained authorization from the Public Prosecutor's Office to carry out these observations particularly in the call center where clerks—civil servants who often have some law training but are not necessarily lawyers—receive calls from police officers. Through special headphones, I could listen to the entirety of the calls in real time. To protect the anonymity and privacy of the people involved in the cases reported to this office, and to respect the conditions imposed on me when I obtained the authorizations to carry out my fieldwork, I do not refer here to specific cases but rather describe typical situations generally. When I describe a specific call, I change some details to make the story anonymous but still representative of the situation I observed. The specific office and the lower criminal courts where I carried out my observations handle all kinds of cases reported by the police, whether or not they involve detentions; however, when offenders are detained, the specific unit of the Public Prosecutor's Office, where I conducted my observations, and these lower criminal courts work together.

Most flagrant criminal offenses treated by the criminal justice system in Chile are not crimes as serious as homicides or rapes but rather what would be called misdemeanors in Anglo-Saxon contexts. For example, a police officer making a traffic stop may discover that someone is driving a car reported stolen 2 weeks ago; a woman calls the police after her ex-husband threatens to kill her; two young men are spotted fighting on the street; local civilians report stolen purses, wallets, and cellphones. Or also, while municipal security guards are patrolling the streets, they detain someone who has apparently just mugged a passerby; neighbors call the police after hearing a man assaulting his partner; someone is driving drunk and is pulled over by the police. All these situations may lead to the detention of people in flagrante delicto, which will ultimately be carried out by the police.Footnote ¹⁰ Legally, the police are the ones who decide whether a person should be detained; that is, whether the conditions described in Article 130 are met. If they decide so, then they must inform a prosecutor within the next 12 h, and a hearing must be held within 24 h from the time of detention. Because of the large number of these cases that require the quick intervention of prosecutors in Santiago,Footnote ¹¹ and as a management strategy, the Public Prosecutor's Office [Ministerio Público] created specialized offices, operating 24 h a day, that are in charge of handling these detentions. Police officers in the field call clerks in this office to inform them about their detention decisions. In these phone conversations, highly trained clerks, in collaboration with prosecutors, arrange the facts temporally and spatially, separate relevant from irrelevant information, and propose a preliminary legal qualification for the case. They also write a narrative describing what happened based on what was reported by the police officer.

Taking Article 130 as a methodological starting point, I follow flagrant criminal offenses as they move from the locations where police “find” them to their treatment at detention review hearings. I did not follow the same individual case through the different stages, but rather tracked flagrant crimes as a category that configures practices through the treatment of different judicial cases. I am interested in flagrant crimes not so much in terms of how they emerge as the result of cognitive operations of classification carried out by particular individuals—although they do. Instead, I am interested in how they facilitate certain situations in which the circumstances that made the crime flagrant in the first place can be obviated through certain practices. Following the logics of case studies as reviewed by Reference MillerMiller (2018), that is, by identifying and documenting in depth a specific phenomenon, my research contributes to the socio-legal scholarship on how technical devices, such as legal provisions and case files, contribute to the production of judicial facts.

RESULTS: HOW TO NOT HAVE TO KNOW?

What happens when the clerks at the local prosecutors' office receive a call from police who have detained someone in flagrante delicto? They do what people usually do in legal systems: they fill in a form known as a logbook [bitácora]. In this form, they describe what police officers tell them. The description of what happened, as written by these clerks, will be part of the file but has little importance evidence-wise because, legally, what counts is the police report [parte policial] itself. Yet, this narrative serves an important control and coordination function between police and this office, stating not only what police report about the case but also when and how they make these reports.

In their calls with the police, the clerks ensure that the place where the criminal offense supposedly occurred falls within their office's jurisdiction and that the length of the detention complies with legal provisions—that no more than 12 h have elapsed between the alleged offense and the detention nor between the detention and the call. At this stage of the criminal procedure, the alleged facts of the offense that led to someone's detention do not matter as much as when and where they happened.

Once the clerk records this information on the computer, along with the preliminary narrative of the case, a file with an internal number is created. A prosecutor then has to make a decision: Will the detained person, at this point called the “defendant” [imputado], go through a detention review hearing at the criminal court? The decision on whether a defendant remains in detention depends on the type of criminal offense of which the person is accused and, except in some very specific cases, is made “routinely” (Reference Lempert and SandersLempert & Sanders, 1986). Did the defendant steal something valued less than a certain amount of money? If so, the offender will be released. Did the defendant threaten or assault someone or refuse to take the breathalyzer test when apparently driving drunk? If so, the offender will remain detained and undergo a court hearing. In every case, the file will include the name of the prosecutor who formallyFootnote ¹² made this decision and, if the defendant remains detained, another clerk will coordinate the scheduling of the detained person's hearing. The detained person will be directly brought to the criminal court by police, and not to this office. A typical call describing such a case might go as follows:

• • Office of the Public Prosecutor [Fiscalía]. Good evening. How can I help you?

• • This is Officer [carabinero] Luis González. I would like to report a procedure with a person arrested for a minor assault [lesiones leves].

• • Go ahead.

• • The victim arrived home from work and discovered that his neighbor had used his parking spot. This is a very crowded street and, apparently, they had somehow distributed the parking spots, and this one was the one for the victim. But it's informal and there's a history of conflict between them because of the parking situation…

• • Ah, ok, so there's a history of conflict here.

• • Yes, so when the victim arrived, he was mad that he could not park his car, so he went to knock at the door of this man [sujeto], the defendant [imputado], and asked him, apparently kindly, to move the car so he can park his…

• • Ah, ok, I see.

• • But this guy did not want to, so he started shouting that he was tired of this neighbor, the victim, always hassling him because of the parking; and then the other, the victim, replied by shouting, too. You know how it is, the whole conflict escalated, and the defendant took a wooden stick he had in his house and hit the victim with the wooden stick; the victim has minor injuries on his arms. At that point, with the brouhaha, other neighbors had arrived and controlled the guy, the defendant, and called us.

• • Ah, ok, does the defendant have injuries, as well? Was it more a fight or an assault?

• • No, it was really an assault. I do not think the victim wanted to fight; he wanted to talk to the defendant; the other neighbor said that this guy, the defendant, is the violent one…

• • I understand. Tell me, what time did this happen?

• • About 7:00 p.m., when the victim arrived home.

• • Time of the arrest?

• • 7:45 p.m.

• • Ok, give me the information on the victim and the defendant, please.

• • [The policeman gives the information on both the victim and the defendant: full name, address, phone number, and, most importantly, personal identity number.Footnote ¹³]

• • Did you check the identity of the defendant?

• • Yes, it has been verified.

• • Did you bring the victim to a health center?

• • Yes, I have the report from the medical center here.

• • Can you read the description of the injuries, please?

• • The report says: “Hematoma on right upper-arm consistent with minor injuries.”Footnote ¹⁴

• • Ok, the defendant will be in the detention hearings tomorrow morning.

At the end of the conversation, the clerk shares with the police officer the number assigned by the system to the case and the name of the prosecutor in charge. Between that call in the evening and the next morning, a clerk in the office will prepare the file for the case, which will include the following elements: the narrative written by the clerk who received the call, the criminal record of the defendant, a medical report describing the victim's injuries, if any, and perhaps most importantly, the police report describing what happened from the perspective of the police, along with the victim's statement.

The mandated documents and information for each case are standardized in the procedures of this unit of the Public Prosecutor's Office, but certain types of cases require additional documents or more detailed versions of the required materials. In verbal-threat cases between members of the same family [amenazas en contexto VIF], some documentation that proves the relationship between the victim and the defendant is required. In incidents that cause damage to property [daños], it is important to include evidence of the estimated cost of the damage. In assault cases, the medical report should be included. When the victim claims that the defendant is making verbal threats [amenazas], the police report needs to repeat exactly what the defendant allegedly said, including all the swear words and insults, which will be read out loud at the hearing, thus making for a jarring blend with the formality of law procedures—“a curious mix of theatre and the mundane” (Reference Sylvestre, Damon, Blomley and BellotSylvestre et al., 2015, p. 1347).

In none of these cases, however, do the documents included in each file directly address the criminal offense itself; rather, they communicate the value of a damaged piece of property, the severity of an injury, the relationship between two people, the wording of a threat. Whether this good, this injury, this relationship, or this threat corresponds to the legal definition of the criminal offense in which the detained person participated is contingent on filing practices. This evidence is constructed through documentary practices that add more documents, but not necessarily more information, about the criminal offense. The material in the file is compiled and edited by the clerk based on what the police said, who base that information on what the victim told them. Even though routine scrutiny of this chain of reporting is a common feature of adversarial procedures, actors collaborate in these cases to act “as if” the police officer had witnessed the assault—therefore, detaining the neighbor in flagrante delicto—and thus allowing such scrutiny to be avoided.

For each case involving detention and a subsequent detention hearing, the staff at this office will review and print the file documents, place them in a paper folder, and assign the case a new number that matches the court's filing system. To highlight key information that will be useful to the prosecutor, who will be at the detention hearing review and who will likely see these documents for the first time only moments before the hearing starts, the clerks affix Post-it notes to relevant sections of the material. At the detention hearings, the file that was carefully assembled some hours earlier becomes the central focus. The case draws primarily on the police report; other documents are rarely shown or referred to in court. At the hearing, the prosecutor reads excerpts of the police report, particularly those parts that correspond to the definition of the crime according to the criminal code. The defense attorney then responds, reading from documents taken from the same file, which is physically exchanged between the two of them throughout the hearing; they pass it back and forth between their desks in front of the judge.

At this point in the judicial proceedings, there is nothing more on which the two lawyers—the prosecutor and the defense attorney—could rely for their arguments; there is no more evidence than what is in that file. When the two lawyers do engage in discussion, it is not about the offense itself but about these types of questions: Is the police report worded in a way that is appropriate to the crime committed? What time was the person detained, and when was the prosecutor informed about it? Do we know the approximate cost of the damage caused by the defendant? Does it matter that the defendant had been accused of, but not convicted for, similar crimes in the past? If the defendant is the owner of the house, where the victim has been living for only a few weeks, must the defendant leave the house if he or she has made threats of violence? Paradoxically, the least important thing in these hearings is what seems to have so unquestionably—so flagrantly—happened.

For example, one paradigmatic type of flagrant crime is a theft in a supermarket or a store, when someone shoplifts something, which is defined as a “theft” [hurto] by the Chilean criminal code. If the commercial value of the stolen goods is lower than a set level, the suspect who is initially detained will be released; if it is higher, the person detained will be brought to the court for a detention review hearing. In both cases, the documents that are part of the file are standardized: a legal statement signed by the security guard of the supermarket or the store describing what he or she personally saw on the security camera footage, and a receipt detailing the value of the stolen items are enough. During my fieldwork, I heard about cases and saw detention reviews of people who had allegedly stolen bottles of liquor, diapers, car fresheners, blue jeans, cosmetics, backpacks, bedsheets, cheese, meat, coats, and a drill, among many other disparate things. Once a young man was detained for stealing a heater: “He took and appropriated, for profit [con ánimo de lucro] and against the will of its owner, a Toyotomi heater valued at $150,000 pesos [USD$210], crossing the cash register's area without paying its value,” said the prosecutor at the review hearing, with no one raising an eyebrow. While his grandmother wept next to me in the court's hearing room, I could not help but wonder how the defendant managed to steal something of that size. In any case, what exactly happened and whether the defendant did indeed commit the criminal offense, at least at this stage in the judicial treatment of the case, are beyond the point.

At the unit of the Public Prosecutor's Office where I carried out my observations, clerks also receive calls about situations in which a person is not detained by virtue of Article 130. For example, a family returning home after a weekend away finds their house has been robbed; a woman walking casually down the street has her wallet grabbed by someone streaking past; and a daughter tells her mother that she had been abused by a member of the family some months ago. Or a police officer calls to report the case of a woman who had allegedly been raped the night before; or the case of a mother who apparently saw her brother and her youngest daughter kissing and touching each other in the home living room; or the case of a man who had been defrauded by someone who pretended to have kidnapped her sonFootnote ¹⁵; or the case of a woman who, having recently gotten a job as a housekeeper and later left alone in the house, allegedly orchestrated the robbery of all the goods inside.

In all these cases, police called to make a report and share the names of the presumed culprits, creating a file in the Public Prosecutor's Office system and eventually making them the subject of further investigation. Yet in none of these cases did the police detain the alleged offender. “Why?” I would ask, thinking about all the other situations where police called after they had already detained the suspects. “Why didn't the prosecutor tell the police to detain the suspect? Why haven't the police done so already?” I would ask. The legal answer to these questions is that prosecutors cannot order a person to be detained: that is the prerogative of the police. The sociological answer to these questions, however, involves a broader understanding of what can and cannot be supported by applying the legal fiction of the detention in flagrante delicto: the chain of authority, truth, and trust that is produced by subsequently telling the story of what happened from victims and witnesses to police, and later from police to clerks and prosecutors, does not operate in these cases. These cases will not be treated as flagrant criminal offenses and, unlike those cases that do, clerks will ensure that more information is gathered. In other words, prosecutors will need to know more before appearing in front of the judge. In these last cases, therefore, they will give instructions to the police about what kind of investigative actions should be carried out because prosecutors are the ones legally in charge of guiding criminal investigations in Chile. The results of the investigation, in any case, will be analyzed by another unit of the Public Prosecutor's Office, and not by the one I studied.

While in other cases there are occasions, such as trials, for iteratively and recursively returning to the facts (Reference SchefferScheffer, 2010)—to challenge them; to question them; and to uphold and refute signatures, countersignatures, stamps, and seals, sometimes rather than “facts” (Reference SureshSuresh, 2019)—detentions in flagrante delicto launch a series of practices in which the criminal offense itself—the theft, the assault, the threat—is not discussed. In the case of flagrant criminal offenses in Chile, the treatment of cases allows actors to avoid discussion about the apparent symmetry between the case file and the criminal offense. The existence of the documents, and not that much their content,Footnote ¹⁶ is what upholds the facts that are qualified as a flagrant criminal offense. In the detention review hearings, their handling and indexation—as well as their preparation beforehand by the clerks—allow judges, prosecutors, and defense attorneys to not have to know more.

The idea that the suspect was caught in flagrante delicto supports the drafting of “a script for a particular kind of collaboration” (Reference RilesRiles, 2011, p. 59), the flagrant character of the criminal offense articulates and connects people, things, and practices in such a way that the facticity of the criminal offense can be overlooked and put permanently “on hold.” Article 130's agonistic attempt to describe all the possible situations in which a person can be found so evidently committing a crime illustrates the complexity of the practical dilemma confronting actors in the criminal justice system: It is highly unlikely that a police officer will encounter someone at the very moment he or she is committing a criminal act—reaching into the pocket of a distracted passerby, punching someone in anger, threatening a neighbor, shoplifting something from the supermarket, or mugging someone in the park. Therefore, rather than describing someone caught in flagrante delicto, “in blazing crime,” as the Latin expression defines it, the application of Article 130 launches a series of procedures that allow actors to build legal facts through documentary practices and organizational arrangements without actually having to refer to what happened.

CONCLUSIONS: CRIMINAL LAW BUREAUCRACIES AND THEIR EFFECTIVENESS

Police, prosecutors, judges, defense attorneys, and clerks at the different offices involved can avoid the question of “How can I be so sure?”—the one that opened this text—in the case of flagrant criminal offenses in Chile through specific “material-semiotic networks of controversy and dispute” (Reference McGee and McGeeMcGee, 2015, p. 65). Applied initially in situ by police but traveling through different physical and symbolic places—the street, the prosecutors' office, and the courtroom—and mobilizing ideals of police discretion, prosecutorial and judicial efficiency, and justice's fairness and rights—Article 130 provides a practical way to put into motion the criminal justice system. As a tool in the Chilean criminal justice system allowing prosecutors, defense attorneys, and judges to presume a certain proximity between “the truth” of what happened but was experienced by very few people and its legal treatment, the detention in flagrante delicto works as a technicality, “a technique for working with and in the meantime” (Reference RilesRiles, 2010, p. 803). Actors in the criminal justice system know that defendants caught in flagrante delicto were not caught red-handed but rather were arrested by virtue of a legal provision that defines a flagrant crime in broad terms, allowing police to respond in many different ways, from running in hot pursuit of an alleged criminal to knocking on the door of someone accused by the neighbors of having done something illegal the night before.

How do these wide-ranging situations all eventually fall into the same practical category of flagrant criminal offenses? What do they have in common? They allow actors in the criminal justice system to not have to know. When police, prosecutors, clerks, defense attorneys, and judges collaborate around criminal offenses committed in flagrante delicto, they are dealing with facts considered to have happened recently, in which the system worked promptly to capture the alleged culprit; in these cases, facts are considered “fresh”, and so are witnesses' and the police's recollection of them. And this swiftness to detain a person and the quick and timely intervention of law and order extend to the way in which legal procedures move forward. However, this speed also means that only a limited amount of information can be collected. The facts may be fresh, yet there is no time to gather all the potential evidence that might be available. Thus, prosecutors use what information they have, which, paradoxically, does not relate directly to the criminal offense itself: the flagrant character of the facts that legally justified the detention of the defendant is not directly addressed by the procedures. Flagrant criminal offenses' facticity is therefore produced through practices that make it possible to avoid directly referring to the alleged flagrant facts.

We know that the law's effectiveness results from the quality of the connections made between people, places, and times, thereby producing facts that are legible for bureaucracies—“the effectiveness of form in generating the effect of effectiveness” (Reference RilesRiles, 2001, p. 172). In criminal law, however, “getting to know” usually involves the work of detectives and police officers investigating and following clues. Unlike higher courts responsible for sophisticated legal reasoning, such as the one analyzed by Reference LatourLatour (2010[2002]) in the French Conseil d'État, or administrative courts trying to determine levels of credibility and risks in asylum cases, such as the ones studied by Reference GoodGood (2007) in the United Kingdom, there is something upsetting about criminal law, particularly in those cases—such as flagrant criminal offenses in Chile—that are presumed to be temporally very close to the facts. At least in principle, “criminal law does require of actors to assume the burden of proving that something did, indeed factually happen” (Reference Van Oorschot and SchinkelVan Oorschot & Schinkel, 2015, p. 506, italics in original). But how do they do it in the case of flagrant criminal offenses? My findings suggest that they do not need to prove it. The evidence regime associated with flagrant criminal offenses in Chile, which is neither exclusively rhetorical nor material but rather the result of a combination of practices, turns evidence into a question of “methodology” rather than of “epistemology” (Reference EngelkeEngelke, 2008), one of following the same script for collaboration.

Finally, my findings show the heuristic productivity of the analysis of technicalities for a socio-legal approach to criminal law. Used as a method rather than as a theory (Reference Riles, Cowan and WincottRiles, 2016), the analysis of technicalities in the application of criminal law helps us interpret people, situations, interactions, and documents that otherwise would have been reduced to variables or to gaps between law-in-the-books and law-in-practice. When this approach is used to answer research questions traditionally associated with administrative or public law, and thus with the state—its power, its sovereignty, its authority—we are able to grasp the complexities of the workings of the criminal justice system in practice, rather than as areas of opposition between the state, understood as an abstract entity, and laypeople. By following cases of flagrant criminal offenses in Chile through their treatment in the criminal justice system, we see that, despite intentions or perceptions, they are constructed through a series of practices that give facticity to the alleged facts precisely by not having to know more about them. These practices are located both at the heart of the rule of law and the margins of society–affecting the poor and dealing with the bulk of low-level criminal offenses, such as the one I describe at the beginning of this text–and are carried out through the mundane work of police officers, clerks, prosecutors, defense attorneys and judges. What seems to be a mere procedural aspect of the detention of a person by virtue of Article 130 of the Chilean Criminal Procedure Code contributes to the creation of a practical category of criminal offenses in which it is possible, for actors in the criminal justice system, to “not have to know.”