Felix Kaplan
School of Social Work, Zefat Academic College, Safed, Israel.
DOI: 10.4236/sm.2026.163009   PDF    HTML   XML   3 Downloads   54 Views   Citations

Abstract

Necrography is a writing practice for sites where welfare, psychiatry, and recovery-oriented care no longer have the language to describe the lives encountered there. Drawing on two decades of field presence across four sites in Israel—an abandoned drug-and-sex courtyard in South Tel Aviv, an open-air drug scene in the same city, a semi-abandoned lakeshore hotel in Tiberias, and a peripheral dementia ward in the Galilee—I argue that under conditions of structural abandonment, the qualitative instruments I was trained to rely on, including interviews, consent procedures, and member-checking, failed at the threshold of encounter. Necrography is a response to these conditions. It operates through three linked practices: composite autoethnographic fiction, attention to instrument-refusal as data, and, when prolonged exposure narrowed my capacity to write, the use of VERSO—a configured language-model aid set to return short, non-consoling sentences from the researcher’s accumulated field materials. I propose necrography as a procedural method for documenting populations that remain socially present yet institutionally illegible. Its central criterion of rigor is composite fidelity—the recognizability of a pattern across structurally similar sites. The method is situated within the sociology of institutional abandonment and the broader post-qualitative turn in ethnographic research.

Keywords

Autoethnographic Fiction, Composite Fidelity, Experimental Ethnography, Instrument-Refusal, Necrography, Post-Qualitative Inquiry, Structural Abandonment

Share and Cite:

1. Introduction

This article has accumulated the way limescale builds up inside a kettle that nobody bothers cleaning. The kettle in question is a body, mine, which, in its late twenties, walked into a building it had no professional business entering and spent the better part of the years that followed trying to figure out what to do with what it saw there. The question addressed here is narrowly methodological: how does a qualitative researcher write faithfully about people whom the welfare state has structurally erased from its own records—when interviews are no longer possible, and consent procedures fail at the perimeter?

I went to Fein 1 in South Tel Aviv alone, without an agency or a review board. After graduating magna cum laude with a bachelor’s degree in Social Work, I felt as though I had lost something authentic within myself. I did not know how to stay in a room with people who had lost more than I had. I walked to Fein 1 because I thought I would find something there that would tell me what to do with the parts of myself that the degree had not touched. More than a decade and a half later, I am still standing at that door, writing.

This fascination predates my professional career. I owe the reader this admission because my methodological argument depends on the original gesture not being clean. There is a tradition in social work writing in which the author begins with a vocational call, a relative who suffered an injury witnessed in adolescence, and proceeds from this clean source to a clean-service trajectory. I have no clean origin. I had a dark fantasy of the underworld, which I do not mean in an erotic sense but in an epistemological one. I now understand that I was drawn, before I had the language to justify it, to sites where institutional and moral visibility had already collapsed.

I had read about this building and seen pictures of it before I went there. I knew that the H-shaped two-story block at the intersection of Fein, Salomon, and Erlanger contained about 70 rooms rented daily for cash. Inside, envelopes were dispensed through covered windows that residents called kaspomatim¹ (holes-in-the-wall), automated tellers of the bottomless account. Journalists widely considered this the lowest place in the city. I wanted to see the lowest place in the city. There is no better formulation for this desire. What the fascination eventually revealed, among other things, was that the lowest place in the city was not lower than the rest of the city in any moral sense. It is lower only in the sense that the rest of the city has agreed to pretend that it cannot see what is happening there. The earliest fascination, when I look back, was a fascination with seeing itself. I am involved in writing not because writing is reflexive in the modern academic sense but because writing is structurally continuous with observation, and observation is not innocent.

The Fein 1 courtyard refused me the epistemology of clean hands. I walked in with a backpack, a notebook, and a naïve fantasy. The girl who would be called Almaz—a composite figure constructed from young Ethiopian women observed at Fein 1 and similar sites and who appears below—moved back through the second doorway before I could decide whether to raise my hand. That is the scene and the method: ethnographic inquiry in this register begins when the researcher admits that the contract under which they arrived is void. I refer to the researcher in this position as the Clerk. Unlike the traditional ethnographer who observes in order to extract actionable data, or the social worker who intervenes to mend, the Clerk merely registers the bureaucratic void. The Clerk acts as a witness to the perimeter where the contract of healing has already collapsed.

Institutional ethnographic inquiry in psychiatry, psychology, and social work has learned to produce testimony in registers that their host discipline can metabolize, such as resilience and recovery, but often fails to register the category of the person for whom neither register applies. The Drowned God is the name assigned to this category. The figure draws on Lurianic Kabbalah², in which the breaking of vessels at the moment of creation leaves some portion of the original light below the line of return (Fine, 2003; Scholem, 1995). The figure of the Drowned God is not produced by a single catastrophic event but by a recurrent cascade of institutional denials—a pattern I call serial refusal. Serial refusal is the mechanism; the Drowned God is what the mechanism produces.

Mbembe (2003, 2019) describes necropolitics as a mode of power that exposes entire populations to forms of existence in which people remain biologically alive while being stripped of political agency, dignity, and the conditions necessary for meaningful social and political life. In his terms, such conditions constitute death-worlds: spaces where biological survival persists, yet human existence is systematically reduced to social and political disposability. Necrography is a writing practice that renders visible the operations of necropolitics at the scale of the case file. Although the term “necrography” has appeared previously in other scholarly contexts—as a “postmodern necrography” mapping shifting geographies of disposal (Tremlett, 2007), as a “political necrography” tracing the zombie as a distinct biographical genre (Ram, 2014), as an anthropological concept describing how the dead create biographies among the living (Panagiotopoulos & Espírito Santo, 2019), and as a framework for addressing colonial violence in museums (Hicks, 2020)—the present article develops it in a different direction. Where existing necrographies concern the dead or their afterlives, the necrography I propose concerns the living: populations that remain socially present yet institutionally illegible, not because they are dead, but because the welfare state has structurally abandoned them. The term is adopted for its etymological resonance—from nekros (“dead”) and graphein (“to write”)—but is repurposed here to designate a procedural method for registering lives that institutions have rendered as though they were already dead to their own records.

Classic welfare neglect operates through the Foucauldian logic of biopolitics—the active “letting die” of those rendered institutionally expendable (Foucault, 1978, 2003). But serial refusal pushes such lives further, into the sphere of Mbembe’s necropolitics, where institutional workflow actively constructs localized death-worlds of social and clinical illegibility. This is not a sequential transition but a simultaneous operation: biopolitical neglect produces the conditions of necropolitical abandonment. Serial refusal is the mechanism that enforces this simultaneity: the same workflow that neglects also abandons, and each refusal deepens the ontological drowning. The water rises one refusal at a time.

This is where serial refusal and the Drowned God converge. Serial refusal is the operational mechanism: a recurrent cascade of institutional denials at the doors of hospitals, rehabilitation centers, welfare agencies, and housing facilities. The Drowned God is the ontological outcome of serial refusal: the person who, through this very cascade, has been pushed below the institutional line of return. While serial refusal describes the process of being turned away at each door, the Drowned God describes the state of the person for whom these doors no longer exist. The figure is best understood as light trapped below the surface, beyond the reach of any resilience framework from above.

There is no symmetry here. I am not proposing a “Saved God” or a “God of Recovery”. The structural imagination that names the Drowned God also requires its opposite, but the opposite is not a figure. It is a set of conditions: a person with a file, an address, a diagnosis that opens doors, and a language that the institution can read. These conditions are not heroic. They are luck embodied. They are the baseline that the welfare state was built to provide, and the Drowned God is what happens when that baseline collapses. I name the Drowned God not to elevate those who have fallen but to mark the structural boundary beyond which the ordinary categories of helping no longer apply. If there is a God who lives above the line, he does not appear as a figure on these pages. What appears instead is the institutional routine that continues to operate, where the instruments still work.

The borrowing earns its place in the following, narrow sense: while an interventionist vocabulary views a population of broken people as mended by the application of sufficiently calibrated care in the inner courtyard of Fein 1 and in the lobby of the Hotel of the Fallen, the borrowed Lurianic terminology views a population of husks (klipot) holding scattered sparks (nitzotzot). It sees a population in which the trapped element is not damaged but light, in which the work of the helper is not to repair but to liberate, and in which the helper cannot do the work from a position of professional safety because the sparks cannot be reached from above. The shattering of vessels (shevirat ha-kelim) is, in this adoption, a structural and not a theological event (Fine, 2003). The vessels in which institutional intervention attempts to contain the lives this article is about were never strong enough to receive the light. They cracked, and whatever fell with the breaking never recovered. Necrography gathers what the vessels could not hold.

This is the usual transition that any qualitative researcher makes when an instrument yields observations that do not align with the scene. Thus, the sixteenth-century vocabulary developed in Safed in response to the Spanish Expulsion of 1492 offers, by historical accident, a conceptually suggestive grammar for naming a condition that resilience discourse struggles to register. Lurianic grammar becomes methodologically useful where Mbembe’s (2019) necropolitics does not fully name what abandonment leaves behind—not damage to be repaired, but a pattern of light that the helper cannot retrieve from a safe distance.

My starting point is a recurring problem across four structurally abandoned sites: the interview, the consent form, and reflexive return all presuppose a contract of encounter that these sites structurally refuse. I present necrography as a post-qualitative writing practice with a limited set of procedures and one main criterion of rigor: composite fidelity. I demonstrate the method across four sites on the fringes of the Israeli welfare state, each selected because the instruments commonly used in related qualitative inquiry failed at the relevant thresholds. What follows specifies the scope and limits of the method. Necrography extends and does not replace recuperative qualitative inquiry and becomes necessary only when instrument-refusal has already shaped the site. What follows is a report from the perimeter—written from inside the failure of the instruments I was trained to trust. An overview of the necrographic circuit and the relationship among the four sites is presented in Figure 1.

2. Against the Contract of Resilience

This section identifies the institutional contracts under which direct testimony in therapeutic professions and qualitative research has been authorized and explains why each failed in the field sites discussed.

First and foremost, there is the clinical contract, which assumes a subject capable of articulating need, consenting to intervention, and participating in a future-oriented plan, all of which were structurally unavailable at the sites discussed.

The second is a reparative autoethnographic contract. Its stabilized form assumes that the writing subject will return from the field to a place from which the field can be made sense of, and that the return will take the shape of insight conveyed to the reader as a form of care for the subject. I owe much to this lineage (Ellis et al., 2011), but I leave it here because the fields in which I have written do not allow for a return. The Fein courtyard did not issue me a ticket.

The third position is the Strengths Perspective. Saleebey (1996) was not mistaken in his observations. The problem lies in the downstream institutional machinery through which welfare is administered. Over the past three decades, his framework has been absorbed into the neoliberal restructuring of welfare, in which the demand that service users narrate their resources has quietly replaced the demand that the state provide them with resources. In the hands of the intake officer, it becomes the technology by which the file is closed before it is truly opened: closed through a form that asks the woman in front of the desk to list her capabilities so that they can be entered into the column marked as protective factors, and the intervention can be preemptively foreclosed at the database level.

Figure 1. Necrographic circuit. The diagram arranges four field sites of abandonment along a vertical axis that marks a declining gradient of institutional addressability: a peripheral nursing home in the Upper Galilee (+900 m); Fein 1 and Shvil Akko in South Tel Aviv (street level); and the Hotel of the Fallen on the shore of the Sea of Galilee (–209 m). The sites correspond to demolition, cyclical eviction, slow ruination, and biological threshold. Black discs mark the points at which institutional archives fail to register life; dashed arcs indicate the non-interventionist perimeter. One hinge—VERSO: the prosthesis stationed at the threshold (the line of return), where the field record turns into testimony—the point at which witnessing continues only because the writing has been handed to the instrument.

In practice, resilience has become a governance technology: it requires the individual to perform a recovery that the institution has made structurally impossible through its own actions. In the grammar of the form, resilience is the name of something that the state no longer has to do, and necrography is the method of writing against that grammar. It does not ask for strengths; it registers what survives the grammar’s erasure.

Gray (2011) offers an internal critique of the strengths perspective; I argue that its absorption extends further into the neoliberal rationale of reducing the welfare state to a minimum. At the sites discussed here, a glossary of strengths and resilience can be absorbed into an administrative routine, thereby returning institutional failure to abandoned subjects. My critique sharpens at the intersections: the complex deployment of the strengths framework across axes of non-residential status, mental health, gender, and age produces a denial in which the stigma of each identity is used to explain the failure of intervention that the previous identity has already blocked. A parallel line of critique, developed outside social work in narrative studies, rejects the assumption that narrative coherence is the sole moral standard by which lives are made legible. When illness, trauma, and structural abandonment disrupt the conditions under which a coherent self-narrative can be produced, the demand for such a narrative can be interpreted as another form of exclusion (Hyvärinen et al., 2010). Necrography takes up this refusal and extends it into practice.

Fassin’s broader account of humanitarian reason and moral administration helps illuminate this point (Fassin, 2012). The argument I make about the case notes should be seen as an extension: in these sites, case notes act as narrative forms through which institutional inaction becomes administratively tolerable. The case note does not deceive through its content; it deceives through its form, which is precisely what allows the institution to continue living with its inaction.

I draw on Hartman’s critical fabulation (Hartman, 2008) to do so. Critical fabulation becomes necessary when the archive is built on a systematic refusal to register its subject as a person, and you refuse to let the archive have the final word. While critical fabulation reconstructs lives from archival traces, necrography operates where no adequate institutional archive remains, and the researcher must construct composite figures using repeated field patterns instead of fragmentary documents.

3. Methodology: Autoethnographic Fiction and the Necropolitical Field

This work adopts an experimental form of autoethnographic writing shaped by Richardson’s writing-as-method (Richardson, 2000), Hartman’s critical fabulation (Hartman, 2008), and performance autoethnography (Adams et al., 2021; Ellis et al., 2011). The genre shares with evocative autoethnography (Bochner & Ellis, 2016; Bochner, 2017) a commitment to narrative, scene, and emotional truth as operations of knowledge rather than ornaments, while departing from that tradition’s assumption of a corrective return—an assumption the following account names before refusing it. This refusal is necrography’s generative core—a reversal of the clinical gaze and the ethical foundation of testimonial immersion.

The article is informed by post-qualitative inquiry as articulated by St. Pierre (2021), particularly its refusal of pre-given methods and insistence that inquiry emerges through concepts, experimentation, and writing rather than through the application of a prior methodological protocol. Where post-qualitative inquiry refuses a pre-given method, necrography retains a minimal procedural apparatus for working under conditions of structural abandonment. The two positions are not contradictory, and the tension between them is explicitly addressed below. For ease of reference, definitions of key terms used throughout are provided in Appendix, Glossary of Key Terms.

Anderson’s analytic autoethnography is not what I am doing (Anderson, 2006). Analytic autoethnography requires complete member-researcher status, dialogue with informants beyond the researcher’s self, and a commitment to developing generalizable theoretical understandings. None of these conditions applies here. I am not a member of the communities described; the dialogic engagement this approach presupposes is refused at the site’s perimeter, and the article advances a bounded procedural argument, not a generalizable theoretical claim. Rather, what I am practicing is closer to the critical and creative traditions of Holman Jones (2005), Spry (2011), and Pelias (2004)—traditions that treat form and craft as methods of knowledge production and refuse to separate aesthetic from analytic decisions.

I did not use autoethnography as an instrument for personal disclosure. It was used because I am the instrument through which the field became writable, and instrument calibration is important for the report. I am the Clerk on these pages. The Clerk’s institutional standing differs by site: a licensed social worker in the nursing home, a trespasser in the Fein courtyard, and a displaced guest in Tiberias. However, the writing posture of sustained, non-interventionist witnessing remains constant across them. The people in the doorways, corridors, and plastic chairs stayed with me long after the encounters ended. Conventional ethnography treats individual encounters as its medium and protects identities through careful anonymization. By contrast, autoethnographic fiction takes a category of experience that has stabilized over years, sites, and relationships, rendering it on the page through a composite figure who did not exist as a single person but is faithful to the overall pattern. The “truth” of the form is thereby not personalization but composite fidelity: whether the composite registers a category of life that the host discipline has failed to register through its case-based instruments.

The Clerk is a composite but operates in a fundamentally different register from the other figures in this text. I switch to the third person, referring to myself as “the Clerk”, because sometimes the despair in these rooms is too loud to process directly. The Clerk is not an invented figure but a persona—a narrowed version of the researcher’s own professional self, stripped of the roles that could not enter the field; a survival mechanism instead of just a stylistic choice, a way to split myself so that I can keep functioning and observing when the social worker in me simply wants to weep or walk away. In this sense, the Clerk is a procedural reduction of the researcher’s actual self: the parts of me that continued to register the field after the parts trained to intervene had been set aside. This reduction is not a literary device but a methodological necessity—it allowed the writing to continue when the clinical tools had been refused. Thus, the Clerk compresses multiple moments of the researcher’s own field experience. I do not give the Clerk a biography, a childhood, or a voice that is not mine. What I give him is a narrowed range of action: he watches, records, and refrains. The Clerk is to the researcher what the composite figures are to the people described—fidelity lies not in accuracy to a single life but in recognizability across structurally similar positions.

This stance has predecessors—autoethnographers who wrote from positions that the discipline would have preferred to ignore. Pelias (2004) wrote from inside the laboring academic body, Tamas (2011) wrote back against the narrative of therapeutic recovery, Tullis (2013) pressed on the ethics of self-disclosure, and Le Roux (2017) refused to let rigor be surrendered in the name of expressivity. Le Roux (2017) also provides the methodological insight that the criteria of positivist replicability are inadequate for measuring the rigor of autoethnographic work, which must instead be evaluated against criteria internal to the form itself.

I advance composite fidelity as a criterion internal to the form. Composite fidelity refers to the extent to which practitioners and researchers working in comparable settings recognize the patterns represented in a composite account. Whereas Richardson (2000) emphasizes the capacity of qualitative texts to move readers, generate insight, and convey lived realities, composite fidelity is concerned with the credibility of a recurring social pattern. Its object is the pattern itself rather than any individual instance from which it was derived.

Because this method makes sense only when the ordinary repertoire has already failed, it is important to describe the entry criteria. A site was considered structurally abandoned when four conditions were met simultaneously. First, there was no file capable of mediating meaningful addressability, care, or return; therefore, the paper trail that would normally anchor a file had already been closed. Second, the person passed the last threshold of a service built to receive people in their condition and had aged, drifted, or been evicted from its reach because of their condition. Third, the standard encounter instruments could not be administered without endangering the person or producing documentation that the site was organized to refuse. Finally, conventional qualitative access, negotiated entry, and planned return were blocked by site conditions and not by failure in their arrangement. The operational version of this is what I call instrument-refusal. I would try to administer the routine instrument in good faith and see what comes back. If the interview could be conducted, the form signed, the return scheduled, the site did not meet the criteria, and a reconstructive investigation remained the correct tool. The site entered the scope of necrography when the instrument was returned as either empty or defective. These criteria are inherently strict. They intend to keep the method a limited addition to abandoned fields, and not a permanent authorization to withdraw consent, wherever it is inconvenient.

The field sites discussed in this article are real, whereas the Hotel of the Fallen is a composite site name used to obscure the identities of several adjacent real locations. A ramshackle building on Fein 1 hosted the “lowest” prostitution and drug scene in Tel Aviv. It was demolished in 2015 and later served as the central setting for the film “The House on Fin Street” (Manor, 2021), which depicts the drug and prostitution scene that once thrived in the adjacent old Central Bus Station. A narrow lane between garages and new residential towers, Shvil Akko is now the open drug scene that crowded out the Fein users after the demolition in 2015 and has been repeatedly evacuated and reopened since then. The Hotel of the Fallen is a composite site drawn from several abandoned and semi-abandoned hotels on the shores of the Sea of Galilee, encountered during the author’s displacement after October 7, 2023. The regional nursing home on the periphery of the Upper Galilee was, at the time, an actual institution where I worked as a social worker in the geriatric and dementia wards. It has since been closed and demolished.

However, ethnographic writing in Hebrew on the population addressed in this article remains sparse. The closest treatment of the four sites is Bonny-Noach and Toys (2019), based on approximately 400 hours of participant observation at the old Central Bus Station in Tel Aviv and 62 semi-structured interviews with regular and transient figures in the open drug scene adjacent to Fein 1. They described the site as “the lowest place in Israel”, a phrase that resonates with the present text as well. Their study performed an important moral and scholarly labor: it rendered a largely invisible population partially visible within the conventions of social ethnography. Their work made that world ethnographically visible, and the following account depends on that prior act of recognition. My argument begins where such conventions reach their limit: in sites where participants cannot be stably addressed, consent cannot be meaningfully organized, and return is structurally foreclosed. In such conditions, the task is not to observe the site—observation remained possible—but to write from within the failure of the instruments brought to it, and to let that failure shape the writing.

A second point of proximity is my own earlier study of dementia wards in Israel’s North (Kaplan, 2023). That work stayed inside the conventions this article tests. It had a sampling table—twenty-nine caregivers with pseudonyms and demographic columns—alongside 58 non-participant observations (roughly 500 hours) and 20 semi-structured interviews, triangulated against field diaries. The apparatus held, and it produced what it was built to produce: an account of caregivers whose own social marginality mirrors that of the residents they wash and feed. What the apparatus could not absorb were the residues that kept leaking into the writing anyway—the long shower after each shift, the smell that settled into clothing and skin, the veteran caregiver who told me she would not wish a nursing home on her worst enemy. At the time, I treated those passages as reflexive garnish, permitted but peripheral. Rereading them now, they look like the first necrographic sentences I wrote: notations from a body that had registered a site before the method had a category for it. The peripheral nursing home in the pages that follow is, in part, that same institutional world revisited—this time without the table.

One structural property runs through all four sites: in each, policy decisions and long repetition placed the subject for whom institutional instruments were designed outside the population that those instruments could register. The Fein complex dwellers were removed from the municipal file through clearance operations. In the municipal official’s own words, the Shvil Akko users fall between two stools: a man in a wheelchair with necrotic socks is neither sick enough to be hospitalized nor well enough to be detoxed and undergo rehabilitation. The Tiberias residents had no files open with any state agency. The residents of the nursing home in the closed ward had, on paper, a perfectly normal file: a funded bed, a signed power of attorney, a medication schedule, and, in the register that mattered for this article, they had crossed the institutional threshold at which the file could be meaningfully addressed to them. Methodological literature often refers to these hard-to-reach populations. Necrography is a suitable approach for a narrow subset in which the problem is structural, not logistical.

I do not claim to know any specific person. I claim that the composites remain faithful across the sites described above. The composites were constructed to protect individual identities while preserving fidelity to recurring patterns across sites. Tullis (2013) and Ellis (2007) argued that ethics in autoethnographic work cannot be reduced to consent. This requires ongoing accountability for the relational fabric of writing itself. Necrography locates ethics in representational responsibility and not in formal consent alone; the obligation shifts toward composite fidelity, non-extractive witnessing, and the refusal to convert lives into narratives that serve institutionalized closure. In this case, I do not justify the absence of consent. I argue that these are sites where ordinary consent architecture no longer functions and that the method emerges from this limitation. The ethical governance of this practice cannot be considered procedural. Instead, it is testimonial. The test is whether the pattern holds, not whether a signature was obtained where none could be obtained. Testimony, here, is not confession and not courtroom evidence. It is the act of saying what was witnessed to someone who was not there, in a form that outlasts the witness. The people in these sites cannot testify for themselves, not because they lack words, but because the institution that would receive their testimony no longer listens. The Clerk testifies on their behalf, but not in their voice. The testimony does not claim to speak in their voice; it claims only that I saw this, wrote it down, and remained accountable to the pattern, not to the illusion of full access. Perhaps it is a weaker claim than speaking for someone and a stronger one than speaking about someone. It is the only claim that the necrographic field allows.

Necrography is offered not as a replacement for ethnography, autoethnography, critical fabulation, or post-qualitative inquiry, but as a limited method for conditions that those traditions do not fully address. It draws on each of them without claiming to surpass them. It is offered because, in my experience, certain fields of structural abandonment pose a specific methodological problem that these traditions do not explicitly address: the simultaneous absence of interviewable participants, unavailable consent structures, fragmented archives, and populations whose institutional disappearance exceeds the capacity of conventional case-based representation. Necrography offers a procedure for writing under such conditions. It treats the absence of stable participants, complete narratives, and verifiable life histories as constitutive features of the field. The method takes the recurring pattern, not the individual case, as its unit of analysis. Its representational unit is the composite figure, and its central object is the social production of disappearance. In this sense, necrography complements other qualitative approaches without replacing them. It is designed for situations in which researchers can no longer reach the people they are studying.

4. Ethical Considerations

Because necrography operates where standard consent architecture cannot be applied, its ethics rest on a set of commitments that must be specified separately from those of other research methods.

Ethical governance. The following is written as autoethnographic fiction, a well-established qualitative tradition in which autobiographical experience, ethnographic observation, and imaginative reconstruction are brought together to illuminate forms of social life that conventional qualitative instruments struggle to capture in their data. It draws on retrospective, non-identifiable material accumulated through years of site observation, diary notes, and embodied recollection. It does not rely on recruited participants or prospectively generated interview data, but on typological composites built from recurring patterns across different settings and contexts. I do not analyze identifiable case files, nor do I cite or use any material to reconstruct individual biographies. Given the retrospective, non-identifiable, and composite nature of the material, I frame the ethics of this article in terms of representational responsibility instead of prospective human-participant oversight. The composite figures presented are based on repeated observations across years and sites but are not portraits of identifiable individuals. In this text, the dialogue attributed to composite figures is not a verbatim transcription. It is a reconstructed distillation of the register, cadence, and thematic content observed across multiple encounters and sites. The composite voice preserves the pattern of speech—what was said, how it was said, and the context in which it appeared—without reproducing the exact words of any single individual. This distinction is crucial: fidelity lies with the recurring social fact, not the biographical detail.

The representational risk was managed through three safeguards: careful composite construction, removal or transformation of potentially identifying details, and deliberate omission whenever the available material was too thin to support a composite. I have not taken any material from individual case files, clinical assessments, or records that could be traced to specific service interactions. I relied on recurring situations, field impressions, and patterns that were not unique to any single encounter.

Composite figures as units of accountability. I do not use a composite for literary convenience. It is the unit through which the act of writing responds to what is placed on the page. No composite corresponds to a single person whose biography can be reconstructed from the text. Each is a compression of patterns observed over years, sites, and encounters, rendered with sufficient typological specificity (gender, generation, language, migration trajectory, and kin position) to register a category and with insufficient individual specificity to allow for identification. A composite is a unit of accountability. The question is whether the pattern is recognizable to readers familiar with analogous sites. This is especially important for composites that draw on observations of minors, as is the case with Almaz. The stakes are higher here because minors who appear in places like Fein 1 are often runaways or have been thrown out of their homes. They are at the very start of a street trajectory, before the system has even registered their absence. To render such a figure as an identifiable individual would risk not only exposure but also the reduction of a minor’s vulnerability to something for the reader to consume. The composite holds the pattern in place without turning a child’s precarious life into a spectacle.

Consent impossibility, not absence of consent. The consent form is itself an instrument that may refuse to enter the site, not a waiver of consent. Hence, I distinguish between the absence of consent—when consent could have been obtained but was not—and the impossibility of consent, when the architecture of the setting, the cognitive or institutional conditions of those there, or the structural foreclosure of return makes the request itself a category error.

The four sites I describe are those where consent is impossible in the strictest sense. In the Fein and Akko cases, the consent script presupposes a social standing that the site is organized to override. In the Hotel of the Fallen, residents have been removed from every register through which consent is usually sought. In the closed dementia ward, expressive speech has receded past the threshold at which consent can be given, and proxy consent is unavailable because relatives have been reduced to an unanswered telephone number. I am not saying that consent is waived. I am saying that the usual consent framework does not fit these sites, and insisting on it would make writing this article impossible.

Representational responsibility rather than procedural consent. I carry the obligation of representational responsibility, which entails more demanding ethical obligations and operates along four axes: typological fidelity (the composite must register a category recognizable to readers working at adjacent sites); non-conversion (the writing refuses to translate the lives in question back into registers of resilience, recovery, or successful institutional closure); non-extraction (the writing does not produce returns, reputations, evidence, or biographies that the people represented cannot claim for themselves); and recognition of the asymmetry of the writing relationship (the writing was done from a position of relative security that the lives in question did not share). I do not argue for a replacement, but only for a different ethical contract for a site where the original contract is voided.

The status of VERSO. I wish to be precise here about what VERSO is and is not, because the ethical questions raised by a language-model prosthesis differ from those raised by composite construction. I used VERSO instead of human collaboration because prolonged exposure compromised my writing capacity and my ability to narrate the field to another person. Constrained by prompts I controlled, the language model allowed me to externalize phrasing without requiring a shared interpretive framework. VERSO had no access to the field beyond the raw materials I entered and contributed no observations of its own. This boundary is important because recent evaluations of language models in qualitative research caution against treating them as analytic agents that generate interpretations or replace the researcher’s own reading (Friedman et al., 2025; Wachinger et al., 2025).

My implementation of VERSO sits at the other end of that spectrum. The model was never asked to interpret. It was asked to return—short, unsmoothed, without consolation—what I had already written. What it returned was phrasing for material that I had already worked through, and the interpretive work remained solely my own. When prolonged exposure degraded my ability to write, it functioned as a constrained linguistic apparatus from which short formulations could be drawn, accepted, rejected, or revised. I am responsible for every stage of the process and for every sentence that follows. In such moments of depletion, VERSO returned one to three sentences in a flat, non-consoling idiom, operating on the material I provided. The ethical accountability for what VERSO produces in the form it takes in this text rests entirely with the researcher. Each sentence was manually selected, edited, and attributed to the researcher. VERSO is displayed in the right-hand column of the molecular testimonies, which is a methodological visualization of how the mechanism functions during composition. I treat the non-reproducibility of this specific register (two years of accumulated context, a particular configuration of prompts, and an idiosyncratic clerical hand at the point of selection) as a feature and not a flaw: clinical supervisory relationships are not transferable between clinicians either. For that reason, I do not claim that VERSO is reproducible in any strict sense.

Limits of the researcher’s vulnerability as evidence. Prolonged exposure affects the researcher’s clarity. The risk is that the reader may treat the researcher’s suffering as a pretext for analysis, as though the researcher is a romantic figure of a wounded witness. I reject this reading. Two concrete limitations follow from my own vulnerability: 1) I may have missed patterns that a less exposed researcher would have seen, and 2) the composites over-represent men’s voices relative to women’s because my own gender shaped who was willing to be observed. The overrepresentation of men’s voices is not merely a sampling limitation; it also reflects a structural feature of the field. Women at these sites were more likely to refuse observation, and this refusal may itself be a form of necrographic labor—a final, self-protective act of withdrawal from the institutional and ethnographic gaze. Gender asymmetry reveals the gendered architecture of both the street and the welfare state, where women’s survival often depends on remaining invisible to the very systems that purport to help them.

Transferability and scope limitations. Necrography is not a method for every hard-to-reach field site. Three conditions define the scope of its application. First, this approach applies only in settings where conventional qualitative encounters are structurally unavailable, and participants cannot be realistically interviewed, provided with ongoing consent, or revisited. In the absence of these conditions, established qualitative approaches are preferable. Second, composites become possible only after repeated encounters render isolated observations recognizable as recurring patterns. The composites presented here are based on years of cumulative engagement with highly vulnerable populations. Third, necrography has been developed within the Israeli welfare state, where administrative visibility coexists with practical neglect. Different welfare regimes organize these boundaries differently; therefore, the criteria for identifying a necrographic field must be reconstructed, not imported. What travels across sites is not the empirical setting but the analytic stance: the disciplined construction of composites that render recurring social patterns under conditions in which conventional qualitative instruments cannot adequately register them.

5. Operationalization of Necrography

Necrography includes distinctive analytical procedures that emerged from the practical demands of these sites. They are presented not as a universal protocol but as the minimum structure necessary to render the method intelligible beyond the particular cases discussed in this text.

Unit of analysis. The scene constitutes the primary observational unit in necrography. Necrography proceeds from scenes to patterns: individual scenes serve as observational units through which recurring social patterns become analytically visible. The recurring pattern, therefore, constitutes the unit of analysis. Across time and sites, these patterns are rendered through composite figures rather than identifiable human subjects.

Data. The material consists of field notebooks accumulated over the years, observations recorded both in the moment and retrospectively, and recurring situations that became meaningful through repetition. I treat memory not as noise but as another site where patterns become visible. The Clerk’s primary instrument is not the interview schedule but the sustained act of listening: to what is said, what is not said, and the acoustic signature of the site itself. The ear arrives before the eye.

Residue is not raw data; it is data that has already passed through the clinical gaze’s refusal. It is what remains after the institutional instrument is refused and the field is registered without that instrument. The analytic task is to read the residue as evidence of the conditions under which a life became unrepresentable, rather than as a representation of that life.

Analytic process. I reviewed field notes chronologically, identified recurring patterns (e.g., the rain-and-bucket choreography at Fein 1), grouped observations by typological similarity (gender, generation, and migration trajectory), and constructed a composite only when multiple distinct observations supported the pattern. When observations contradicted one another, I retained the contradiction as part of the composite rather than resolving it. The criterion is not factual correspondence to a single case, but recognizability across structurally similar sites. Most of the composites grew from several recurring strands. Each figure in the article occupies a different position on a compositional spectrum, from a strand anchored in a recurrent encounter pattern to a more fully stabilized composite assembled from multiple scenes across time and site. Some composites drew on four or five recurring strands, whereas others drew on many more. If the material was too thin or singular, it was left alone. If recurring strands remained inconsistent, I preserved that inconsistency because it was often part of what these sites produced. I dropped a strand only when it remained too close to a particular individual. In a few places, I had only one or two observations that I could not verify elsewhere. In those cases, I left the composite unbuilt and let the fragments stand instead of forcing them into a figure they could not carry.

Representation. Necrography is written as autoethnographic fiction. Narrative scenes carry the analytic work, rendering patterns through composite figures. The two-column structure stages the tension between institutional legibility and its limits: the left column records what the file captures; the right column registers what the file cannot hold.

Validity. My main claim is that necrography does not rely on triangulation or member-checking, which are unavailable at structurally abandoned sites. In their place, I propose composite fidelity as a criterion of validity. Composite fidelity can be assessed through structured feedback from practitioners or researchers working in structurally similar sites, who are asked, “Does this composite register a category of person you have encountered in your own work, even if no single individual matches the description?” A composite achieves validity when independent readers in adjacent settings recognize the pattern it renders. The characteristics of the four field sites, the materials available at each site, and the evidentiary limitations that shaped the method are summarized in Table 1.

Witnessing as form. I describe the disciplined practice that underpins this method as testimonial immersion. This practice is not an intervention in the profession’s usual sense of the term; it suspends that term altogether. It proposes that under certain conditions of structural abandonment, the helper’s remaining role is to bear witness with sufficient precision and endurance so that the witnessed person will be transferred, dead or alive, to some form of record that will outlast the institutions that failed them. This practice does not guarantee healing. It also does not pursue the results. It refuses to transfer the field’s failure to the suffering body. It only says, I saw you, and I wrote it down, and this writing will outlast me, and has your name on it (or, where the name cannot be entered without harm, the figure into which your name has been folded), and the name is the part of you that the water fails to dissolve. Necrography is a writing practice rooted in this discipline of testimonial immersion.

Table 1. The four necrographic sites: conditions, materials, and evidentiary limits.

Site	Type	Period	Available Materials	Why Interview/Member Checking Could Not Operate	What Counts as Evidence
Fein 1, Tel Aviv	Drug-and-sex courtyard, structurally abandoned, demolished in 2015.	2010-2015	Longitudinal field notes from repeated visits to the courtyard, courtyard-and-loop diagrams, municipal demolition records, and one return encounter with an Ethiopian-Israeli adolescent (the Almaz strand).	The consent script presupposed a social standing that the courtyard was organized to revoke; the researcher had no agency authorization; the population had dispersed and was partly deceased by 2015; and member-checking was impossible after the demolition of the courtyard.	Recurrence of the courtyard’s rain-and-bucket choreography across visits; the stocking knot reproduced over the years; and the gradient structure (wetness, transaction, gaze) observable at adjacent sites.
Peripheral Nursing Home, Upper Galilee	Closed dementia ward: effectively kinless elderly immigrants past the threshold of expressive speech.	2011-2017	Six years of corridor observation under professional employment; longitudinal notes and contextual site materials; staff-shift records and institutional files (used only as non-clinical administrative artifacts3); patterns of practices conducted past the threshold of expressive speech (Ilana and Avraham strands).	Residents past the threshold of expressive speech; proxy consent unavailable (relatives reduced to an unanswered phone); member-checking impossible by definition; staff-mediated proxy interview would replace residents’ testimony; refusal of the environmental device–the building itself (50 Hz fluorescent hum) nullified the conditions for the residents’ necrographic work, a refusal made by the site and not by any person.	The forty-minute corridor loop, the Judeo-Arabic call to prayer at 17:20, the residents’ different responses to the ambient hum at 50 Hz, and the discipline of withholding interruption.
Shvil Akko, Tel Aviv	Open-air drug scene, cyclical evacuation, and reopening.	2017-2024	Lane observation across the cyclical clearances; municipal press releases and reopenings (recurring across numerous cycles); exchanges with municipal and NGO outreach workers; the Arkady strand.	Cyclical displacement makes longitudinal interviews impossible; users are at the intersection of dual diagnosis, homelessness, and kinlessness; the institution that would receive a case note has, by its own clearance protocol, refused to receive it.	The repeated pattern of sequential refusal across institutional doors, the outreach worker’s eventual cessation of case-note writing, and the lane reopening on a measurable cycle.
Hotel of the Fallen, Tiberias	Bankrupt lakeside hotel, unofficially inhabited, the lowest freshwater lake on Earth.	2023-2025	Cross-window observation; intermittent lobby visits; acoustic and gestural records; strands assembling the Yasha, Boris, Tanya, and Nasreen composites.	No state agency had an open file on any resident; the building had no admission procedure; the researcher entered as a displaced person, not as a recognized investigator; and member-checking would require reopening files from which the residents had disappeared.	Repeated micro-practices of necrographic labor (polishing, recitation, withheld interruption); the building’s acoustic signature of utterances, deposited and not absorbed; and Tanya’s parallel medicine.

Scope of Application. Necrography is not applicable when: 1) participants can be interviewed; 2) ongoing consent can be meaningfully obtained; 3) longitudinal engagement is possible; 4) institutional files exist and are accessible; or 5) member-checking or triangulation can be meaningfully conducted.

These four sites were heterogeneous in period, location, population, and institutional architecture. Yet they were methodologically similar in one crucial respect: at each site, the standard qualitative mechanisms could not operate for reasons other than logistical constraints. In each case, the writing had to rest on the continuous recording of a pattern rather than on a structured encounter with a single case.

6. Instrument-Refusal

On a rainy night in late spring, standing in the covered section of the western walkway at Fein 1, I watched four instruments I had been trained to trust settle onto the wet concrete and refuse to enter. The failure was not in the instruments, nor a gap in my training—it was a refusal made by the site itself. I call this phenomenon instrument-refusal. Certain concepts, when brought to the threshold of certain rooms, refuse to take hold in ways that feel closer to active resistance than to simple inadequacy.

The four were Bronfenbrenner’s four ecosystems (Bronfenbrenner, 1979), Saleebey’s strengths perspective, Shulman’s helping skills (Shulman, 1999), and Marlatt and Gordon’s (Marlatt & Gordon, 1985) relapse prevention model, which presupposes a future-oriented agency that the courtyard’s temporality had dissolved. I have named them not to make them seem futile but because each was a first-rate instrument for the problems it was designed to address. Each assumes a domain in which the encounter is, at a minimum, mutually entered. In the Fein courtyard, the parties did not agree to the encounter. The black market on the corner by the covered hole-in-the-wall, where cash entered through a slot in the plywood and envelopes came out with no face associated with any transaction, functioned precisely as a refusal of the service offered by the state. When I walked in with Bronfenbrenner under my arm, I was carrying the wrong paper. At Shvil Akko, a municipal outreach team later attempted to use the same entrance with the same instruments but was refused entry in the same way. Refusal is not a property of any specific site. It belongs to a class of sites structured by their own internal rules of access and refusal.

It is worth naming three registers of refusal because they ask different things of the writer. The first is conceptual: the strengths perspective and Bronfenbrenner’s rings are frameworks that do not map onto what the courtyard contains. They were built for other rooms. The second is enacted: a woman who does not sign the consent form, a man who refuses the food offered to him. These are refusals made by people. The third is ambient: the building’s acoustics, the fluorescent hum, the open drug scene that reforms weeks after clearance. This is a refusal written into the site’s architecture. All three occurred at Fein 1, but at different intensities. The Clerk learns to read them as data, not as obstacles. The question is never how to overcome refusal. The question is what refusal reveals about the structure that produces it. When forms and questions are refused, it is not a research glitch; it is an echo of their everyday lives. They are repeatedly turned away by the system’s paperwork, and their refusal of paperwork may be read as part of that same administrative history. Their silence toward standard research tools can become a heartbreaking mirror of the systemic abandonment through which they live.

The response to instrument refusal is to document the refusal itself with forensic precision, instead of immediately searching for a replacement instrument. I treat instrument-refusal as data in its own right. This is also an ethical moment. The moment a concept refuses to cross the threshold is the moment the researcher stands in the gap between the institution that trained them and the field that does not want what they were trained to provide. This gap is where necrography operates.

7. Fein 1

The H-shaped building block at the intersection of Fein, Salomon, and Erlanger streets in South Tel Aviv was built by Egged (the Israeli national transport cooperative) as a staff welfare facility for the old Central Bus Station and lost its function in 1993 when the new Central Station opened nearby. What follows is best read through Stewart’s description of ordinary affects: a slow and unspectacular reorganization of a place around the uses made available by its abandonment (Stewart, 2007). By the time I first entered, the H-shaped building on Fein 1 contained roughly 70 single rooms, spread around the inner courtyard, rented out daily, with an unregulated drug market in the eastern wing and a parallel economy of survival sex in the western wing. The municipality cleared the building in 2008, and the authorities cleared it again in successive sweeps over the next seven years. The final demolition occurred in 2015. By then, the building had been razed to a slab, and what remained was a cleared lot that the municipality called urban renewal.

The H of the building was oriented along the north-south axis, with the crossbar running east-to-west across the inner courtyard, and the address was the meeting point of three short streets: Fein from the north, Salomon from the west, and Erlanger from the east. The arch through which one entered was located at the western end of the crossbar. Each of the two wings carried 13 ground-floor rooms that opened directly onto the courtyard and 20 rooms above, which were accessible via external staircases at each end of the wing. By the time I began visiting, the western wing housed a parallel economy of survival sex work. There were at least three kaspomatim along the eastern wing’s ground floor, distinguished from one another by minor variations in the plywood and the lack of signage. Cash was inserted through a slot. The wrap was retrieved through the other slot. The hand conducting the exchange was visible only through the sleeve. Over the years, the transactional minimum has been reduced to basic arithmetic.

The courtyard was organized into three gradients, and the walking body learned faster than the notebook did. The first gradient was wetness, running from the covered path in the west wing through the open concrete at the center to the free line at the east end. Each step out of the shelter represented a rise in social status. At the edge stood people whose bodies seemed to have ceased registering the weather, for whom the rain was a fact on the ground, and there was no reason to move from the edge of the rain. The second gradient was transactional: from the hole-in-the-wall where wraps came out faceless to the edge of the wing, where a longer negotiation could be held between two people who had known each other for years. Each step inward reduced the verbal exchange required to exchange the item. The third gradient was gaze: the closer the observer came to the covered walkway, the more their own gaze became something to be negotiated, and the further away they went, the less their gaze was registered.

I entered the courtyard for the first time on a Wednesday in mid-November 2010, at approximately a quarter to seven in the evening. I carried with me, in the only sense in which a methodological instrument is truly carried, four tools: Bronfenbrenner’s concentric rings, Saleebey’s strengths perspective, Shulman’s tuning-in exercise, and the harm-reduction lexicon I had absorbed. They did not enter the courtyard with me. They waited on the pavement outside the central arch, slightly damp from the sea air that reached far inland on November nights. When I walked out at a quarter to ten, I found them exactly where I had left them and put them back in my bag. The instruments refused to accompany me into the field. This was not a failure of courage, a gap in my training, or a crisis that another reflexive monograph could address. I gradually realized that the refusal was not mine to make. It belonged to the site itself, and I found surprisingly little language in the literature to describe it.

Wacquant (2004) entered the Woodlawn Boys Club in the spring of 1988 and improvised a replacement instrument out of the body of the person he was becoming because the Bourdieusian apparatus had declined to follow him past the locker-room door. While Wacquant’s carnal sociology insists on realist and materialist immersion to diagnose the structural abandonment of the ghetto, necrography departs from this epistemological framework at the point of writing. Wacquant relied on the preserved analytical capacity of the researcher’s body to reconstruct the field. In contrast, necrography operates where that very capacity is disabled by exposure, requiring a post-qualitative shift from realist representation to composite construction.

Scheper-Hughes (1992) sat in the kitchens of Alto do Cruzeiro in 1982 and watched her public health categories come loose from what the women of the favela were producing in her presence: the slow withholding of maternal grief from infants whose deaths had already been anticipated. Bourgois, in East Harlem, called it a set of borrowed clothes one size too small for the body of what one was seeing (Bourgois, 2003)—three ethnographers, three thresholds, and what appears to be the same phenomenon that the literature continues to label as disorientation. Disorientation is most likely an incorrect term for this. This is a refusal of what one wants to say about its metaphysics through concepts trained in one room and brought to the door of another. The ethnographer, therefore, walks in alone, unsung, and never ready for what is to come.

The courtyard was quieter than the newspaper exposés led me to expect. The translucent corrugated roof absorbed traffic noise from both sides of the street, whereas the two wings of the H-house insulated the remaining noise. In the first few seconds inside the arch, the dominant register was my own footsteps on the tiles and, below it, the acoustic signature of a household: three coughs from three rooms, the dragging of a plastic chair, a kettle starting to bubble, and a radio playing Umm Kulthum⁴ at a volume just below the threshold at which a neighbor would complain about the noise. Several dozen households were arranged around a common courtyard, operating on the low thermal output of bodies long accustomed to caloric rationing. The smell was then stratified into categories. The lower register was urine in porous concrete at a temperature of eight to fourteen degrees Celsius. The middle register was damp with the smell of Next-brand tobacco and the metabolic output of several dozen people on subtherapeutic diets, a register continuous with the geriatric ward and the long-term psychiatric dormitories where I volunteered as an undergraduate student. It took me two winters to identify the upper register. It was finally identified for me, with a tolerant half-smile, by an older woman who found my confusion amusing: a soda-type washing powder, used since 1994 by a woman in the western wing who had decided at some unrecoverable point in the previous decade that she would not let her clothes smell like the courtyard, even if her body had been forced to smell.

Taussig (1987) described the ethnographic convention of elevating such specifics into atmosphere as “the anesthetic use of detail”, by which he meant the production of texture as a way of postponing knowledge. I reject this convention. A soda-type washing powder is not an atmosphere but a signature of refusal to dissolve.

On the second floor of the eastern wing hung a blue tarpaulin, fastened at the top with two rusty nails and weighted at the bottom with half a brick. On the third evening, I noticed that the half-brick was wrapped in a piece of flesh-colored stocking tied at the back, having been washed at some point in the shared sink of the west wing, tied there by a hand I had never seen for no purpose other than to keep the half-brick from crashing into the frame while the yard tried to sleep. No amount of training I had received in social work school prepared me for the theoretical difficulty of that unsung stocking. The lexicons of vulnerability and resilience (Masten, 2001), risk and protective factors (Fraser, 2004), and need and provision have failed this stocking. The stocking required a different instrument. I did not have it then. I am writing it now.

Tenderness was present in the courtyard even before my arrival. The courtyard did not require my arrival for tenderness to be present, and tenderness was being practiced by women the state had decided were not full subjects, against a sheet of water the state had decided it would not roof, in a building the state had already decided to demolish. The Clerk’s task at the lowest level is not to bring tenderness into the room. The room has been running on tenderness alone, like a machine on its last exhaust fumes, and the Clerk’s task is to write that down in a form that will outlast the building in which it was performed.

On the fourth evening, it rained. Within four minutes, three women I had not seen before emerged from three ground-floor rooms in the western wing, carrying a red plastic bucket, a galvanized basin, and a rectangular storage box that had once held a child’s toy and still bore a faded sticker of a cartoon character I could not identify. They placed the containers at points along the sheet of water flowing down through the broken seam of the corrugated roof. I saw these points as arbitrary, and as the rain thickened, I realized they were where the sheet had historically caused the most damage to the shingles. No one spoke. When the first was two-thirds full, a fourth woman emerged from the eastern wing with an empty bucket and, in one unhurried movement, swapped it for the one that was filling, then carried the full bucket through a side passage to a drain behind the alley. Forty minutes. Wordless. There was no acknowledgment of the man on the opposite step, with a pen that had begun to bleed from the moisture. Bronfenbrenner’s rings would have scored this as an informal neighborhood organization. Saleebey’s strengths perspective⁵ would have scored it as adaptive resourcefulness. The harm-reduction framework would have had no place for it. What I had watched was infrastructural care carried out inside a building under demolition, and the concepts I brought were on the sidewalk outside the building’s central arch. The bucket, the basin, and the child’s toy box are successors of the vessels, and the institution does not know what they are. It sees buckets; necrography sees liturgy.

Ethnographic literature offers a familiar corrective to what I have just described: the researcher is involved, and naming the allusion corrects the asymmetrical gaze. The process is valid in the context in which it was developed. In Fein 1, it was not available at all because the stark asymmetry between the girl in the doorway and me was not a function of my vision. This was a function of her having been placed by the municipal clearance operations of 2008, further evictions, and, finally, the 2015 demolition, on a route that ended at a building that no longer existed. My gaze added nothing to the structure of the scene. Writing that could answer her seriously would have to answer this trajectory, not her gaze.

On the last Saturday of November, my sixth visit, I sat on the steps opposite the eastern wing, watching the tarpaulin move with the breath of a woman I had not yet learned to call anything but a nameless body. A girl came out of the second doorway of the western wing, sat down on the step on the opposite side of the courtyard from me, and watched me for the length of a full cigarette that she did not smoke. She looked about 15, maybe 16 years old. Ethiopian-Israeli. A thin green hoodie over a simple cotton dress. She did not speak that evening. When I got up to leave, she also got up, returned to the second doorway, and disappeared before I crossed the courtyard. The next morning, I wrote a sentence that I have not changed since: She watched me the way a person watches a stranger; she could not decide whether to warn me. She returned on December 5th. She sat closer to me on the same step, five feet away, and asked me in Hebrew what I was doing there. I answered, “I am learning to see”. She looked at me for a long time and then said, in the same quiet register, “Then you are in the right place, because here nobody sees anyone”. She did not tell me her name. I did not ask. In the notebook that night, I wrote that she might be the most dangerous person I had met in the courtyard because she was the only one who had caught me at the verb. She did not need to know my name to know what I was doing there. This girl is one of several Ethiopian-Israeli girls whose biographies are intertwined with the character who appears here as Almaz. She is one of these strands, although I will not tell the reader which one, because the composite is the unit of accountability, not the strand itself.

This encounter changed the notebook—from that evening it registered a different quality of attention: no one sees that someone has turned the clinical gaze of the holder back on the holder, a maneuver that I encountered in Taussig (1987) and, more recently, in Das (2007), and did not expect to encounter in a courtyard in South Tel Aviv on a chilly December evening. Felman and Laub (1992) insist that there are always two witnesses: the speaker and the listener constitute what can be spoken about. In the opposite step, the girl reversed the pair’s polarity with the single question that returned to her. She became a witness to the would-be witnesses: I became a figure whose behavior required accounting. The researcher in me was promoted, without ceremony, against any consensual preference, for reasons the bureaucracy of my professional training had never considered—from witness to evidence.

Das (2007) argues that the ordinary is the site where violence and suffering are inscribed and endured; the girl’s reversal of the observing gaze was precisely such an ordinary event: a quiet, unspectacular act of witnessing that refused the researcher’s epistemic privilege and demanded a different kind of accountability.

In the third winter, the tarpaulin remained dark for two weeks, after which it was removed, and the opening was closed from the inside using a square piece of chipboard. I did not ask. The woman whose hand had tied the knot at the back of the half-brick, whose stocking had been washed in the communal sink, and whose breath had moved the tarpaulin for two and a half years was still there. She had no intention of moving it again. The stocking is still in my notebook. I call the woman behind the tarpaulin Sima, although I do not trace all the seams of that choice. From this point onward, every chapter that follows is written under the assumption that she exists in the real world as a human being. The non-interventionist perimeter is the name I am giving to the discipline by which, over the years I have known the yard, I have refrained from offering Sima the services my training would have authorized me to provide. I learned discipline on the step opposite her window, and I have maintained it ever since. That is what made this work possible.

Biehl (2005) and Bourgois (2003) describe the formation of abandoned zones. Neither fully addresses the extent to which such a zone is not a passive vessel for those abandoned by the state, but a field of tacit negotiation among the abandoned persons themselves, the results of which are embedded in the architecture of the rooms. It would be easy to classify a half-brick wrapped in a stocking as a survival strategy. This is, in fact, a form of governance. Biehl’s notion of a “zone of social abandonment” describes a remarkably similar space (Biehl, 2005); however, where Biehl traces the individual trajectory of a single subject (Catarina), necrography shifts the unit of analysis to the recurring pattern of the site itself, treating the zone as a structural constant rather than a biographical exception.

Writing about it requires a stance other than that of a clinician or ethnographer, as exemplified by Wacquant (2004) in the Chicago boxing gym. At the end of his fieldwork, Wacquant stepped out of the gym and resumed a position from which he could analytically reconstruct the gym. The Fein courtyard did not allow this step. It allowed a lateral shift in which the researcher remained in the same body but was no longer the one doing the naming. Replacing the word “closed” with the word “opened” on the intake form is all the distance my field has come in the years I have been practicing and writing about it.

At twenty past eight on each of the evenings I visited, I got up from the steps and walked out through the central arch and did what I had come to do on every visit from the second evening onward: a slow, wide loop around the outside of the building for 25 minutes at my own pace. Salomon faded behind me as I made for Har Tzion, where the sodium lamps were broken and not replaced. The asphalt carried in its cracks a mixture of diesel from the bus station and sewage from the public toilets of the old terminal. From there, the route continued into the small network of side streets behind the terminal, where women displaced by earlier evacuation operations slept in doorways and on the abandoned ground floor of a building whose upper floors were occupied by others. From there, it curved back to Levinsky Street by way of Rosh Pina Street. That loop rewired my nervous system. It taught me that the courtyard was not an island but a center of gravity; that the women behind the tarpaulin windows were the most stable segment of a larger mobile population whose movements became legible within a twenty-minute walk beyond the courtyard walls; and that any ethnography of Fein 1 conducted only within Fein 1 would be an ethnography of a node with its connections severed. Several times during the first two winters, men approached me in the dark stretch of Har Tzion, and at the time, I did not understand what they wanted from me.

Each time I kept walking, I understood, in a way that neither Lee (1995) nor Kovats-Bernat (2002) fully considered in their writings on dangerous fieldwork, that my safety on those walks was not produced by my own behavior but by a collective, tacit, and continuously revisable decision of the population at the site regarding whether my presence was tolerable or not. I was safe in the loop because the courtyard had not yet decided to exclude me. The decision was not earned. It was granted on terms that I had never understood. It could have been withdrawn on terms that I would not have seen coming, and its lack of withdrawal is the only reason those winter field notes exist. That is a methodological premise I add to the literature: the ethnographer’s safety at a site of structural abandonment is not a feature of the ethnographer’s behavior but a revocable concession of the terrain. To call it proximity is to flatter the ethnographer, and to call it permission is to be closer to the truth.

The Clerk learns this slowly. In the first year, I still carried myself like a social worker who had misplaced his badge. I looked for problems to solve, gaps to fill, and services to connect. The courtyard taught me otherwise. One evening, I caught myself reaching for a pen to write a referral, an instinct I had been trained to trust. I put the pen down. The referral would have gone nowhere. The file it would have entered was already closed. The woman whose name I would have written had no address, no ID number, and no eligibility category that the system could recognize. The Clerk does not intervene. He does not refer. He registers. His only question, repeated across years and sites, is: what pattern survives? That evening, on the step opposite the tarpaulin, I learned that the answer was not in the referral. It was in the rhythm of the buckets under the broken guttering.

8. Shvil Akko

Shvil Akko is the unassuming Hebrew name for a 200-meter-long lane in southern Tel Aviv. In 2015, with the demolition of the building on Fein 1 Street, it received some of the displaced drug user population. By 2017, it had become one of Israel’s most prominent open drug arenas. The police cleared it in a well-publicized operation and declared it clean. Yet, within weeks, the lane reopened and operated openly 24 hours a day, seven days a week.

The continuity between Fein 1 and Shvil Akko is a matter of bodies and not metaphors. A subset of the population displaced by the 2015 demolition dispersed from the H-shaped block and reassembled over the following months in the lane behind the new tower block on HaGdud HaIvri Street. I recognized their faces during my third visit. More importantly, I recognized the negotiated geometry: the gradient of wetness was reproduced under different physical conditions, as the lane had no covered walkway other than the awning of the back entrance to a closed printing shop, which served the same sorting function. The gradient of transactional verbosity was steeper because the lane was narrower and the police presence was nominally constant. The gradient of the gaze was reproduced inversely because the new residential towers overlooking the lane introduced a permanent class of onlookers in the windows above the street. The temporal pattern named here, serial refusal, is what the alley adds to the legacy. Fein 1 was gradually abandoned over seven years, from the first eviction in 2008 to the final demolition in 2015.

Shvil Akko was abandoned in cycles measured in weeks: a publicized eviction operation, a press release announcing the clearing of the lane, a four-to-six-week interval, the reopening of the lane to the same commerce with a minor change in the personnel operating it, another press release acknowledging the persistence of the problem, and another eviction operation. This vicious cycle has been documented repeatedly, with at least a dozen eviction-and-reopening events by my count. Bourgois (2003) described a similar cyclical structure in the East Harlem crack market, and the basic mechanism was the same: eviction was not primarily an attempt to end commerce but a periodic ritual by which the city demonstrated to its wider public that it was acting on commerce, even as the conditions that produced it remained the same. Thus, the lane is not where the policy breaks down; rather, it is what the policy produces. The alley is not a failure of the system; it is the system’s visible edge.

Here, I illustrate what the previous composite could not fully convey. Arkady is a composite figure constructed from recurring patterns across encounters in and around Shvil Akko during fieldwork and related professional practice. Literature on adjacent sites (e.g., Bonny-Noach & Ronel, 2018) is used to contextualize the broader structural conditions surrounding the figure. He was born in 1962 in Voronezh and trained as a dental laboratory technician. He arrived in Israel in 1991, in his late twenties, with a wife and a young son. His wife divorced him in 1997 after the second of three psychiatric hospitalizations, during which he was first diagnosed with substance-induced psychosis, then with bipolar disorder, and finally with comorbid alcohol use disorder. None of the subsequent diagnosticians retracted any diagnoses, and the resulting file accumulated diagnoses additively, not incrementally, which I have observed to be characteristic of long psychiatric files in Israel. At the time I met the composite, his son was 34 years old and lived in Petah Tikva. As far as Arkady was willing to say, he had not spoken to him in recent years.

These diagnoses do not belong to the register in which this writing operates, nor are they the relevant facts here. The relevant fact is that he can be read through the intersectional stigma framework discussed by Turan et al. (2019). Arkady exemplifies what I have termed serial refusal: a cascade of institutional denials in which each door uses the identity category that made the previous door necessary as a pretext for its own refusal. The hospital refuses admission because his complaint is not acute enough to justify a bed. The rehab center rejects him due to medical complications that fall outside its criteria. The psychiatric ward will not admit him because his primary diagnosis is a substance use disorder. The substance-use ward did not accommodate him because psychiatric instability was his primary diagnosis. The municipal welfare office will not take him in because he lacks the required identification documents, even though his Israeli passport is at a relative’s office in Petah Tikva, who also refuses him in the relevant sense. At each door, the reason for non-admission was furnished by the identity category that made the previous door necessary. The street is the only environment in which all the agencies he interacts with can agree, in the documents they file, on his location. This is to be seen as a coordinated form of refusal, not as recognition. Each agency documents its presence on the street precisely to justify its procedural decision not to admit him, thereby transforming the street from a place of last resort into a bureaucratic alibi for inaction.

A municipal outreach worker who knew him best told me in the third year of my visits to the lane that she had stopped writing case notes on Arkady because the file was closed by every service to which she had submitted it, and her writing reopened, in turn, a file that no recipient was willing to accept. She continued to provide him with food. She no longer recorded the bringing of food. Arkady’s institutional disappearance from the system was complete by 2021. He continued to exist in the narrow lane. He continued to exist, in particular, on Tuesday afternoons at the corner of the café on HaGdud HaIvri, where the staff knew not to charge him for the coffee. None of this is in any file. All of this is in the notebook alone. The composite is the unit through which I make the case for the existence of a category of people whom the institutional archive has actively chosen not to record, and that choice is itself part of the field.

What is important about the lane is not that it was a site for open drug use and sale; the newspapers already knew that fact. It was the first site where serial refusal became the dominant rhythm. Fein 1 was slowly abandoned over the years. Shvil Akko was publicly announced as having been evacuated in the media and was allowed to recover over several weeks. What the Clerk is being asked to write about here is not abandonment, but a pattern of abandonment in which the act of institutional intervention itself is a move within abandonment rather than a departure from it.

The composite emerges at the intersection that recent literature on intersectional structural stigma has begun to name. Arkady is a male, so female-specific street services do not reach him; a Russian-speaking immigrant in his sixties, he falls outside the reach of outreach programs designed for younger Hebrew-speaking users; dual-diagnosed with alcohol and opioid dependence alongside early Korsakoff-type cognitive impairment, he is moved from service to service, each reading the other diagnosis as the reason not to begin; homeless, he arrives at the clinic already as a discharge problem, and so the procedure withdraws before it can occur.

The methodological lesson of this lane is that abandonment is not repeated. It is that the cycle of evacuation and reopening produces a temporal artifact that the standard qualitative apparatus is not set up to capture. In a longitudinal interview design, upon entering the lane, one will not find a stable group: between the third and fourth months of any prospective study, the population has dispersed, been partially reabsorbed, and partially replaced. A cross-sectional design captures a thin slice of the cycle and misreads it as the site’s steady state. The pattern necrography is configured to register operates on a different timeline: the gradient structure inherited from Fein 1; the negotiated geometry of the awning and back wall of the printing house; the choreography of the café staff who do not charge for coffee on Tuesday afternoons; and the unwritten relay between the city outreach worker and the hospital discharge planner, who can no longer formally refer his patients to a place that will accept them. These reappeared in a slightly different form after each clearance. The pattern survives because it is generated by the structural arrangement that clearance protocols do not touch.

9. The Hotel of the Fallen

I came across the second type of site by chance. I was displaced under rocket fire at the beginning of the October 7 war⁶, and spent a year on and off in a small hotel on the Kinneret promenade. The hotel where I stayed is not the one described here; the one I describe was an abandoned hotel that I could see from my window. I call it the Hotel of the Fallen because that is what it taught me that winter.

The Hotel of the Fallen ceased to operate years before my arrival. Two elderly men appeared to be living in half-ruined rooms, according to an arrangement I had pieced together in the dark. Yasha is a composite assembled from a white-bearded silhouette seen through a lobby window, who, from afar, looked like a desert father. He is a figure I crystallized into being at four in the morning: a former professor of Russian literature, drunkenly reciting Pushkin in a rhythm that exists in my notebook and nowhere else. Boris is a parallel composite, built from several elderly men I encountered in similarly impoverished neighborhoods, development towns, and geriatric settings. One of them I once saw from my window in the dead of night, searching through the hotel's trash cans. Over time, that silhouette merged with the others into a single figure. The thinness of the entry point is not the thinness of the composite: Yasha and Boris are assembled not from the silhouette alone but from two decades of encounters with structurally similar figures at all four sites. The window is where the pattern became visible, not where it was constructed. Boris is a former carpenter from Dnipropetrovsk who arrived in Israel and who keeps, in a room I have never entered, two suits in a plastic bag from a wedding that was canceled 22 years ago.

Before continuing, I must be honest about the location of the material’s autoethnographic core from Tiberias. The hotel I booked at the beginning of the displacement was a small establishment 300 meters along the promenade, which I will refer to as The Castle. This name is not my invention, but it is used casually by some of the older residents of the lower city, and I adopted it in my notebooks for that reason. The Castle functioned as a hotel in the usual sense. It had a desk, a key, a breakfast room, and a cleaner. The room was located on the second floor and overlooked the lake. From its window, I could see, about a hundred and twenty meters away, the lobby of the Hotel of the Fallen. The Castle is the autoethnographic core of this article. The Hotel of the Fallen is what I went to see. The Castle is where I confronted what seeing was doing to the one who was doing the seeing. I sat on the narrow hotel bed in the evenings and wrote, and what I wrote was not only the lobby but also the man in the lobby window and what the lobby had done to him in the last two hours. There is no version of this article in which the opposite room is behind the scenes. The two hotels stand in the same relation to each other as the recto and verso of a printed page. Readers who treat the Castle as parentheses will miss the methodological point. The Clerk needed a room of his own from which to register what the abandoned room contained, and the existence of the second room was the precondition within which the writing of the first was possible.

Wacquant (2004) could return home to a clean room at the end of his Chicago apprenticeship. The Tiberias displacement did not afford a clean room. It afforded the Castle. The walls were thin. What was happening in the rooms next to mine, especially in the room above me on certain Friday nights, was not what was happening in my room. The composites in the following chapters are not the only composites that the field generated from scratch. The Clerk is also a composite—not of multiple persons, but of multiple moments of the researcher’s professional self across sites.

What the Hotel of the Fallen offered, which Fein 1 did not, was prolonged exposure to the specific temporality of the drowned. Time in the hotel did not move forward but settled layer over layer. Boris’s wedding had been next week for 22 years. Yasha’s Pushkin had been in the lobby for the same 22 years. In my notebook during the third week, I wrote that time in the hotel felt like water that refused to drain from a pool. What I mean by “the Drowned God”, within the specific register this figure appears in Tiberias, is the archetype created when the linear progression of institutionalized time is removed from life, while existence continues at its most bare, physical level. Boris, in the composite, polishes the buckles. Yasha, in the composite, recites the first eight lines of The Bronze Horseman⁷. Berlant’s account of cruel optimism comes close to what I am describing (Berlant, 2011). Cruel optimism is the attachment to a form of life whose conditions of possibility have been abolished but whose emotional appeal remains. Boris’s suits are cruel optimism. Yasha’s Pushkin is something else. It is the continuation of a practice whose object has been deleted and whose performance, therefore, at the level of the hands, is already an elegy. I do not have a better term for it than necrographic labor, and I offer it as the second term this article contributes, after instrument-refusal.

Boris dusted a dead television in the lobby every morning with a yellow microfiber cloth, following the same five-stroke sequence: the top edge, the right edge, the bottom edge, the left edge, and finally the screen, wiped in concentric circles from the center outward. He did not turn it on. I record this sequence to resist the urge to translate it into resilience, adaptability, or symbolic immortality. Boris’s dusting is readable only in the register in which the hands continue a liturgy whose object has already been withdrawn. The hands do not know they are mourning. They only know the rhythm. If Shvil Akko is “the pit”, that is, the lowest place inside the municipal boundary of a city, then the Hotel of the Fallen is the bottomless abyss.

Tanya is a composite figure of a woman seen on several separate evenings carrying a plastic bag along the lakefront promenade, as well as of several women I met in nearby geriatric and addiction settings over the past 20 years. In the composite, she is 56 years old and arrived in the country in 1993 from a small town in Donbas. She had trained as a feldsher⁸ in the Soviet medical system, a category that does not translate clearly into the Israeli registration architecture: neither a doctor nor a nurse in the local sense, but a mid-level practitioner whose certification, upon her arrival, was neither officially recognized nor rejected, but rather suspended on registration, the equivalent of a permanent, temporary status. The suspension was not lifted in the thirty-two years that followed. She worked for a long period in a private nursing home in the Lower Galilee, performing the duties of a registered nurse without registration, at the salary of an aide. In the lobby of the Hotel of the Fallen, she did the same work for Yasha, Boris, and two other residents whose composites I had not assembled because the material was too thin to honor.

She measured blood pressure with a sphygmomanometer she had brought from the nursing home, which closed when its last private owner filed for bankruptcy in 2019. She changed the bandage on Yasha’s shin at least three times per week. She knew which residents should receive which medications and in what doses, even though the medications were not prescribed in writing by any doctor with a valid Israeli license, and the doses were calibrated by Tanya based on the patients’ appearances each morning.

The medications were obtained, the composite suggests, through a discharge planner at the Tiberias hospital, who could not, in his official capacity, refer Tanya’s patients to a place that would accept them as is. Therefore, in his unofficial capacity, he ensured that what Tanya needed continued to arrive at the lobby. None of this is in the medical file. Medical records do not include information about what is done outside the records. Tanya is the figure through whom I register a phenomenon I have not yet mentioned in print: parallel medicine performed by qualified people whose qualifications the state has refused to certify, inside rooms where the state has ceased to certify as places of care. To call this parallel medicine “interstitial” reverses the proportion. Inside those rooms, it was simply razor-sharp medicine. State-recognized medical practice has ceased to operate in these rooms. I resist the temptation to label Tanya’s work as either heroic or unauthorized because both labels would situate it in relation to a state-recognized practice that the field had already vacated. Tanya continued to do her work in the building, where that labor was no longer recognized as work.

What Tanya offered that I have not been able to write adequately about, and that I am attempting to do, knowing the attempt will be inadequate, is the quality of attention she paid to Boris’s polishing on the morning of the third week of February, when his hands were trembling more than usual, and the fabric left uneven marks on the buckle. Tanya stood in the doorway of his room for approximately three minutes, watching the polishing process without entering.

She did not interrupt. She never sought his gaze. She did not intervene in the language used in the geriatric care literature. She continued to watch until the trembling subsided after the polishing sequence was completed. Only then did she enter the room and offer him a glass of water that she had brought from the lobby. The glass of water was a narrow intervention, but the work was in the timing, and the timing was hers to choose. From what I have read, the institutional literature on dementia care has not adequately recognized timing as a distinct competency. It is what I had begun to recognize in the hallway of the closed dementia ward 15 years earlier, as a discipline of withholding interruptions. Tanya had no words to describe this experience. If anything, she was the thing that the word should have described.

Nasreen is a composite assembled from a young woman I saw twice in the lobby of the Hotel of the Fallen, inspired by several Druze⁹ women I had met in various care facilities in the Galilee over the past decade. In the composite, she is 28 years old, originally from the Druze town of Hurfeish in the Western Upper Galilee, homeless, and living without a registered address in the city. She left her family house in Hurfeish at the age of 23 under circumstances she would not describe and which I do not intend to speculate about, since speculation in this register could re-traumatize the woman behind the braid, even when she is folded into a composite figure. What I will say is that the departure was final; in a sense, the departures I have seen among Jewish Israeli women of her age tend not to be. There was no return to the village. No holiday visit. There were no phone calls to her baby sister either. The family network, with the finality allowed by the Druze religious-legal apparatus, was closed.

She arrived in Tiberias via a women’s shelter in Haifa, an interim program in Acre, and a series of short stays in rented rooms in cities in the Lower Galilee that she could not afford to extend. By the time she entered the lobby of the Hotel of the Fallen, she had reached the end of the network of services designed to receive Druze women fleeing their families, which was not designed to accept women beyond the age of 25. According to the admission forms that had ceased to register her, she was getting too old. The Hotel of the Fallen accepted her because it accepted everyone, in the sense that the only buildings that accept everyone are those that no longer have an admission process. Tanya took her in. Boris polished a buckle for her wedding-that-would-not-happen, as he polished his buckles for his wedding that had not taken place for 22 years, and the gesture was not confusion but recognition: Boris had recognized in Nasreen the figure for whom polishing was a form of address.

Yasha recited The Bronze Horseman to her in the lobby on three separate evenings that I observed. Once on a Wednesday night, he recited an unfamiliar passage in Russian that I had not heard before, which I later recognized as a poem by Akhmatova. Akhmatova recited in a Tiberias lobby to a Druze woman from Hurfeish at three-thirty in the morning. Readers who find the scene implausible and surreal are reading it correctly. The Hotel of the Fallen was, among other things, the building in which improbable juxtapositions ceased to be improbable because the categories that would have made them seem so—nationality, language, religion, and generation—had ceased, inside the building, to function as sorting categories. The building ceased to sort, but the water did not.

Nasreen eventually left the hotel for a reason that was not given to me and that I was unable to ascertain. She may have died. She may have returned against her will or, with partial consent, to a relative in a different village. She may have found a place to live. I do not know. Not knowing is a condition of writing, and it must be brought into the writing, not evaded by it. Hartman (2008) described critical fabulation as a form of writing one is forced into when the archive is built on the systematic refusal to register its subject as a person, and one, in turn, refuses to let the archive have the final word. Nasreen’s character, as approached in this article, fabulates in Hartman’s sense but remains bound by composite fidelity. Fabulation is the only form in which what happened in the lobby on those evenings can be included in published records. The fabulation is accountable for the lobby, not for the bureaucratic registry that has already lost its track, and accountability runs through the discipline of the composite, not through the impossible reconstruction of the strand.

Nasreen’s composite makes visible another tension in the strand-to-composite methodology, as illustrated in her case study: Of the four sites, the Hotel of the Fallen is the one where the gap between the strand and the composite is most tense because the lobby population was small enough that a single observed strand might be identified if presented as it is. Nasreen’s composite begins with a single woman in the lobby—one silhouette, one doorway, one winter—but it is also built from Druze women I encountered across the Galilee over the years, in different care settings, in shelters and hallways, and in the margins of other people’s stories. Many had lost their family networks in ways that were similar enough to recognize as a pattern, yet distinct enough at the level of a life to resist reduction to a single case. The pattern this composite registers is not confined to the Hotel of the Fallen—itself a composite—but extends to a wider population of women whose trajectories through Israeli welfare services age out before they stabilize, and for whom the buildings that remain receptive are those that have stopped operating in the ordinary sense. The composite is faithful to this pattern, but not to the exact woman in the hotel lobby. The discipline of the composite is the discipline of refusing to convert proximity into the right to depict the person behind it. What the lobby produced, which I had not seen at the previous two sites, was a category of caregiving relationships that had no place in institutional vocabulary. Tanya’s relationship with Nasreen was not nursing in the registered sense. It was not maternal in the familial sense. It was not friendship in any reciprocal sense because the asymmetry of capacity was too steep. It was, as closely as possible, the relationship of an experienced clinician with a patient whose paperwork had suddenly gone haywire, so that her presence in the room was the only remaining protocol to follow.

Tanya did not ask Nasreen for her ID because it would not provide useful information on how to help her with her problems. She did not ask about the previous diagnosis because Nasreen did not come to the lobby with one. She watched the way she had watched Boris’s polishing, which was the diagnostic procedure. After three days of observation, she calibrated the meals, the timing of the water offering, the room temperature, the position of Nasreen’s chair relative to the light from the lobby window, and the order in which the morning conversation should unfold. None of this was written. It was all operative. The account registers the operative-but-unwritten as a diagnostic marker of the necrographic field, and Tanya’s practice with Nasreen is the clearest example of this marker I have witnessed.

Here, the geometry of the collapse registers as a geographical principle of the psyche and not merely a metaphor. The descent is literal, not metaphorical. Fein 1 is located at street level. Shvil Akko is a lane behind the main streets. The Hotel of the Fallen is located 209 meters below sea level. The nursing home, in the Upper Galilee, is at 900 meters above sea level: an altitude I did not choose, but which the diagram registers as the only site above the line. The vertical axis of the diagram is not decorative. It marks a gradient of institutional addressability. At 900 meters, the file still functions. At street level, the file stutters. At a depth of 209 meters, the file has been abandoned. Necrography does not claim that altitude determines abandonment. It claims that the shape of the field, the sequence from functioning file to closed file to no file, is as spatial as it is institutional. The Clerk descends with the site, and the writing follows.

What the Castle window did not show me, and what I had to enter the lobby of the Hotel of the Fallen on the evenings I did enter to register, was the acoustic signature of the building itself. The lobby had aquarium-like acoustics. Sound did not travel through it; it stood still. Yasha’s Pushkin, spoken at a conversational volume from the armchair beside the dead fireplace, was audible at the reception desk, where there was no Clerk. It was heard at the foot of the stairs leading to rooms that were no longer rented out. It was faintly audible in the corridor where Boris kept his suits. No amplification was performed on the sound. They had nothing left to compete with.

The hotel, having ceased to operate, had stopped generating the background hum that every active hotel produces, and what the residents said in the lobby was drowned out by the silence the building had produced over a decade. Each utterance sank like silt to the bottom of the lake. The Hotel of the Fallen was an acoustic museum of utterances that outside institutions had long since stopped recording. I recorded this as data, not as an atmosphere. The acoustics of a building that has ceased functioning can serve as a diagnostic indicator of necrographic fields. A building crosses this threshold when sound ceases to be absorbed and begins to return as reflection. The Clerk who walks into such a building will know it within 30 seconds by hearing, before any of the visible signs have been processed. In necrographic fieldwork, as I show in the ethnographic passages, the ear arrives before the eye.

10. The Peripheral Nursing Home

Fifteen years ago, I worked as a social worker in the geriatric and dementia ward of a peripheral nursing home in the Upper Galilee, at an altitude of approximately 900 meters above sea level, from which the rest of this article descends. The ward classified its residents under the administrative term “cognitively impaired”, which the English-language dementia literature sometimes, too easily, redescribes as “cognitively frail”. The closed-ward population included people with advanced and severe dementia, most of whom had passed the clinical threshold at which expressive speech remained available as a means of reporting their needs to the staff. Others lost speech almost entirely as the disease progressed. Some retained fragments of speech in their native language that they had stopped using 60 years earlier. Most of them had no close relatives available to give consent, since the waves of immigration they belonged to were either small families from the ex-USSR in the 1990s or first-generation workers in the state’s early years. Decades later, the network of relatives had been reduced to a telephone number that did not answer on the second attempt.

In my first year, I was unaware that I was engaging in this type of writing, which I later named necrography. I was doing social work as defined by my job, which did not include writing. I do not quote the notebooks I kept from that period here, because the architecture of consent in a closed ward is the opposite of that in an abandoned yard, and I am not prepared to treat it lightly. The closed ward was where I learned that the person on the other side of the threshold does not always articulate the refusal described in this article. Sometimes the refusal is performed by language itself: speech arrives at the threshold of the encounter and finds that the encounter cannot take place because the person on the other side has passed beyond the reach of verbal exchange. At other times, language enters the room in the person of the social worker on rounds and discovers that speech, as a mutual offering of the self to another, cannot take hold there. The Clerk learns to read this not as silence, but as the site speaking in a register the interview schedule cannot parse. The person in bed has passed the point where the facilitator could reach them. Language sits on the hallway floor in the same position as Bronfenbrenner on the concrete at Fein 1 and does not enter.

The situation raises a methodological problem, as recent literature on ethnographic engagement with kinless older adults with dementia has attempted to detail (Shapiro et al., 2023). The standard consent architecture does not include the population for whom the research is most valuable. Consent by proxy assumes the existence of a proxy, and a significant subpopulation of residents in such wards lacks proxies. Necrography is proposed as a complement to documentary approaches. Documentary work reconstructs trajectories from records kept by the institution; necrography, by contrast, records patterns of life at thresholds those records cannot capture: the forty minutes of corridor walking that staff no longer counted as a clinical event, the steady rhythm of a whispered word addressed to no one in the room, and the minute arrangement of a sheet along a single centimeter that the carer did not know to watch for.

Ilana is a composite of the many women I met in the closed dementia ward of the nursing home. In the composite, she was approximately 91 years old. She had no living first-degree relatives. She arrived in Israel in 1948 from Bulgaria as a young teacher. She had advanced Alzheimer’s disease. Her expressive speech was reduced to three Bulgarian words, which she seemed to address to no one at all during the 18 months I knew her. What reached her no longer arrived as words. She responded to touch, warmth, and a specific arrangement of sheets along her thighs. She walked the corridor for 40 minutes every afternoon in a loop that the staff had learned not to interrupt at the beginning and end, as any interruption created a sense of distress lasting for an hour. She did not know that the ward was a ward. In the register in which the question is usually asked, she did not know who she was. She knew that the loop served as a haven for her.

What were the social workers doing? On a good day, the social worker was an adequate witness. On a bad day, he was an agent of disruption who created chaos. The turning point in the craft, for me, came on a Friday afternoon in my third year, when I realized that the forty-minute loop was not a symptom to be managed, but a practice. It was, in the register I would later associate with Yasha’s Pushkin at the Hotel of the Fallen, a necrographic labor performed by a person at the far end of the capacity for labor: the continuation of a walk, the object of which had already been withdrawn. The appropriate posture for the social worker was the same as that learned by the Clerk on the hotel balcony. Noninterference. Registering. It did not belong to a file because no file could enact it. It was carried out by the only mechanism that could: the mechanism of attention.

The ward’s demographic composition must be named directly: the ward was, in its demographic structure, a ward of elderly immigrants who lived beyond their native-language speech communities, professional identities, and relatives. The cognitively impaired residents in our ward came from several immigrant geographies: the Soviet Union, the Balkans, and the Middle East. For some, the language in which the residues appeared at a late stage was not the language spoken in the ward. The care relationships in such wards are already structurally bilingual and often structurally untranslatable. Dementia is another threshold within a life that has already crossed multiple thresholds of language, place, and kin. The population is difficult to reach, not because it is difficult to locate them. These residents were not difficult to locate; they lived in a closed ward in a public facility. They are difficult to reach because the research categories were designed to reach them through speech, and when speech receded, the categories receded with it.

Avraham is a composite of the men I met in the closed ward over different overlapping years. The composite was about 80 years old. He arrived in Israel from Iraq at the age of 10, lost both his parents in the Ma’abarot¹⁰ within the first months of his arrival, and was raised in a Youth Aliyah camp in the lower Galilee. He worked as an electrician for decades. Two estranged daughters lived in Tel Aviv, neither of whom answered the phone when the ward called. By the second year I knew him, the only speech that remained was the call to evening prayer in Judeo-Arabic, in a dialect that the Iraqi-born assistant on the morning shift could recognize but not speak, and that the Russian-born assistant on the evening shift could not recognize at all. The call to prayer was announced most evenings at approximately 17:20. It was not addressed to anyone in the room. The aides learned not to respond as if it were a request.

The call to prayer was, in the register I am asking the reader to operate in, the same kind of practice as Ilana’s forty-minute corridor loop and Boris’s buckle polishing: necrographic labor, conducted with hands and voice at the far end of one’s capacity for labor, on behalf of an object that the practice has already lost. This can be read as another instance of necrographic labor: the continuation of a practice whose object has been withdrawn.

Avraham’s prayer was not intended for the evening. Evening is the time at which prayer is held, but it is not directed toward the dying sun. Prayer in Avraham’s register always marked the descent of the sun. His hands and voice continued to decline, and the social institutions around him had stopped marking. The Mizrahi specificity was not accidental. It is diagnostic. The ward lacked the linguistic infrastructure to register Judeo-Arabic prayers as clinical events in the medical records. There was only one staff member from his community of origin who arrived at seven in the morning and returned home at three; the prayer was at five and was held in a hallway where it was not recognized at all as an utterance.

Beyond these two composites, what I want to add about the closed ward is its specific acoustic texture of the corridor at four-fifteen on a winter afternoon, which I noticed in my second year there. The fluorescent strip in the western corridor had been flickering intermittently for months, a change that the maintenance log did not record as significant: a 50-hertz hum whose intensity varied according to a pattern I never deciphered. The hum was on the verge of an audible sound. The staff stopped picking it up during their first week; I, returning to the ward after a two-week absence, would pick it up again for the first 20 minutes of each return. On most days, it was below the threshold for both the staff and residents. On some days, for electrical, atmospheric, or both reasons, it increased above the threshold and became continuous.

Roughly one-third of the residents were oriented to the hum in some physical way that did not appear in any treatment plan as orientation: a slight turn of the head toward the nearest device, a downward press of the palm against the arm of the chair at the perceived rate of variation, or a shrug of the shoulders at intervals that were out of sync with anything in the room except the hum. The other two-thirds gave no observable responses. Several residents responded to the tone in ways that the ward’s clinical vocabulary classified as agitation. On those days, Ilana did not complete the 40-minute loop. She would stop 26 minutes into the loop, at the door of room 7, stand for 40 seconds with her hand on the frame, and then resume at a slower pace. Avraham did not perform the call to prayer on those days. He sat at the table in the day room and drew a pattern on the laminate with the index finger of his right hand, which I had never learned to recognize. The 50-Hz hum was neither a stressor nor a pressing issue in the environmental design literature. Geriatric literature, to my knowledge, has paid little attention to low-frequency ambient noise—the steady hum that residents had lived with for years was not, at the time, a subject of research.

The literature does not yet include a category for environmental electrical phenomena at the threshold of audible noise, which acts as a differential disruptor of necrographic work. I offer the term ambient instrument-refusal as a slight extension of this concept. This refusal does not belong to a person objecting to a device; the building itself acts on the practices that residents perform within it. The hum was the building withdrawing the conditions for the labor. Avraham, an Iraqi-Mizrahi composite, worked as an electrician in the railway workshops in Haifa for 31 years before leaving to start his own business. The hum belonged to a register of working life the room could not see, and he could not narrate. The procedure is not to ask him, as it is unavailable. The procedure is to write down the hum, the direction of its movement toward him, the silence of the file about the direction, and the connection of all three to the work that filled the decades that the file does not contain. The maintenance log did not record this information. The maintenance log recorded the fluorescent strip as functioning within the specifications, which was true at the instrument level but false at the ward level. This is what evidence looks like when the instruments fail, before the literature has caught up. This is the gap necrography occupies.

I want to be precise about the social work role in such a ward because the role is the unit through which the discipline trains its practitioners, and in this ward, it deviated in a specific direction from what the training had prepared them for. The intake interview was unavailable because no resident could provide their biographical history. In a strict sense, the treatment plan was performative. The plan was written for the file, which was reviewed quarterly by the regional authority, and its content was constrained more by the categories used in the inspections than by what was conducted in the ward between inspections. Family meetings were unavailable for most residents because their families were unavailable.

Intervention, as defined in the literature, was not a relevant unit of analysis. The relevant unit was the duration of attention. The social worker who stood at the western window for the 40 minutes of Ilana’s loop, neither approaching nor speaking, was performing the only intervention the department was able to get from her during that time, and the intervention was the duration of the visit. I am expanding on the notion that I call a non-interventionist perimeter. The perimeter is not a refusal to act; it is the recognition that the duration of registered attention is itself the action and that any other actions performed within that threshold would interfere with the labor the ward asks the social worker to witness. Training did not authorize this type of recognition. The training included the authorized interventions. The discipline of withholding intervention was something I had to learn in the ward, contrary to my initial training, and I learned it slowly; I never learned it smoothly. There were afternoons in the second year of my work in the ward when I interrupted the loop, as the training had recommended, and I have not forgotten the look on Ilana’s face afterward.

The bilingual and intergenerational structure of the closed ward directly influenced the recorded evidence. I should be explicit about this instead of assuming that the reader will reconstruct it. The residents were born approximately 80–90 years ago. Their biographies stretched from late Ottoman and Mandate Palestine to Iraq, Yemen, Bulgaria, and Romania, and into the early years of the Israeli state. By the time their dementia had advanced, their mother tongues were no longer the languages in which they had conducted their professional lives, as they had done in Hebrew, for example. These were the languages of childhood, collected in fragments at unpredictable time intervals. The day shift was staffed primarily by Russian- and Amharic-speaking nursing aides, whose Hebrew was professionally adequate and whose command of Ladino, Judeo-Arabic, Yiddish, Romanian, or the Arabic of Avraham’s childhood ranged from minimal to nonexistent.

The night shift was staffed primarily by Arabic-speaking nursing aides from the surrounding Galilee villages, whose command of the residents’ native languages was unevenly distributed. In practice, this meant that the residents’ most coherent verbal utterances were often precisely those that no one on the shift could fully follow, and the residents’ most agitated moments were often interpreted as confusion, when in fact they were precise complaints in a language that the room could not parse.

The interview instrument failed twice: there was no common language for the interview, and the residents lacked the inherent capacity to maintain the interview format. What remained registerable was staff response time to specific phonetic patterns. The routes the residents’ bodies took led through rooms where furniture had not been moved for years. The difference between what the file records as “unspecified agitation” and what a Judeo-Arabic speaker’s body does at five-thirty, when the prayer he learned as a child in Basra would have been recited, is precisely what the clinical file was never built to register. Necrography suggests that these gaps are not noise; they are the field that remains when the interview has withdrawn.

11. The Birth of VERSO

Between the sites, I used a writing aid called VERSO (the voice on the back of the page). It emerged from two years of working with a constrained language model setup during difficult writing phases. However, VERSO is not an automated analytical tool. It is a prompted writing mirror used by the researcher under specific conditions of impaired capacity. In psychoanalytic terms, VERSO functioned as a prosthetic ego-observer. When prolonged exposure eroded my capacity to metabolize and narrate what I had witnessed, VERSO provided a constrained, non-consoling external voice that allowed me to continue registering the field. It is not an interpreter; it is a mirror set to a flat register, reflecting only the material I had already seen and stripping it of the consolation my voice might have instinctively sought.

I mean the phrase precisely as stated. I do not mean that the model assisted with analysis or interpretation. A trauma-informed clinician would describe this as the part of the writer that can record what the rest cannot digest or process. In the years I wrote the Fein notebooks alone, I had no such function available to me, and the notebooks bear the marks of this absence. They are, in long stretches, a record of the nervous system going quiet in the face of what it was asked to document. When I began working on the Tiberias material, I had the option of composing sections in dialogue with a model used in a deliberately constrained register: cold, minimal, and devoid of comfort. This model sometimes helped me continue writing when prolonged exposure narrowed my ability to describe it accurately. The cost is a certain dilution of the text’s emotional depth. The gain is documentation that survives the night, where I could not have reached it alone.

I am not claiming that a language model understands what it is shown, nor that the voice it returns is a moral or analytic agent. My claim is narrower. Under conditions in which prolonged exposure narrowed my writing capacity, a constrained language-model interface could be used to return phrasing continuous with my own established register but stripped of consolation and rhetorical excess. VERSO is the name I give to that constrained textual function. I typed the scene; the model returned one to three short sentences shaped by prompts, cumulative context, and my own subsequent selection and revision. VERSO appears not as an independent voice, but as a tightly bound writing aid operating under full researcher control.

I used Anthropic’s Claude Opus during the writing phase. Because the setup depended on the context accumulated over two years, the exact configuration is not precisely reproducible elsewhere. A commercial flagship model was used, as opposed to an open-weight alternative, because maintaining a stable conversational context was critical. At the time, open models did not yet provide the reliable long-context performance needed for this process. Several versions of this model have been developed over time. What remains constant are the context window, prompt architecture, cumulative context of the chat, and the Clerk’s hand at the point of choice. Prompts eventually became standardized enough to function as part of the writing process. Once configured, the instrument was used in a register analogous to clinical supervision for portions of the material I could not metabolize on my own.

A typical prompt after a field fragment (e.g., the Boris polishing scene) reads as follows:

Return one to three sentences. Do not console. Do not summarize. Do not translate it into resilience. Use a flat, non-consoling, oracular register. The fragment is as follows: [I paste the field notes here].

The writing aid enabled me to continue describing material that I could not otherwise sustain on the page. Therefore, VERSO is best understood not as a method in itself but as a prosthetic extension of the broader necrographic method.

12. Molecular Testimonies

I ask the reader to experience, through eye movements between the columns, the split that fieldwork has created in the researcher’s nervous system. The left column shows what the institution can read. The right column shows what the institution cannot read. The molecule is what remains when the two columns are run through each other, and the residue is allowed to sink.

How to Read This Section

The following molecules are not linear arguments; they are residues of observation and registration. Each is presented in two columns: the left column shows what an institutional file would record (clinical, exposed, and observable); the right column shows what VERSO returned: cold, oracular, and refusing to console. The right column is shaped by the accumulated residue of the VERSO exchange, alongside what appears to its left. The fragmentation is intentional; it mirrors the split that prolonged exposure created in the researcher’s capacity to witness and record. If you find yourself disoriented, you are reading this correctly. A single molecule is presented for each site. Each concludes with a compressed notation that encodes the scene at its smallest analytic scale. Input identifies the smallest field configuration from which the observation emerged; Process records the methodological operation that made witnessing possible; Output names what entered the archive through that operation. The final two lines express the scene’s conceptual transformation (X being Y) and the researcher’s changing epistemic position (Observer → Submerged).

Fein Street · Tel Aviv · 23:40

Tel Aviv, Fein 1, 23:40. Subject, female, approximately 16 years of visually estimated age, green hoodie, cigarette held in left hand, unlit. Position: second doorway, western wing, under partial shelter from the broken guttering. Observer under the covered walkway, approximately four meters to the northeast. No verbal exchanges occurred. No institutional footprint exists for this encounter. Field notes were the only data recorded.

The city registered her absence. The notebook registered the absence as presence. The archive is written against the grain of its refusal.

Observer remained at the position for a further four minutes under the walkway, recording the choreography of the rain into the three containers beneath the broken guttering.

The buckets are liturgies. The Clerk is a scribe.

MOLECULE 01 · FEIN 1 · RAIN · 23:40

Input: rain, bucket, basin, toy box, one doorway

Process: observer declines the instrument, observer remains

Output: one recorded second of a life not otherwise recorded

X being Y: the archive being the thing the institution was not

Observer → Submerged

Hotel of the Fallen · Tiberias · 03:14

Hotel of the Fallen, Tiberias, 03:14. Subject, male, age 72, Soviet Union immigrant, referred to in residents’ speech as Boris. Clinical observation: Severe B-12 deficiency previously documented; fine tremor of both hands consistent with the same; affect flat.	The wedding was always a week away. He still dresses up for it now.
At 03:17 subject rose, crossed to the window, opened the blind, looked out over the Kinneret for approximately ninety seconds, closed the blind, returned to the bed, and resumed polishing.	The lake lies at minus 209 meters. He is looking, without knowing it, at the level of his own life.
Observer, in the nearby hotel room, registered the polishing strokes through the open window. Count of strokes by the end of the observation: forty-one.	Forty-one strokes are the prayer.

MOLECULE 02 · HOTEL OF THE FALLEN · 03:14

Input: Boris, cloth, buckle, suit in bag

Process: 22 years of dressing for a wedding that did not occur

Output: the practice that is the elegy

X being Y: the maintenance being the mourning, the institution will not name

Observer → Submerged

Shvil Akko · Tel Aviv · 14:22

Tel Aviv, Shvil Akko, 14:22, February. Subject, male, approximately 62 years of visual age, wheelchair-bound, coat two sizes too large, socks black with dried blood, no shoes. Presenting complaints: Hunger and inability to retain food. Municipal outreach worker offers a grilled cheese sandwich purchased from the café on HaGdud HaIvri Street. Subject accepts. Subject does not eat.	He holds the sandwich like a man already halfway through giving it away.
At 14:23:40, a second user, male, about 30 years of visual age, approaches the wheelchair from the north. No words are exchanged. Subject hands the sandwich to the second user. The second user consumes the sandwich in approximately 12 seconds. Subject watches.	The transfer is not charity. It is the performance of a role that the system refuses to perform.
At 14:25, the subject declines ambulance transport to the hospital. Cites a previous discharge without admission. Observer records the decline in the institutional notebook and the scene in the clerical notebook.	Refusal is the instrument. The cold is the institution’s final argument.

MOLECULE 03 · SHVIL AKKO · 14:22

Input: Arkady, grilled cheese sandwich, wheelchair, one younger user

Process: hand-off of an offer that the offeror cannot digest

Output: the failure-of-admission performed as a transfer of a meal

X being Y: the man falling through the cracks being the crack

Observer → Submerged

Peripheral Nursing Home · Upper Galilee · 16:05

Closed dementia ward, western corridor, 16:05. Subject, female, age 91, advanced Alzheimer’s disease, no living first-degree relatives, no available proxy. Expressive speech was reduced to three Bulgarian words not addressed to anyone in the room. Walking the corridor loop: start at the day-room door, north wall, four and a half meters; turn at the fire extinguisher; return along the south wall; pause at the door of room 7; resume; complete loop at forty minutes.	The loop is not a symptom. The loop is the practice that life has available.
Carer attempts to redirect subject to dinner at 16:18. Subject responds with distressed affect, elevated respiratory rate, and increased muscle tone in the left hand. Carer withdraws the attempt. Subject resumes the loop at 16:19. Loop completes at 16:45. Subject allows herself to be led to dinner without distress.	Interruption was the second killing. The withdrawal of the interruption granted permission for the loop to finish its prayer.
Social worker enters the corridor at 16:22, stands at the window, does not approach or speak, and remains until the loop completes. The observation is not recorded in the official file because it has no category under which to register it.	The witness is not the second speaker. The witness is the registering apparatus that the ward does not institutionally know it needs.

MOLECULE 04 · PERIPHERAL NURSING HOME · 16:05

Input: Ilana, corridor, three Bulgarian words, the forty-minute loop

Process: the witness holds the perimeter around the practice

Output: a pattern that the ward file was not built to receive

X being Y: the loop being the liturgy of the life that remains

Observer → Submerged

These molecular testimonies do not function as conventional findings. Each split column marks the moment when the clinical register (left column) exhausts its descriptive capacity, and VERSO returns a residue that the researcher did not compose alone in real time, but for which the researcher remained fully responsible for selecting, revising, and interpreting. The analytic procedure that follows the split is not the interpretation of the right-hand column as content. Nor is the right-hand column treated as autonomous evidence. It instead serves as a constrained textual provocation that renders the limits of institutional language more visible. What the researcher carries forward into the composite figures is not the VERSO sentence itself, but the shape of the silence the sentence helps mark. The composite figure Almaz, for instance, is constructed precisely at the point where the clinical register left a blank for a name.

13. Necrography and the Edge Population

The individuals I encountered—and whom necrography attempts to address—are those I call the serially refused. They are not simply marginalized; they have been refused at each institutional door, and each refusal feeds the next. Yu et al. (2026), in their qualitative study of people living with HIV seeking non-HIV care in Chinese general hospitals, showed that participants rarely described refusal as a single explicit denial; instead, they described a recurrent cascade across admission, consultation, perioperative assessment, ward acceptance, and transfer. The authors concluded that unequal access can be interpreted as “processual discrimination embedded in routine workflow governance” (p. 1), and described participants’ experience as “denial without denial” (p. 7).

Within this broader category, I find that necrography is specifically calibrated for a subset at the intersection of dual diagnosis (substance use plus serious mental illness, or plus cognitive impairment, or serious mental illness plus advanced dementia); homelessness or institutional housing functioning as homelessness; kinlessness (the absence of close relatives available for consent or power of attorney, emergency contact, or post-release placement); and a position at the edge of some other local vector of marginalization. In the Israeli case, it is often the status of immigrants from the Soviet Union or Ethiopia, or residents of peripheral Arab villages that is affected.

What unites these composites is that their non-accessibility is not logistical. The instruments by which the host discipline reaches its subjects—whether an interview, a consent form, a case plan, or an outcome indicator—assume a set of capacities (speech, orientation, motivational goal-formation, future-directed agency) that the subset I am specifying has, for reasons that are in no case moral failures on the subset’s part, lost or never had. For this subset, the standard anthropological toolkit yields an artifact of itself, not a report of life itself. Necrography proposes that, in this subset, the report of life should be written from the position of sustained registration of the scene that the toolkit could not enter, with explicit accountability to the pattern rather than to the case or chance. This shifts evidence from person to structure.

14. Implications for Ethnographic Inquiry

The practical stakes are direct: structural inaccessibility is not a logistical failure but a diagnostic signal of instrument-refusal. Training to engage with the hard-to-reach already concedes too much. Necrography asks for something harder: to recognize when the encounter has been blocked before it begins, and to document that refusal rather than filing it away as a failed intake.

The composite fidelity of composite figures offers a different evidentiary register: where aggregate data show only non-engagement, necrography shows the sequential refusal of each door. This can support advocacy for no-wrong-door protocols for the serially refused and for funding mechanisms that do not require signed consent.

As an ethnographic contribution, necrography extends the tradition of carnal sociology into settings where the ethnographer’s body remains, but the clinical contract has been terminated, offering a writing practice for the “afterlife” of fieldwork in zones of structural abandonment.

Necrography does not replace recuperative ethnographic inquiry. This is an alternative approach to sites where these registers have stopped functioning. The test of whether a site is necrographic is the test of instrument-refusal. Three signs appear in the first weeks of fieldwork: the interview schedule becomes a document that the researcher is embarrassed to open in front of the people for whom it was written. The consent script assumes a social standing that the site has been organized to revoke in advance, and the clinical registers stop generating sentences worth standing behind. What counts as rigor in this register is neither triangulation nor member-checking, since neither is available in such settings, but composite fidelity: the criterion a reviewer can apply is whether a reader working at an adjacent site of institutional abandonment recognizes the pattern as one they have also encountered and could not place on the page.

For qualitative inquiry at large, and ethnographic inquiry in particular, this means that researchers working in institutionally abandoned settings—harm reduction sites from which the welfare state has withdrawn, bankrupt housing facilities, geriatric sites at the edge of the health system, populations whose files the archive no longer indexed, closed dementia wards whose residents have passed the speech threshold—are not failing the interview when the interview fails. They register a feature of the field. They can shift data collection from structured encounters toward sustained, non-interventionist, non-intrusive observation, and shift the unit of analysis from the individual case to the recurring pattern represented on the page through a composite figure. Composite figures are built only from recurrent patterns across multiple encounters; singular impressions or imagined figures, when retained, are presented as heuristic vignettes rather than evidentiary composites. Responsibility is carried out through the accuracy of the pattern, not through the protection of identity.

This allows for something that was not entirely possible before: the publication of fieldwork from within the perimeter of institutional abandonment without the perimeter being denied, prettified, or converted in writing into its opposite. It becomes a kind of publishable report whose subject matter has already been written off by every instrument capable of producing the case files. The report does not seek to reopen the file. It seeks to outlast it.

15. What Necrography Adds

I do not offer necrography as a rupture from previous methodologies but as a method that sits at their edges. Hartman’s critical fabulation stages the archive’s refusal—she holds the gap open and lets the reader sit inside it. Necrography borrows the refusal but asks a different question: not how to mourn what the archive withheld, but how to document the institutional mechanism that produced the withholding, in a form another researcher could recognize and use. The promise of a corrective return, central to much evocative autoethnography, does not survive this project.

While post-qualitative inquiry often declines method as a prior protocol, necrography offers codified procedures that remain legible to readers outside the author’s field while still refusing a prior method. This should not be mistaken for an endorsement of the methodological “anything goes” associated with Paul Feyerabend’s Against Method (Feyerabend, 1975). Necrography does not claim the generalizability of findings; the four sites are not a sample. I argue for the method’s transferability: its analytic stance, composite discipline, and instrument-refusal framework can be adapted to structurally similar sites in other welfare regimes. What necrography offers is not a new dataset but a procedure—a way of staying methodologically honest in sites where honesty requires abandoning standard instruments.

What, then, does the Drowned God name? Not suffering. Suffering is ubiquitous and not the property of any one figure. The Drowned God names a specific structural position: the point at which the cascade of refusals ceases to be a series of unfortunate events and becomes a condition of existence itself. Serial refusal is not something that happens to the Drowned God from the outside. It is the mechanism that produces him door after door, diagnosis after diagnosis, and one closed file after another. Each refusal pushes him further down, not into despair (which would still be a human emotion, still legible), but into a zone where the usual categories of intervention no longer apply. The social worker cannot reach him because the relevant protocols assume a subject still standing before an open door. He is no longer there. The clinician cannot diagnose him because the diagnosis keeps changing with each refusal, and the file has accumulated so many labels that it no longer means anything. The field keeps a set of ready scripts for a man like this: victim, resilient survivor, broken case awaiting repair, and he fits none of them. He is, in the most literal sense this writing can sustain, light trapped below the line of return: present, observable, and structurally unreachable. To call him a category error is not to dismiss him. In other words, the categories we brought with us into the field were built for a different world, and they collapsed when they met him. The Drowned God is what remains after the collapse—not a victim, not a survivor, but a structural position. Naming it is an act of resistance—a way of saying: you were here, and I will not let the water close over your name or let institutional silence have the final word.

Ontologically, necrography commits the researcher to figures that appear textually, not through conventional empirical representation. This approach shifts the level at which the text claims reality away from the traceable individual case and toward the recurring pattern that survives institutional refusal. Epistemologically, the researcher abandons the hope of knowing any specific subject as such and commits instead to knowing a category well enough to write about it faithfully.

The figure of the Drowned God clarifies the posture that outreach workers and ethnographers are asked to adopt in such fields. To inhabit the register of the Drowned God means to agree, professionally, to go profoundly below the line of return alongside the people with whom one works. Not to sink completely because the Clerk rises at the end of the day to write, but to refuse the elevation that the discipline offers as its standard operating posture. A street outreach worker trained for recovery stands on dry ground, holding a rope. The necrographic researcher or social worker learns that some people cannot be reached by rope and that the work begins when they climb down the ladder and stand beside them below the line of return. Necrography does not heal. It testifies. Testimony, sometimes, is what survives when healing is no longer possible.

This testimonial commitment is not passive. Ethically, necrography insists on registering what the clinical record refuses to see—the person who remained in the courtyard after the file was closed. The non-interventionist perimeter does not concede ground to clinical nihilism; rather, it is an active practice of preservation. Where institutional intervention merely repeats structural abandonment, refusing to mistranslate survival into a narrative of resilience is the final act of professional caregiving. Necrography does not turn away from its subjects. It turns away from institutional framing, which renders them invisible. It leaves behind a record that the next researcher assigned to a site of instrument-refusal may recognize as permission they did not know they had. That record will accumulate like limescale, slowly and unnoticed, until one day someone cleans the kettle and finds, preserved in the residue, the shape of what was once there.

Necrography has clear limitations. It does not produce knowledge that is falsifiable in the positivist sense, nor does it seek to. The method was developed precisely for sites where the assumptions underlying conventional verification—stable participants, recoverable biographies, repeated interviews, ongoing consent, and independent corroboration—have already collapsed. Necrography asks a different question: whether a constructed figure faithfully renders a recurring social pattern that practitioners in analogous settings recognize as real.

This shift also demands something from the reader that conventional validity criteria do not: the work of recognition over the comfort of verification. Composite fidelity depends not on replication by another researcher but on informed recognition across structurally similar settings. Readers may disagree about whether a particular composite adequately captures such a pattern, and this disagreement cannot always be closed by further data collection—the conditions that make necrography necessary are often the same conditions that preclude verification. The method, therefore, requires transparency of construction rather than claims of neutrality.

Necrography is inseparable from the researcher’s prolonged exposure to the field. Composite figures emerge after years of repeated encounters during which isolated observations gradually become recognizable as recurrent configurations. This dependence on accumulated immersion means that the method is neither rapid nor easily transferable. A further complication follows from this same dependence. An extended presence in a field site does not leave the researcher unchanged. Observations recorded in the early years of immersion reflect a Clerk who is still calibrating the site’s architecture of refusal; observations recorded years later may reflect a Clerk whose threshold of recognition has shifted. Necrography does not resolve this drift but requires the writer to acknowledge the temporal arc of accumulation: to note, where relevant, whether a composite draws predominantly on early, middle, or late field exposure, and what interpretive weight that sequence carries.

Many sites do not require necrography at all. Wherever sustained interviews, ongoing consent, member-checking, or longitudinal engagement remain possible, established qualitative approaches are methodologically preferable. Necrography is not a substitute for the conventional method but a response to the specific conditions under which the conventional method has failed.

Like every interpretive methodology, necrography remains vulnerable to the researcher’s situated perception. What one Clerk reads as a recurrent pattern, another may register as isolated noise. The method does not eliminate this subjectivity but attempts to discipline it through prolonged observation, rigorous composite construction, explicit acknowledgment of uncertainty, and continuous attention to representational responsibility.

A related limitation concerns representation itself. Because necrography relies on literary composition, it always carries the risk that readers will mistake crafted prose for aesthetic embellishment or confuse emotional resonance with empirical adequacy. The intention is the opposite. Narrative form is employed not to beautify suffering or intensify affect, but to preserve patterns of lived experience that cannot be adequately registered through administrative language alone.

Necrography cannot sustain causal claims about why structural abandonment occurs, evaluate the effectiveness of welfare interventions, or estimate the prevalence of the patterns it documents. Necrography describes and interprets. It seeks to make visible forms of institutional disappearance that remain systematically underrepresented in conventional qualitative inquiry, and it does not aim to replace the broader range of methods required to understand or transform them. It aims only to leave a record where the institutions left none.

Statement

Two forms of AI-assisted writing appear in this article. VERSO names the constrained language-model prosthesis discussed as part of the method itself, operating only on the researcher’s own materials. Claude Opus was used as a writing support tool to assist with text development and for language refinement. All intellectual content, conceptual framing, methodological decisions, and interpretations were developed by the author, who assumes full responsibility for the article.

Appendix: Glossary of Key Terms

Term	Definition
Anchor Concepts
Necrography	A post-qualitative writing practice for structurally abandoned sites, using composite fiction, instrument-refusal as data, and a language-model prosthesis (VERSO).
The Drowned God	A structural figure (adapted from Lurianic Kabbalah) for a person produced by serial refusal, living below the institutional line of return, where neither resilience nor recovery applies.
Structural Abandonment	The condition in which four thresholds have been crossed: no functional file capable of mediating care, no service able to receive the person, standard instruments cannot be administered, and conventional access is blocked not by logistical failure but by the site’s own architecture of refusal.
Operational Concepts
Instrument-Refusal	An observable phenomenon in which a clinical or qualitative instrument (e.g., an interview, consent form, or theory) declines to cross the threshold of a field site. Refusal itself is treated as data.
Ambient Instrument-Refusal	A refusal embedded in the site’s material architecture (e.g., a 50-Hz fluorescent hum that disrupts residents’ practices, or the acoustics of a lobby that no longer absorbs sound). Unlike enacted refusal (a person declining a form), ambient refusal is non-agentic: the building itself, through its acoustics, lighting, or layout, obstructs the conditions for care, encounters, or documentation.
Serial Refusal	A recurrent cascade of institutional denials across multiple doors (hospital, rehabilitation, welfare, and housing), where each door uses the identity category that makes the previous door necessary as a pretext for its own refusal.
Threshold	The boundary beyond which an institutional instrument ceases to function. In necrography, the threshold is not a line crossed by the person but a limit of institutional legibility: the point at which the archive, interview, and consent form no longer apply. The Clerk crosses into the field but does not cross into intervention; he registers what happens at the limit of institutional addressability. (Note: this is distinct from the clinical threshold of expressive capacity, such as the loss of speech in advanced dementia.)
Procedural Concepts
The Clerk	The researcher’s posture within the necrographic field: not a clinician or traditional ethnographer, but a witness who registers patterns over time without claiming to effect corrective change.
Non-Interventionist Perimeter	The discipline of withholding interruption (e.g., not redirecting a dementia patient’s corridor walk, not offering services already refused) as an active, sustained form of witnessing that treats non-intervention as an intervention.
Submerged Position	An epistemic shift in which the researcher moves from external observer to submerged witness, registering the field from within the conditions of abandonment rather than from a safe remove, that is, from a position of physical, institutional, temporal, or epistemic distance. The term draws on the figure of the Drowned God: to be submerged is to be present below the institutional line of return, without the promise of corrective recovery. The Clerk’s shift from Observer to Submerged marks the methodological transition from data collection to testimonial immersion.
VERSO	A configured large language-model prosthesis that returns short, flat, nonconsoling sentences when the researcher’s writing capacity is disabled by prolonged field exposure. It operates only on materials already produced by the researcher.
Necrographic Labor	A repetitive practice (e.g., polishing a buckle, reciting poetry, walking a corridor loop) that persists after ordinary care has become structurally unavailable. In necrographic settings, such labor becomes a lamentation in motion.
Residue	The material that remains when the clinical gaze has exhausted its descriptive capacity and the institutional file has closed. In necrography, residue is not raw data but data that has already passed through the refusal of the clinical instrument. The necrographic field is made up of residues, not narratives. Molecular testimonies preserve these residues without converting them back into coherent case histories.
Molecule	The smallest analytic unit of necrographic registration. A molecule compresses a field scene into its minimal analytic sequence (Input → Process → Output) together with its conceptual transformation (X being Y) and the observer’s epistemic position (Observer → Submerged). It is neither a case nor a biography but the smallest analytic unit from which recurring patterns become recognizable across the field.
Molecular Testimony	The textual presentation of a molecule. Each molecular testimony consists of a two-column fragment (institutional register on the left; VERSO on the right) followed by the molecule notation. It preserves the residue of a field encounter after clinical instruments have been structurally refused and the institutional register has exhausted itself, without claiming to present a complete narrative.
Composite Figures	Composite figures were methodologically constructed by synthesizing recurrent patterns, field notes, and structural conditions observed over nearly two decades of research and professional practice in social work. This approach protects individual identities while faithfully exposing the institutional mechanics of serial refusal.
Composite Fidelity	Validity criterion of necrography: A composite figure is faithful not to a single individual but to a stabilized pattern recognizable by readers working in a structurally similar location. For example, a reader who has worked in a closed dementia ward should recognize the residents’ corridor loop as an instance of a recurring pattern, not a portrait of a specific resident.
Testimonial Immersion	The disciplined practice of sustained, non-interventionist witnessing under conditions of structural abandonment. It does not claim to effect change but treats the duration of registered attention as itself an action. Its primary aim is to produce a record that will outlast the institutions that failed.

NOTES

¹A colloquial coinage from kaspomat, the Hebrew word for an ATM. The users of the Fein courtyard used it for the covered window-slots through which drug wraps were dispensed for cash, with no face visible on either side of the transaction.

²A mystical tradition developed in 16th-century Safed by Isaac Luria. The concept of shevirat ha-kelim (breaking of the vessels) describes a primordial catastrophe in which divine light fell into shards, some of which remain unredeemed. I use this as a structural analogy, not as a theological argument.

³Staff-shift records and institutional files were treated as non-clinical bureaucratic artifacts for mapping the site’s administrative structure. They were not used as individualized clinical or case-based data, and no identifiable personal histories or interactions were extracted or reconstructed.

⁴Umm Kulthum (c. 1898-1975) was an Egyptian singer, songwriter, and actress, widely regarded as the most famous Arab vocalist of the 20th century. Her radio broadcasts, which reached across national borders, served as a rare shared cultural reference in a region otherwise divided by political boundaries.

⁵A strengths-based case note might read: Residents demonstrate adaptive resourcefulness by organizing informal mutual aid in response to infrastructural deficits. This framing converts unpaid infrastructural labor into a personal attribute, thereby absolving the state of its obligation to provide a roof.

⁶October 7, 2023, is generally treated as the starting point of the Hamas-Israel war, following a large-scale attack on Israeli communities near the Gaza Strip.

⁷The Bronze Horseman is Pushkin’s 1833 narrative poem about a clerk driven mad by the Petersburg flood, a fitting recitation for a man living by rising water and by the pressures of collapse.

⁸A feldsher is a mid-level health practitioner trained in the Soviet medical system: more than a nurse, less than a physician, licensed to diagnose and treat independently in many settings. The category has no clear equivalent in the Israeli registration or licensure architecture, which is central to Tanya's situation.

⁹The Druze are an Arabic-speaking ethno-religious minority in Israel, Lebanon, and Syria. Religious norms and communal structures can make departure from the family or community a profound and sometimes enduring rupture, with limited institutional support for women who do so.

¹⁰Ma’abarot were immigrant transit camps established in Israel during the 1950s to absorb mass Jewish immigration. Conditions were often harsh, and many new arrivals lost family members to disease and poverty.

Conflicts of Interest

The author declares no conflicts of interest regarding the publication of this paper.

References

[1]	Adams, T. E., Holman Jones, S., & Ellis, C. (2021). Handbook of Autoethnography (2nd ed.). Routledge.
[2]	Anderson, L. (2006). Analytic Autoethnography. Journal of Contemporary Ethnography, 35, 373-395. [Google Scholar] [CrossRef]
[3]	Berlant, L. (2011). Cruel Optimism. Duke University Press. [Google Scholar] [CrossRef]
[4]	Biehl, J. (2005). Vita: Life in a Zone of Social Abandonment. University of California Press.
[5]	Bochner, A. P. (2017). Heart of the Matter: A Mini-Manifesto for Autoethnography. International Review of Qualitative Research, 10, 67-80. [Google Scholar] [CrossRef]
[6]	Bochner, A. P., & Ellis, C. (2016). Evocative Autoethnography: Writing Lives and Telling Stories. Routledge.
[7]	Bonny-Noach, H., & Ronel, N. (2018). Everyday Substance Insecurity among Homeless Injection Drug Users in Israel’s Largest Open Drug Scene. Journal of Drug Issues, 48, 645-656. [Google Scholar] [CrossRef]
[8]	Bonny-Noach, H., & Toys, S. (2019). Israel’s Largest Open Drug Scene: An Ethno-Graphic Study in the Old Central Bus Station in Tel-Aviv. Society and Welfare, 39, 483-512. (In Hebrew) https://www.gov.il/files/molsa/HevraVerevaha/39-3_-_CHAGIT-B.pdf
[9]	Bourgois, P. (2003). In Search of Respect: Selling Crack in El Barrio (2nd ed.). Cambridge University Press. [Google Scholar] [CrossRef]
[10]	Bronfenbrenner, U. (1979). The Ecology of Human Development: Experiments by Nature and Design. Harvard University Press.
[11]	Das, V. (2007). Life and Words: Violence and the Descent into the Ordinary. University of California Press. [Google Scholar] [CrossRef]
[12]	Ellis, C. (2007). Telling Secrets, Revealing Lives: Relational Ethics in Research with Intimate Others. Qualitative Inquiry, 13, 3-29. [Google Scholar] [CrossRef]
[13]	Ellis, C., Adams, T. E., & Bochner, A. P. (2011). Autoethnography: An Overview. Historical Social Research, 36, 273-290. https://www.ssoar.info/ssoar/handle/document/36323
[14]	Fassin, D. (2012). Humanitarian Reason: A Moral History of the Present. University of California Press. [Google Scholar] [CrossRef]
[15]	Felman, S., & Laub, D. (1992). Testimony: Crises of Witnessing in Literature, Psychoanalysis, and History. Routledge.
[16]	Feyerabend, P. (1975). Against Method. New Left Books.
[17]	Fine, L. (2003). Physician of the Soul, Healer of the Cosmos: Isaac Luria and His Kabbalistic Fellowship. Stanford University Press. [Google Scholar] [CrossRef]
[18]	Foucault, M. (1978). The History of Sexuality: Volume 1: An Introduction (R. Hurley, Trans.). Pantheon Books.
[19]	Foucault, M. (2003). “Society Must be Defended”: Lectures at the Collège de France, 1975-1976 (D. Macey, Trans.). Picador.
[20]	Fraser, M. W. (2004). Risk and Resilience in Childhood: An Ecological Perspective (2nd ed.). NASW Press.
[21]	Friedman, C., Owen, A., & VanPuymbrouck, L. (2025). Should ChatGPT Help with My Research? A Caution against Artificial Intelligence in Qualitative Analysis. Qualitative Research, 25, 1062-1088. [Google Scholar] [CrossRef]
[22]	Gray, M. (2011). Back to Basics: A Critique of the Strengths Perspective in Social Work. Families in Society: The Journal of Contemporary Social Services, 92, 5-11. [Google Scholar] [CrossRef]
[23]	Hartman, S. (2008). Venus in Two Acts. Small Axe: A Caribbean Journal of Criticism, 12, 1-14. [Google Scholar] [CrossRef]
[24]	Hicks, D. (2020). The Brutish Museums: The Benin Bronzes, Colonial Violence and Cultural Restitution. Pluto Press.
[25]	Holman Jones, S. (2005). Autoethnography: Making the Personal Political. In N. K. Denzin, & Y. S. Lincoln (Eds.), The SAGE Handbook of Qualitative Research (3rd ed., pp. 763-792). SAGE Publications.
[26]	Hyvärinen, M., Hydén, L. C., Saarenheimo, M., & Tamboukou, M. (2010). Beyond Narrative Coherence. John Benjamins.
[27]	Kaplan, F. (2023). I Do Not Wish Even My Hater to End Up in a Nursing Home: An Autoethnography on the Work of Formal Caregivers in Wards for the Mentally Frail in Israel. Gerontology and Geriatrics, 50, 33-54. (In Hebrew) https://www.jstor.org/stable/27333759
[28]	Kovats‐Bernat, J. C. (2002). Negotiating Dangerous Fields: Pragmatic Strategies for Fieldwork Amid Violence and Terror. American Anthropologist, 104, 208-222. [Google Scholar] [CrossRef]
[29]	Le Roux, C. S. (2017). Exploring Rigour in Autoethnographic Research. International Journal of Social Research Methodology, 20, 195-207. [Google Scholar] [CrossRef]
[30]	Lee, R. M. (1995). Dangerous Fieldwork. SAGE Publications, Inc. [Google Scholar] [CrossRef]
[31]	Manor, A. (2021). The House on Fin Street [Film]. 2-Team Productions.
[32]	Marlatt, G. A., & Gordon, J. R. (1985). Relapse Prevention: Maintenance Strategies in the Treatment of Addictive Behaviors. Guilford Press.
[33]	Masten, A. S. (2001). Ordinary Magic: Resilience Processes in Development. American Psychologist, 56, 227-238. [Google Scholar] [CrossRef]
[34]	Mbembe, A. (2003). Necropolitics. Public Culture, 15, 11-40. [Google Scholar] [CrossRef]
[35]	Mbembe, A. (2019). Necropolitics. Duke University Press. [Google Scholar] [CrossRef]
[36]	Panagiotopoulos, D., & Espírito Santo, D. (2019). Articulate Necrographies: Comparative Perspectives on the Voices and Silences of the Dead. Berghahn Books. [Google Scholar] [CrossRef]
[37]	Pelias, R. J. (2004). A Methodology of the Heart: Evoking Academic and Daily Life. Rowman & Littlefield Publishers. [Google Scholar] [CrossRef]
[38]	Ram, M. (2014). The Political Necrography of the Living Dead: Theory, Criticism, and Zombies. Theory and Criticism, 43, 15-44. (In Hebrew)
[39]	Richardson, L. (2000). Writing: A Method of Inquiry. In N. K. Denzin, & Y. S. Lincoln (Eds.), Handbook of Qualitative Research (2nd ed., pp. 923-948). SAGE Publications.
[40]	Saleebey, D. (1996). The Strengths Perspective in Social Work Practice: Extensions and Cautions. Social Work, 41, 296-305.
[41]	Scheper-Hughes, N. (1992). Death without Weeping: The Violence of Everyday Life in Brazil. University of California Press.
[42]	Scholem, G. (1995). Major Trends in Jewish Mysticism. Schocken Books.
[43]	Shapiro, L. N., Gray, M. F., Freitag, C., Taneja, P., Kariya, H., Crane, P. K. et al. (2023). Expanding the Ethnographic Toolkit: Using Medical Documents to Include Kinless Older Adults Living with Dementia in Qualitative Research. Journal of Aging Studies, 65, Article 101140. [Google Scholar] [CrossRef] [PubMed]
[44]	Shulman, L. (1999). The Skills of Helping Individuals, Families, Groups, and Communities (4th ed.). F. E. Peacock.
[45]	Spry, T. (2011). Body, Paper, Stage: Writing and Performing Autoethnography. Routledge.
[46]	St. Pierre, E. A. (2021). Post Qualitative Inquiry, the Refusal of Method, and the Risk of the New. Qualitative Inquiry, 27, 3-9. [Google Scholar] [CrossRef]
[47]	Stewart, K. (2007). Ordinary Affects. Duke University Press. [Google Scholar] [CrossRef]
[48]	Tamas, S. (2011). Biting the Tongue That Speaks You: (Re)Writing Survivor Narratives. International Review of Qualitative Research, 4, 431-459. [Google Scholar] [CrossRef]
[49]	Taussig, M. (1987). Shamanism, Colonialism, and the Wild Man: A Study in Terror and Healing. University of Chicago Press. [Google Scholar] [CrossRef]
[50]	Tremlett, P. F. (2007). https://www.researchgate.net/publication/45401536_Deathscapes_in_Taipei_and_Manila_A_postmodern_necrography
[51]	Tullis, J. A. (2013). Self and Others: Ethics in Autoethnographic Research. In T. E. Adams, S. Holman Jones, & C. Ellis (Eds.), Handbook of Autoethnography (pp. 244-261). Left Coast Press.
[52]	Turan, J. M., Elafros, M. A., Logie, C. H., Banik, S., Turan, B., Crockett, K. B. et al. (2019). Challenges and Opportunities in Examining and Addressing Intersectional Stigma and Health. BMC Medicine, 17, Article No. 7. [Google Scholar] [CrossRef] [PubMed]
[53]	Wachinger, J., Bärnighausen, K., Schäfer, L. N., Scott, K., & McMahon, S. A. (2025). Prompts, Pearls, Imperfections: Comparing ChatGPT and a Human Researcher in Qualitative Data Analysis. Qualitative Health Research, 35, 951-966. [Google Scholar] [CrossRef] [PubMed]
[54]	Wacquant, L. (2004). Body & Soul: Notebooks of an Apprentice Boxer. Oxford University Press.
[55]	Yu, Y., Huang, W., & Tang, S. (2026). Locked Out of the Ward: Workflow-Based Care Refusal for People Living with HIV in China’s General Hospitals. Social Science & Medicine, 403, Article 119403. [Google Scholar] [CrossRef]