# Two Countries in One Body *A registration form, a language in two scripts, and the route no checkbox can hold* Kanwar Partap Singh Gill, MD · November 5, 2026 · The Human Pulse Before most visits now there is a form. It may be on a clipboard or, more often, on a tablet handed across the front desk, and somewhere in its first few screens it asks a person to say who they are in a handful of taps: race, ethnicity, preferred language, country of birth, sometimes religion, marked optional. The questions take less than a minute. They are answered in a waiting room, with one eye on the door, by people who have come about something else entirely. I want to begin by defending them. Those boxes are how a health system discovers whether some groups of patients wait longer, are treated differently or do worse, and without them a good many inequities would stay invisible. The clearest recent example I know concerns an instrument most people have had clipped to a fingertip. In December 2020, a short report in the *New England Journal of Medicine* compared readings from pulse oximeters with oxygen levels measured directly in arterial blood, and found that the oximeters missed dangerously low oxygen in Black patients nearly three times as often as in White patients. The device had been in universal use for decades. The discrepancy was discoverable only because the hospitals involved had recorded, for each patient, a category on a form. Had nobody ticked the box, the oximeter's failure would have looked like nothing at all: scattered individual misfortunes, each explicable, none of them a pattern. So a checkbox is a crude instrument, and when the question is fairness a crude instrument is far better than none. My interest is in what happens next: in the distance between a category that is accurate and a person who is complete, and in the way information gathered for one purpose can slide, without anyone deciding it, into another. ## One Question and the Next A category can be useful for one question and inadequate for the next. If a health system wants to know whether people who speak different languages receive explanations they understand, collecting language information has a clear purpose, and the purpose is served by counting. If a clinician wants to know which explanation this particular person will find useful today, the category is only a starting point. The second question requires a conversation that the first cannot replace. The distinction matters because data travels. A field collected to examine fairness across thousands of people does not stay in the statistician's spreadsheet. It appears on a clinical screen, beside a name, during a visit, where it begins to suggest things it was never collected to suggest: assumptions about diet, about belief, about family, about risk. Nobody announces this move. The field simply sits there, and a busy reader draws a conclusion from it, as busy readers do. Some background information is genuinely relevant to a clinical decision. But its relevance has to be established for the decision at hand. It cannot be presumed because the information happened to be on the screen. Population measurement and individual care are not enemies. They answer different questions with different instruments, and each becomes dangerous when it is asked to do the other's work. A count can show that a group is being failed, and cannot say who an individual is. A conversation can learn who an individual is, and cannot, on its own, show that a group is being failed. Most of what follows concerns the border between them. Some would go further and do away with the boxes altogether, on the ground that recording race at all lends a social invention the authority of a medical fact, and that a clinic which asks the question has already half-answered it. I feel the force of this, and one of the cases I come to below supports it. But the argument proves less than it seems to. Ceasing to record a category does not stop people from being treated according to it. The receptionist, the triage nurse and the physician still have eyes, and their assumptions do not need a form in order to operate. What disappears is the ability to check. A health system that stops counting cannot discover that its oximeters fail some patients, or that interpreters are rarely called for others, or that one group waits longer than another for the same referral. Refusing to measure does not take the measuring tape out of anyone's hand. It only ensures that nobody can see where it is being held. The better course is harder: to keep the count, and to guard, with some ferocity, the border between the count and the person. ## Who Drew the Lines The categories on that screen have a history, and it is worth knowing, because a category always looks more natural than it is. The race and ethnicity categories in American medical records generally follow federal standards, which the Office of Management and Budget first issued in 1977 and revised in 1997. In March 2024 it revised them again. The [new standards](https://www.federalregister.gov/documents/2024/03/29/2024-06469/revisions-to-ombs-statistical-policy-directive-no-15-standards-for-maintaining-collecting-and) combine race and ethnicity into a single question that allows more than one answer, add Middle Eastern or North African as a seventh minimum category, and require federal agencies to collect more detailed categories by default; under Asian, the first examples listed include Chinese and Asian Indian. The revision gave agencies until 2029 to bring their existing forms into line. Not every clinical form in the country is bound by these rules, and many systems keep their own lists, but the federal categories are the grammar most of them were built on. Further back, the lines were drawn more crudely and with heavier consequences. In the early twentieth century, the United States census [recorded people from India](https://apa.si.edu/now/has-the-census-evolved/) as "Hindu," whatever their religion, so that many of the Punjabi Sikh farmers then working in California's valleys were counted under the name of a faith they did not hold. The same habit of naming appears in a document that every student of this history eventually reads. In 1923 the Supreme Court decided *United States v. Bhagat Singh Thind*. Thind was a Sikh who had served in the United States Army during the First World War, and the question certified to the Court [began](https://www.law.cornell.edu/supremecourt/text/261/204): "Is a high-caste Hindu, of full Indian blood, born at Amritsar, Punjab, India, a white person" within the meaning of the statute. The Court answered no. Justice Sutherland wrote that the statute's words were "to be interpreted in accordance with the understanding of the common man," and set aside, in so many words, "the speculations of the ethnologist." I read that opinion with a particular discomfort, because the question describes its subject as born at Amritsar, the city where I grew up, and names him by a religion that was not his. But the discomfort teaches something more general than grievance. The category in *Thind* was not discovered. It was decided, by people, for a purpose, on the stated basis of what ordinary people supposed. Categories have always been made that way, and remade when the purposes changed. For several decades after the "Hindu" years there was no separate census category for people from India at all. "Asian Indian" appeared in 1980. The lives the categories could not hold went on regardless, and sometimes they produced names of their own. In the early decades of the century, in the Imperial Valley and elsewhere in California, a good number of Punjabi men, most of them Sikh, married Mexican and Mexican American women; the historian Karen Leonard traced these families in *Making Ethnic Choices* (1992). Their children were often raised Catholic, grew up speaking Spanish and English and sometimes a little Punjabi, and were known to their neighbours as Mexican Hindus. The name was wrong about religion, as the census had been. It was right about something no field of the period could record: that a family could belong to two places at once and make a home out of the joining. Nobody designed that category. The people inside it lived first, and the word came after. On a form today I would tick Asian, and on a detailed version Asian Indian. The answer would be correct. It would also tell its reader nothing about Khadur Sahib, Amritsar, Little Rock or Fresno, and nothing about whether the person who ticked it prays in a gurdwara, a mandir, a mosque, a church, or nowhere in particular. That is not a failure of the box. It is a fair description of what boxes are for, and a reminder that the purposes for which they are drawn deserve as much scrutiny as the answers people give. ## The Tape of the World No one has described both the value and the injury of social measurement more exactly than W. E. B. Du Bois, and he could, because he practised the one and suffered the other. Du Bois was a sociologist who counted. *The Philadelphia Negro*, published in 1899, was built on a house-to-house study of a single city ward, and for the Paris Exposition of 1900 he and his students at Atlanta University prepared a set of hand-drawn charts of Black American life that are still startling in their clarity. His premise, as I read it, was that a people whose condition went unmeasured could more easily be ignored, and he set out to measure it. Yet it was the same Du Bois who wrote, in *The Souls of Black Folk* in 1903, the most famous account in American letters of what it is to be measured by others. He called it double-consciousness: "this sense of always looking at one's self through the eyes of others, of measuring one's soul by the tape of a world that looks on in amused contempt and pity." The phrase *the tape of a world* has stayed with me for years. A tape is a measuring instrument. It is not wrong about inches. What makes it an injury is whose hand holds it, and what it is held against, and what the holder intends to conclude. I want to be careful here. Du Bois was writing about the experience of Black Americans under the color line, a condition of a specific and brutal kind, and I do not claim it, or suggest that a Punjabi doctor in California lives inside it. I borrow from him only a distinction that his own career embodies. Measurement in the hands of people trying to see an injustice is one thing. Measurement read as a verdict on the person measured is another. The same number can serve both. Medicine has made both uses. For many years, the most widely used equations for estimating kidney function from a blood test included a coefficient that raised the estimate for patients recorded as Black. A social category was being treated, inside a formula, as though it measured something about the kidney. In 2021 a joint task force of the National Kidney Foundation and the American Society of Nephrology recommended a new equation that does not use race. The pulse oximeter and the kidney equation sit side by side in my mind as the two faces of the same box. In one case the category exposed a failure that would otherwise have been invisible. In the other, the category was itself installed as a piece of biology, and the error ran the other way. None of this means that ancestry and origin are clinically irrelevant. Some conditions do run more often in families from particular parts of the world. The place where a person grew up shapes which infections they met as children, which vaccines they received, what they ate, what illnesses were common around them. That is one good reason a form asks where a person comes from. But the category opens a question; it does not answer it. *Asian Indian* is not a genotype. *Punjabi* names a language and a region. *Sikh* names a faith. *Indian* names a nationality. These are four different kinds of fact, and a careless screen can make them look like one. What matters clinically is usually something more particular that only the person can supply: what illnesses ran in their own family, where they actually lived and for how long, what they actually eat. The box is the beginning of that history. Treating it as the history is how the tape gets into the wrong hand. It helps, I find, to notice that *where a person comes from* runs on at least four clocks. There is ancestry, measured in generations and not chosen by anyone now living. There is birth, which happened on one day in one place. There is upbringing, the years in which a body met its first infections, received or missed its vaccines, and learned what to eat and what to fear. And there is residence, where the person has lived lately and lives now, which shapes exposure and access more directly than any of the others. A form usually asks about one of these and lets its reader assume the rest. Two people of Punjabi descent, one raised in a village near Jalandhar and the other in a California suburb, may record the same ancestry; their upbringing and their residence, which often matter more to the clinical question, have almost nothing in common. A history taken properly asks about each clock separately, and allows the answers to disagree. ## Two Scripts, One Field The next screen asks for a preferred language, and for a great many people, in Punjab and wherever Punjabis have settled, the honest answer is Punjabi. But Punjab itself has been divided between two countries since 1947. On the Indian side, Punjabi is written in Gurmukhi. On the Pakistani side it is written in Shahmukhi, a Perso-Arabic script that runs from right to left. A person from Lahore and a person from Amritsar, cities a short drive apart, can share a language, its jokes and proverbs and its particular music, and still be unable to read each other's handwriting. A field that says *Punjabi* does not say which. A handout printed "in Punjabi" can be faithfully translated and remain unreadable to the person it was printed for. If the title of this essay wants a literal case, here is one: two countries in one language. The arrangement is not unique to Punjabi, and anyone who grew up in north India knows its mirror image. Hindi and Urdu, in their everyday spoken forms, are close enough that speakers of each can follow most of what the other says, yet they are written in entirely different scripts, Hindi in Devanagari from left to right and Urdu in a Perso-Arabic script from right to left. Their formal registers also draw on different reservoirs, Sanskrit on one side and Persian and Arabic on the other, and printed material tends to be written in the formal register. The spoken bridge between them can therefore vanish on the page. A person who would follow a conversation without difficulty may be unable to read a single line of the leaflet that summarises it. And the difficulty does not end with script. Speaking and reading are different abilities, and a single field quietly treats them as one. A young woman born in California may speak Punjabi at home every day and read it hardly at all. Her grandfather may read Gurmukhi easily and English not at all. Their neighbour may have learned to read in Hindi, in Devanagari, or in Urdu, and speak Punjabi with all of them. Each of these people would honestly tick *Punjabi*. Each needs something different on paper, and possibly something different in the room. "Preferred language," on inspection, contains at least three questions: what language a person wants to speak, what they can comfortably read, and whether they would like a trained interpreter present. The professions that do this work keep the distinction in their vocabulary: interpreting is spoken, translation is written, and the two are different skills. A form that asks one of these and assumes the other two has not been careless out of malice. It has simply mistaken a person's language for a single fact. There is a further complication that any translated leaflet runs into. The Punjabi actually spoken about health in many homes, on both sides of the border and abroad, is already bilingual. People say they have *sugar*, meaning diabetes, or that their *BP* is high, or that the *report* has come back; the English words have been absorbed so thoroughly that they no longer feel foreign to the people using them. A translator working conscientiously toward a formal Punjabi may replace them with older or more learned terms that are perfectly correct and much less familiar, so that the translation ends up further from the reader's own speech than the English it replaced. The field says Punjabi. The Punjabi a person speaks may be one that no style guide would approve, and it is the one that works. The interpreter deserves a closer look, because the role is so often misunderstood, and the most penetrating account of the misunderstanding I know is a short story. In the title story of Jhumpa Lahiri's *Interpreter of Maladies* (1999), Mr. Kapasi drives tourists to the Sun Temple at Konarak, and he has another job: in a doctor's office, translating for Gujarati-speaking patients whose complaints the doctor cannot understand. He thinks little of the job; he had once hoped to interpret between dignitaries and nations. His passenger, Mrs. Das, an American of Indian parentage visiting with her husband and children, finds the job romantic, and comes to believe that a man who interprets maladies might interpret hers. She confides a secret she has kept for years, about the paternity of one of her sons, and asks him to suggest some remedy for the way it makes her feel. Mr. Kapasi, flattered and then dismayed, asks whether what she feels is really pain, or guilt, and she is offended, and the brief intimacy between them collapses. The story is about many things, but a physician reading it notices first what Mrs. Das gets wrong about the work. She assumes that someone who can carry words between two languages must also be able to diagnose the person speaking them. Mr. Kapasi's actual job is narrower and more valuable. He makes it possible for a patient and a doctor to understand each other. He is not the doctor's assistant, and he is not the patient's confessor, and the moment he is asked to be either he is lost. The codes of ethics that govern medical interpreting in the United States ask for exactly that line: accuracy, impartiality and clarity about the limits of the role, so that an interpreter renders faithfully, does not substitute his own judgment, and says so when a word will not carry. Lahiri's story shows what happens when the line is blurred by a stranger's longing. ## The Language of the Task The word *preferred* raises a quieter difficulty, because it assumes a single answer that holds for every purpose. Many bilingual people use different languages for different kinds of talk. One language may be the language of numbers, forms and appointments, learned at work and at school; another may be the language of the family table. A person might want a laboratory result explained in English, the language in which they have read about the condition, and the decision about it discussed in Punjabi, which is how it will be talked over at home that evening. Or the other way round. They may not know which until the moment arrives. There is a temptation here that I want to resist, which is to assign each language a fixed emotional territory: English for facts, the mother tongue for feelings. That makes a pleasing pattern and is often false. Some people find it easier to say frightening things in a second language, precisely because it holds them at a little distance from the words. Some people who are entirely fluent in English for every purpose of their working lives will still want a trained interpreter for one particular conversation, a consent to surgery or a discussion of a serious diagnosis, because the stakes have changed and they want no doubt at all about what was said. And some people decline an interpreter even when their English is limited, for reasons that make perfect sense: in a small community, the interpreter on the telephone may turn out to know their family. None of this can be deduced from a birthplace, a surname or a box. The practical conclusion is modest. Ask, and ask again when the task changes. *Would you like to do this part in Punjabi or in English?* *Would you like this written down, and in which script?* *Would you like an interpreter for this conversation, even if we did not need one last time?* Questions of this kind take seconds, and they move the decision to the only person entitled to make it. A clinician who shares a patient's language can be tempted to skip them, and to switch languages unasked, as a kindness. Sometimes that is exactly the kindness wanted. Sometimes it is a presumption, received as a suggestion that the person's English was not good enough. The difference is the question. I would want the person to have room to correct the inference even when the category itself is accurate. Yes, Punjabi is the preferred language; this particular document may still be easier to read in English. Yes, the family came from a certain region; the person's own history may include other places and circumstances that matter more to the question at hand. A good record would hold both the category and the correction, and would let the correction win. ## A Route a Box Cannot Hold The Kashmiri poet Agha Shahid Ali, who grew up in Srinagar and spent most of his writing life in the United States, opened one of his best-known early poems, "Postcard from Kashmir," with a couplet that I have never been able to read without thinking of a registration form: "Kashmir shrinks into my mailbox, / my home a neat four by six inches." The postcard is a homeland made small and portable and tidy, and the poem goes on, with characteristic wit, to confess the speaker's love of neatness, and to suggest how far the neat image will be from the place itself. A decade later, in *A Country Without a Post Office* (1997), the tidy rectangle has been replaced by its opposite. Kashmir in those poems is a place of curfew and violence where the post office has been shut and letters go undelivered, and the poet writes toward a home that cannot write back. Between the two books lie two ways a home can be lost while remaining entirely real: by being made too small to hold, and by becoming unreachable. Ali gave much of his later energy to a form that has its own bearing on this subject. He championed the strict ghazal in English, the form in which every couplet must be complete in itself, a whole thought in two lines, and the poem is held together not by a story but by a recurring refrain and rhyme. A ghazal does not progress from beginning to end in the way a narrative does. Its couplets can be read in almost any order and still belong to one another, because the same word keeps returning to claim them. I have come to think that a life lived across several places is often better described that way than as a story with a departure and an arrival. It is a series of whole couplets, each complete where it was written, bound by something that recurs. The cultural theorist Stuart Hall, writing on diaspora in 1990, put a related idea in the vocabulary of theory. Cultural identity, he argued, is a matter of becoming as well as of being; it is "not an essence but a positioning." The phrase is useful in a clinic because a position can be described only by the person who occupies it, and only for now. A box records an essence, or at least something that looks like one. A conversation can ask where a person is standing. Punjabi has its own small grammar for this. The word for a foreign land, ਪਰਦੇਸ, *pardes*, is built on the word for one's own country, ਦੇਸ, *des*, with a prefix meaning *other*. In that word, at least, *abroad* cannot be named without first naming *home*. A form that asks for a country of birth has no such grammar. It records a fixed point and lets every other place go unmentioned, where the language at least insists that the far country is always measured from somewhere. My own route is a short one to state and a long one to explain: family roots in Khadur Sahib, a childhood in Amritsar, training in Little Rock and in Fresno, and Fresno, now, as home. Which parts of that does a form answer adequately? Religion, if the form asks and offers the option: Sikh, and the box is enough. Preferred language is harder. Punjabi was my first language, and I also speak Hindi, Urdu and English, and the form allows one. What needs a conversation is everything that connects the points: what each place gave, which of them a person means when he says *back home*, and why the answer may change with the question being asked. The title of this essay borrows its shape from Du Bois, from his "two warring ideals in one dark body," and I should say plainly where the borrowing stops. The places in my own life are not at war inside me, and I would not want to dramatize a fortunate route into a divided soul. But there is a sentence a little further on in Du Bois's chapter that seems to me, with every difference granted, to speak to anyone who belongs to more than one place: in the merging he describes, the Black American "wishes neither of the older selves to be lost." A good form, and a good clinician, should require nothing more drastic of anyone. Not a choice between countries. Room for both. ## The Gift and Limit of Recognition Imagine two people who tick exactly the same boxes: Asian Indian, Punjabi, Sikh. One is a farmer near Yuba City whose family has worked the same ground for four generations and who has lived nowhere else. The other moved from Ludhiana three years ago and is still learning where the good mangoes are sold. The form cannot tell them apart. A clinician who shares something of their background may not be able to either, at first, and a shared background gives no advance knowledge of either one. Recognition is a real gift. A familiar name, a greeting that needs no translating, a shared reference to a place can let trust form faster than it otherwise might, and in medicine trust formed quickly is worth a great deal. Studies of care given in a patient's own language, by clinicians who speak it, have generally found better understanding, and nobody who cares about access should talk that advantage down. But recognition can also produce a confidence it has not earned, and the hazard is wider than the shared language I have written about [elsewhere](../punjabi-as-a-first-language-of-medicine/index.html). A shared background tempts the listener to supply what was never said: to assume a diet, a family arrangement, a degree of religious observance, a view about who should be told a diagnosis first, or a set of resources and obligations, when the person in front of him may hold none of these. The mechanism is simple and easy to miss. Faced with a stranger, a careful clinician asks who lives at home, what the person eats, who should be told what, and what the person believes about the illness. Faced with someone who seems familiar, the same clinician may feel he already knows, and the questions quietly drop out. Familiarity can make a history shortest in exactly the places where it should be longest. Recognition also has a cost that runs in the opposite direction, and it is easy for a clinician who enjoys being recognized to forget it. Communities that are small and closely connected can make privacy precious. A person may be glad to hear a familiar language in the examination room and still be reluctant to say certain things to someone who might know an aunt, worship in the same place or move in the same circles. For some patients, a clinician from outside their community is not a second-best arrangement but a deliberate choice, made for the sake of being able to speak freely. That choice deserves the same respect as the opposite one. The farmer near Yuba City and the man from Ludhiana may differ more from each other in their beliefs, their circumstances and their wishes than either differs from a stranger. Mr. Kapasi and Mrs. Das are a fair warning here. The Das parents were born and raised in America; to Mr. Kapasi they look Indian and behave like any foreign tourist, and he is as opaque to them as they are to him. They share an ancestry and almost nothing else, and each of them builds a private fiction about the other out of that shared surface. A clinician who shares a patient's background is in the same danger, with more at stake. Recognition only opens the conversation. Neither of the two imagined people owes the room a representative account of Punjab, and each of them may carry a detail that matters only to them. ## The Account the Person Can Revise What I would want, in the end, is for identity to be treated in a record the way a good clinician treats a history: as information supplied by a person, attributed to them, and open to their correction. Federal standards moved in this direction in 1997, when they first allowed people to report more than one race, and the 2024 revision went further when it folded ethnicity into the same question. A record could go further, in small ways. It could note who supplied a category and when. It could allow a person to say that a field does not describe them well, and to say so in their own words. It could separate the language a person speaks from the script they read and the help they want. None of these would make the form longer by much. All of them would make it more honest about what it knows. Federal statistical practice has long preferred that people identify themselves instead of being classified by someone looking at them, and the preference is wise. A category assigned by observation records the observer's guess. In a busy registration area that guess may be made in a second, from a face, a surname or a turban, and once entered it can sit in the record for years, informing people who never met the person and never learned that the answer was not his own. The point is not to make forms endlessly detailed. No form can finish the work of knowing a person, and a form that tried would become an interrogation. It is to keep the boundary visible between the information a person supplies and the conclusions others draw from it. A category can help a system notice whom it is failing; the oximeter's story shows that it can save lives by doing so. The encounter must still allow the individual to say who has actually arrived. At the end of registration, the tablet thanks the person for completing it, as tablets now do, and the screen goes blank. Everything it asked was reasonable, and every answer may have been true. What the person actually carries in with them, two scripts or one, a *des* that may have moved, a route across more than one country, a set of couplets bound by a word that keeps returning, is still out in the waiting room, holding a number. It will have to be asked for in the visit, and the person will have to be allowed to answer in more than one way, and to change the answer the next time. ## Evidence and Editorial Note **Documented findings.** Public biography: family roots in Khadur Sahib; upbringing in Amritsar; training in Little Rock and Fresno; practice in California since 2010 with Fresno as home; Punjabi as first language, with Hindi, Urdu and English; a Sikh. Office of Management and Budget, "Revisions to OMB's Statistical Policy Directive No. 15," *Federal Register*, March 29, 2024 (standards first developed in 1977 and revised in 1997; single combined race and ethnicity question allowing multiple responses; Middle Eastern or North African added as a seventh minimum category; detailed categories collected by default, with Chinese and Asian Indian among those listed under Asian; existing collections to comply by March 28, 2029) (linked). That race and ethnicity fields in American medical records generally follow these federal categories, while not every clinical form is bound by them: standard reference knowledge of health data standards. Census classification of people from India as "Hindu" regardless of religion in the early twentieth century, the period without a separate category, and the introduction of "Asian Indian" in 1980: Smithsonian Asian Pacific American Center, "Has the Census Evolved?" (linked); sources differ on the exact census years of the "Hindu" classification, so the essay does not specify them. *United States v. Bhagat Singh Thind*, 261 U.S. 204, argued January 11–12 and decided February 19, 1923, opinion by Justice Sutherland: the certified question and the quoted phrases "the understanding of the common man" and "the speculations of the ethnologist," from the opinion text on Cornell's Legal Information Institute (linked); Thind's Sikh faith and his service in the United States Army during the First World War: standard reference knowledge. Punjabi Sikh farmers in California in the early twentieth century, and the Punjabi Mexican families of the Imperial Valley and elsewhere known locally as "Mexican Hindus," as studied in Karen Isaksen Leonard, *Making Ethnic Choices: California's Punjabi Mexican Americans* (1992): standard reference knowledge, described without quotation. The long-standing preference in federal statistical standards for self-identification over observer classification, and the 1997 revision's provision allowing more than one race to be reported: standard reference knowledge of Statistical Policy Directive No. 15. Sjoding MW, Dickson RP, Iwashyna TJ, Gay SE, Valley TS, "Racial Bias in Pulse Oximetry Measurement," *New England Journal of Medicine* 2020;383(25):2477–2478, December 17, 2020 (in two cohorts, occult hypoxemia undetected by pulse oximetry nearly three times as often in patients who identified as Black as in those who identified as White): text of the letter, PubMed Central PMC7808260. The 2021 recommendation of the joint National Kidney Foundation and American Society of Nephrology task force to adopt a kidney-function estimating equation without a race coefficient: Delgado C and colleagues, "A Unifying Approach for GFR Estimation: Recommendations of the NKF-ASN Task Force on Reassessing the Inclusion of Race in Diagnosing Kidney Disease," *American Journal of Kidney Diseases* 2022;79(2):268–288, published online September 23, 2021 (PubMed abstract). W. E. B. Du Bois, *The Philadelphia Negro* (1899) and the charts prepared with Atlanta University students for the 1900 Paris Exposition: standard reference knowledge; *The Souls of Black Folk* (1903), chapter I, "Of Our Spiritual Strivings" (double-consciousness; "measuring one's soul by the tape of a world that looks on in amused contempt and pity"; "two warring ideals in one dark body"; "wishes neither of the older selves to be lost"): public-domain text, Project Gutenberg edition opened. Jhumpa Lahiri, "Interpreter of Maladies," title story of *Interpreter of Maladies* (1999): plot described without quotation from standard reference knowledge of the story. Agha Shahid Ali, "Postcard from Kashmir," in *The Half-Inch Himalayas* (1987), opening lines quoted; *A Country Without a Post Office* (1997) and his advocacy of the strict ghazal in English: described from standard reference knowledge. Stuart Hall, "Cultural Identity and Diaspora" (1990), in Jonathan Rutherford, ed., *Identity: Community, Culture, Difference*: the phrase "not an essence but a positioning" and the idea of becoming as well as being, from standard reference knowledge of the essay. The partition of Punjab in 1947; Punjabi written in Gurmukhi in India and in the right-to-left Perso-Arabic Shahmukhi script in Pakistan; Hindi and Urdu as closely related spoken languages written in Devanagari and in Perso-Arabic script, with formal registers drawing on Sanskrit and on Persian and Arabic; the absorption of English words such as *sugar* (for diabetes), *BP* and *report* into everyday Punjabi talk about health; the professional distinction between interpreting (spoken) and translation (written); the principles of accuracy, impartiality and role boundaries in American codes of ethics for medical interpreters; and the general finding of studies of language-concordant care that patients understand more: standard reference knowledge. The meanings of ਦੇਸ (*des*) and ਪਰਦੇਸ (*pardes*) and the latter's formation with a prefix meaning "other": standard reference knowledge. **Vendor representations.** None. No registration software or record system is named. **Recollection and tradition.** The two people in the section on recognition, the family members differing in what they speak and read, and the patients choosing between languages are hypothetical and illustrative; they do not describe real patients. No remembered scene is described. Nothing is asserted about the author's own travel, circumstances or experience of moving between countries beyond the public sequence of places, and the essay describes no particular clinic or practice. **Analysis and forecast.** The defense of demographic categories, the distinction between population measurement and individual care, the readings of *Thind*, Du Bois, Lahiri, Agha Shahid Ali and Hall, the comparison of a life across places with the ghazal, the division of "preferred language" into speaking, reading and interpretation, and the proposals for a revisable record are the author's interpretation as of November 6, 2026. **Independence.** This essay is not sponsored by any institution, company, health system or technology developer. It is personal reflection, not medical advice; readers seeking care should contact their own physician's office. ## Related reading - [The Accent in the Room](https://personal.kpsgill.com/essays/the-accent-in-the-room/) - [The Word for Worry](https://personal.kpsgill.com/essays/the-word-for-worry/) - [When I Look Back From the Next Thirty Years](https://personal.kpsgill.com/essays/when-i-look-back-from-the-next-thirty-years/)