Dementia in India: A Reanalysis of Cohen and Brijnath

I. Two Records of One Kind of Mesocosm

The two book-length ethnographies that have shaped the anthropology of dementia in India are a decade and a half apart. Lawrence Cohen’s No Aging in India (1998) rests on fieldwork in Varanasi between 1988 and 1990, extended to clinics, old age homes, ashrams, police bureaux, a pharmaceutical firm and archives across India and beyond. Bianca Brijnath’s Unforgotten (2014) rests on ten months in Delhi in 2008, during which she conducted forty-six interviews with twenty families caring for a relative with a diagnosis of dementia (Brijnath 2014: 23). Brijnath calls Cohen’s book “exemplary” and notes that no ethnography of dementia care in India had followed it (2). She also records that sathiyana, the “going sixtyish” at the centre of Cohen’s material, “was hardly mentioned” by her interlocutors (8). The two books are therefore usually read as a sequence, from senility heard through idioms of heat, weakness and the Bad Family to dementia handled through diagnoses, attendants, medicines and love.

This article rereads both books through Ontologically Grounded Theory (OGT), which treats a qualitative record as a selective multisymbolic deposit of living coordination. OGT starts from the ontology of Living Value Theory, whose generative principle is that life is recursive mediation in a mesocosm. Human living is coordinated through five irreducible mediations: multisensorial embodiment, multispecies being-with, multiversal dwelling, multimateriality and multisymbolization. It proceeds with counterparts whose aspects are nonrecursive, selfrecursive or interrecursive, and it can move across five levels of recursivity, from coordination that needs no attention (L1), through felt salience (L2), articulation (L3) and stabilized categories (L4), to the repositioning of those categories (L5). Applied to a single study, OGT asks what the record preserves, what it leaves unspecified and where its concepts cease to track what they describe. Applied to two studies, it adds a prior question. Before any difference between the books is read as a difference between Varanasi in 1989 and Delhi in 2008, it must be established whether the two records could have registered the same things at all.

The argument is that Cohen and Brijnath built different instruments. Cohen’s was tuned to voice: to how old people were heard, by whom, and with what consequences for the moral standing of their families. Brijnath’s was tuned to handling: to the work through which families fed, cleaned, medicated, presented and protected a diagnosed relative. Sorted with care, the differences between the books turn out to be more often differences of instrument than of mesocosm, and several of the apparent changes dissolve into continuities registered under different names. What both instruments missed, though tuned so differently, is the most telling result: the absorbed coordination of the person living with dementia across whole days and nights, and the found conditions of heat, darkness, season and distance within which households care.

II. Two Instruments

Cohen describes his fieldwork as organized in “three overlapping spheres”: a study of old people and their families in several neighbourhoods stratified by class, a study of institutions for widows, renunciates and the dying, and a study of medical, religious, legal and political practitioners (Cohen 1998: 41). The four core neighbourhoods were Nagwa Harijan basti, a Chamar settlement south of Assi Ghat; the Bengali quarter in the old city; and the middle-class colonies of Ravindrapuri and Nandanagar. Entry into the record depended on residence, friendship and the referrals of neighbours, and no diagnosis was required. Cohen interviewed in seventy-five households in Nagwa alone, where he counted “at most a dozen” persons of seventy-five or older (238). He reconstructed ages through questions about floods, marriages and Partition (239) and wrote down remembered conversations each evening (10). His medical training gave him a clinical eye; his own position, as he notes, also made him “another motevala”, a man of means who came to take something (232). In the colonies, the instrument met resistance. Younger family members “were rarely willing to have me meet or even see the old relative”, and most old people he was sent to ask about “were asleep or too old or too sick to meet me” (190–191). The signal that this instrument registered best was voice: anger, repetition, silence, and the ways listeners heard them.

Brijnath’s families were recruited through the Delhi chapter of the Alzheimer’s and Related Disorders Society of India and two public hospital outpatient departments: thirteen through home visits with the society’s volunteers, four through its functions and three through the clinics (Brijnath 2014: 22–23). Every family had a diagnosed member; fifteen were middle class and five affluent (23). Interviews lasted about an hour, visits two to four hours, and she reports some 250 hours of observation in all (14, 27). Her arrival in a household was an occasion. People with dementia “were bathed, powdered, perfumed and dressed in their good clothes before I had arrived”, and families asked her at once how the person seemed to her (31). She was persistently taken for a doctor, and in time she accepted the role, accompanying families to clinics and prompting medication reviews (27–28, 31). A carefully designed consent protocol for people with dementia “was ineffective and never used”; instead, she observes, they “consented or rescinded in their own fashion”, verbally or through their bodies (31). The signal this instrument registered best was carework: feeding, cleaning, medicating and the presentation of the person to the world.

Table 1. The two research instruments

III. What Each Book Discovered

Cohen’s achievement is to have shown that, across the Varanasi neighbourhoods, change in an old person was heard before it was seen. Voices were described “most often as angry, as repetitive, or as the nonvoice of silence, sadness, or loneliness”, each with a positive and a negative moral reading (Cohen 1998: 175), and whether an old voice counted as wisdom or as bakbak depended “frequently on the politics of the hearing” (176). His concept of the familial body records that “the old body is not only embedded in its family but is, in a very real sense, appropriated by and becomes its family”, so that it “was not stigmatized as ‘itself’ but as the body of the family” (177). From this follows the explanatory weight of the Bad Family, articulated in Nagwa as weakness measured in chapatis, “She is weak because we are weak” (230), and in the colonies as tension, balance and adjustment (193–196). His concept of the dying space captures how a weak old age became a space of withdrawal at the threshold of the household (181–186). Around these findings he assembled others that remain striking. Of six patients referred as potentially demented at the Banaras Hindu University psychiatric clinic, “none could be diagnosed as demented by DSM-III-R criteria” (214); psychiatrists at a national institute found that “dead persons were more amenable than live ones to the elicitation of senile pathology” (39); and a pharmaceutical director launching an antidementia drug told him, “Frankly, we don’t want doctors to do these tests” (217). His account of Somita Ray in Calcutta shows the diagnosis at work. A long conflict between Somita and her daughter-in-law was relocated, through a psychiatrist recently visited by drug representatives and a CT scan that “allows for the possibility of Alzheimer’s”, into Somita’s brain, which relieved the daughter-in-law of blame without securing Somita a place in any institution: “No one here cares about Alzheimer’s” (299).

Brijnath’s achievement is the most intimate record of dementia care in India yet published. She follows diagnosis as a process of seeing, showing and being seen; medication as hope, side effect and discipline; the economy of care in medicines, diapers and attendants; food as seva, discipline and pleasure; stigma as accumulated small slights; and death, which for her families mostly came in hospital. Her governing theme is love. Her account of diagnosis is especially acute on language. Families who could not read a scan relied on the doctor’s identification of its “dark spots”, and took up neurological vocabulary in code-switched form, “cell-vell weak hai” (48); chronic illness was spoken of in a present tense that fused person and condition, “mental hai, sugar hai” (142). Her most consequential finding is that no family she worked with considered letting a relative who had stopped eating die without artificial feeding, and that every family whose relative reached the late stages chose the nasogastric tube (129–130). Shivbaksh Chand, the husband who had become his household’s cook, put it plainly: “Whether you understand this as seva or as her being with me, being together, I want her to eat” (130).

Both books record far more than their arguments use, and in both the surplus lies chiefly in passing descriptions, asides and accounts of method. That is where the reanalysis concentrates.

IV. Two Fields of Visibility

Mapped separately, the two records show the skew that OGT expects of most qualitative evidence. Multisymbolization dominates both, being-with is strongly represented, and dwelling is barely investigated. The expected pattern is less informative than the particular shape each record gives it.

Cohen’s record is overwhelmingly a record of reception. It documents in great detail how listeners heard old people, which categories they used and what was at stake for them in doing so. That is a record of the listeners’ recursivity, of what became salient to them (L2) and how they stabilized it (L4). It documents the recursive organization of the old people themselves much less, and the gap matters most in the colonies, where Cohen was often kept from meeting the person concerned. Brijnath’s record is far denser on embodiment, but the bodies it documents are mostly bodies being handled. Excretion, swallowing, bathing, turning and medication side effects are recorded as the content of carers’ work and as the source of carers’ pain, while the selfrecursive organization of the person with dementia, appetite, fatigue, sleep, ease and discomfort, enters incidentally. Lakshmi Kumari Kochar, aged ninety-two, claimed to be eighteen, fifty and eighty on different visits, and “as her age increased, so did her appetite” (Brijnath 2014: 121). The record notes the coupling and moves on.

Being-with is strongly represented in both, in different registers. Cohen documents being-with chiefly as reception, and with great subtlety. When Harinath Prasad, a retired schoolteacher in Nagwa, angrily made a drunken tea-stall owner tally the days of a leap year, the bystanders insisted that he had been sahi, correct, and the stall owner, who had all but abandoned his own parents, said: “I am his son. He was teaching me” (Cohen 1998: 175). The record also contains a sustained case of multispecies being-with that the argument reads chiefly as symbolism. Mashima, a Bengali widow in her seventies, walked the lanes each morning with chapatis for a pack of dogs that “growled furiously at any passerby who appeared in their dog’s-eye view to threaten Mashima”; “Dogs, Mashima told me, were her children” (263–264). Her daily route was a coordination with responsive nonhuman others that kept her mobile, linked her to the households that gave her chapatis and protected her in the lanes. Brijnath documents being-with as love and seva, and also as co-presence that asks little of anyone. At a respite centre in Guruvayoor, men lay back on the porch to be shaved, eyes closed, “jugulars exposed”, and even noisy residents fell quiet (Brijnath 2014: 166); she borrows the phrase “being alone together” for the solitudes that care produced inside homes (151).

Multisymbolization dominates both records, and it does so twice over. The mesocosms studied were dense with symbolic deposits: proverbs about sixty and seventy-two, epic stories of fathers and sons, tonic advertisements and newspaper photographs in Cohen; diagnostic reports, scans and English medical terms embedded in Hindi in Brijnath. The records add their own layer. Cohen’s practice of writing down remembered conversations each evening (Cohen 1998: 10) placed much of his material in reference to absence before analysis began, and his argument re-presents it through structuralism, psychoanalysis and the anthropology of the relational self. Brijnath organizes her chapter on the voices of people with dementia through the form of a Bollywood film, an L5 frame laid over episodes of song, touch and dance.

Multimateriality is represented with surprising precision in both books, often in passing. Cohen records the charpoy at the threshold, the television moved into a bedroom so that old parents need never leave their bed (182), latrines built in Nagwa with municipal subsidy while the drainage that would have made them usable was never repaired (225), and the bare room in which a pathologist’s mother lay: “a large and spotless room whose only contents were a mattress and an old woman lying on it” (201). Brijnath records Sita Aggarwal’s thin mattress on the floor beside her husband’s high bed with three mattresses (94), and the Tandon house, on a first floor “up a winding staircase, behind high concrete walls”, behind grilled windows and a gate locked from inside, as Sheila Tandon’s world “shrank to her bathroom and bedroom” (145). Such descriptions document who was placed where, and how the made world narrowed the lives it housed.

Dwelling, the coordination with found conditions that no one made, is the least investigated mediation in both, and it appears in the records in characteristic places. In Brijnath, Delhi’s 43-degree heat, its dust storms and its chronic shortages of water and electricity are described as conditions of her own fieldwork, together with the observation that the elite were “buffered” from them by generators, water tanks and air conditioners (17–19). That observation names exactly what OGT calls material buffering: found conditions reach human living through made arrangements, and the distribution of those arrangements decides who is exposed. The record never turns the observation onto the households it studies, so we learn which of Brijnath’s families had generators and air conditioning only by inference from their class. Found conditions enter the accounts of care in single clauses. Sita, describing eight days during which her husband neither urinated nor defecated, adds: “It was terrible, terrible and it was cold” (128). Shivbaksh cooked rajma in winter and bottle gourd in summer (123). The outpatient halls had “no air conditioning or heating; only the proverbial dust” (46). In Cohen, the main road of Ravindrapuri was “shadeless and unusable as common space” (191), “colonies at night were relatively empty and vulnerable places” (192), and a friend of the Mukherjee household said that Amita Mukherjee’s mother had not been the same since robbers struck her in the face while she was alone in the house (198). In Nagwa, Vipat, said to be the oldest man in the settlement, lay on his charpoy day and night, his time “punctuated only by meals and long, slow walks the half a kilometer or so to the banks of the Ganga to relieve himself” (237). Read beside the unusable latrines, that sentence documents a daily coordination of excretion that ran through terrain, distance and river because a material arrangement had failed. Neither book asks how heat, darkness, season or monsoon altered wandering, sleep, agitation or incontinence, or how much of what families experienced as the course of an illness was coordination with a climate.

Selection operated through recruitment and through the research occasion as well as through the authors’ attention. Brijnath’s route made diagnosis a condition of visibility, so that every person with dementia she met had already been translated into a clinical category, and households that still spoke of sathiyana would mostly not have reached her. Cohen’s route made voice and neighbourhood reputation the conditions of visibility, so that he met many old people who were heard as hot-brained or weak and very few who would have met clinical criteria. The research occasion also shaped both records, and in strikingly similar ways. Cohen’s colony families kept old relatives out of sight or declared them asleep (190–191). Brijnath’s families bathed and dressed them for display and asked for her verdict (31, 141). Both are forms of the same coordination, the management of what a visitor with institutional standing would see, and both authors had the candour to report it. Brijnath also recorded the point at which presentational care relaxed and “certain bodily enactments of normality dropped off”, revealing urine bags, catheters and mussed hair (143). Cohen subjected his own record to a late reconsideration, acknowledging that “Indians, when push comes to shove in much of this book, are written of through a relational, fluid, and soft language of anger and affect, and Europeans and Americans through a monadic and hard-wired one of memory and cognition” (292). In OGT’s terms, affect and cognition are both L4 abstractions. The asymmetry that actually shaped his record lay between voices heard at length and coordination rarely observed.

V. Reconstructing the Coordination of Care

Traces of absorbed coordination

The coordination that most needs reconstruction is the one both records preserve least: what people living with dementia were doing, without comment, while others talked about them. It survives mostly as traces, and several kinds recur.

The most reliable is a discrepancy between an attributed category and an incidental description of action. Cohen describes the mother of Amita Mukherjee, a widow of eighty-three in Ravindrapuri, three times. In one passage she “rarely left her room”, and mother and daughter agreed that she was weak and immobile, yet when Cohen visited in the daughter’s absence the old woman “negotiated the entire house with relative ease” (Cohen 1998: 182). In another she lay on her bed, “getting up only to go to the bathroom”, while visiting relatives laughed that “She can walk to the market, and she does; she doesn’t obey us!” (197). The record holds an L4 status, “bed-gifted” in Cohen’s phrase (182), alongside fragments of the L1 coordination that contradicted it, and it cannot settle which described most of her days. Brijnath had assumed that Helen Meena Chand could neither walk nor manage her affairs, and found her mobile and able to pass on messages (Brijnath 2014: 161). When Shivbaksh, who described himself and his wife as having dementia, “but she more than me”, asked Helen the time, she answered, “it’s daytime, around noon” (75).

A second kind is a competence that appears when a stake is engaged. When Brijnath tried to retrieve a prescription she had left with Helen, Helen’s voice “grew firm and solid, her speech rapid and incisive: ‘Why give me something only to then ask for it back?’” (161). A third is coordination restored by a change in another’s response. Meenakshi Ranjarajan asked her daughter, “Where are we going?”; her daughter answered “Home, sweet home” and spoke the first line of Payne’s poem, and Meenakshi recited “all five verses and thirty lines” (167). A fourth is a public decorum carried at L1. Hari Prasad, who rarely spoke and easily got lost, communicated by “folding his hands and saying Namaste or shaking hands”, and that absorbed competence, together with the density of his neighbourhood, let his daughter present his condition as the forgetfulness of age and kept the neighbours helpful (147–148). In Calcutta, Cohen reports, a family considered it very important that a retired engineer, despite anger and forgetting, “could still greet visitors appropriately” (Cohen 1998: 182). A fifth kind is the researcher’s own change of approach. Brijnath’s early questions modelled on the Mini-Mental State Examination produced “polite masks of half smiles” and the retort “What kind of a question is that to ask someone of my age!” (Brijnath 2014: 160–161). She then “just began to be with them”, holding hands, stroking hair and, with a woman of eighty at a day-care centre in Kochi, “dancing to silent music” (162–163). Her account of that change documents both the distance between the articulation her questions demanded and the coordination available, and some of the coordination itself.

Some traces are interrecursive exchanges that the record describes precisely and cannot interpret. During an interview in the Singh household, Harinder Singh repeatedly laid his hand on his daughter Gurneet’s leg; each time she quietly moved it back to his armrest, and each time he returned it (149). Brijnath offers readings in terms of disgust and the spatial order of the household and states that “this remains conjecture” (149). The exchange is a sustained L1 coordination between father and daughter, and the record rightly leaves its meaning open. Each of these traces is confined to a sentence or two. Together they show that the coordination of people living with dementia was more varied, more competent and more dependent on others’ responses than either book’s categories allowed.

Cohen’s record also documents recursive fluidity, the capacity to move between levels in ways that preserve coordination, without naming it. A low-caste woman from Nagwa sat each morning by a pump on Assi Ghat and shouted her possession while demanding money; a neighbour asked her, “why are you screaming? People will give you much more money if you sing”, and the next day she sang bhajans and received more (Cohen 1998: 250). Siranji, a woman in her eighties, cursed her granddaughters within the family compound and kept her quarrels inside its walls: “Siranji for her part knew when not to speak” (248). Her granddaughter Hoshila, her “chief tormentor” at home, dressed her, wrapped a shawl around her and led her “slowly and lovingly” through the lane to a rickshaw (248). The household that heard Siranji as hot-brained also sustained her through coordination that needed no comment.

One episode in Brijnath allows all five mediations to be followed through a single breakdown. Sita Aggarwal recalls that in January of the previous year her husband went eight days without urinating or defecating, refused even water and screamed with pain, and that after an enema “he couldn’t stop for two days, his system was so blocked” (Brijnath 2014: 128). The episode is embodied, a selfrecursive process of retention and release that no one could direct. It is coordination with a responsive other who could neither report nor anticipate, so that Sita checked, asked and checked again. It is multimaterial: the catheter “wasn’t on then”, and the clean-up needed scrubbing, Dettol and soaking. It is multisymbolic in the protective formula “touch wood” with which she frames it. And it is dwelt: “it was cold”. The record specifies four of these contributions well and the fifth in three words.

Counterparts and the bracketing of response

Two interventions recorded by Brijnath change the type of coordination through which care proceeds. The first is the nasogastric tube. Feeding by hand is interrecursive coordination of considerable delicacy, as in Govind Ballabh Tandon’s nightly practice of waking around three in the morning, when his wife Sheila was most responsive, to feed her ice-cream: “She knows that I am feeding her from my hands. She responds to me” (134). The tube makes nourishment independent of the person’s response. The second is sedation. A volunteer of the dementia society described families content for doctors “to give antipsychotic drugs and make the person very numb and drowsy. That way the problems of wandering, the problems of lying, the problems of repeating oneself are gone” (73). Antipsychotics remediated the carer’s burden by reducing the person’s responsiveness, and Brijnath records their costs: unsteady gait, drooling, falls, and Nandini Pillai coming home to find her father “sitting under the tap waiting for me to come home to lift him up” (73). Many women carers responded by quietly reducing doses or seeking alternatives: only six administered medication as a matter of course, while the rest sought alternatives (n=6) or “reduced the dosages of their own volition” (n=5) (74). These carers were weighing one stake against another, a manageable body against a responsive person. Cohen’s material shows a counterpart of a different kind: the tonic and the antidementia tablet, nonrecursive commodities whose relevance in Varanasi households came from what giving them signified. A Calcutta psychiatrist explained the prescription of a drug few believed in: “We can … only keep trying”, adding that the children needed to give something (Cohen 1998: 220).

Stakes, burden and displacement

Cohen identifies the central stake of senility talk precisely: what was at stake when an old voice was heard as wrong was “the moral coherence of her or his family” (177). Brijnath documents carers’ stakes with great richness: being a good daughter, protecting a marriage, a son’s marriage prospects, fear of starvation, love. Ajit Singh’s family approached three or four prospective brides who refused him because of his father’s illness (Brijnath 2014: 148). Both records document the stakes of the person living with dementia much less, and chiefly through others. Brijnath preserves wishes that families set aside: Rajesh Menon wanted to work, “not for money”, but to “spend time. Engage” (164), and a “peaceful death in Trivandrum” (172); Helen said, “I am ready to go up” (172). Cohen preserves Siranji’s demand for “powerful pills from a good private doctor” and her fear of hospital (Cohen 1998: 246–247), and Ganga Jali’s hunger (249).

The recursive burden of anticipation and repair fell on particular bodies. Seventeen of Brijnath’s twenty primary carers were women, and in twelve of the fifteen families coping with incontinence women alone did the cleaning (Brijnath 2014: 24, 127). Sita Aggarwal described the anticipatory work: “I don’t know when he’s going to do it, so who can I call to clean it?” (128). When Harinder Singh was admitted to a neurology ward, only his wife was allowed in; she was refused a rubber sheet and told to take him to the toilet, while their son waited for hours in the foyer with food the guard would not let him bring up (70). Attendants recruited from a resettlement colony worked shifts that became fourteen- to sixteen-hour days with commuting (110). In Cohen, burden appears through household fission, with old parents usually ending up with the youngest son (Cohen 1998: 179), and through the labour of neighbours, children and strangers who returned the lost.

The most successful coordination in Brijnath’s book is Shivbaksh’s day, from tea at five through cooking, sweeping, rest, the evening milk set for yoghurt, dinner and his single nightly peg of whisky, to a cup of milk each at ten (Brijnath 2014: 123). Brijnath explains that the complexity of daily food preparation became evident to her only through this “anomaly” (122), a man doing what women did, or supervised, in eighteen of her twenty households. Successful coordination became visible when it was displaced onto an unexpected person. The same holds for neighbourhoods. When Omar Khan wandered out “without any shoes”, local children brought him home, “Aunty he was here, aunty he was there” (146); when Hari Prasad failed to return from buying dal, the ration shopkeeper sent his assistant out on a scooter (147). These episodes document, at the moment of their engagement, networks of being-with that usually ran without remark and that kept wandering from becoming disaster.

Tempo

The mediations did not share one tempo, and both records register conflicts of tempo in passing. A pathologist in Varanasi described his mother: “She often sleeps in the day. At night she’s up, and goes around the house” (Cohen 1998: 200). Tandon reorganized his own sleep around his wife’s three o’clock responsiveness (Brijnath 2014: 134). Outpatient doctors saw fifty to a hundred and fifty patients in a two-hour session, and people with dementia “rarely returned to the OPD following diagnosis”, their relatives refilling prescriptions that went unreviewed for years (46, 71–72). Carers’ days were structured by a “tri-cyclical performance of preparation, serving and flavour in each meal” (124–125), while the days of people with dementia were “defined by others: prayer times, bathing time, mealtimes, siestas, evening walks and bed” (163). The question of whose rhythm governed whose runs through both books, and neither takes it up.

VI. Sorting the Differences

With both instruments reconstructed, each apparent difference between the books can be tested by asking whether the other record could have registered it. Table 2 sets out the result.

Table 2. Differences between the books, sorted

VII. The Limits of the Evidence

The three forms of absence must be kept apart. Some matters are absent from the research account: dwelling as a condition of care in both books, the night in most of both, poor households in Brijnath, and advanced dementia in most of Cohen’s neighbourhoods. Others are documented without sufficient evidence to establish how coordination proceeded. Brijnath records that a clean-up could take three hours with a father “in that panic mode”, and that his aggression rose with electrolyte imbalances and his sleep improved with a bowl of curd at night (Brijnath 2014: 127–129). The record cannot specify what bodily state, temperature, light or sequence of demands made a given episode escalate. A few matters are demonstrated nonoccurrences. No family in Brijnath’s sample considered withholding artificial feeding (129). In Varanasi in 1989, old people were virtually never brought to the psychiatric clinic (Cohen 1998: 216). The psychiatric nursing homes of Calcutta would not consider Somita Ray despite her diagnosis of Alzheimer’s, and only very expensive facilities would take her (299). Only this third kind establishes that something did not happen, and only for the populations the records cover.

The traces of absorbed coordination reconstructed above are, by the same discipline, traces. Each rests on a clause or a short passage, and what they show is that the categories applied to people living with dementia did not exhaust their coordination. They cannot show how often, or across how much of a day, that coordination ran.

VIII. Fit, Misfit and Practical Knowledge

Cohen’s concept of the familial body tracks a recurrent pattern in his material with real mesocosmic fit across being-with and multisymbolization: stigma and explanation attached to the family before they attached to the old person. His analysis of weakness in Nagwa reaches further. “Difference was embodied as weakness, not pollution: caste was a medical condition” (Cohen 1998: 232). Kamzori tracked lifelong undernutrition, labour until collapse, the loss of common land and the transactional order of caste, holding embodiment, being-with, materiality and political economy in a single view. His observation that categories such as weakness, balance and Alzheimer’s work “as much against meaning as for it” (230; see also 301–302) identifies protective L4 operations that deflect blame from those who must go on living together. The familial body reaches its limit where it absorbs the old person’s bodily, material and environmental situation into the signification of family relations, so that a failing body, a crowded flat and a walk to the river become hard to hold apart.

One of Cohen’s cases shows an attribution of memory loss serving a household stake. In Chittupur, a Dalit settlement near Nagwa, the old patriarch Mausaji kept asking his daughter-in-law for more rotis, and she found that “he was hiding them under his bed” (243). When he said bitterly, “Khana nahin hai”, there is no food, his family replied that he had lost his memory when he “got a wind” six years earlier and so wrongly believed he had not eaten (244–245). Memory loss, Cohen concludes, was “a second-order problem”, invoked “to explain a discrepancy between Mausaji’s and his family’s memory of a history of food transactions” (245). Cohen’s analysis has excellent fit here: an L4 category of cognitive deficit settled an interrecursive contest over food in favour of those who controlled the hearth.

Cohen also records an evaluative contrast that OGT can make precise. At the Nava Nir old age homes in Calcutta, residents planned meals, managed budgets and controlled their own medication. Cohen, who had worked as a nursing aide in Boston tying patients to their chairs (6), “was and remain[s] enthralled by the independence of Nava Nir residents”, while Calcuttans he interviewed found the homes “pathetic places precisely because of the perceived lack of total dependence” (117). The same coordination was judged by two sets of criteria, one valuing self-direction and the other valuing the dense dependency through which seva is shown. Neither is the coordination’s own measure of success, and the episode shows how easily an observer’s criteria become the criteria of an account.

Brijnath’s concept of love tracks the coordination between spouses with real fit, and her refusal to sort deaths into good and bad (Brijnath 2014: 174–176) guards against evaluative misfit. The concept is less able to distinguish whose stakes a coordination meets. It explains why families chose the tube and dismissed talk of dying, but it cannot weigh those choices against the wishes of Rajesh or Helen, which the record preserves. Some of the book’s most valuable knowledge lies outside its concepts. When a carer telephoned late one night because her father had stopped eating, Brijnath advised her to put a rubber sheet under him, to offer something nourishing every few hours instead of full meals, not to argue with him and to let him talk about old memories (130). Women carers who reduced antipsychotic doses on their own judgement were applying a practical theory of their own. Both are remediations that lower recursive demand on everyone involved, and both track the coordination of care more closely than the book’s theoretical apparatus does.

IX. What a Third Study Would Observe

Read through a comparative OGT, No Aging in India and Unforgotten describe less a passage from senility to dementia than two partial records of urban Indian households living with old age and cognitive change, produced by instruments tuned to different signals. Cohen heard how old people were heard, and documented the moral economy of households, castes and classes with an acuity no later work has matched. Brijnath watched how people with dementia were handled, and documented the bodily and material labour of middle-class care with an intimacy Cohen’s method did not allow. Sorted with care, their differences show real change in the institutional and material deposits of care, considerable continuity under different names in how families experienced and explained what was happening, and a set of absences that both instruments share.

A third study would be designed around those absences. It would observe whole days and nights across the seasons, in households of different classes, including the poorest. It would investigate material buffering directly, asking how heat, power cuts, monsoon, darkness and distance reach people living with dementia and their carers through the arrangements each household can afford. It would follow feeding, medication and sedation as changes in the type of coordination, attending to what each secures and what each brackets. It would give attendants and servants their own voice. And it would take as its main object the coordinations that keep trouble from arising: the neighbour who brings someone home, the decorum that holds in public, the hand offered and withdrawn, the routine that runs so well that no one mentions it. Both books show where such coordinations surface in a record. A study built to observe them would no longer have to find them in passing.