Primo Levi, Camp Power, and Terror Capitalism: A Conversation with Darren Byler
What does Italian writer and Holocaust survivor Primo Levi have to tell us about life in reeducation camps in Xinjiang today? What role does labour play in these facilities? What is terror capitalism and how does it relate to other frontiers of global capitalism? Can there be such a thing as ‘benign’ surveillance? These and other questions are at the centre of this conversation with anthropologist Darren Byler, who in recent years has emerged as a leading voice in documenting the mass detention of Uyghurs and other Muslim minorities in northwestern China. The impending publication of three books of his—In the Camps: China’s High-Tech Penal Colony (Columbia Global Reports, October 2021), Xinjiang, Year Zero (co-edited with Ivan Franceschini and Nicholas Loubere, ANU Press, December 2021), and Terror Capitalism: Uyghur Dispossession and Masculinity in a Chinese City (Duke University Press, January 2022)—offers us an opportunity to revisit some key aspects of his work.
Ivan Franceschini: We are often warned to avoid facile historical comparisons when we discuss contemporary matters, yet history offers uncanny lessons when it comes to the Xinjiang camps. In your books, you repeatedly cite the great Italian writer and Holocaust survivor Primo Levi to highlight a few common patterns between the situation in the camps in Xinjiang as described by survivors and his experiences of being interned in Nazi lagers during World War II. This is something I would like to explore further in this conversation. The first obvious point of comparison is in the dehumanisation of prisoners. In particular, you mention this passage from Levi: ‘Some of them beat us from pure bestiality and violence, but others beat us when we are under a load almost lovingly, accompanying the blows with exhortations, as cart-drivers do with willing horses.’ One of the most infamous memories and legacies of the lagers are the numbers tattooed on the wrists of prisoners. How does this dehumanisation occur today in the camps in Xinjiang?
Darren Byler: As I was grappling with the accounts I heard from detainees and that I saw detailed in internal police documents, I turned to Primo Levi’s work as an example of the kind of writerly method that I wanted to use in the book and less because I thought the Holocaust was the best example with which to compare the system. While death from neglect, disease, and abuse does appear frequently in the Xinjiang camps, the intentional mass killing of the Nazi death camps appears to be absent. So, in some ways, historical examples from China or the Soviet Union might have been better analogues when it came to camp ideologies and management, and the detainee camps in occupied Iraq might have been a better point of comparison for contemporary detainment of racialised, religious others. Levi’s work stands out though because he writes as a careful observer of systems of power. By noting the small details and the bureaucratic norms of camp life, he shows how extraordinary violence is made possible. The goal in my writing was to make careful, rather small, truth claims that add up to something larger.
Levi, and perhaps Hannah Arendt (1964), have done some of the best thinking about the implications of dehumanising systems, and that is something I wanted to consider in this book as well. I wanted to understand what mass internment means for the contemporary moment, particularly what it means to use advanced technology as part of the process. To be clear, I’m writing from the position of someone who has lived in the society I write about for two years and have been thinking about for a decade, but that is not the same as someone like Levi, who directly experienced the things he is describing, or Arendt, who is from the community that was targeted. So, I’m writing from a greater distance than both of them and, of course, with infinitely less skill.
So that is why I turned to Levi’s work. But, in doing so, certain things that people told me stood out in ways that they might not have otherwise. One of them has to do with a Kazakh herder who I refer to as Adilbek. He really emphasised how the camp experience felt to him as someone who had spent a lifetime with sheep and as a Muslim who had avoided pork his whole life. He said the guards struck them hard if they wanted to punish them, and with gentler blows if they were just trying to hurry them along, ‘like a farmer herding sheep’—a line that was uncanny in its resemblance to the lines you quote from Levi. The guards, too, most of whom were Kazakh and Uyghur, also came from farming life and had the same halal traditions. So, in their effort to exert their authority over the detainees, they referred to them directly as dogs or pigs, particularly when they were beating them over challenges to authority. Numerous detainees from across the region spoke about this usage of dehumanising names, though Adilbek was the most explicit in describing the different types of beatings.
Much of the dehumanisation in the camps comes from deindividuation and imposed scarcity. In every camp, people were given a uniform and had their head shaved or closely cropped. In many cases, they were forced to use buckets as toilets in the presence of other detainees and in front of high-definition surveillance cameras. The smell of the buckets permeated the cell. The detainees were permitted to shower very infrequently and given a change of clothes even less often. Showering is timed so detainees often fought with each other over access to water and soap. They are forced to sing ‘Red’ songs before meals as loudly as possible. Often, food is withheld until they sing well enough. Many people became ill due to poor nutrition, lack of sanitation, hours of sitting on plastic stools or benches, beatings, and lack of medical attention. While they were not assigned a permanent number, in the cells, they responded to roll call by shouting out the number they had been assigned within the cell. Numbers are typically assigned based on who is the ‘class monitor’ [班长] of the cell and the time of their arrival in the cell; the last to arrive is given the highest number and least favourable sleeping arrangements. If there are more people than can sleep hands to feet on a narrow bunk, the detainees are forced to take turns sleeping and ‘standing guard’, which introduces an additional antagonism in the social order of the cell.
IF: One second similarity can be found in what Levi called the ‘grey zone’, which, sadly or perhaps aptly, today is also the name of one of the main outlets spreading misinformation about what is going on in northwestern China. In The Drowned and the Saved (1986), Levi discusses at length this ‘grey zone’ of ‘prisoners who to some extent, perhaps for a good purpose, collaborated with the authority’, which ‘was not thin, constituting instead a phenomenon of fundamental importance for the historian, the psychologist and the sociologist’. In his words (which I am quoting in my translation, here and below): ‘There is no prisoner who does not remember that, and who does not remember his surprise then: the first threats, the first insults, the first blows did not come from the SS, but from other prisoners, from “colleagues”, from those mysterious characters who also wore the same striped tunic that they, the new arrivals, had just put on.’ In your books, you discuss at length the ways in which the reeducation camps in Xinjiang produce the conditions for these types of dynamics to emerge—in fact, you dedicate entire chapters of In the Camps to the collaborators; you also discuss how relationships between prisoners themselves are often characterised by violence. Can you tell us a bit more about this dimension of the camps? Who are the people who end up sustaining the economy of the camps and what are their motivations?
DB: Much of the discipline in the camp spaces came from other detainees within the cells, Uyghur and Kazakh guards, Muslim and Han camp instructors and ‘life teachers’ [生活老师], Han and Uyghur Civil Affairs Ministry bureaucrats who monitored detainee progress, judges, and prosecutors from all ethnicities. I see all of these individuals as being in the position of ‘swarms of low-level functionaries’ that Levi describes as the primary actors in the grey zone. Because they are operating in or around the extralegal space of the camp, they have extraordinary power over the detainees under their charge, but they are not the architects of the system—the camp commanders, the system designers, the party secretaries who set the quotas, and so on. The Muslims involved in the system face maximal coercion—as a constant rhetoric of rooting out and destroying ‘two-faced’ people is held over them at all times.
Han civilians—who work in the camps or evaluate the trustworthiness of their Muslim neighbours—also face the threat of loss of promotion, losing status if they appear unsupportive of the campaign, and potentially arrest (though this appears relatively rare) if they take too much of a role in resisting the system. On the other hand, those who do participate are often rewarded for their perceived sacrifice with promotions, commendations, and bonuses. Here there is an echo of Arendt’s thinking regarding dulling of conscience [Arendt 1964]. If there are few dissenting voices, little information that contradicts the regime of truth within a society, it is relatively easy to accept the norms that are given without much reflective thought. One of the things I’m trying to think through is the way advanced technology, which has been used to diagnose ‘terrorism crimes that are not serious’ [Government of China 2019] throughout Xinjiang’s Muslim population, aggregates and expands this type of unreflective truth-making. In many cases, it appears as though technology-led policing creates scripts of action that diminish the need and possibility for thought.
IF: This links to the question of solidarity, which you discuss in Terror Capitalism. In his recollection of the Nazi camps, Levi also remembered that ‘one entered hoping at least in the solidarity of your companions in misfortune, but the hoped-for allies, save some special cases, were not there; there were instead a thousand sealed monads, and among this a desperate struggle, hidden and unceasing’. What about solidarity both between those within the camps and between those inside and outside, including Han people?
DB: My sense is that Xinjiang—in and out of the camps—is also largely a society of ‘a thousand sealed monads’. Inside the camps, people do compare notes at least initially as they try to make sense of what just happened and what will come next. So, at first, there is a lot of whispered discussion regarding why people had been detained. But in interrogation and over time inside the cell, it appears as though guards and interrogators often turn detainees against each other. During interrogations, detainees are pushed to confess a kind of ‘middle guilt’—not too guilty that they could be sent to prison but also not too innocent that they did not deserve to be in the camp (professing one’s innocence is construed as being resistant or having a ‘bad attitude’). Declaring this guilt often requires that they name others who are more guilty than them: the ones who first invited them to a Quran study group on WeChat or showed them how to use a VPN [virtual private network], and so on. The privileged position of ‘class monitor’ carries with it the responsibility of informing on resistant or failing detainees in the cell and doling out punishment for lack of discipline. Over time, as detainees become more desperate, they often become more suspicious of each other, so some of the initial solidarity begins to dissipate and people are often simply silent.
Outside the camps among Han populations there is certainly ethnonationalist solidarity in the campaign. When I was there last, in 2018, Han ‘volunteers’ who had been sent down to villages to monitor Uyghur families referred to each other as comrades and really saw themselves allied in shared struggle against the terrorism-in-hiding that they projected on to Uyghurs. In some cases, they saw themselves as helping Uyghurs in a kind of paternalistic way—that they were saving Uyghurs from themselves.
Among certain segments of the Han population, particularly those who lived in proximity to Uyghurs before the waves of market-oriented economic migrants in the late 1990s and 2000s, there is a good deal of sympathy for the Uyghurs and Kazakhs, but in most cases, such people are not in positions to act outside their immediate social sphere. Many of the leaked documents that have shed light on the inner workings of the system come from Han individuals who want to furtively expose what has been done in their name.
Among Uyghurs and Kazakhs, in some cases, the camp system has produced greater solidarity across social class and ethic difference, as many wealthy, urban citizens who were previously less exposed are also missing loved ones. But it has also produced a great deal of alienation. The family members of detainees report being ostracised by their communities, and many spouses divorced as a result of one of them being taken. All of this produces the sealed monads Levi describes.
IF: Levi’s oeuvre also offers some fundamental insights into the role of labour in concentration camps, beyond the infamous moniker ‘Arbeit Macht Frei’ that could be found at the entrance to Auschwitz and other Nazi concentration camps. Again, in The Drowned and the Saved, he writes that ‘unpaid work, that is slavery, was one of three purposes of the concentration system; the other two were the elimination of political adversaries and the extermination of the so-called inferior races’. He notes how, in the earliest lagers, work served purely persecutory purposes and had no use whatsoever in terms of production. This was because ‘for the Fascist and Nazi rhetoric, which in this is heir of the bourgeois rhetoric, “work dignifies”, and therefore the undignified adversaries of the regime are not worthy of working in the usual sense of the term. Their work must be punitive: it must not leave space to professionality, it must be that of beasts of burden, pull, push, carry weights, bend your back on the land. This is useless violence as well: perhaps only useful to put down the current resistances and to punish the past ones.’ What purposes does work serve in reeducation camps in Xinjiang?
DB: Many former detainees noted that being allowed to work inside the camp or being transferred to work in a factory from the camp was a great relief from the drudgery and violence of life in camp cells and classrooms. Simply having something to do and a bit more bodily autonomy felt humanising. Because the factories where former detainees are transferred (in distinction from factories where farmers are assigned) are managed with the support of camp personnel, many detainees still viewed them as a carceral space. In some cases, they were locked in cubicles as they worked, and they were not permitted to move beyond allotted boundaries without permission. The guards and managers viewed them as dangerous and often still treated them as criminals rather than as professionals or employees.
Those who were placed under a form of house arrest likewise talked about how they were forced to perform their submission to the Civil Affairs Ministry workers who monitored them. One former detainee who was assigned to work as a cook and cleaner in a Civil Affairs unit [社区] said the two Han women who monitored him expected him to come to every flag-raising ceremony and sing Red songs on command. He said he learned to always keep a smile on his face and say, ‘Okay! Okay!’ [行!行!] to everything they asked. He said it felt like he was their pet. Among the former detainees I interviewed, none of them felt able to ask about compensation for their work. In general, they were made to feel that they should just be grateful that they were no longer locked in camp cells. And, in some cases, they were told quite explicitly that they could be detained again at a moment’s notice.
Forms of status coercion that come from detainability—which, as Erin Hatton (2020) points out, refers not to the threat of losing a job or income, but the threat of being labelled untrustworthy—are extreme.
This means that work is primarily about extracting maximal profits from deeply controlled workers, which I refer to in Terror Capitalism as a permanent underclass. That is, at least in the near term, such workers are indefinitely excluded from the right to freely choose where they work and under what conditions. This is the case for both former detainees and Uyghur and Kazakh villagers who are simply deemed unproductive or ‘surplus labourers’ [剩余劳动力]. State documents show that individuals who fall into the categories of assigned work are graded based on their trustworthiness [Qapqal County Social Security Bureau 2018]. Those who are given a certificate validating their trustworthiness are typically those who are sent along with police officers and cadres who ensure their discipline to work in factories across the country. Those who fall into normal or untrustworthy categories are sent to work in locations in Xinjiang or into a range of different types of ‘training’. The camps—or ‘concentrated closed training and education centres’ [集中封闭培训教育中心]—are one track used in this training. In all cases, a securitised form of the ‘dormitory labour regime’ (Smith and Pun 2006) used in relation to migrant workers across the country is used by factory managers to manage coerced Muslim workers. For former detainees working in camp-associated factories, rather than Muslim surplus labourers working in eastern China, the security measures used in this ‘reeducation labour regime’ are greatly enhanced with forms of biometric tracking, checkpoints, and locked enclosures.
IF: Interestingly, Levi also takes care to point out how work could serve as a defence in the camps for those who found the way to be assigned to jobs in line with their skills. These people, he wrote, ‘by finding their usual activity, recovered at the same time, to a certain extent, their human dignity’. For others, labour was a relief in that it distracted them from the thought of the worse things to come. Does the economy of the Xinjiang camps leave space for this kind of valorisation of labour?
DB: Certainly, working provides an escape and a kind of valorisation, especially for those who had skills they had cultivated before detention. For instance, Vera Zhou, a University of Washington geography student who was pressed into work as an unpaid English tutor for a police officer’s family after her detention, talked about how having a social role again helped her to regain her confidence as an individual and understand the injustice she had experienced. In general, because work was offered only to those who had proven their submission and knowledge by passing Chinese-language and legal knowledge exams, it means that detainees have passed a certain threshold in their reeducation. So, it gives them the feeling that the worst is in the past and that they now have more security. However, the reeducation labour regime operates within a highly coercive system of surveillance and legal indeterminacy that prevents demands for equitable wages or autonomous work conditions. Workers are always in a highly dependent position due to their detainability and the infrastructure that enforces it.
IF: There is no question concentration camps now have quite a long history. As Andrea Pitzer has brilliantly shown in her One Long Night (2017), the appearance of these structures is intimately linked to new technological and scientific advances towards the end of the nineteenth century and has little to do with the political system of a country. Concentration camps were first conceived by the Spanish in Cuba in the late 1890s, expanded by the British in South Africa during the Boer War, normalised by all warring factions during World War I, and finally manifested in the extreme variants of the Soviet gulags and the Nazi lagers, before lapsing into the more familiar forms of ‘black’ detention sites that became common in Latin America in the 1970s. Yet, Levi and most authors who have discussed different iterations of concentration camps in recent history frame their writings in terms of state violence, authoritarianism, or even totalitarianism. Instead, you chose to discuss the camps in Xinjiang as a form of ‘terror capitalism’ within the context of ‘global capitalism’. Some might see this framing as counterintuitive and accuse you of downplaying the agency of the Chinese authorities. Why do you think capitalism, or ‘terror capitalism’, is the best frame to make sense of what is going on in Xinjiang? But, most importantly, do you believe the camps in Xinjiang represent a historical rupture from the other forms of mass detention that appeared during the long twentieth century? And, if so, how?
DB: In analysis and critique of camp systems (Andrea’s work is an exception to this) there is often a reflex to place the blame for them on the singular evil of key leaders. While political leadership and ideology certainly play a pivotal role—the camps in Xinjiang likely would not have been possible without people like Xi Jinping, Chen Quanguo, and Zhu Hailun and the effect of the Party-State system—my feeling is that limiting analysis to such actors risks ignoring the economic and material conditions that gave rise to the underlying antagonism and fuel its ongoing momentum.
So, for me, the story of the camps and surveillance state goes back to the 1990s when the drive for natural resources to fuel China’s export-oriented economy is what turned the Uyghur-majority south of the region into an internal settler colony. The Uyghur response—both real and imagined—to their dispossession is, in turn, what spurred the state’s adoption of War on Terror rhetoric and tactics post 9/11. This war space offered a place to build and experiment with new population-control technologies that have helped propel the exponential rise of computer vision and digital surveillance industries into a wide range of domains. At the same time, the camps and surveillance systems have become a carrier of the regional economy, attracting companies interested in subsidised, cheap, and subservient labour in low-skill domains such as garment manufacturing. That is, there is an economic and political call and response that have produced the dynamic we see in Xinjiang today.
And, importantly, this dynamic is interlinked with dynamics that pervade global economic and political systems not only in terms of supply chains, policing tool development, and insurgency theory, but also in terms of what frontiers of contemporary capitalism look like. Part of what I want to show in the book is the way the term ‘terror capitalism’ should be thought of in relation to recent scholarship focused on particular frontiers of global capitalism such as disaster capitalism and surveillance capitalism [Klein 2007; Zuboff 2019]. It also, importantly, focuses attention on the way these global novel formations should be thought of as articulated to the ongoing imperialist and colonial processes of racial capitalism [Robinson 1983].
Disaster capitalism theorises the economic and political complex that responds to emergent disasters associated with the climate crisis—a type of ‘counter-disaster capitalism’ that justifies itself by masking the role of capitalism and political ideology in the production of disasters. Similarly, surveillance capitalism, or ‘dataveillance’, hides in plain sight, masking the production of controlled digital subjects and the exploitation of precarious, often racialised, workers in a patina of ‘smart’ convenience and logistical speed. These new global formations are connected to racialised capitalism, which describes the way economic frontier-making is premised on capturing the land and labour of ethno-racial others—a type of capitalism that justifies itself by masking the production of race and differential value.
Terror capitalism is related to these framings of contemporary global systems and their histories. On the one hand, it is a type of ‘counterterrorism capitalism’—premised on the prevention of the disaster-like terrorist event. However, naming it ‘counterterrorism capitalism’ would obfuscate the way the capitalist formation itself produces the figure of the terrorist as its object of investment. Terrorists and terrorism are not a priori givens, but rather historically contingent phenomena. Treating them as concrete objects that can be ‘countered’ works to mask a novel form of global racialisation that is attached to Muslim bodies as a result of the Global War on Terror [Mamdani 2002; Rana 2011; Brophy 2019]. In this sense, terror capitalism—particularly in a domestic, internal, colonial context like northwest China—combines the war-like machinery of terrorism ‘disaster’ prevention, corporate and police dataveillance with the racialised apparatus of capitalist-colonial frontier systems. Terror capitalism is about the production of the terrorist as a juridical category and object of a technological gaze, and in turn the expropriation of the land and labour of those who are produced as potential terrorists.
Framing the system as an expression of terror capitalism produces a different object of critique than framing the system as a new form of state terror. Terror capitalism, unlike state terror, focuses attention on the linkage between state and private industry investment in the mundane activity of capitalist frontier-making. Even more importantly, it refocuses attention away from the methodological nationalism of ‘state authority’ narratives, which dominates much social science discussion of contemporary China and, in doing so, frustrates an easy slippage into new Cold War binaries—which frame China as a site of illiberal despotism and human rights crises, ignoring the way racial capitalism and surveillance systems that target unprotected citizens are generated in the United States and elsewhere. None of this refocusing is to deny the role of the Chinese Communist Party in mandating and incentivising private industry to take on their roles in the systems; it is simply to say that such a focus has been made the dominant frame of analysis and my research shows that it does not fully capture the political and economic logics and material effects of the system.
IF: Digital surveillance is a fundamental topic in all of your books, since, as you make clear, the latest advances in surveillance technology are what made the Xinjiang camps possible. While pointing out how surveillance systems support systemic racism and dehumanisation by making targeted populations detainable, even in the West, you also explain that there are significant differences between what is going on in northwestern China and elsewhere. ‘In Xinjiang’, you write in In the Camps, ‘the networks of cameras are much denser and supported by checkpoints and data surveillance, and every resident of the region has submitted their biometric data to the authorities in a comprehensive “public health” initiative. Because of the fidelity and scale of the dataset on which their algorithms are trained, the tools of the Xinjiang authorities are much more finetuned and invasive.’ If technology is held to be neutral and the difference is only in degrees, it seems that banning trade with Chinese companies engaged in these fields or demanding that universities stop cooperating with Chinese actors in such fields, while important, is just clutching at straws. In his treatise about post-pandemic politics, Benjamin Bratton has recently argued for the need for an epistemic change in the direction of a ‘positive biopolitics’, which would involve a substantial reevaluation/reframing of the term ‘surveillance’. Do you think that is possible and advisable?
DB: Advanced technology definitely has a role to play in a wide range of issues confronting the contemporary world—certainly in terms of the climate crisis and global social inequity. The answer to the surveillance state and digital capitalism is not simply banning algorithmic tools. Rather, it demands technology design for human liberation, social justice, and the protection of vulnerable populations. It means clear and enforceable limits on what kinds of data can be collected and for what purpose. Refusing to support Chinese companies that are involved in the camps, or boycotting US companies that support US policing and military systems—both of which are things I support as a short-term tactic from a moral justice perspective—will not in themselves produce the structural changes that are necessary to make surveillance work in a fundamentally different way (such as what Bratton suggests).
As a social scientist, my research method is itself a type of surveillance—a granular-level process of observation and participation in the phenomena I want to understand. In the past, this method, ethnography, was placed in service to imperial projects and, in Iraq and Afghanistan, counterinsurgency theory. The problem with surveillance is not the method or technology per se, but rather has to do with who is looking, what are they looking at, and for what purpose. In an era of digital enclosures and big tech, the totality of life is now the object of surveillance and the ‘who looks’ is defined by those who control these enclosures. For protected citizens like myself, this situation is disconcerting but not dire. For unprotected citizens like Uyghurs, or undocumented immigrants in the United States, these technologies are life defining. Technology design and regulation must start from that position, rather than that of protected citizens. Ultimately, that is what I hope these two books push readers to consider.