# What the Chart Cannot See

*A field guide to what is seen in the room and only partly written down*

Kanwar Partap Singh Gill, MD · November 5, 2026 · The Examination Room

The title should not be taken literally. A chart can say that a man walks with a limp, that a question took a long time to answer, that a daughter is worried about her father's drinking. Any of these can be written down, and careful clinicians write such things down every day. The difficulty lies elsewhere: in what the written account leaves out, in what its next reader quietly supplies, and in how easily a description made on one afternoon hardens into an assumption on another. A record does not fail to see. It sees once, in one light, through one pair of eyes, and then it goes on saying what it saw.

In April 1934 a young artist named Roger Tory Peterson published *A Field Guide to the Birds*. He had learned to watch birds in New York, in the Bronx County Bird Club, where the ornithologist Ludlow Griscom could name a species in the instant it crossed his binoculars. The bird books then on the market, as an exhibition at the Linda Hall Library tells the story, served the collector, who identified a bird with a rifle and a ruler, by its anatomy and its measurements in the hand. Peterson wanted a book in which living birds could be "run down by impressions, patterns, and distinctive marks." His plates showed birds as people actually see them, at a distance or in flight, and small arrows pointed to the marks that separate one species from its nearest neighbours. His publisher, Houghton Mifflin, was nervous about a bird book in the Depression and printed only two thousand copies. [They sold out within a week.](https://lindahall.org/experience/digital-exhibitions/drawn-from-nature/05-the-making-of-a-field-guide)

A medical record has more in common with the collector's book than with Peterson's. It is precise, cumulative and measured; it holds the specimen still. What follows is a small guide to the other kind of knowledge, the kind gathered in the field of the examination room. It differs from the birding kind in one large respect, which the rest of the guide will keep returning to. A field mark on a bird identifies the bird. A field mark in a clinic identifies only a question.

## How to Use This Guide

Every entry below is either general clinical knowledge or an imagined example, and the imagined ones are flagged where they appear; none describes a particular person.

A guide to people has to begin with the ways in which people are unlike birds, and those differences shape everything that follows. A warbler does not change its nature between one sighting and the next. It does not read the guide, and it does not object to its entry. People do all three. The man described last spring as a heavy drinker may have stopped in the summer. The woman whose chart calls her anxious may read the word on her telephone in the parking lot and wonder who decided it. The young father recorded as declining a vaccine may have changed his mind, or may never have declined it at all. A species is a stable category with a stable set of marks. A person is a particular life whose marks mean different things at different times, and sometimes mean nothing.

Peterson's arrows answered a precise question: which of the birds that look like this one is it? Once the arrow had done its work, the bird was named and the looking could stop. In the clinic an observation almost never permits the looking to stop. Its proper use is to begin a conversation that could not have begun without it.

So the entries come with three cautions, and the cautions are the substance rather than the small print. An observation is a reason to ask, never proof of what someone is thinking or feeling. What was seen should be written in plain words and kept apart from what it was taken to mean. And the person in the room is entitled to revise what the chart says about them. In each entry, the line I would ask a reader to linger on is the one field guides call *similar species*: the other things the same mark might be. Medicine has a grander name for that line, the differential, and it is where most of the honesty in an observation lives.

## Bell's Clerk

The art of reading people at a glance has a famous ancestry in medicine, and it is worth meeting before the entries, because it shows both what the art can do and what it is tempted to claim.

In the late 1870s a medical student at Edinburgh named Arthur Conan Doyle was chosen by the surgeon Joseph Bell to be his outpatient clerk. Doyle described the work in his memoir *Memories and Adventures* (1924). He arranged the outpatients, made "simple notes of their cases," and showed them in, one by one, to the large room where Bell sat among his dressers and students. The clerk, in other words, kept the chart. And the clerk noticed that his chief "often learned more of the patient by a few quick glances than I had done by my questions."

The case Doyle thought one of Bell's best went like this. A man came in, and Bell, without asking him anything, told him that he had served in the army, had not long been discharged, had been in a Highland regiment, as a non-commissioned officer, and had been stationed at Barbados. Each statement was met with "Aye, sir." Then Bell explained himself to the students. The man was respectful but had not removed his hat; soldiers do not, and a man long out of the army would have learned civilian ways. He had an air of authority, and he was plainly Scottish. As for Barbados, his complaint was elephantiasis, "which is West Indian and not British." To the audience it all seemed miraculous until it was explained, and then simple. When Doyle came to invent a detective a few years later, he recalled, he thought of his old teacher's "eerie trick of spotting details."

Sherlock Holmes inherits the trick and the theatre. At his first meeting with Watson, in *A Study in Scarlet* (1887), he says, "You have been in Afghanistan, I perceive," and later lays out the chain: a medical type with a military air, a face darkened by the tropics but wrists that are fair, a haggard look, a left arm held stiffly. In "A Scandal in Bohemia" (1891) he puts the doctrine into a sentence that medical teachers have been quoting ever since: "You see, but you do not observe." Watson has climbed the stairs to their rooms hundreds of times and cannot say how many there are. Holmes knows there are seventeen.

I admire this tradition and would not want a physician who had none of it. But the anecdotes share a structure worth noticing. In each, the observer speaks first and the subject confirms. *Aye, sir.* The method's successes are recorded and its misses are not, and Doyle, to his credit, let one clause into his account that the legend has generally left out: Bell's results were sometimes very dramatic, "though there were times when he blundered." Nobody tells the stories of the blunders. A tradition built from anecdotes is a record of hits.

Holmes, moreover, lives in a world arranged by his author. The tanned face could not belong to a man home from a sailing holiday, because Doyle put no such man in the story. In a clinic the similar species are always present. A sunburnt face belongs to a farmworker, or a golfer, or a grandfather who sat too long at a grandson's football game. Holmes himself, in the same Bohemian story, gives the warning I would keep before the trick: "It is a capital mistake to theorise before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts."

One more detail in Doyle's memoir belongs in a field guide. When he took over Bell's outpatients, Bell warned him that he would need to know Scottish idiom, and Doyle, "with the confidence of youth," said he had it. On one of his first days an old man told him he had "a bealin' in his oxter." Doyle had no idea what this meant. It meant an abscess in the armpit. The clerk's notes could only be as good as his grasp of the words, and the first observation in any consultation, before the hat or the wrists, is of language.

## Entries

**The gait.** *Field marks:* the walk from the waiting room to the chair, seen before a single question has been asked. A hand that brushes the door frame, or the edge of the counter, as though confirming that it is there. A shorter step on one side. A hesitation at the threshold where carpet gives way to tile. *Similar species:* new shoes; a long drive; a polished floor; a sore foot that has nothing to do with balance; the caution of someone who once fell and has not forgotten it. *Notes:* Walking speed can be measured with a stopwatch and a marked corridor, and physical therapists have argued, in a 2009 white paper by Stacy Fritz and Michelle Lusardi, that it deserves standing as [the sixth vital sign](https://scholarcommons.sc.edu/sph_physical_therapy_facpub/4). The argument is a good one. A number of that kind has the great virtue of comparability: next year's walk can be set beside this year's. The chart can hold the number. It holds the hand on the door frame only if someone decides that it belongs there, and the hesitation at the threshold almost never. What no number can do is choose among the similar species. That takes a question, and the best questions about a walk are often addressed to the person's day rather than their legs: whether the stairs at home have a rail, where the bathroom is at night, what happened the last time they fell, if they did.

**The pause.** *Field marks:* a silence before an answer, longer than the question seemed to need. *Habitat:* questions about alcohol, money, sleep, mood, safety at home, and the number of tablets actually taken. *Similar species:* the pause of translation, whether through an interpreter or inside a bilingual head; the pause of a hearing aid with a tired battery; the pause of grief; the pause of someone deciding whether this doctor can be trusted with the truth; the pause of someone counting honestly. *Notes:* Imagine a man who is asked how many drinks he has in a week, and who looks at the ceiling for so long that the question begins to feel, even to the person who asked it, like an accusation. The observer's mind has already supplied a reading: reluctance, perhaps concealment. Then the man says, "Do you mean a can or a glass? I'm working it out." The pause was arithmetic. He was being more careful than the patient before him, who said "socially" without a flicker. A pause invites a gentler question or a clearer one. It should never be written down as evasion, because evasion is a judgment about motive, and the observer has seen only a silence.

**The bag of bottles.** *Field marks:* a plastic bag of medicine bottles, brought from home and emptied onto the desk. *Notes:* The medication list in the chart records what was prescribed. The bag records what was dispensed and, with care, what was taken: one bottle still full months after it was filled, another nearly empty too soon, two bottles of the same medicine under different names from two pharmacies, a bottle prescribed to a spouse. Few documents in medicine are as candid as a kitchen cabinet tipped onto a desk, and few are as easily misread. The full bottle may mean the medicine was never taken; it may mean a newer bottle is waiting at home; it may mean the tablets were being halved to make them last. *Similar species,* here, are the stories behind the bottles, and the bottles alone cannot tell them apart.

**The quick yes.** *Field marks:* every part of a plan accepted at once, without a question, often with the eyes on the door. *Similar species:* genuine agreement; trust; deference to a physician's title; exhaustion; a parking meter running out; the wish to go home. *Notes:* The plan recorded as agreed may not be the plan that will be followed, and the record will not know the difference until the next visit, when it will usually call the gap by a word I would rather it did not use. A question that gives permission to doubt, such as which part of this seems hardest to do, tends to reveal more than asking whether everything is clear. "Is that clear?" has only one polite answer.

## The Answer That Has Changed

*Field marks:* today's account differs from the one in the record. A medicine listed as taken is not being taken. A symptom described as new is said to be years old. A history told one way is now told another.

Imagine a woman of sixty whose chart, from three years ago, says *never smoker*. Today, asked by a different nurse in a different way, she says that she smoked for about ten years in her twenties and stopped when her first child was born. Nothing in the encounter suggests that she is lying now, or that she lied then. What are the possibilities?

The question may have been asked differently. "Do you smoke?" is in the present tense, and the honest answer to it, from a woman who stopped thirty years ago, is no. The earlier form may have offered a box that someone ticked to move things along, or a default that nobody changed; electronic forms have a way of turning silence into an answer. She may have understood smoking to mean a lifelong habit and not counted a decade that ended before her children were born. The question may have been put in her second language, or through a relative who translated it as he understood it. She may, three years ago, have been ashamed of an old habit in front of a young doctor, and feel differently now. Or memory may simply have done what memory does, which is to file the distant past under a heading that no longer quite fits the facts.

None of these is deception, and several of them place the error in the record rather than in the woman. Yet the discrepancy matters clinically; a history of smoking, even an old one, can change which screening a person is offered. So the right response is neither suspicion nor a tactful silence. It is to say plainly what the chart says and ask what is true now: "The record from a few years ago has you as never having smoked. Is that right, or did we miss something?" The second clause matters. It places the likely error where it most often lies, with the people who asked and wrote, and makes it easy for her to correct them.

Then comes the part this guide is really about, which is the writing. The new note could simply replace the old entry, so that *former smoker* appears and *never smoker* vanishes. That is tidy, and slightly dishonest. Better to let the record show its own revision: former smoker, about ten years in her twenties; earlier record listed never-smoker; she explains that she understood the earlier question to be about current smoking. A future reader then knows not only the fact but its history, and will not be puzzled, years later, by two entries that disagree. Familiarity with a chart can be a peculiar way of being unprepared for the person it describes. A chart that shows how it was corrected is less likely to leave the next clinician unprepared in the same way.

## The Daughter Who Looks Away

*Field marks:* imagine a father asked how much he drinks. He answers, and the answer is not alarming. His daughter, in the corner chair, looks at the floor. *Range:* also found in sons, spouses, friends and the neighbour who drove. *Similar species:* disagreement; embarrassment at the question itself; fatigue; a message arriving on her telephone; a history of her own, with drink or with this father; grief at hearing aloud a number she already knew was true.

This is the hardest entry in the guide, because two people are in it and the observer stands in a different relation to each. The father is the patient. The daughter is a source, a witness, a companion, sometimes a caregiver who knows more about his evenings than anyone in the building; she is also a person with her own perspective, her own fears and her own stake in how the story is told. Her look is information. It is not testimony, and the chart should never turn an interpretation of her face into something she said.

There are two easy errors here, and they point in opposite directions. The first is to assume that the companion is the more accurate witness, because she is younger, or sober, or speaks the clinician's language more fluently, or simply because families are presumed to know. Sometimes she does know more. Sometimes she knows a version shaped by a quarrel, or by a worry that has outgrown its evidence, or by her own exhaustion. The second error is to treat the companion's account as interference, an obstacle between the clinician and the patient's own words. Both errors share a premise, that one account must be true and the other false. More often they are two partial views of the same evenings, taken from different chairs.

The father's agency matters, and so does his privacy. What he would tell a doctor alone he may not want said in front of his daughter, and the reverse is just as true. One way through is to offer each of them a few minutes alone, and to offer it as routine, something done for everyone, so that neither hears it as an accusation. In many families a parent's appointment is a family occasion, and a request that the daughter step outside can feel like an insult unless it is plainly ordinary. The question can also be put openly to both: "You both know his evenings better than I do. What would each of you say?" That sentence does two things at once. It invites the daughter's view without treating her face as a statement, and it lets the father hear that his own account is not being set aside.

And the record? If the daughter says, in her own words, that she is worried about her father's drinking, the chart can say so, and should say who said it, with her estimate attributed to her and his to him. If she says nothing and looks away, the chart has very little to say, and should say little. *Daughter present* is true. *Daughter appeared to disagree* is a guess wearing the clothes of an observation. The gap between those two sentences is the gap between a record that preserves what happened and a record that preserves what someone thought about what happened, and the second kind travels farther than it deserves.

## Korolyov on the Road

The title of this guide flatters its author, and I should say so before going further. *What the chart cannot see* implies a division of labour in which the physician in the room sees the real person and the record holds only data. Literature has long been fond of that division. It is not true, or not true enough to build on.

Henry James, in "The Art of Fiction" (1884), gave the observer's creed its best-known form. Experience, he wrote, is "a kind of huge spider-web of the finest silken threads suspended in the chamber of consciousness," and he told the aspiring novelist: "Try to be one of the people on whom nothing is lost!" To show what he meant, he recalled an English novelist who had been praised for her portrait of French Protestant youth. Her opportunity had consisted of a single moment in Paris: going up a staircase, she passed an open door and glimpsed some young Protestants in a pastor's household, seated at table round a finished meal. "The glimpse made a picture; it lasted only a moment, but that moment was experience." From it, James says with admiration, "she turned out her type."

For a novelist this is a gift, and James was right to prize "the power to guess the unseen from the seen." For a clinician writing a chart it is precisely the danger. A glimpse that turns out a type is what a record must not do, because the type will outlive the glimpse, and the person will be met through it.

Chekhov, a physician who never stopped writing stories, understood both sides of this. As a writer he is the master of the observation without its gloss. In "The Lady with the Dog" (1899), after Anna Sergeyevna has wept and said that he will despise her now, Gurov "cut himself a slice and began eating it without haste," a slice of watermelon, and there follows "at least half an hour of silence." Chekhov does not say what Gurov feels. He gives us the watermelon, which says more than an adjective could, and leaves the meaning to the reader. As a clinical note, it would be a model: the observable act, its duration, and no diagnosis of the soul.

In "A Doctor's Visit" (1898), Chekhov gives us a doctor who works the other way, and he does it so quietly that a reader can miss it. Another essay on this site, about [concerns raised at the door](../the-patient-who-says-one-more-thing/index.html), reads this story for what staying the night allows a doctor to hear. I want to read its first pages, before the doctor has met his patient at all. Korolyov, a professor's assistant from Moscow, is driven out to a factory to see the owner's daughter, and on the road he passes the workers going home. Chekhov tells us what he sees: "in their faces, their caps, their walk, he read physical impurity, drunkenness, nervous exhaustion, bewilderment." He has never been inside a factory. He has read about them, and has always thought that behind their quiet walls there must be ignorance and egoism on one side and unhealthy toil on the other. Here is a man reading field marks, gait included, and what he reads is his own opinion. When he meets his patient, the habit continues; watching her sob, he decides privately that what she needs is a husband. By morning he has built a large theory of the place and its unhappiness, and when she looks at him, it is "clear to him, too, that she thought so herself."

Whether she did, Chekhov never quite says. Korolyov is sympathetic and intelligent, and he may be right about her. But the story lets us see that his reading of her began as a reading of factories, formed on the road before she had spoken. Had he written a chart, it would have been a moving document, and it would have been partly about him.

I do not offer Korolyov as a villain. I offer him because he is the observer at his best and still mistaken in the way the best observers are, through the coherence of a view formed before the patient spoke. It is the best argument against my title. Human observation is not a clear window onto the person. It is coloured by mood, fatigue, fondness and dislike, by what the observer has read and what happened in the previous room, by his own class and language and expectations. The record, for all its flatness, has one great advantage over memory: it does not improve the story afterwards. A clinician may remember a patient as "always anxious"; the notes may show three calm visits and one frightened one. In that contest the folder is often the more faithful witness, and it can correct the physician's vanity as surely as the physician can correct the folder's dullness.

So the chart is not the villain of this guide either. What the chart cannot see is real. So is what the observer cannot see, and that includes, much of the time, the observer.

## What Belongs in the Folder

Should the chart see everything? The question sounds rhetorical until one tries to answer it. A person may disclose something because it matters to the conversation at that moment, to explain why she has not slept or why the tablets ran out, without expecting every shade of it to become a permanent description, available to whoever opens the file in the years to come. Some details are clinically important and deserve careful preservation. Others are provisional, or private, or meaningful only within the exchange that produced them. The ability to record more does not settle which is which.

Return to the pause. It can be observed. Its meaning cannot honestly be entered without further inquiry. "Hesitated before answering" is already a selection from a much richer encounter. "Appeared evasive" adds an interpretation that may travel much farther than the evidence supporting it. A later reader will meet the interpretation without the room, the question, the man counting cans, or any chance to hear him explain.

That such sentences travel is no longer only an intuition. In a study published in 2018 in the *Journal of General Internal Medicine*, Anna Goddu and colleagues gave 413 medical students and residents at an urban academic medical centre one of two chart notes about a single hypothetical patient, a young man with sickle cell disease. One note used stigmatising language; the other described the same facts neutrally. Readers of the stigmatising version rated the patient less favourably, and the residents among them chose less aggressive treatment for his pain. A second study, published in *Health Affairs* in 2022, used machine learning to search more than forty thousand history and physical notes at another urban academic medical centre for negative descriptors of patients, words such as *resistant* and *noncompliant*. Black patients had 2.54 times the odds of White patients of having at least one such descriptor in their notes. The first study shows that one clinician's words shape the attitudes and decisions of the next. The second shows that the words are not evenly distributed.

So before a sentence enters the folder I would want it to pass a few plain tests. Is it relevant to this person's care, now or foreseeably? Is it verified, and if it is not, does it say so? Does it need to be preserved, or did it belong to the conversation rather than the file? And would the patient recognise it, if she read it, as a fair account of what she said and did? The last test is no longer hypothetical. Under federal rules that took effect in 2021, patients in the United States can, with limited exceptions, read their clinical notes, often on their telephones. I do not think this should make notes timid. It should make them accurate, which is a different thing.

The chart's limits can also protect. Imagine a man who mentions, to explain his insomnia, that his marriage is in difficulty. He has told a doctor something true and relevant. The record needs to know that there is strain at home bearing on his sleep; it does not need the substance of the quarrel, the name of anyone involved, or the clinician's private view of who was in the wrong. Imagine a woman who explains, in tears, why she dreads a particular examination. The record should carry a note that this examination requires care, consent and time; the next clinician does not need the memory itself, and she may not want it retold. American law already recognises that some records deserve higher walls; the records of specialised substance-use treatment programmes, for one, are held under a separate federal confidentiality rule. Ordinary judgment can build smaller walls of its own.

There is an opposite failure, and an honest guide should name it. Vagueness can be a way of avoiding responsibility. *Psychosocial stressors* may protect a patient's dignity; it may also deprive the next clinician of what she needs in order to help. The governing virtue is proportion: enough context to help a later reader ask a better question, and not so much that the sentence answers the question for her.

## The Right to Amend

The chart is often imagined as a verdict. It is better understood as an account that other people, including its subject, are entitled to answer.

The mechanism exists, and the essay here on [notes that patients can now read](../when-patients-started-reading-their-notes/index.html) describes it from the patient's side. Under the federal privacy rule, a person may ask a healthcare provider to amend information about them in the record. The provider may decline if the information is accurate and complete; but the person may then file a written statement of disagreement, which is appended or linked to the disputed entry, together with any rebuttal. What I find quietly admirable in this procedure is what it does not do. It does not erase. The original sentence remains, and beside it stands the person's objection. At the point of dispute, the record becomes two voices instead of one.

Most corrections never need the formal procedure, and a clinic that sent every misunderstanding through it would have failed long before. The same principle can work inside the ordinary sentence. Uncertainty can be left visible: *patient reports*, *per daughter*, *unclear whether*, *to clarify at next visit*. These phrases are sometimes mocked as defensive hedging. Used well, they are the record admitting how it knows what it says. A sentence that names its source can be checked. A sentence that has forgotten its source can only be believed or ignored.

Return to the woman whose smoking history changed. If the new note preserves the earlier entry, the correction and her explanation, it has done in miniature what the amendment rule does formally. It has let her answer the chart. And it has kept a piece of useful history, the fact that the record was once wrong in this particular way, which may help someone notice the same mistake in another file.

## What a Machine Might Notice

Software now reads records more thoroughly than any clinician has time to, and it is beginning to listen in examination rooms as well. It is fair to ask whether some of what the chart cannot see will soon be seen by something else.

Some of it will, and that is welcome. A program can measure walking speed from a sensor more reliably than a glance. It can notice that a prescription has not been refilled in four months, which is the bag of bottles in digital form. It can find that a sentence about a patient has been carried forward unchanged through a dozen notes, or lay a weight trend beside an appetite that nobody has asked about. These are detections of patterns in recorded data, and they can help a clinician ask a better question.

It is another matter to claim that software can read what the daughter's look meant. Faces have attracted a large commercial ambition of exactly that kind, and the science has not kept pace with it. In 2019 a review in *Psychological Science in the Public Interest*, led by Lisa Feldman Barrett, examined the evidence for the common assumption that emotional states can be read from facial movements. The authors found that people do sometimes scowl when angry and smile when happy, more often than chance would predict; but that the ways people express emotion vary substantially across cultures, situations and individuals, that the same configuration of the face can express more than one emotion, and that a scowl often communicates something other than an emotional state at all. A program that labelled a lowered gaze as *disagreement* would be doing, with confidence, what this guide forbids the clinician to do even tentatively.

The most promising use of pattern-finding in this territory may point the other way. The study of negative descriptors used a machine to read clinicians' sentences, not patients' faces. That seems to me the right direction for the instrument: not to watch the person in the chair more closely, but to help the people who write about her see the habits in their own prose.

## A Last Entry, for the One Who Writes

**The observer.** *Field marks:* tired in the late afternoon; still carrying the diagnosis from the previous room; fond of this patient, or wary of him; holding a theory formed on the road in, as Korolyov held his; sharing a language with the patient, which brings its own risk, since a shared language conceals the moment when a word has been translated into the listener's meaning. *Similar species:* insight; intuition; the experienced eye. *Notes:* The observer cannot be removed from the guide, because every entry in it was written by one. The most that can be asked is that he appear in it, visibly, as a source with his own range of error.

Wordsworth, in the lines written above Tintern Abbey in 1798, describes a state in which, "with an eye made quiet by the power / Of harmony, and the deep power of joy, / We see into the life of things." He meant a near-mystical condition, the body's own motions almost suspended, and I would not press it into service as clinical method. I borrow only the first half of the phrase and leave the second. An eye made quiet is an eye that has not yet decided. It has not finished the sentence or chosen among the similar species. That is the eye I would want in the room. I would trust no eye, my own included, that claimed to see into the life of things and then wrote down what it found there.

The essay from which this series grew, [The Primary Care Office After the EHR](../the-primary-care-office-after-the-ehr/index.html), proposes three tests for any technology that enters the clinic: fidelity, equality and embodiment. This guide belongs mostly to the first. Fidelity here runs in two directions. The record should be faithful to the person, which means proportionate, attributed, and willing to show its corrections. And whoever reads the record should stay faithful to the possibility that the person has since moved on from it, disagreed with it, or simply exceeded it.

Picture, finally, an examination room between visits: a closed folder on the desk, and beside it an empty chair. For the moment the folder holds everything there is to know, and it holds a good deal of it well: the dates, the numbers, the medicines, a note about the rail on the stairs at home, a correction appended to an old entry, a daughter's concern in her own words. In a few minutes the chair will hold more information than the folder ever has, and almost none of it will arrive already written down. What I would ask of the folder is modest. It should help the next person in the room ask a better question. It should never tell her that the question has already been answered.

## Evidence and Editorial Note

**Documented findings.** The essay relies on the author's public biography only in general terms: a family physician practising in California since 2010, whose first language is Punjabi. Roger Tory Peterson's *A Field Guide to the Birds* (Houghton Mifflin, April 1934); his apprenticeship in the Bronx County Bird Club under Ludlow Griscom and Griscom's identification with binoculars rather than a rifle; earlier books built for collectors on anatomical differences and measurements; the quoted phrase about impressions, patterns and distinctive marks; arrows to key field marks; a first printing of 2,000 copies, sold out within a week: Linda Hall Library, *Drawn from Nature*, "The Making of a Field Guide" (linked). Arthur Conan Doyle's account of serving as Joseph Bell's outpatient clerk at the Edinburgh Infirmary, the notes he made, Bell's quick glances, the discharged Highland soldier from Barbados, the admission that Bell sometimes blundered, the "eerie trick of spotting details," and the "bealin' in his oxter": Doyle, *Memories and Adventures* (1924), Project Gutenberg text. Holmes's lines from *A Study in Scarlet* (1887) and "A Scandal in Bohemia" (1891): Project Gutenberg texts. Henry James, "The Art of Fiction" (1884), as collected in *Partial Portraits*: Project Gutenberg text. Chekhov's "The Lady with the Dog" (1899) and "A Doctor's Visit" (1898), quoted in Constance Garnett's public-domain translation, *The Lady with the Dog and Other Stories*: Project Gutenberg text. Wordsworth, "Lines Composed a Few Miles above Tintern Abbey," *Lyrical Ballads* (1798): Project Gutenberg text. Walking speed proposed as the sixth vital sign: Fritz and Lusardi, white paper, *Journal of Geriatric Physical Therapy* 32(2), 2009 (catalogue record linked). Goddu and colleagues, "Do Words Matter? Stigmatizing Language and the Transmission of Bias in the Medical Record," *Journal of General Internal Medicine* 33(5), 2018: randomised vignette study of 413 physicians-in-training; PubMed abstract and published correction. Sun, Oliwa, Peek and Tung, "Negative Patient Descriptors: Documenting Racial Bias in the Electronic Health Record," *Health Affairs* 41(2), 2022: 40,113 history and physical notes; adjusted odds 2.54; PubMed abstract. Barrett, Adolphs, Marsella, Martinez and Pollak, "Emotional Expressions Reconsidered," *Psychological Science in the Public Interest* 20(1), 2019: PubMed abstract. The right to request amendment, denial where information is accurate and complete, statements of disagreement, rebuttals and linkage to the disputed record: 45 CFR 164.526, eCFR text. The site's essays on late concerns and on patients reading their notes, which treat "A Doctor's Visit" and the amendment rule from other angles, are linked. Patients' access to clinical notes under federal information-blocking rules from 2021, and the separate federal confidentiality rule for records of specialised substance-use treatment programmes (42 CFR Part 2): standard regulatory knowledge, stated in general terms. The three tests of fidelity, equality and embodiment: the author's flagship essay (linked). Observations about gait, pauses, companions, medication bottles, smoking histories and agreement are general clinical knowledge, stated without statistical claims. **Vendor representations.** None. Software that summarises records or listens in examination rooms is described in general terms; no product is named or assessed.

**Recollection and tradition.** Every entry is general or hypothetical. The man counting drinks, the woman whose smoking history changed, the father and daughter, the man with insomnia and the woman who dreads an examination are imagined illustrations, introduced as such; they do not describe real patients, and no remembered encounter of the author's is described. The remark that a parent's appointment is a family occasion in many families is a general observation, not a claim about all Punjabi or Sikh families. The essay does not assert that records cannot document observation, or that any expression, pause or gesture proves a diagnosis, motive or state of mind.

**Analysis and forecast.** The field-guide framing, the comparison between records and specimen-based identification, the readings of Doyle, James, Chekhov and Wordsworth, the tests proposed for what belongs in a record, and the suggestions about questioning, attribution and correction are the author's interpretation and preference as of November 6, 2026; they are not documentation, legal or compliance guidance. Remarks about what software may detect are general and not a forecast about any product. **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

- [A Record Repeated Is Not a Record Corroborated](https://personal.kpsgill.com/essays/a-record-repeated-is-not-a-record-corroborated/)
- [The Older I Get, the More I Listen](https://personal.kpsgill.com/essays/the-older-i-get-the-more-i-listen/)
- [Two Countries in One Body](https://personal.kpsgill.com/essays/two-countries-in-one-body/)