I. The Paradox of Imagining Hospitals Otherwise
Janina Kehr and Fanny Chabrol close their review of hospital anthropology in the Annual Review of Anthropology with a question: “How can hospitals become otherwise, emancipated from the violence and exclusion produced by racial, gendered, and classist technocapitalist forces?” They follow it with another: “What would a hospital that centers care, healing, and human dignity look like?” (Kehr and Chabrol 2026: 227). These are serious questions, asked by two scholars whose work on hospitals I respect, and they deserve a serious answer. My answer begins from a simple historical observation. Hospitals have already become otherwise, many times over.
Tuberculosis sanatoria spread across Europe, North America and the colonial world, and then within a few decades largely disappeared. County asylums were built as self-contained communities, reorganized and in most cases closed. Isolation hospitals lost the functions that had once justified their existence. During the influenza pandemic of 1918, fraternity houses, schools and disused colleges became hospitals for a few weeks and then became something else again. COVID-19 produced another round of rapid assembly, adaptation and closure. Meanwhile St Bartholomew’s Hospital in London has kept its name and its site since 1123 through changes so complete that almost nothing done there today resembles what was done there at its founding. The central puzzle of this article follows from the contrast. Why does an anthropology committed to imagining institutional transformation attend so little to the transformations that have already taken place?
Kehr and Chabrol do not ignore history. They write about colonial legacies, about closures driven by austerity and disaster, about Fanon’s reforms at Saint-Alban and Blida-Joinville, about the Black Panther clinics. My claim is narrower. Their historical references rarely become explanations of how particular hospital arrangements emerge, persist, change and disappear. I will argue that this is no accidental omission. It follows from three linked commitments in the review: taking the hospital as an already identifiable institutional object, interpreting its political complexity through expansive concepts such as haunting and entanglement, and treating emancipatory possibility as the privileged horizon of inquiry. Together these commitments allow the review to show that hospitals are politically significant without explaining the coordinations that make any particular hospital work, harm or change.
The corrective I propose comes from Living Value Theory, which begins from continuously changing living coordination across five irreducible mediations: multisensorial embodiment, multispecies being-with, multiversal dwelling, multimateriality and multisymbolization (Ecks 2026a). An institutional identity such as “hospital” is an L4 symbolic stabilization within that coordination, and the durability of the identity must never be confused with the temporalities of the coordination it designates (Ecks 2026e). The guiding proposition of what follows is that an anthropology of hospitals must explain what hospitals have become before it can say much of value about what they might become.
My principal test case is Saiba Varma’s The Occupied Clinic: Militarism and Care in Kashmir (2020), which Kehr and Chabrol cite at the point where their argument about care and violence is most exposed. I use it twice and for two different purposes. The first is analytical: Varma’s ethnography draws distinctions that the review’s synthesis erases. The second concerns the conditions under which that ethnography was produced, which became the subject of a public controversy in 2021. I was close to that controversy when it broke, and I write about it here only on the basis of the public record. I state that proximity at the outset for the same reason I will argue that the positions from which ethnographic knowledge is produced must be stated.
II. How a Literature Review Preselects Its Conclusions
The review deserves a fair reconstruction. Kehr and Chabrol organize two decades of hospital anthropology as a movement through five problematizations: hospitals as cultural sites, as political and economic terrains, as haunted infrastructures, as more-than-human environments and as more-than-disciplined sites. Within this arc there is much valuable ethnography. They discuss Livingston’s account of improvisation on a Botswanan oncology ward, Street’s analysis of diagnostic instability in Papua New Guinea, the brokerage of ward boys and cleaners in Bangladesh, kin who feed and nurse patients where hospitals provide neither, and the political economy of partnership hospitals and clinical trials. They rightly reject the assumption that hospitals everywhere are “identical clones of a global biomedical model” (van der Geest and Finkler 2004: 1995). None of this is in dispute. My concern is with what happens when this differentiated scholarship is assembled into the review’s larger theoretical synthesis.
The decisive passage comes early. “This review is by no means exhaustive,” the authors write, “as we have chosen not to include a large corpus of ethnographies in which hospitals figure primarily as institutional backgrounds for the understanding of biomedicine, its interventions, technologies, treatments, and patient experiences. We also do not review recent collaborative projects between anthropologists and health care workers that documented how hospital services adapted during and after the COVID-19 pandemic.” Instead they focus on literature that engages with “hospitals as such,” as they are “diversely built, inhabited, maintained, worked in, transformed, destroyed, closed, imagined, experienced, or judged” (Kehr and Chabrol 2026: 219).
Every review must draw boundaries, and this one is defensible as a way of tracing a particular intellectual trajectory. Its consequences for the review’s conclusions are nonetheless considerable. The two bodies of work set aside are precisely those most likely to show what hospitals accomplish and how they actually change: ethnographies of treatment and its effects, and studies of how services adapted during the most rapid wave of hospital transformation in living memory. The list of what “hospitals as such” undergo includes being transformed, destroyed and closed, yet the review selects its literature according to how hospitals have been conceptually framed rather than according to which empirical processes would explain those transformations. The question of what hospitals could become is then posed on a foundation from which the evidence about what they have become has been partly removed.
The movement from evidence to classification compounds the problem. Each of the review’s headings directs interpretation. What does it mean to call a hospital haunted? What distinguishes a more-than-human hospital from a hospital not so described, given that every hospital houses microbes, machines and medicines? Which empirical differences become visible through these categories, and which disappear? I do not suggest that the categories are empty. Haunting can name something real about how unresolved deaths and colonial pasts enter clinical life, and attention to microbes and devices has produced excellent work. The difficulty is that the categories increasingly decide how findings are read without supplying criteria for deciding which interpretation a given finding warrants.
A simple diagnostic makes this visible. Replace “hospital” with “university,” “school” or “prison” in the review’s theoretical sentences. Universities, too, “crystallize the reproduction of the unequal value of lives” while being “potential sites of the otherwise” (Kehr and Chabrol 2026: 219). Schools have colonial histories, prisons are haunted, all three depend on nonhuman infrastructures, all contain forms of violence, and all can be imagined otherwise. The substitution proves nothing false. It reveals how little of the synthesis is specific to hospitals, and it poses the challenge that runs through the rest of this article: what does the theory explain about hospitals that it would not explain equally well about almost any other institution?
III. The Hospital as an Assumed Institutional Object
Kehr and Chabrol repeatedly acknowledge that hospitals exceed their walls. Following Solomon’s (2022) account of the traffic through a Mumbai trauma ward, they write that “the contours of hospitals become blurred to the wider world,” and they conclude that “what and where a hospital is and whom and what it moves, transforms, or disciplines can thus not be taken for granted” (Kehr and Chabrol 2026: 226). Yet this recognition does not overturn the analytical premise that the hospital is the object whose wider connections are to be traced. The hospital remains the grammatical and theoretical subject, which is haunted, entangled, more-than-human and finally summoned to become otherwise.
Living Value Theory reverses the order of explanation. The primary phenomenon is living coordination: people falling ill, kin recognizing that something is wrong, journeys made or abandoned, clinical skills enacted, drugs acting on bodies, oxygen and electricity delivered, patients returning to households that must now carry their convalescence. None of these processes is initially contained within an institution. A hospital is the L4 stabilization through which a particular configuration of them becomes recognizable as an enduring form and, in the case of a named organization such as St Bartholomew’s, as a particular identity capable of bearing responsibilities, budgets and a history (Ecks 2026e). This distinction allows the hospital to be taken seriously as a consequential achievement without treating it as a self-contained entity.
Consider an older man in a Scottish town who becomes breathless over several days in January. His wife notices that he is slower on the stairs, a general practitioner visits and measures his oxygen saturation, an ambulance takes him across icy roads, he receives oxygen and antibiotics on a medical ward, and he returns home five days later to a daughter who has come to stay. Where does the hospital’s contribution begin and end? Its legal, clinical, spatial, financial and administrative boundaries need not coincide, and none of them coincides with the processes that constitute his recovery. The hospital’s formal identity makes certain distinctions actionable, such as admission, discharge, clinical responsibility and staffing, while those same distinctions can obscure how his outcome depended on coordinations reaching far beyond the institution.
Here lies my ontological disagreement with the review. Kehr and Chabrol begin with the bounded institution, discover that its boundaries do not hold, and describe the resulting connections through concepts such as entanglement. I begin with living coordination and ask how institutional boundaries come to be stabilized within it, what they make possible and what they conceal. The difference matters because an analysis that starts from the institution must spend its energy reconnecting what its own starting point has separated.
The category “hospital” also hides a historical problem. A medieval foundation for the sick poor, a nineteenth-century sanatorium, a modern intensive care unit and a pandemic facility in an exhibition hall do not form one comparable series. The institutional designation stabilizes their apparent sameness, while historical inquiry exposes their differences. The difficulty is therefore deeper than heterogeneity. A relatively stable designation can conceal extraordinary changes in the processes to which it refers, and an analysis that takes the designation as its object inherits that concealment.
This returns us to the substitution test. Kehr and Chabrol’s concepts travel easily between institutions because they do not require reconstruction of the coordinations that distinguish one institutional form from another. An account grounded in living coordination is more demanding. It must establish why a specific institution emerged, which vital stakes it helps living beings pursue, which coordinations sustain it, and what makes it viable or dispensable. A general theory of institutions ought to make the differences between institutions more intelligible rather than dissolving them into a shared vocabulary.
IV. The Missing History of Actual Transformations
History enters the review mainly as inheritance. The section on haunted infrastructures attends to how colonial violence, unresolved deaths, spirits and injustices continue to shape hospitals. “Anthropologists have grasped the different afterlives of this technopolitical modernity through the notion of ‘hospital hauntings,’” the authors write, “affirming that haunting is historical and that the past leaves traces” (Kehr and Chabrol 2026: 223). Some of this scholarship is empirically important. The emphasis nevertheless produces a particular picture of institutional time, in which the past predominantly enters as something that survives, returns or remains unresolved. A hospital’s past also consists of institutions that no longer exist, treatments that were abandoned, functions that migrated elsewhere and forms of coordination that became unnecessary. Haunting is one historical process among several and cannot serve as a general account of institutional temporality.
The tuberculosis sanatorium is the clearest counterexample. Sanatoria assembled a distinctive configuration: separation on rural or upland sites chosen for air and light, prolonged rest alternating with graduated exercise, a rich diet, open verandas, closely supervised routines and residence lasting months or years. They were historically specific responses to particular therapeutic and epidemiological possibilities, and their varied local forms served purposes of segregation and moral training as well as treatment (Bryder 1988). Effective chemotherapy changed those possibilities. In Madras, the Tuberculosis Chemotherapy Centre randomly allocated patients to treatment at home or in a sanatorium and found that home treatment with chemotherapy, supported by a domiciliary service that supervised the taking of drugs, achieved results close to those of the sanatorium, even though home-treated patients had poorer diets, more crowded housing and less rest (Tuberculosis Chemotherapy Centre, Madras 1959). The trial tested directly whether a therapeutic stake still depended on a particular institutional configuration, and found that it no longer did to the same degree. Drugs alone did not close sanatoria, since the timing and manner of their decline depended on finance, staffing and policy. The question that matters is what altered the viability of that configuration, and the answer is historical and mesocosmic rather than ideological.
Psychiatric institutions show transformation without any simple emancipation narrative. The county asylum, which Goffman (1961) made the paradigm of the total institution, was reorganized and in Britain largely closed after the Hospital Plan of 1962 (Ministry of Health 1962). Some forms of confinement were reduced, other forms of compulsory treatment and professional authority persisted, some patients gained better arrangements and others fell into serious gaps in care (Scull 1977; Barham 1992). What changed was the organization of particular therapeutic and custodial coordinations.
The most striking source for this history in South Asia is Saiba Varma’s own article “Disappearing the Asylum” (Varma 2016). It examines how the government psychiatric hospital in Srinagar was selected to become a national “Centre of Excellence,” one of eleven among thirty-seven public mental hospitals across India, as part of reforms promoted by the WHO and the Movement for Global Mental Health (Varma 2016; 2020: 89). The reforms aimed to convert custodial institutions into research and training centres, to shorten stays, to expand outpatient care and to move treatment into the community. Varma shows that this modernization depended on keeping long-term, chronically ill patients in custodial conditions with diminishing resources, and that the asylum was not abolished so much as split in space, time and knowledge. The book supplies the mechanism. Psychiatrists now spent most of their time in the outpatient department and in a short-stay “family ward” with a horizon of thirty days, and rarely visited the long-term wards unless there was an emergency (Varma 2020: 125). Shorter stays were achieved in part through electroconvulsive therapy, often administered without anaesthesia, which “freed up bed space, and enabled shorter hospitalizations,” while every effort was made to discharge patients quickly, “even if family members and patients desired more institutional care” (Varma 2020: 122–23). “Shock accelerated institutional treatment,” Varma concludes (2020: 123).
This is exactly the kind of investigation into actual institutional transformation that a review titled “Hospitals Otherwise” ought to have placed at its centre. A hospital became otherwise; the transformation had identifiable causes in global policy, national funding and professional aspiration; it was realized through specific material and therapeutic means; and its burdens fell unevenly on long-stay patients and on families. Kehr and Chabrol cite Varma’s book and her co-authored work on hospital hauntings, but not this article. The contrast is revealing. The same long-stay patients appear in Varley and Varma’s (2018) essay as figures who “haunt psychiatric modernization projects,” and it is in this register that the review takes them up. Selecting different work by the same anthropologist changes what becomes visible about hospitals: in one register the left-behind patients are spectral residues, in the other they are the predictable consequence of a reform whose mechanism can be reconstructed.
Emergency hospitals complete the picture. At the University of Pennsylvania in October 1918, two fraternity houses became an influenza hospital within days and were returned to use as naval barracks less than three weeks later (University of Pennsylvania Archives 2018). The NHS Nightingale Hospital London was built in nine days in 2020 against an anticipated catastrophe in critical care. It admitted fifty-four patients with outcomes comparable to established intensive care units, and was then placed on standby because lockdown and the expansion of capacity elsewhere meant the feared demand never arrived (Collins et al. 2021). The significant question about the Nightingale is how anticipated need, existing clinical capacity, materials, staffing and public authority made such rapid assembly possible, and how a future that did not happen governed its entire existence. Whether it was a success or a failure in some general sense matters far less.
Tuberculosis sanatoria, psychiatric hospitals and pandemic facilities followed radically different trajectories. No single political aspiration or general theory of institutional transformation explains them. Each requires reconstruction of how particular coordinations became possible, compelling, sustainable or dispensable. The question is never simply whether hospitals can change, but why particular hospital forms changed while others persisted.
V. Differential Temporalities: Why Hospitals Change
There is no institutional clock. The coordinations a hospital stabilizes unfold at very different speeds: a patient’s oxygen saturation changes within minutes, a nurse’s clinical judgement develops over years, a protocol may be revised in months, a building endures for generations and a name for centuries. The variation runs within mediations as well as between them. Materials can be temporary, as the beds in a converted fraternity house were, or nearly permanent; symbolic classifications can be revised overnight or persist for centuries; embodied practices can survive the institutions in which they were formed. No mediation has a characteristic speed, and institutions share no single temporality because the coordinations they stabilize share none (Ecks 2026d).
Observing that things change at different speeds explains little by itself. The explanatory work lies in specifying how a development in one coordination alters the viability of another. In tuberculosis care, chemotherapy altered the rationale for prolonged residence, but buildings, nursing establishments, funding arrangements and public health administrations did not change in step. Treatment capabilities changed before estates were repurposed; allocations of beds and money persisted after their therapeutic rationale had weakened; and the practices of supervised treatment and case notification outlived the institutions in which they were formed. Sanatoria could therefore close long after the reason for their existence had eroded. In Srinagar, the relation ran the other way. A national grant and a global policy model arrived quickly, while the therapeutic means to realize short stays humanely, above all anaesthetists who could make ECT safe and legal, did not. The reform was implemented by recombining old means with new aims, and the long-stay wards remained as a slower, older layer beneath the new identity (Varma 2016; 2020: 122–25).
Conversely, emergency hospitals could be assembled rapidly because they recombined capacities accumulated over much longer histories. The Penn wards could open within days because physicians, interns and medical students had been trained for years before the epidemic arrived; the Nightingale depended on a national health service, intensive care medicine, ventilator manufacturing, an oxygen supply industry and military logistics. A new institution did not require its capacities to be newly created. The temporality of assembly differs from the temporalities of what is assembled.
Rate of change must also be distinguished from reversibility. A hospital fitted into an exhibition hall can be dismantled as quickly as it was built, because its materials are temporary and its staff can return to the hospitals they came from. A purpose-built sanatorium on a remote hillside could lose its rationale within a decade while remaining costly to maintain and hard to repurpose. Fast change can be easily reversible, and slow accumulation can make even a rapid loss of rationale difficult to undo.
These distinctions matter against explanations that privilege political discourse. Hospitals persist for many reasons. Some arrangements continue to serve compelling vital stakes; others persist because of sunk investments, professional careers, legal responsibilities, geography or the absence of alternatives. Some transformations are blocked, others delayed, others astonishingly fast. Ideology can be central in particular cases, as the politics of austerity documented in the review’s own section on political and economic terrains shows, but it cannot be the universal explanatory principle.
Stability and transformation are also simultaneous. A hospital can keep its name, site and legal identity while almost every aspect of its clinical practice changes, as Srinagar’s hospital did when it became a Centre of Excellence while remaining, for its patients, the p?gal kh?n? where people were once chained to their beds (Varma 2020: 4). It can lose a function while retaining staff, buildings and budgets. What survives a closure is an empirical question: a sanatorium can disappear while its building becomes a rehabilitation centre, an asylum can close while inpatient psychiatry continues elsewhere, and an emergency hospital can be dismantled while its expertise returns to other wards. Institutional change is the ongoing reconfiguration of living coordinations whose temporalities differ, and never the movement of a stable object from one condition to another.
The practical consequence is that historical explanation must precede political prescription. Anyone who wants to change a hospital must understand the temporalities and constraints of the coordinations to be altered. A symbolic reform may be announced quickly and enacted slowly or not at all; a therapeutic innovation may transform care without any new institutional ideology; a building may remain in use long after its medical purpose has disappeared. An account of differential temporalities explains both rapid change and inertia, and for that reason it offers a sounder basis for reform than the assertion that alternatives are possible.
VI. Care, Harm and Violence: The Failure of Entanglement
Kehr and Chabrol repeatedly describe hospitals as places where care and violence are entangled. Their abstract states that hospitals “simultaneously reproduce and aim to alleviate unequal entitlements to life,” and the review returns throughout to “care and harm,” “healing and wounding,” and “liberation and violence” occurring together. That care and violence can coexist is not in question. Coexistence, however, does not establish the nature of their relationship, and a hospital anthropology needs at least seven distinct relations to be kept apart.
The first is iatrogenic harm, in which medical intervention itself injures. The second is institutional coercion, in which patients are detained, restrained or treated without consent. The third is the treatment of harm inflicted elsewhere, when hospitals receive the wounded and traumatized of violence occurring beyond their walls. The fourth is the political restriction of clinical attention, in which medicine treats suffering while excluding its political causes from its remit. The fifth is organizational or economic harm, in which shortages, corruption or inaccessible services cause preventable suffering. The sixth is care deployed as an instrument of rule, in which states or armed forces provide treatment in order to win loyalty or gather intelligence. The seventh is care that resists violence, as when clinicians shelter, document or treat in defiance of the forces that harm their patients. These relations differ in cause, direction, agency and consequence. One institution may display several at once, and that does not make them the same phenomenon.
The word “entanglement” does little analytical work unless the processes being connected are specified. Does violence produce a need for care? Does treatment itself become violent? Does political authority constrain treatment? Does clinical classification obscure the political causes of suffering? Does medical practice oppose an existing violence? These questions cannot be answered by declaring care and violence entangled. The concept allows a relationship to be asserted precisely where its character most needs explaining.
The concept also lacks a negative case. What would count as a hospital that is not entangled with politics, inequality, violence, history or nonhuman processes? If every hospital is entangled by definition, finding entanglement in a particular hospital discovers nothing. A concept whose presence is guaranteed in advance loses its power to discriminate, and a useful theoretical concept must clarify which differences matter.
The cost appears when unlike evidence is placed under one heading. An unnecessary treatment that causes injury demands investigation of clinical decision-making, bodily consequences and institutional responsibility. A psychiatric clinic treating distress produced by militarization raises different questions, since it may relieve suffering while confining clinical attention to symptoms. Neither case establishes the other. When both are presented as evidence of the entanglement of care and violence, the theoretically dramatic conclusion is purchased at the expense of empirical precision. The review’s own paragraph on war shows how quickly this happens. It moves from care and violence becoming “disturbingly close” in Kashmir’s psychiatric clinic, to the “blurring of lines between violence and medical care” in the targeted destruction of Iraqi hospitals, to the “systematic destruction of hospitals” in Gaza (Kehr and Chabrol 2026: 222). A clinic in which militarism shapes the conditions of treatment, a health system deliberately dismantled by war, and hospitals destroyed as targets are three different relations between medicine and violence, and the paragraph presents them as degrees of one.
What Living Value Theory requires instead is an investigation of what is being coordinated, which vital stakes are pursued and by whom, how the different mediations participate, and what kinds of harm or benefit result for whom. Its value does not lie in the capacity to describe everything as coordination. It lies in using that ontology to reconstruct specific processes with distinguishable consequences. Complexity does not justify vagueness, and the more heterogeneous an institution is, the more analytical differentiation it demands.
VII. Kashmir: The Decisive Test of Empirical Specificity
Kehr and Chabrol introduce Varma’s work in a single sentence: “In the Kashmiri region occupied by India, in Pakistan, humanitarianism and militarism are entangled in the face of massive trauma and an epidemic of mental illness, creating disorienting worlds for doctors and patients alike in the psychiatric clinic, as Saiba Varma has shown in her moving ethnography” (Kehr and Chabrol 2026: 222). The geographical slip is the least of the problems, though it is telling. Varma’s fieldwork took place in Srinagar, in the Indian-controlled Kashmir Valley, and her book opens with a map distinguishing the territories controlled by India, Pakistan and China. I do not want to exaggerate the review’s claim. It does not say that the psychiatric clinic commits the violence from which its patients suffer. The problem is that the sentence, and the argument it serves, give the reader no way of knowing which relation between care and violence Varma’s evidence establishes.
Varma’s book itself is far more differentiated. It does advance a strong thesis: Varma writes that the boundaries she had “unconsciously drawn between medicine and militarism dissolved,” that “medicine was not just a remedy for violence but part of its repertoire,” and that she had to tell “a story of violence through medicine” (Varma 2020: 8–9). She also uses the word “entanglements.” But she immediately distinguishes three forms of care: militarized care, meaning practices of the Indian state, military or police that further an imperial project in the language of protection; humanitarian, NGO or clinical care, which aims at neutrality but becomes “distorted by violence and militarism”; and everyday care, the noninstitutional hospitality, feeding and attendance of kin (2020: 12). She then separates three registers in which militarism and care are related: rhetorics of love that justify military presence, the use of medical and psychological techniques as instruments of counterinsurgency, and the unintended seepage of militarism into spaces meant to be neutral (2020: 13–14).
Read with these distinctions, the book documents most of the seven relations I set out above, each in its own setting. Treatment of harm inflicted elsewhere fills the outpatient department, where a woman who has lost three sons reports years of dag, restless pain, and leaves with another prescription for a generic antidepressant (2020: 4). The political restriction of clinical attention appears in the psychiatrist’s translation of that pain into “somatic features ... characteristic of trauma victims,” made without time or inclination to explain it to the patient (2020: 5), and in NGOs that encourage Kashmiris to see themselves as patients and victims rather than political subjects (2020: 26). Care as an instrument of rule appears in the police-run De-Addiction Centre, housed within the Police Control Room, which senior officers established to restore the image of a force associated with brutal counterinsurgency, and whose patients feared being recruited as informers (2020: 54–55). Iatrogenic harm and coercion appear in the administration of unmodified ECT, which psychiatrists defended as “some ECT is better than no ECT” after national legislation had moved against it, and in quick discharges made against the wishes of patients and kin (2020: 122–25). Organizational harm appears in curfews and strikes that disrupted the flow of drugs and staff, and in a culture of impunity that bred corruption (2020: 8–9). Care that resists violence appears in the doctors of the early 1990s who supported the armed movement and sheltered militants in hospitals (2020: 76), and in the everyday hospitality with which the book ends. A psychiatric clinic treating the victims of a military occupation, a police clinic dispensing treatment as counterinsurgency, and a hospital in which an effective treatment is administered in a form that patients associate with torture are three different situations with different causes and different remedies.
The ECT case shows why the differences matter. Patients’ fear of karant was no superstition. Electric shock had been used extensively in interrogation, and for patients such as those Varma describes, the presence of shock in both clinic and interrogation centre undermined any clear demarcation between them (2020: 115–21). Yet the clinic’s reasons for using ECT came from elsewhere: soaring demand, chronic shortages, the absence of anaesthetists and the policy goal of short stays. To call this an entanglement of care and violence is accurate and uninformative. To reconstruct it is to see a therapeutic technology whose material form, given without anaesthesia, coincided with an instrument of torture, deployed by clinicians pursuing a globally endorsed reform under conditions of scarcity, and encountered by patients whose understanding of shock had been formed by militarized violence. The harm here has at least three sources, and each would require a different intervention.
The possibility that the review’s synthesis passes over is that of effective care under political constraint. Psychiatrists in Srinagar could relieve real suffering while lacking any authority to challenge the conditions that produced it, and they could at the same time reproduce troubling classifications and practices. These possibilities must be investigated separately. A form of medicine is partly constituted by the domains of suffering it recognizes as actionable and by those it excludes from its remit (Ecks 2026c), and the politically restricted ontology of the Srinagar clinic is consequential without turning every therapeutic encounter into an instance of violence.
The general lesson can be put plainly. An anthropology that cannot adequately distinguish a hospital treating victims of violence from a hospital perpetrating violence has surrendered a fundamental part of its capacity to explain what happens in hospitals. Kehr and Chabrol do not deny this distinction anywhere. Their synthesis simply does too little to preserve it, and in citing Varma it discards the very distinctions her ethnography was built to draw.
VIII. The Conditions of Ethnographic Knowledge: The Varma Controversy
The second use of the Kashmir case concerns a different question, and the two must not be conflated. The first asks what Varma’s evidence shows; the second asks under what conditions that evidence was produced.
On 14 September 2021, an anonymous account on Twitter stated that Varma’s father had served in India’s external intelligence service, the Research and Analysis Wing, and had worked in Kashmir in the 1990s (Tahir 2021). A statement by sixteen Kashmir scholars followed on 21 September. Its signatories rejected the idea that a daughter should be punished for her father’s actions, but said that they had not been told of the connection despite years of professional relationships with her, and voiced concern about a possible breach of ethics towards the vulnerable patients among whom she worked (Kashmir Scholars 2021). On 30 September, fifty-three scholars of South Asia signed a further statement arguing that informed consent should have included disclosure of her family’s ties to the security apparatus, especially for psychiatric patients in Kashmir, and calling for “transparency, accountability, and oversight” (Statement from Academics 2021). Varma responded on 1 October. She wrote that her father had worked in Kashmir in the early 1990s, when she was ten, and had held no posts relating to Kashmir since; that he never enabled connections or contacts in the field, beyond arranging a car to meet her at Srinagar airport early in her fieldwork, and never received her data; that she had told interlocutors that her father worked for the central government and, when asked, named the Cabinet Secretariat, through which they “recognized that my father worked for intelligence agencies”; and she apologized to colleagues for not telling them earlier, while standing behind her scholarship (The Wire 2021). The commentator C. M. Naim later criticized the statement as ad hominem and argued that critics should have shown how the omission affected her analysis (The Wire 2021).
Naim’s challenge is the right one, and it can be met without any speculation about motives. The methodological issue is more specific than positionality in general. Patients and clinicians who suspected or knew something of the researcher’s family might have interpreted her differently, and this could shape what they disclosed, withheld, avoided or emphasized. Varma’s own response confirms that this happened. Some interlocutors stopped speaking with her and others “became more guarded”; some chose not to associate with her; others spoke because they saw her as a messenger who could carry their critiques to the Indian state and to Indian and American audiences, or because her background made their critique “more powerful.” She also explained that her family background informed the design of the research: she avoided eliciting trauma narratives, relied on secondary sources for accounts of state violence, worked mainly in outpatient departments and did not sit in on one-on-one therapy (The Wire 2021). The book itself records the general condition. “As an Indian researcher,” Varma writes, it was difficult to access internal critiques of the Kashmiri movement (2020: 52); and in the most unsettling sentence of her chapter on disturbance she reports that she “also worried about the virtual map circulating about me: Who was I spying for? Did people trust me?” (2020: 87).
Several of these choices were protective, and I do not suggest that a researcher who declines to elicit trauma narratives from psychiatric patients has done anything but exercise care. The point concerns the evidence. An ethnographic encounter is interrecursive: interlocutors orient to what they take the researcher to be, and that orientation shapes what can be said. A retrospective explanation, however sincere, cannot reconstruct how each participant understood the researcher’s position, how those understandings changed over months of fieldwork, who declined to take part, or which topics were quietly avoided. No general pattern can be inferred from the available information. What can be said is that conclusions about care, trauma and state violence were drawn from a field whose conditions of disclosure were themselves shaped by the relationship at issue, and that the book’s account of violence rests in part on sources chosen at one remove because of it. The conditions under which something could be said are part of the evidence needed to interpret what was said. This is the central claim of Ontologically Grounded Theory, which treats prevented conversations, withdrawn interlocutors and unasked questions as consequential nonoccurrences rather than as gaps to be passed over (Ecks 2026b).
Kehr and Chabrol cite Varma’s work three times, in each case approvingly, and nowhere mention the controversy. That silence does not show that they were unaware of it or that they concealed it. It is consequential nonetheless, because their review repeatedly invokes colonial power, epistemic injustice, ethical responsibility and the political conditions of medical knowledge, and closes by calling solidarity with people’s political struggles an “ethnographic responsibility” (2026: 227). Their use of Varma offers no corresponding examination of how the conditions of ethnographic knowledge may themselves have been shaped by the structures of political power under discussion. The political conditions of medical care are invoked, while the political conditions of the ethnographic knowledge being cited remain unexamined.
IX. Epistemic Justice, Vital Stakes and Political Imagination
The review ends by asking what hospitals oriented toward care, healing, dignity and health justice might look like. These aspirations are not objectionable, and Living Value Theory is not indifferent to justice. The difficulty is that the desirability of an objective explains neither why a particular institutional transformation would occur nor how it could be realized. The review moves too easily from an account of what hospitals are to an account of what they ought to become.
The vital stakes of the people who coordinate hospital life are more various than the analyst’s priorities. A patient may be struggling for relief from pain, a nurse to prevent deterioration overnight, a surgeon to finish an operation safely, a family to secure a bed, a cleaner to keep a ward safe under precarious employment. These activities cannot be understood by assuming in advance that epistemic justice is their organizing objective. Epistemic recognition can certainly be a pressing stake: a patient who is not believed, or a clinician who challenges the exclusion of patients’ knowledge, may pursue it with great urgency. The Srinagar clinic provides its own example in the young man who politely refused ECT and walked out with visible relief, overturning for a moment the expectation that patients obey (Varma 2020: 115). But whether and where such stakes are pursued is something to be investigated rather than assumed.
A proposed reform changes a hospital only insofar as it changes living coordination. That may require new clinical practices, shifts in professional authority, new infrastructure, different financing, new forms of patient participation or responsibilities reaching well beyond the hospital. Declaring that a hospital will become emancipatory or just does not realize those aims. Srinagar shows the gap. A reform designed to end custodial care and humanize treatment was announced, funded and partly implemented, while its realization depended on unmodified ECT and left the long-stay wards largely unvisited. Political visions can become vital stakes in their own right; they can mobilize people and establish new practices, as Fanon’s reforms and the Black Panther clinics did. That is an empirical process to be explained, and no vision is guaranteed realization by its moral attractiveness.
There is a final risk that the review does not confront. A theory committed to epistemic justice can itself substitute its own categories for the concerns of the people it studies, and in doing so reproduce the epistemic injustice it opposes. This does not happen in every instance of critical hospital anthropology. But the responsibility of the anthropologist extends beyond identifying morally desirable futures to understanding the vital stakes already being pursued, including those suppressed, rendered inarticulable or blocked by existing arrangements. The ambition to improve hospitals must never become a reason for failing to understand them.
X. Conclusion: Explaining Hospitals Rather Than Imagining Them
Kehr and Chabrol are right that hospitals can be otherwise, but that is not the discovery their argument needs. Hospitals have already become otherwise across radically different histories. Some transformations arose from new treatments, some from epidemics, others from changes in finance, political struggle, professional reorganization or the redistribution of care into households. Sometimes these processes reinforced one another, and sometimes they obstructed one another. No general vocabulary of haunting, entanglement or emancipation can substitute for reconstructing them.
Rather than a fixed institution awaiting the political imagination that will transform it, the hospital is an L4 stabilization within continuously changing living coordination. Its identity may persist while its practices change beyond recognition; its buildings may outlast its therapeutic rationale; its essential functions may migrate to clinics and homes; and a new hospital can be assembled in days by recombining capacities accumulated over generations. These processes become intelligible through the differential temporalities of the mediations that hospitals coordinate, and through the vital stakes of those who depend on them. The same discipline is required when studying harm. A hospital inflicting violence, a hospital treating the victims of violence, a hospital operating under a politically restricted definition of illness, and a clinic run by the police to win hearts and minds are doing different things, and Varma’s ethnography, read on its own terms, shows how each can be told apart. The conditions under which such an ethnography was produced belong to its evidence, and an anthropology of political conditions cannot exempt its own sources from that scrutiny.
The task of hospital anthropology is not to demonstrate that hospitals could be different, because their histories already establish that. The task is to explain what particular hospitals do, which forms of living coordination sustain them, why they change, and why some possibilities become viable while others remain unrealized. An anthropology that substitutes theoretical possibility for historical explanation, and conceptual entanglement for empirical differentiation, risks losing sight of the very hospitals it seeks to transform.
References
Barham, P. 1992. Closing the Asylum: The Mental Patient in Modern Society. London: Penguin.
Bryder, L. 1988. Below the Magic Mountain: A Social History of Tuberculosis in Twentieth-Century Britain. Oxford: Clarendon Press.
Collins, G.B., Ahluwalia, N., Arrol, L., Forrest, N., McGlennan, A., O’Brien, B., Proudfoot, A., Trainer, M., Schilling, R., Sullivan, E., Westwood, M., Wragg, A. and Knight, C. 2021. ‘Lessons in Cognitive Unloading, Skills Mixing, Flattened Hierarchy and Organisational Agility from the Nightingale Hospital London during the First Wave of the SARS-CoV-2 Pandemic.’ BMJ Open Quality 10(3): e001415.
Goffman, E. 1961. Asylums: Essays on the Social Situation of Mental Patients and Other Inmates. Garden City: Anchor Books.
Kashmir Scholars. 2021. ‘Statement on Anthropologist Saiba Varma.’ Medium, 21 September.
Kehr, J. and Chabrol, F. 2026. ‘Hospitals Otherwise.’ Annual Review of Anthropology 55: 217–233.
Ministry of Health. 1962. A Hospital Plan for England and Wales. Cmnd 1604. London: HMSO.
Scull, A. 1977. Decarceration: Community Treatment and the Deviant: A Radical View. Englewood Cliffs: Prentice-Hall.
Solomon, H. 2022. Lifelines: The Traffic of Trauma. Durham: Duke University Press.
Statement from Academics. 2021. ‘Statement from Academics on Anthropologist Saiba Varma’s Research.’ 30 September. Reprinted in Raiot, 3 October.
Tahir, M. 2021. ‘Behind the Books.’ Kashmir Observer, 22 September.
The Wire. 2021. ‘Does Nondisclosure of Familial Proximity to “Security State” Compromise Research on Kashmir?’ The Wire, 3 October (updated 10 January 2022). Includes the Statement from Academics, Saiba Varma’s response of 1 October 2021, and a comment by C. M. Naim.
Tuberculosis Chemotherapy Centre, Madras. 1959. ‘A Concurrent Comparison of Home and Sanatorium Treatment of Pulmonary Tuberculosis in South India.’ Bulletin of the World Health Organization 21(1): 51–144.
University of Pennsylvania Archives. 2018. ‘Penn and the 1918 Influenza Epidemic.’ University Archives and Records Center, archives.upenn.edu.
van der Geest, S. and Finkler, K. 2004. ‘Hospital Ethnography: Introduction.’ Social Science & Medicine 59(10): 1995–2001.
Varley, E. and Varma, S. 2018. ‘Spectral Ties: Hospital Hauntings Across the Line of Control.’ Medical Anthropology 37(8): 630–644.
Varma, S. 2016. ‘Disappearing the Asylum: Modernizing Psychiatry and Generating Manpower in India.’ Transcultural Psychiatry 53(6): 783–803.
Varma, S. 2020. The Occupied Clinic: Militarism and Care in Kashmir. Durham: Duke University Press.