The number of disappearance: trajectories in the tally of victims of forced disappearance in Latin America

ABSTRACT The “metric turn” is shaping human rights knowledge, governance and politics globally. This article seeks to contribute to the emergent analysis of numbers in human rights matters from a Latin American perspective. We explore a phenomenon that is hard to count; the number of victims of forced disappearance –persons who are kidnapped and murdered, and whose bodies are disposed of. And we study it in three contexts of political institutional violence on the continent – the dictatorship in Chile (1973–1990), the armed conflict of Colombia (1958–), and México's dirty wars (1964–1998) and narco-conflicts (2002–). Focusing on numbers' liveliness, we draw from interviews, institutional documents and archive analysis to examine the trajectory of the number of forced disappeared persons and how it mobilizes and is shaped by human rights concerns. Transitivity is crucial in the trajectory and liveliness of numbers. Challenging the view that numbers only prove effective when decontextualized, we show that transitivity is a mathematical, cognitive, and political achievement, and identify the effects of transitive but also of referential and provisional numbers in the human rights field. Particularly, how they contribute to the acknowledgement and handling of human rights violations, pushing for the production of accountability regimes.

El número de la desaparición: trayectorias del conteo de víctimas de desaparición forzada en América Latina RESUMEN El "giro métrico" está configurando el conocimiento, la gobernanza y la política de derechos humanos a nivel global. Este artículo busca contribuir al emergente análisis de los números en materia de derechos humanos desde una perspectiva latinoamericana. Examinamos un fenómeno que es difícil de contar; el número de víctimas de desaparición forzada -esto es, el secuestro y asesinato de personas seguido de la disposición de sus cuerposen tres contextos de violencia política institucional en el continente -la dictadura en Chile (1973)(1974)(1975)(1976)(1977)(1978)(1979)(1980)(1981)(1982)(1983)(1984)(1985)(1986)(1987)(1988)(1989)(1990), el conflicto armado de Colombia (1958-), y las guerras sucias (1964-1998) y narco-conflictos (2002-) de México. Abordando los números como procesos y centrándonos en su vitalidad, mediante análisis de entrevistas, documentos institucionales y archivos, examinamos la trayectoria del número de desapariciones forzadas para explorar su capacidad de movilizar y ser moldeado por asuntos de derechos humanos. La transitividad es crucial en la trayectoria y vitalidad de los números. Desafiando la opinión de que los números solo resultan efectivos cuando se descontextualizan, mostramos que la transitividad es un logro matemático, cognitivo y político e identificamos los efectos en el campo de los derechos humanos de números transitivos pero también de aquellos referenciales y provisionales. En particular, evidenciamos cómo contribuyen al reconocimiento y tratamiento de las violaciones a los derechos humanos, impulsando la producción de regímenes de rendición de cuentas. emblematic cases, and legal semantics (Fakuda-Parr 2011, 79;Langford and Fukuda-Parr 2012, 222;Allier, Vilchis, and Vicente 2020). However, numbers have also counted human rights atrocities, especially in Latin America, a region whose recent past is marked by military coups, armed conflict, and authoritarian governments which, through a variety of legal and clandestine bureaucracies and institutions, have violated the human rights of thousands of people. Since the 1970s activists, professional and official agents in the field of human rights in Argentina, Chile, Colombia, Peru, Bolivia, Uruguay, Guatemala, and Mexico, have quantified massacres and genocides, political imprisonment, and forced displacement of populations, among other criminal acts. 1 In various cases, in this region and elsewhere, the transitional processes of justice implemented by governments have begun counting atrocities committed as part of the process of recognition of and reparation to the victims. Globally, since the 1990s, international organizations and development agencies have used quantification, production of indicators, and other factualization technologies and measurements to recognize and denounce repressive practices, and to exert pressure on nations to comply with human rights as a principle of democratic governance. In our era of evidence-based governance, numbers are especially useful to foster regulatory strategies (Merry 2016, 10), ranking states according to human rights treaty obligations. The massification of the internet, as the best register of digital artifacts, and cooperation in innovation between human rights organizations and science, allowed information platforms to develop (Forensic Architecture, Metrics for Human Rights, or WITNESS). These initiatives have widened the possibilities of production and visualization of data, usually combining numbers with images and maps. A new form of activism based on statistics has come to dispute public statistics and to pluralize the "statistical reality" (Bruno, Didier, and Prévieux, 2014). Numbers are "seductive" (Merry 2016), appealing to a diverse public and spurring social mobilization to produce change. Using visualization techniques (Rall et al. 2016), they communicate widely, contributing to the generation of political commitment.
In the field of human rights, statistics have penetrated the practices of denunciation, identification, reparation, and the creation and evaluation of mechanisms, policies, and practices that ensure the abuses are not repeated. Numbers are used at local, national, and international levels, shaping human rights knowledge and governance globally, in what Langford and Fukuda-Parr (2012) termed the turn to metrics in human rights.
However the literature concerning the production, uses, and effects of the numbers of human rights atrocities is incipient and little is known about the controversies and negotiations over numbers' formulation or the consequences they may have on the phenomenon they represent. How numbers assemble credibility as they travel, for example, from local Bosnia to the international Hague (Toom 2020, 366) or what their impact is on regulation, responsibility, and governance of human rights problems (Langford and Fukuda-Parr 2012), is little known. Available literature describes numbers' paradoxical role in the human rights arena as "necessary and complicated, reparative and dehumanizing" (Nelson 2015, 24). In a field that pursues accountability, in which the recourse to horror and impunity sows mistrust, quantifications provide legitimate means for all parties to assess their responsibilities (Langford and Fukuda-Parr 2012, 223). Regarded as uncontroversial and unambiguous, numbers hold the promise for "uncompromising and reliable objectivity" (Nelson 2010, 89), releasing atrocious and often contentious facts from controversy, while helping to produce "stable populations" and "conceivable subjects" (Gatti 2017). "Counting does the vital human rights work, based on the notion of a universal equivalent (…) of making every death equal to anyone else's" (Nelson 2010, 89), but it presupposes stable, conspicuous, fully present populations, that can be standardized, individualized, and established (Verran 2010, 46). Political and para-political violence, however, produces a field of ambiguity, movement, and instability (Walters 2014). Analyzing the numbers of victims of sex trafficking, Martínez (2020) sustains that the methods used to count his reference subjectsthe citizen (Desrosières 2004) "grate" with mobile populations and "break" with complex phenomena, clandestinity, illegality and diffuse frontiers. Guesstimates, over-estimation, and false numbers populate the field of human rights. Following human bones, mass graves, shovels, lists, DNA samples, and mud, Toom (2020) has shown how the number of Srebrenica genocide victims is produced and contested in a complex sociomaterial entanglement he calls an "ontologically dirty knot": never finished, situated, controversial.
In this article, we examine the number of people subject to forced disappearance, persons who are kidnaped, killed, and whose bodies are hidden in secret mass graves or disposed of in the seas and forests. An event particularly complicated to quantify as it involves counting absent units and heterogenous references: death tools, the number of people looked for, people officially disappeared, bodies to identify, graves, clandestine sites of exhumation, exhumed human remains (which in Mexico are not measured in bodies but in kilos of bones). A number that is particularly useful as it inscribes an event whose victim is silenced. We focus on Latin America, the continent of disappeared people, and we trace this number for six decades in three contemporary contexts. 2 The first case is the last dictatorship in Chile (1973)(1974)(1975)(1976)(1977)(1978)(1979)(1980)(1981)(1982)(1983)(1984)(1985)(1986)(1987)(1988)(1989)(1990) where the military and the police, with the collaboration of many civilians, undertook executions, forced disappearances, massive detention, persecution and torture of left-wing sympathizers, and anyone who represented a threat to the new regime. Human rights were systematically violated in over 1132 detention centers causing at least 3000 deaths, a third of them disappearances, 40,000 prisoners (94% of whom were tortured), and thousands of exiles. The second case is Mexico where the dirty war as a countersubversive practice  and the narco-conflicts (2002-) have to date left approximately 98,000 disappeared. The third case is Colombia, with multiple types of on-going armed conflict (Trejos, Bravo, and Badillo 2021) characterized by the bloody territorial interaction between political actorsarmed left-wing organizations, the state and illegal actors, paramilitary organizations, and agents of narco-traffickingcausing, according to sources, between 80,000 and 120,000 disappeared people.
Transitivity is crucial in the trajectory and liveliness of numbers. Following those who posit that the qualities of numbers such as impartiality, objectivity, precision, systematization, intersubjectivity, and trust are "socially crafted historical achievements" (Lampland 2010, 381; Porter 1995) dependent on a complex set of social relations, materialities, and technologies (Latour 1987;Daston 1995) and on extensive negotiation among interested parties (Wise 1995), we suggest that the transitivity of the number of forced disappearances is based on a social convention that assumes this number is legitimate, avoids questioning the practices that produced it, and allows its use for some time and between various sites. Likewise, when the number is disputed and incapable of maintaining its meaning between contexts and times, it can still confront human rights abuses. We suggest that the arithmetic is probably never precise nor the only way to count the disappeared, and that the lack of a stable or permanent figure does not necessarily imply a failure in the counting, challenging the view that numbers only prove effective when decontextualized. Numbers may be mathematically referential or provisional, and still become symbols or "moral reminders," acting in "inventive ways," "more like numerology than arithmetic" (Guyer et al. 2010, 38), for example, as kinship metaphors (Urton and Nina Llanos 1997) participating in roles and relations like "minus one," the name given to the relative each extended Guatemalean family lost in the 36 year conflict (Nelson 2010, 2). Or, as we will see here, as red numbers, signaling indignation and the limits of admissibility, and calling for state action; or iconic numbers (Verran 2012) representing individuals or occurrences of a specific kind that overwhelmingly surpass them.
Three processes of quantification of forced disappearances allow us to show that in post-dictatorship Chile governmental policy validated a number of disappeared persons produced and made visible by civil society during the military regime. Consequently, a single number has been available and has participated for four decades in different transitional justice processes, until recently when it was reopened by the state to adapt Chilean nomenclature to the international system of human rights. In Mexico, the number is not stabilized because it shifts between various units of measurement. However, a reference number can still dispute the distribution of the sensible, work as evidence, and generate indignation and repudiation. Colombia is between these two poles; although the number of forced disappeared is provisional because, among other reasons, armed conflicts are on-going, it has contributed to the installation of a bureaucratic humanitarian architecture of governance of this massive crime and generates growing public indignation (Fassin 2011, Gatti, Peñaranda & Martínez 2018).
The article begins by defining quantification and its association with categorization, and presenting the notion of number trajectory as a time and space-sensitive way of exploring numbers' ability to mobilize human rights concerns. Then it follows the trajectory of the number of forced disappeared persons in each national context, with synthetic outlines illustrating the processes that lead to three numbers: one that transits, one referential, and one provisional. Finally, we discuss some of the implications of numbers' circulation in the human rights field in relation to the concerns they help to constitute.

Quantification in the human rights field
We approach numbers as agential and relational rather than passive tools (Holtrop 2018), and therefore we are concerned with their capacity to mobilize human rights issues as they act in the world.
To count implies to objectify and recognize. Objectification provides certain entities with existence; recognition endows them with value. If through the first operation a number puts things in order, through the second it participates in the representation of that order in a specific form (Verran 2010, 113). In human rights, numbers not only organize information and produce knowledge, they also give visibility to people and experiences of violence, announcing "the existence of atrocious occurrences, indicating their political importance and relevance" (Andreas and Greenhill 2010, 1) and calling for their inclusion as part of a "policy of public legibility" (39). Quantification is a technique of realization that accompanies the phenomenon's production; a way of knowing and intervening through factualization via technification (Rose 1991).
Quantification presumes categorization, compilation, and rejoinder (Bowker and Leigh Star 1999;Langford and Fukuda-Parr 2012, 223). Categorization enables organizing elements or individuals in the social space (Ruppert 2012, 39), operating as a convention of equivalency, by defining elements of "the same kind." It creates links between those elements, enabling the conformation of the part/whole relation (Desrosières 2004, 236;Verran 2010). The taxonomic work is substantial for denouncing human rights violations, because it supports the production of a continuous and hierarchical social order (Durkheim and Mauss 1903; Bernasconi 2019) and a common language for enunciating and acting upon these extraordinary occurrences (Desrosières 2004, 271). This repertoire comprises categoriesvictim, perpetrator, witness, relativeand practicesdisappearance, abduction, forced displacement, or torturetypified through the identification of patterns developed alongside information production and the denunciation of these crimes. Moreover, in the social imaginary, the "alchemy of quantification" is transferred to and operates upon the phenomenon that it enumerates. Consequently, the different repressive practices begin to appear before public opinion as discrete, traceable, countable, and denounceable events. "At the heart of quantification is aggregation and replication" (Langford and Fukuda-Parr 2012, 223). Indeed, in their cardinal dimension (Verran 2010), numbers bring us closer to the magnitude of a set of elements whose finitude is difficult to specify. What emerges is a practice, an "ostensible," "hard fact" (Latour 1990;Díaz 2020), a pattern of victimization. Quantity producing quality; death of six million jews = genocide (Nelson 2015, 27).
Transitivity is crucial in the trajectory and liveliness of numbers. It refers to a number's capacity to take a "life of its own" (Crump 1990;Lampland 2010; Mennicken and Espeland 2019) acting as a "technology of distance," affecting upon the events it is referring to in many sites and at the same time (Latour 1987, Robson 1992. Being able to translate human rights demands and to provoke action at a distance (Latour 1987;Robson 1992), numbers pertain to the repertory of instruments with which denunciation of human rights violations circulate. Consider Colombia, where the number of eight million victims has been an instrument to demand the right to peace. To move closer to the production and circulation of numbers, to understand why they emerge, how they are legitimized, what social processes they prepare, and what power relations they cross, we use the term "numbers' trajectories," a time and space sensitive concept to approach how numbers move, change, and assemble credibility (Toom 2020) as they act in concrete social, political, and historical settings. We understand numbers as evolving and never complete entities (Holtrop 2018;Merry 2016), becoming "ontologically more diverse and heterogeneous" (Toom 2020), as they move around performing different tasks, shaping, and in response, to human rights violations.

How many victims of forced disappearance? Numbers' trajectories
Generative and transitive number: constructing enforced disappearance as a knowable reality in Chile The interfaith organization Cooperation Committee for Peace in Chile (COPACHI) provided legal and social assistance to those persecuted by Pinochet's regime. After Pinochet forced COPACHI to close in January 1976, a new organization resumed its work under the structure of the Catholic Church: the Pastoral Solidarity Vicariate, popularly known as the Vicaría. In 19 years both organizations documented 47,252 cases, half of the total cases documented in human rights organizations during dictatorial rule. 3 In October 1973, when COPACHI staff began receiving denunciations about people disappearing after being arrested, they had no idea that these were operations of the system of disappearance. Despite having been detained in front of witnesses, some detainees' names did not appear on the official lists that the military Junta issued to report the people in its custody. The documentation of individual cases denounced to COPACHI by witnesses and relatives enabled its staff to identify common features among detentions, and classify them as similar occurrences. The frequency of the cases and the recurrence of patternsthe ordinal and cardinal dimension of numbersrevealed that these were not isolated events. The category "disappeared" and that of "unlocated (detainee)" first appeared in the organization's public documents in December 1973. By then, it had recorded 224 people in that condition, from a total of 1487 case files of victimized subjects.
The dictatorship, however, denied the existence of the disappeared. In March 1974, when COPACHI had registered 547 "disappeared and unlocated" persons, it filed a mass habeas corpus on behalf of 131 of these people. The Santiago Court of Appeals dismissed the motion based on the nonexistence of this category of persons, "since there only exist prisoners who are recognized as such by the government" (Recurso de Amparo Rol 289-74 3). 4 The dispute between the crime's visibilization by the victims' relatives and human rights organizations (counting) and the dictatorship's denial (discounting) fueled the investigation, denomination, and enumeration of disappeared persons throughout those years. Early on, COPACHI defined and differentiated the phenomenon of disappearance from similar situations. A chronological analysis of the organization's periodic reports shows that the compilation of information and increasing registry which established distinctions from other repressive occurrences, led to the characterization of forced disappearance. In 1975 it found that the loss of all traces of a person after arrest represented "more than half of the cases of detentions." After "days, weeks, or months," a person might "reappear" because s/he (i) had been released, (ii) formally arraigned or indicted, (iii) recognized by an official list as having been executed, or (iv) found dead at the forensic medical service or another facility. However, the cases in which "people reappeared after five or six months" were "quite rare" (COPACHI 1975, 34). The time counted since the disappearance of persons arrested and held at unknown locations and the calculation of the "average time of disappearance," enabled staff to establish further distinctions (Figure 1). In 1975 the notion of "disappeared detainees" was coined, suggesting state responsibility for both crimes: detention and kidnaping.
In November 1976, the military government issued decrees 2343 and 2344 concerning the release of detainees arrested by virtue of state of siege. This applied to those arrested without charges, accused of illegal conduct, or simply considered dangerous to national security. The most notorious mass detention centers were closed. Yet the persons whose disappearances had been reported were not among those released. The specific information about each personsex, age, work, political affiliation, date, hour and place of kidnaping, authors, witnesses, etc.enabled the elaboration of victims' profiles and the time period spanning the trajectory of the crimes (Figure 2).
In April 1978 the military junta decreed the Amnesty Law to protect authors of crimes committed from September 11, 1973to March 11, 1978 In response, the Association of Relatives of Disappeared Detainees initiated a "long strike" which only ended when the military regime promised to investigate those "alleged disappeared persons." The Vicaría began to systematize each case. Starting with the "most proven," the staff completed 478 files of disappeared detainees out of the total 675 it had registered, and sent them to the Interior Ministry between June and October of 1978, in a direct impugnation to the regime. It also bound the files to produce the seven-volume book ¿Dónde están? (Where are they?) (Figure 3). But the dictatorship never investigated a single case.
After the political transition (1988)(1989)(1990) and before its closure (1992), the Vicaría published an updated version of ¿Dónde están? with 984 detailed cases from its archive of 1100 disappeared detainees, entitled "Libros Rojos" (Red Books) (1993). In the post-dictatorship period , through the National Truth and Reconciliation Commission (1991), and the Presidential Advisory Commission for the Qualification of Detained and Disappeared, Politically Executed People and Victims of Political Imprisonment and Torture (2004/2010), the Chilean state recognized this figure, qualifying 1110 people as disappeared detainees by the dictatorship. This activated policies of reparation, memory, and non-repetition.
The number of disappeared persons in Chile played a generative role (Verran 2010): it contributed to the identification, definition, and characterization of an unprecedented repressive practice, installing certainty and predictability, where repression had sought to create obfuscation and deceit. Disappearance became a publicly recognized fact through the documentation and denunciation of victims' relatives and human rights organizations during the dictatorship. In this case, the reality of disappearance and the number of persons disappeared were mutually built, one by the other. After the political transition, the number moved into the field of state accountability, taking part in the official narrative produced by the truth commissions where it was validated and completed. From 2018 onwards when the crimes against humanity committed under dictatorships began being brought to trial, the number transited into the field of justice through citations in legal dossiers about the patterns and methods, based on information provided by the truth commissions which rely on COPACHI and Vicaría archives.
In 2021, this figure was revisited by the Human Rights Undersecretary to update the official number of people detained and disappeared as counted by the Chilean state in its truth commissions, using new evidence from judicial investigations and the homologation of the official count with the international nomenclature, indicated by the 1992 United Nations declaration, concerning the protection of persons against forced disappearances, and adopted in 2006 as an international convention. The update allowed for "a better approximation of the number of victims for whom the state should focus its search efforts" (Ministerio de Justicia y Derechos Humanos 2021, 12). To date, only 303 detained disappeared have been identified through skeletal remains. The collation of the CVs inventory with posterior judicial investigations produces nine wrongly identified cases (seven mistaken, a repetition through misnomer, and one without proven existence) and nine individuals identified as detained disappeared, and dead, despite the fact that the commission only applies the term "dead" to those where the fate of the body is known. This brings the number of detained disappeared to 1092 people. The proposal also includes adjustment in the definition of disappeared detainee including 375 persons hitherto belonging to the category of "politically executed" for whom their bodies are still not recovered. The recount amounts to 1467 victims of "forced disappearance" officially qualified.
In the homologation proposed by the Chilean government, the categories of detained disappeared and politically motivated execution re-meet to produce a new distinction between "executed with" and "without recovery of the body," affecting 375 families who, on the basis of the 1975 distinction, have gathered for decades under the banner of the relatives of the Politically Executed.

Referential and outrageous number of forced disappeared in Mexico
In January 2020, Mexico's federal government acknowledged the existence of 61,637 disappeared and/or missing persons. 5 By April 2021, the Secretariate of Governance (SEGOB) came up with a new figure: 85,053 persons. In June 2021, the National Registry of Disappeared and Non-Located Persons (RNPDNO) accounted for 89,488 persons (see Figure 4). 6 The discrepancy was "officially" explained as the result of an update arising from a study of different sources. Karla Quintana, head of the National Search Commission (CNB), observes a striking detail concerning the total number by the year 2020: 60,053 persons disappeared between 2006 and 2019, while the rest disappeared between 1964 and 2005. This clearly enunciates a differentiation between "figures of the past" and "current disappearances." Most of the latter are associated with the so-called narco wars, characterized by the use of the armed forces to confront drug cartels. By establishing a chronology, numbers also participate in attributing responsibility.
The disappearances that occurred during the "dirty war" (1968-1985 7 ) were only partially documented and visibilized by the Mexican state, which came under international scrutiny. In 2002, the Special Prosecutor for Social and Political Movements of the Past (FEMOSPP) spoke of 532 cases of disappearances but this report was not widely known in its entirety until 2021 when the National Security Archive disclosed it. 8 In 1999, the National Human Rights Commission (CNDH) created a Special Program on Presumed Disappearances (our italics), to give society a concrete response regarding those 532 cases. 9 In early 2006, the 532 cases continued formally open but no progress had been made. Yet, disappearances associated with social and political activism, labeled "numbers of the past," continued to be updated with new names, which Human Rights Watch called "the other disappeared persons." At present, the new patterns of disappearance, associated with extensive military presence throughout the country and its "war" against successive generations of narcotraffic organizations and other organized crime, have not managed to dislodge those 532 "dirty wars" cases from their state of limbo.
Both the 2020 and the June 2021 numbers draw from the last SEGOB report and its reading of "disappearances of the past" and "disappearances of the present." This underscores the challenges associated with "establishing a number" in Mexico. Three elements help to explain this plurality. First, forced disappearances perpetrated by non-state actors organized crime or a combination of police forces at the service of these and hitmenhave increased the problem's scope, transforming it quantitatively and qualitatively, to the point of becoming nearly "immeasurable" by the terms proposed during the "dirty war." Today Mexico faces situations as diverse as the disappearance of women by human trafficking networks and the disappearance of political activists. This plurality affects the creation of equivalence amongst events and prevents aggregation and replication. Secondly, the terminology employed for naming and classifying the problem is debated by government institutions and organizations. On the one hand, the National Search Commission (CNB), the National Disappeared and Unlocated Persons Registry as well as state search commissions refer to "disappeared, unlocated, and located persons." 10 On the other hand, although the term forced disappearance has been used since approximately 1998 11 and is employed by civil society organizations and governmental agencies alike, the week of July 6, 2020 was designated "Week of the disappeared detainee." The term detention/disappearance emphasizes the perpetrators' intent as it is they who exercise the practice with the uncertain whereabouts of the victims. 12 In contrast, the concept of forced disappearance refers to the nonexistence of agency of the victims themselves, once they have been captured by the perpetrators. Moreover, the prevalence of the victims' political identity in the 1960s and 1980s also explains the difficulty in numerical measurement of contemporary forms of forced disappearance for purposes of comparison with those of the past. All this hinders the establishment of patterns and counting, and creates the impression that it is merely a "political problem." Thirdly, the denunciation mechanisms of disappearance in Mexico have been public in naturebut not court-centeredseeking to stir public conscience. Unlike Chile (thanks to COPACHI and the Vicaría), the Mexican judicial system's credibility is low (Fix and López 2001), family members lack incentives and effective access to justice, and, as in Colombia, judicial proceedings have not moved beyond initial stages.
The Mexican case reveals that the plurality, heterogeneity, and diversity that the phenomenon acquires make the trajectory of its numbers complex. Mexico's figures lack stability, indicating the extreme flexibility to which the term disappearance is exposed, as well as the diversity of violences and the plurality of counting patterns the term and the number encompass. However, the referential number supported by the baseline −50,000 unidentified in the country's forensic labs-and iconic numbers (Verran 2010) such as 43, which represents the 43 students of Ayotzinapa 13 and many other victims of forced disappearance the world over (Diaz and Ovalle 2018), 14 highlight the magnitude of this crime and the impunity reigning in Mexico, becoming an outrageous number, denounced by massive protests among the population, pressuring the approval of the 2017 General Law on Forced Disappearance, which created the National Search Commission, a national forensic data base, and territorial commissions focusing on victims.
Provisional figure: the social struggle to maintain the search for disappeared persons in Colombia In 2016, on the occasion of a report on Forced Disappearance by the National Historic Memory Center (CNMH), the institute's director at the time, Gonzalo Sánchez, expressed concern for the country's enormous under-registering of the event. Sánchez attributed it to two factors: (1) "for a long time the disappeared persons were part of registries on abductions or murders, and (2) the chain of violence that sought to guarantee impunity for this crime" 15 (CNMH 2016, 12). However, he contends that despite the under-registry, after comparing and depurating current databases, the initial number of 26,000 victims was superseded and the new figure for 1970 to 2015 can be estimated at 60,630 disappeared persons. According to the Center's official calculations the phenomenon today surpasses the horrific figure of 80 thousand disappeared persons. (Figure 4) His reading sheds light on the fact that both the stabilization of a number as well as the naming and narration of a problem are integral parts of a human rights violation number. To understand the meaning, obstacles, and urgent tasks ahead, a brief reconstruction of the trajectory of this occurrence and the counting is needed. We suggest five "historic vertices" 16 through which the phenomenon has traversed in public visibility, in the explosion of numbers, in its embeddedness in legal-political devices, and in civil action struggles from 1970 to 2021.
The first vertex is that of the phenomenon's public outbreak between 1970 and 1981. This was the period of incipient documentation and fearful denunciation of the problem in a climate of conflict between the state and civil organizations concerning what is understood as, conforms, and typifies disappearance. 17 During this period the primary victims were leftist activists, political leaders, union members, and teachers, mostly young and adult men. The victim profile was consistent with the counter insurgency national security doctrine that prevailed throughout Latin America (CNMH 2016). A second vertex may be identified between 1982 and 1990, coinciding with the generalization of different types of violence in the country, with a prevalence of the practice of forced disappearance above other victimizations during a "dirty war" against all forms of political dissent, unleashed in the systematic extermination of the leftist Unión Patriótica political party, and the persecution of other civil grassroots movements. 18 The figure soared from 269 disappearances in 1982 to 1388 cases in 1990 (CNMH 2016). The third vertex, between 1991 and 1995, is characterized by "a decline" (CNMH 2016) in comparison with other victimization events, but a rise in armed insurgency responsibility in this crime. The fourth vertex, between 1996 and 2005, shows a significant increase, when registries confirm the disappearance of one person every two hours, coinciding with the paramilitary reinstallation and territorial pacification in the country (Jaramillo, Berón, and Victoria 2020). During this period more than 32,000 victims of forced disappearance were confirmed. At this juncture pressure from historic organizations such as Asociación de Familiares de Detenidos Desaparecidos (Association of Relatives of Disappeared Detainees) (ASFADDES) was crucial for the enactment of law 589 that codified this crime, creating mechanisms for prevention, registry, and treatment. 19 Finally, the period between 2006 and 2021, when "disappearances decline but do not cease" and manifest a degree of territorial propagation, 20 also identified in Mexico. In Colombia, 1000 of the country's 1103 municipalities counted disappearances during this period. This historic vertex is characterized by the emergence of post-demobilized groups, and coincides with the signing of the Peace Accord between President Santos' government and the FARC -EP ( Figure 5).
This Peace Accord gave rise to the creation of the Search Unit for Persons Considered Disappeared (UBPD) to establish an integral system to document and register the crime, and also locate the whereabouts of the disappeared persons through regional plans and agreements for territorial searches. Four registry and documentation systems of disappearance, each with different purposes coexist: (i) the Registry of the Observatory of Memory and Conflict pertaining to the National Historic Memory Center, that contributes to historic clarification; (ii) the Victims Unit Registry that guarantees service and reparation to victims; (iii) the National Registry of Disappeared of the Legal Medical Institute, to obtain legal identification of victims; (iv) the information system of the National Attorney General's office, known as the Accusatory Oral Criminal System, for prosecutorial purposes.
Regarding registry and legal prosecutionother entities such as the Prosecutor and the Special Jurisdiction for Peace already exist for these purposesthe proposal calls for this Unit to transit to recognition of the centrality of the victims and their family members in this process. At present, this entity and the Integral Truth, Justice, Reparation, and Non-Repetition System as a whole (arising from the last Peace Agreement) channel information from the country's different disappearance registries to focus on and contribute to the task of "effective clarification" to attain not only the "true" number of disappeared persons, but also to find out how, when and where the disappearance occurred. The fundamental challenge is to concretize a long overdue paradigmatic shift regarding disappearance in Colombia: from a penal search processcentered on the crime and its authorsto a humanitarian, extrajudicial search process centered on the disappeared persons and those looking for them. This shift suggests that the statistic of disappeared people is part of a larger conflict between social organizations and the state to force the latter to take responsibility in the solving of the crimes and the final destiny of the bodiesas a sociopolitical, not merely legal phenomenon.
Despite initial progress, various victims' organizations still dispute the "politics of numbers" (Crisp 1999) of disappearance, mobilizing to ensure greater effectiveness of the new state apparatus. Thus, they have repeatedly demanded that the UBPD determine the whereabouts of victims numbering 120,000, which they insist is the "minimum figure" 21 and not the official 80,000. Among the most active groups are the ASFADDES and Madres de Falsos Positivos (Mothers of False Positives). With the shift of focus to the victims and their families, the demands made by these groups gain legitimacy and resonance. One expression of such politization is the struggle for recognition of their knowledge, which enables them to "resist the catastrophe stemming from forced disappearance, seeking to configure ways of co-inhabiting with it in their daily lives" (Agudelo & Aranguren 2020, 1). Consequently, Colombia's current transitional bureaucratic architecture faces the challenge of taking these political-moral demands seriously. In this sense, the total number of disappeared people in Colombia, as well as remaining the object of formalization, rationalization, and bureaucratization processes, as with all provisional numbers (Lampland 2010), is a highly disputable statistic creating tension between the state and organizations, while it continues to grow as the current government fails to live up to the peace agreement (Bermudez 2022, 3).

Conclusions
In the case of enforced disappearance, it has been argued that neither the phenomenon nor the "person-as such" can be grasped statistically (Edkins 2011); that numbers tend to "fall apart" when they must count liminal subjects. But counting also performs the fundamental human rights work of equalizing every life and death (Nelson 2010) while as an inscription, it contributes to make a complex phenomenon evident. The analysis of the trajectory of the numbers of disappeared persons in three Latin American societies described here, indicates that numbers in the human rights field are not rigid;nor are they conspicuous, singular, or permanent. Because the reality they try to capture is never fully known and they emerge and circulate amidst disputes between governmental and political forces, these numbers are not easily stabilized. Multiplicity, heterogeneity, invisibility, fear, and diverse costs, play against the urgent need to make enforced disappearance quantifiable.
We generally believe that we need singular numbers. Any coexistence is a failure. Yet, as we have seen, numbers belong to processes and relations within which they perform their effects and from which they acquire their value. In Colombia and Mexico the number used to designate forced disappearance tends to be plural, unstable and ambiguous. This occurs because those who denounce and search for disappeared persons tally one number and the State another, generally, much lower; and because different units of measurement and materialities are subjected to the count: people sought or claimed, people officially considered disappeared, unlocated persons, unidentified bodies, mass graves, exhumed human remains, identified bodies. Materiality itself may vanish, such as in the fields of Veracruz, Mexico, when bones themselves disappear. And when the disappearance forms part of more complex crimes (human trafficking, sexual exploitation, narco-wars), multiplicity tensions the number's cardinal property.
However, if we look at numbers' ability to mobilize human rights concerns we can conclude that they do valuable work. As one aspect of management that has demanded the tireless contestation and pressure of social organizations and public entities (Vicaría de la Solidaridad in Chile, FEMOSPP or the Comisión Nacional de Búsqueda in Mexico, the Centro Nacional de Memoria Histórica or UBPD in Colombia), the existence of provisional figures calls for public recognition of the phenomenon, so as to contribute to justice and reparation, with material elements to enable lawyers and the victims' relatives to continue searching. In this sense, provisional numbers have contributed to make enforced disappearance a crime to be known and acted upon, pushing for the production of accountability regimes.
The three cases, with different temporal distances regarding the practice of disappearance, prove that numbers transit through different moments. One is when they paralyze, as a type of fixed figure, and the civil society's legal episteme and political activism, or that of a transitional apparatus such as a Truth Commission in Chile, play a central role. Another moment is when numbers participate or even sustain social mobilization, because they are porous, fragile, and unstable figures, that one must fight to stabilize, or at least to keep on the horizon of public issues, such as in Colombia, a number capable of activating the political-legal mechanism, even functioning as cognitive and political technology; or in Mexico, a number capable of mobilizing and sustaining indignation, denouncing passivity, concealment, and institutional and political impunity.
Forthcoming in-depth studies may expand the question of human rights numbers' liveliness through contentious trajectories by which they affect and are affected by the human rights causes they help to constitute: questions such as the sub-registries, what defies quantification, numbers' interaction with other technologies (testimony, historic narrative, emblematic cases), the overexploitation, manipulation and encryption of numbers, their residual effects and the wear and tear of the denunciation function of numbers: that thin boundary between the mantra and the invisible number.

Disclosure statement
No potential conflict of interest was reported by the author(s).

Funding
This work was supported by the Newton Fund and the National Research and Development Agency of Chile. The article is part of the projects "Political Violence and Human Rights Violations Accountability: Circumstances, Uses and Effects of Forced Disappearance Registration. Lessons from a Comparative Perspective in the Americas" (Newton Prize for the Americas, 2018) and "Beyond the Victim Paradigm" (Fondecyt, 1190834).

Notes on contributors
Oriana Bernasconi is Professor of Sociology, Universidad Alberto Hurtado, Chile. Her work examines configurations of subjects and subjectivities. In the last years, her research has focused on the subjects of political violence and on the role that documentation of human rights violations and archives play in a society's capacity to come to terms with evil pasts. Her work has been published in Subjectivity, Sociology, The International Journal of Transitional Justice, Colombia Internacional, Discourse & Society, and Qualitative Sociological Research, among other journals. Her last book is Resistance to Political Violence in Latin America. Documenting Atrocity (2019).