An examination hall of the old kind is a room built to remove every source of knowledge but one. Bags are left at the door and books are left in the bags. An invigilator walks the aisles to make sure that nothing written has come in on a cuff or the inside of a wrist, and when the papers are handed out and the clock begins, each candidate sits alone with a sheet of questions and whatever he has managed to carry in his head. The arrangement is so ordinary that it rarely strikes anyone as strange. It is in fact a very particular theory of knowledge, enforced with furniture. It holds that what a person knows is what he can produce unaided, on demand, within a time limit, cut off from every other mind and every page.
Students of the sciences in the 1990s, in India as nearly everywhere else, were commonly examined in some version of that room. The years 1994 to 1996, when I was a pre-medical student at Khalsa College in Amritsar, fall in that decade, but I am not going to reconstruct any examination from them. The room matters here as an idea, because for a long time it was also a fair description of medicine. The patient’s words arrived in the consulting room; the findings arrived under the hands; the knowledge needed to make sense of them had been learned years before and was carried in. The chart held fragments. The physician’s head was where the fragments met. It was the place where the whole system, such as it was, came together, and the examination hall was training for that fact.
That is no longer true, or not in the same way. The physician now works inside a ring of other systems: records that remember every value, references that update themselves, software that will answer almost any question typed into it. It is natural to ask whether the head has been demoted, whether the years spent filling it were a toll paid to an older arrangement that has since been dismantled. I think the question is real, and I think the usual answers to it, nostalgic on one side and dismissive on the other, both miss what memory actually does.
The title of this essay calls the mind the first database. I mean “first” in two senses. It came first in history, long before shelves and servers. More importantly, it still comes first in sequence. In every encounter it is the database consulted before any other, and it is the one that decides whether any other will be consulted at all. It is also an odd kind of database. It has a superb schema and a fast index. It keeps almost no dates, it rarely records its sources, and it has no log of what it has changed. Its value and its danger come from the same design, and the rest of this essay is an attempt to say which is which.
A Schema Before the Facts #
In the 1940s the Dutch psychologist and chess master Adriaan de Groot showed chess positions to players of different strength for a few seconds and asked them to reconstruct the board. Masters could do it almost perfectly. Weaker players managed a fraction. It would have been easy to conclude that masters simply had better memories, and in 1973 William Chase and Herbert Simon ran the experiment that tested the idea. They added a control. Alongside positions from real games, they showed boards with the same number of pieces placed at random. On the random boards the master’s advantage largely disappeared. He did not have more memory. He had organised memory. Where the novice saw twenty-odd pieces, the master saw a handful of meaningful groups, a castled king with its pawn shield, a familiar pressure along a diagonal, and he remembered those, the way a reader remembers a sentence rather than its letters.
The lesson of the random boards is the one that matters for medicine. Expert memory behaves like a set of expectations about how things usually go together, and it is only as good as the world’s willingness to resemble those expectations.
Medical education has its own version of the finding. In 1990 Henk Schmidt, Geoffrey Norman and Henny Boshuizen proposed in Academic Medicine that clinical expertise rests less on superior reasoning or on deep knowledge of disease mechanisms than on structures they called illness scripts: organised knowledge built around prototypical and actual patients, rich in information about a disease, its consequences, and the circumstances in which it tends to arise. Students who had not yet seen many patients, they observed, reasoned instead from causal models of pathophysiology, working their way forward from mechanism. The experienced clinician recognises; the student derives. Both are using memory. They are using it organised in different ways.
A script is a schema, and a schema is a set of slots that expect to be filled. This is what lets an experienced reader of illness notice not only what is present but what is missing: the finding the pattern predicts that has not appeared, the detail in the story that sits a little wrong. Nobody notices an absence without an expectation. Recognition, in this sense, depends on vocabulary. To notice that a cough is dry, you need the category of a dry cough and some sense of why it might matter. To notice that a pulse is irregular in a particular way, you need a name for the way. A term learned well is a question built in advance.
I want to be careful here, because there is a sentimental version of this argument that is false. It says that if you pack enough facts into a head, insight will grow among them like moss. It does not. Facts stored side by side are a list, and a list cannot recognise anything. A student who can recite every station of the Krebs cycle may still be unable to say what would happen to the cycle if one enzyme failed; I have written elsewhere about that difference between the names and the thing named. The schema is made by use: by problems worked, cases compared, by the slow discovery that two facts learned in different chapters bear on each other. What the pre-medical sciences gave a student, at their best, was less a stock of facts than a vocabulary of mechanism. Gradient, equilibrium, feedback, saturation, compensation. Each is a way of looking that can be carried from one subject to another, and each tells the eye what to look for before there is anything to see.
So the first service of memory is a strange one. It does not answer questions. It makes questions possible, by giving the mind the categories in which the world can show up as worth asking about.
Is It Perchance I? #
Some sixteen centuries ago, in the tenth book of his Confessions, Augustine of Hippo walked into his own memory as if into a building, and his description of it is still among the most exact accounts of retrieval anyone has written. In E. B. Pusey’s translation he comes “to the fields and spacious palaces of my memory,” where innumerable images are stored. When he calls for something, he reports, some things come at once; others must be sought longer, as if fetched from an inner room; and others “rush out in troops, and while one thing is desired and required, they start forth, as who should say, ‘Is it perchance I?’” He drives these away “with the hand of my heart” until the thing he wanted appears.
Augustine had been a professional teacher of rhetoric, and his memory was trained in the way such men’s were. What he describes is how association works even in a well-ordered and well-trained mind. The index does not return one answer. It returns candidates, ranked by resemblance, and the candidates are eager. They present themselves as if each had a claim.
Punjabi has a precise word for what happens when one of them is wrongly accepted. ਭੁਲੇਖਾ (bhulekhā) is a misapprehension, and in ordinary speech it is the word for mistaking one person for another: a figure seen from behind in a crowd, the gait, the set of the shoulders, the colour of a turban, all exactly right, until he turns around and is a stranger. Mainū̃ bhulekhā pai giā, one says afterwards. A mistaking fell upon me. The grammar is honest. The error is not described as something I did but as something that happened to me, because the resemblance was real. Only the identity was false.
Clinical association works in the same way, and it is worth slowing down to watch it. Imagine a physician who, years earlier, looked after a patient troubled for months by a dry cough that turned out to come from her blood-pressure medicine, from a class of drugs well known to cause such a cough in a minority of the people who take them. It was a satisfying case. The answer was found by thinking, and the cure was the removal of a pill. Now, in this hypothetical, a new patient on the same class of medicine sits in the same chair and describes a cough. The old case rushes forward. Is it perchance I?
It may well be. But a cough in a person who takes that medicine is also a cough in a person who may have asthma, or acid rising from the stomach at night, or an infection that has not finished, or a smoker’s lungs, or a heart that has begun to fail, or something growing that should not be. The memory has done its proper work: it has proposed a hypothesis. It has also done something improper. By the vividness of its own satisfying ending, it has given that hypothesis a head start over the duller possibilities, which have no story attached.
In 1973 Amos Tversky and Daniel Kahneman described a mental shortcut they called availability: people judge how frequent or likely something is by how easily examples come to mind. The shortcut is often sound, because common things are usually easy to recall. It goes wrong when ease of recall is driven by something other than frequency, by recency, drama, emotion or the pleasure of a solved puzzle. Medicine supplies all of those. The case that went badly, the case that was diagnosed brilliantly, the case discussed at length with a teacher: these are the memories that rush out in troops, and they are not a fair sample of the world.
None of this is an argument against association. Without it nothing would be proposed, and a mind that proposed nothing would have to examine every possibility in the textbook from the first page, which would be paralysis dressed as caution. The danger is narrower and more interesting. It lies in the moment when a proposal begins to feel like a finding. A ਭੁਲੇਖਾ lasts exactly until the person turns around. The clinical equivalent of turning around is whatever could show the resemblance to be wrong: the question that would not fit the remembered case, the examination, the result. Memory’s resemblances are invitations to look. They become dangerous only when they are accepted in place of looking.
Without a Date or a Source #
A record in a well-made database carries more than its content. It carries the date it was entered, who entered it, where it came from, and when it was last changed. These fields are dull, and they are the reason a careful reader can trust or distrust what he finds. Memory keeps almost none of them, and keeps the few it does keep badly.
Augustine noticed this too, and his puzzlement about it is one of the most quietly remarkable passages in the book. Of the things he had learned in the liberal sciences, he says, he holds in memory not images but the things themselves, and then he asks how they got there. He has gone over all the avenues of the senses, he writes, and cannot find the one by which they entered. “Whence and how entered these things into my memory? I know not how.” For Augustine the mystery pointed upward, toward an account of knowledge as a kind of recognition of what the soul already contained. For a physician it points somewhere more mundane and more urgent. The knowledge is present. Its provenance is gone.
Psychologists have studied this under the name of source monitoring. In a review published in 1993, Marcia Johnson and her colleagues argued that remembering where a piece of information came from is a separate judgment from remembering the information, made on different evidence and liable to its own errors. People confuse what they read with what they were told, what they were told with what they imagined, and a reliable source with an unreliable one. In memory, a fact from a lecture, a fact from a textbook already several years old on the day it was bought, a remark from a senior in a corridor and a line from a newspaper all tend to shed their labels and settle into one undifferentiated stock called what I know.
The missing date is worse than the missing source, because nothing in the experience of remembering warns of it. A fact learned thirty years ago feels, when recalled, exactly as fresh as a fact learned yesterday, since the recalling is happening now. The feeling of knowing is always stamped with today’s date.
Punjabi, like Urdu, has a word that holds the problem in its two meanings. ਜ਼ੁਬਾਨੀ (zubānī), from the Persian for tongue, means “by heart”: a poem known zubānī is a poem one can say without the book. It also means oral, unwritten, as distinct from documented: a zubānī instruction is one given by word of mouth, with nothing on paper to show for it. The two senses belong together. What we know by heart has the standing of an oral order. It may be accurate and it may be wise, but there is no document to point to when someone asks where it came from.
What fills the space where the date and the source should be is confidence, and confidence turns out to be a poor witness. In a study published in 2013, Ashley Meyer, Hardeep Singh and colleagues gave 118 general internists in the United States four validated case vignettes, two easier and two more difficult, and asked them after each stage of the case for their diagnoses and their confidence. The physicians reached the correct diagnosis in 55.3 percent of the easier cases and 5.8 percent of the harder ones. Their confidence, on a scale of ten, fell only from 7.2 to 6.4. Higher confidence was associated with fewer requests for additional tests. The study has the limits of its design: these were written cases on a screen, not patients in a room, and the physicians were recruited from an online community. But its central finding fits a wider literature, reviewed in 2008 by Eta Berner and Mark Graber, which argued that physicians as a group tend to underestimate how often their diagnoses are wrong.
The mechanism is what matters for this essay. Confidence is the feeling that a memory comes wrapped in, and it is that feeling, not any checked property of the memory, that decides whether the mind goes looking for more. In a database, trust would rest on the date and the source attached to the record. In the mind, familiarity, the sense of I know this, stands in their place, and it impersonates verification very well.
What Changes, and What It Costs to Be Wrong #
None of this means that everything should be checked every time. A physician who looked up the position of the liver before examining an abdomen would be absurd, and a physician who could not proceed without software would be helpless whenever the software was. The real question is which things deserve suspicion, and here the old distinction between knowing and looking up becomes useful again if it is drawn along the right lines.
Some knowledge is structural and changes slowly: anatomy, the broad architecture of physiology, the logic of feedback and compensation, the meaning of a fever or of a pulse that races. Much of what the pre-medical years gave belongs here. Other knowledge is volatile by its nature. Numbers set by committees are the clearest case: thresholds, targets, doses, intervals, the age at which a screening test begins or stops. Lists of drug interactions grow whenever new drugs arrive. Recommendations shift when trials report. These things are not less true than the location of the liver. They are a different kind of truth, a judgment about where to draw a line on a continuous risk, and judgments are revised.
Blood pressure is the plainest example. A generation of physicians learned 140 over 90 as the line at or above which an adult had hypertension. In November 2017 an American guideline from the American College of Cardiology, the American Heart Association and nine other organisations lowered the line to 130 over 80, and the European guidance that followed in 2018 kept the older figure. A physician who carries “140 over 90” in memory is carrying the decision of a committee that later decided otherwise on one side of the Atlantic and not on the other, and the memory does not record the committee. Neither number is a law of nature. Each is a carefully argued choice, and on the page where it was published each carries a date and the names of the people who made it. In memory, both are simply true.
Then the picture complicates itself, as pictures in medicine do. Structural knowledge changes too. For most of the twentieth century peptic ulcer was taught as a disease of acid, diet and stress. The discovery that a bacterium, Helicobacter pylori, plays a central role in gastritis and peptic ulcer, honoured with the Nobel Prize in 2005, is the classic case, and another essay in this series tells it. Concepts do move. The difference is that conceptual change is rare and loud. It is argued over, reported, awarded prizes. Numerical change is frequent and quiet. It happens in a table deep inside a long document that few people read through.
From these distinctions comes something like a working habit, which I offer as reflection rather than rule. It is wise to distrust a memory in proportion to how numerical it is, how recently the field around it has been moving, whether it was set by a committee, whether it was learned once and seldom used since, how much harm it would do if wrong, and how far the patient in front of you differs from the patients on whom that memory was trained. The last of these is easy to forget. A fact learned from adults may not hold for a child, a fact learned from the young may not hold for the very old, and memory does not file its knowledge by the population it came from.
There is a circularity in all this that should be admitted. The decision about what to check is itself made by memory. Nothing outside the head taps the physician on the shoulder and says this one has changed. For that reason the most valuable things a trained memory stores may not be facts at all but warnings attached to facts: a small flag that says this number moves, or this is where people get hurt, or the last time you were certain about this, you were wrong. A memory full of facts and empty of flags is the dangerous kind. A memory with fewer facts and well-placed flags knows when to stop and look.
The Question Before the Search #
Some twenty-four centuries ago Plato’s Meno put to Socrates a puzzle that every search engine has quietly inherited. In Benjamin Jowett’s translation: “And how will you enquire, Socrates, into that which you do not know? What will you put forth as the subject of enquiry? And if you find what you want, how will you ever know that this is the thing which you did not know?” Socrates restates it as a dilemma. A man cannot enquire into what he knows, because he has no need to, and cannot enquire into what he does not know, because he does not know what to look for. He calls it a tiresome dispute, and answers it with his theory that learning is recollection, demonstrated by drawing out of an untaught slave boy the steps of a geometrical proof.
Augustine, in effect, answers Meno from the inside. A woman who has lost a coin, he writes, would never find it unless she remembered it, because she could not recognise it when it turned up. When we search our memory for something forgotten and are offered the wrong thing, we reject it until the right thing comes; and we can do that only because we have not wholly forgotten. “What then we have utterly forgotten, though lost, we cannot even seek after.” William James, in The Principles of Psychology in 1890, gave the same insight its modern form when he described trying to recall a forgotten name. The state of consciousness, he wrote, contains a gap, “but no mere gap. It is a gap that is intensely active.” Wrong names are proposed and the gap rejects them: “They do not fit into its mould.”
An intensely active gap is what a prepared question is. A search box, however capable, answers the question that is typed into it, and the typing requires that the gap already have a shape: a suspicion, a term, a category, some sense of what would count as an answer. A person who has never heard that a particular complication exists cannot search for it. A person who knows it exists but not what it is called will search clumsily. A person who knows its name, its usual company and the circumstances in which it hides will find in a minute what the first person could not find in a day. Retrieval amplifies the prepared mind and does surprisingly little for the unprepared one. This is the deepest reason the years of filling the head were not wasted. They built the shapes of the gaps.
Background knowledge does more than frame the question. It reads the answer. Retrieved material does not arrive with its relevance marked. It takes knowledge to see that a page is out of date, that a study was done in a population unlike the patient, that a recommendation applies to a different stage of the same disease, that two sources which appear to disagree are in fact answering different questions. The search returns text. Memory supplies the judgment of what the text is worth.
A fair objection is that the machines have begun to ask the questions themselves. Records now volunteer warnings about interactions nobody inquired about; software will turn a description of symptoms into a list of possibilities without being asked for any one of them. If the system supplies the question, perhaps the gap no longer needs a shape. Some of this is real and welcome. A prompt that catches what a tired mind passed over is a gift, and it would be vanity to refuse it. But the volunteered answer has a Meno problem of its own. It works from what was entered, and what was entered was chosen by someone’s attention, prepared or unprepared: the symptom that was asked about, the finding that was looked for, the part of the story that was written down. A machine can widen the list of candidates. It cannot by itself tell which of them fits the patient in the room, and what it hands the clinician is, in the end, a fresh troop of possibilities, each one asking Is it perchance I? The sorting that Augustine did with the hand of his heart still has to be done by somebody.
But the same structure that makes search possible also steers it, and here the argument has to turn against itself. In 1942 Abraham Luchins gave people a series of puzzles about measuring quantities of water with jars of different sizes. The early puzzles could all be solved by one long, roundabout formula. When later puzzles arrived that could be solved by a much simpler route, most people who had learned the long formula kept using it, and some failed outright on a puzzle the old formula could not solve. Luchins called the effect Einstellung, a mental set. In 2008 Merim Bilalić, Peter McLeod and Fernand Gobet looked at the same effect in expert chess players, using positions that offered both a familiar, longer winning plan and a shorter, better one. Having found the familiar plan, the players said they were searching for something better. Their eye movements showed that they kept looking at the squares connected with the plan they had already found.
That finding should give any expert pause. The search was sincere. It was also, quite literally, a tour of what the searcher already believed. In clinical life this is the mechanism behind what is sometimes called premature closure: the first well-formed idea controls where attention goes next, and every subsequent look confirms it because every subsequent look is directed by it. The prepared question is the condition of finding anything. It can also be the wrong question, asked with great skill.
There is no formula that cures this, but there is an orientation. It is to ask, at least once, the question designed to disconfirm: what else could this be, and what would I expect to see if I were wrong? Asked in earnest, it deliberately borrows a different schema from memory, so that one shape of the gap is not the only shape available.
Errors and Omissions #
At the foot of countless shop bills in North India, in small print beneath the total, there used to be, and often still is, a short formula: bhūl-chūk lenī-denī, which a Punjabi shopkeeper would read as ਭੁੱਲ-ਚੁੱਕ ਲੈਣੀ-ਦੇਣੀ. Errors and omissions to be taken or given. It is the bazaar’s cousin of the English “E. & O.E.”, and it has a moral grammar worth admiring. It assumes that mistakes occur. It provides for correction in either direction, to the buyer’s benefit or the seller’s. It accuses no one. Correction is written into the transaction in advance, as part of what a bill is, not as a scandal that interrupts it.
The word at its head is ਭੁੱਲ (bhull), and it belongs to the same family as ਭੁਲੇਖਾ and as ਭੁੱਲਣਾ (bhullṇā), to forget. The language files forgetting and error side by side, as if it understood that most of our mistakes are not inventions but lapses: something once known that slipped, or something half known that was taken for the whole.
Medicine has not always been as gracious as the shop bill. The examination hall left behind it a moral residue, and the residue runs in exactly the wrong direction. In the hall, consulting a source was cheating, because what was being tested was the head alone. Many clinicians carry some trace of that feeling into practice, where it becomes a faint shame at looking something up, especially in front of a patient or a junior, as though the reference book were a crib sheet smuggled in on a cuff. But in practice the head is not what is being tested. The patient’s care is. A physician who checks a dose in front of a patient is not confessing ignorance. He is doing the part of the job the examination never measured.
Checking is the easier discipline. The harder one is revising: letting the check change a belief that has been comfortable for years. Here the most useful advice I know comes from Charles Darwin, in the autobiographical recollections he wrote for his family in 1876, published after his death. For many years, he says, he followed a golden rule: whenever a published fact, a new observation or a thought came across him that was opposed to his general results, he made a memorandum of it “without fail and at once; for I had found by experience that such facts and thoughts were far more apt to escape from the memory than favourable ones.”
That is an extraordinary admission from one of the great observers in the history of science, and it is exactly right. Memory is not a neutral store. It edits, and it edits in favour of what we already think. The contrary fact is the one that slips away, and so the contrary fact is the one that must be written down at once. Darwin’s remedy was a habit of writing that made up for memory’s bias.
A clinician can borrow the habit in a modest form. When a check changes what you believed, record the change in a line: what I thought, what I found, where I found it, and when. That line gives the corrected belief what memory never gives anything, a date and a source. It also turns correction into something closer to the shop bill’s lenī-denī, an expected entry in a running account rather than a humiliation. A physician who has kept such lines for years will find that the record reads less like a list of failures than like a map of where the field moved, and of where his own mind was most inclined to hold on.
Extensive, Yet Hazy #
In the same recollections, describing his own mental powers with characteristic flatness, Darwin wrote: “My memory is extensive, yet hazy: it suffices to make me cautious by vaguely telling me that I have observed or read something opposed to the conclusion which I am drawing, or on the other hand in favour of it; and after a time I can generally recollect where to search for my authority.” He added that he had never been able to remember a date or a line of poetry for more than a few days.
I know of no better description of the memory a physician should want now. Extensive in structure: wide enough to recognise, to propose, to notice what is missing. Hazy on the volatile details, and aware of being hazy. Quick to say careful, there is something here. Good at knowing where the authority lives. The examination hall was designed to reward a different memory, one that prized exact recall of exactly the things that change. Yet this is no feeble memory. It has learned its own place in a larger system, and that place is at the front.
So the answer to the question with which I began is neither the nostalgic one nor the dismissive one. The physician’s head never truly held the whole system. Even in the years when it was the only place the fragments could meet, it held a working copy, partial and undated, and the best physicians of those years knew it and kept their books close. What has changed is that the rest of the system is now within reach, which makes the head’s real work easier to see. That work was always to recognise, to propose, to warn and to ask. The years spent filling it built the shapes of the questions. They did not, and could not, make the answers permanent.
Augustine, standing in the palaces of his memory, was overcome by their size and by something stranger. “Nor do I myself comprehend all that I am,” he wrote. “Therefore is the mind too strait to contain itself.” A mind that cannot contain itself was never going to contain medicine. The examination hall was not wrong to send students in with nothing but their heads. It was wrong only if anyone believed that the head was meant to stay alone.
Socrates, once the demonstration with the slave boy is over, offers a sentence that every confident clinician might keep somewhere near the reference shelf. “Some things I have said of which I am not altogether confident,” he admits. But one thing he will defend in word and deed: that “we shall be better and braver and less helpless if we think that we ought to enquire.” That is the right posture for the first database. It is consulted first because it is the one that knows to consult the second. Its knowledge is the condition of every search, and its doubt is the reason the search happens at all. The longer essay at the centre of this series follows the question of where knowledge lives, and who controls it, through three decades of records and software. Whatever the answer proves to be for the systems, the mind at the front of them will matter more as they grow, provided it keeps two things together: enough knowledge to know what to ask, and enough humility to go and check.
Evidence and Editorial Note #
Documented findings. Public biography relied on: pre-medical studies at Khalsa College, Amritsar, 1994–1996, and the author’s work as a physician. Augustine, Confessions, Book X, quoted from E. B. Pusey’s translation (Project Gutenberg text): the “fields and spacious palaces” of memory, the images that “rush out in troops” asking “Is it perchance I?”, the liberal sciences whose entry he cannot trace (“I know not how”), the woman with the lost coin, “what then we have utterly forgotten… we cannot even seek after”, and “the mind too strait to contain itself”. Plato, Meno, in Benjamin Jowett’s translation (Project Gutenberg text): Meno’s question about enquiry, Socrates’s restatement, and his admission “Some things I have said of which I am not altogether confident” with the defence of enquiry. William James, The Principles of Psychology (1890), chapter “The Stream of Thought” (Project Gutenberg text): the “intensely active” gap in recalling a forgotten name. Charles Darwin’s autobiographical recollections, published 1887 (Project Gutenberg text): the “golden rule” of recording contrary facts at once, and “My memory is extensive, yet hazy”. Henk Schmidt, Geoffrey Norman and Henny Boshuizen, “A cognitive perspective on medical expertise: theory and implication”, Academic Medicine 1990;65:611–621 (PubMed abstract): illness scripts and the contrast with students’ pathophysiological reasoning. Ashley Meyer, Hardeep Singh and colleagues, “Physicians’ diagnostic accuracy, confidence, and resource requests: a vignette study”, JAMA Internal Medicine 2013;173:1952–1958 (PubMed abstract): 118 internists, accuracy 55.3% versus 5.8%, confidence 7.2 versus 6.4 of 10, higher confidence associated with fewer test requests; vignette design noted as a limitation. Eta Berner and Mark Graber, “Overconfidence as a cause of diagnostic error in medicine”, American Journal of Medicine 2008;121(5 Suppl):S2–S23 (PubMed abstract). Merim Bilalić, Peter McLeod and Fernand Gobet, “Why good thoughts block better ones”, Cognition 2008;108:652–661 (PubMed abstract): experts’ eye movements after finding a familiar solution. The 2017 ACC/AHA high blood pressure guideline, its date and its eleven sponsoring organisations: Whelton and colleagues, Hypertension 2018;71:e13–e115 (PubMed 29133356); the change in the diagnostic threshold from 140/90 to 130/80 mm Hg, and the 2018 European (ESC/ESH) guidelines’ retention of 140/90, are standard reference knowledge. Also standard reference knowledge: Adriaan de Groot’s chess studies and William Chase and Herbert Simon, “Perception in chess” (1973); Amos Tversky and Daniel Kahneman on the availability heuristic (1973); Marcia Johnson, Shahin Hashtroudi and D. Stephen Lindsay, “Source monitoring” (Psychological Bulletin, 1993); Abraham Luchins’s water-jar studies of Einstellung (1942); the 2005 Nobel Prize in Physiology or Medicine to Robin Warren and Barry Marshall for Helicobacter pylori; the cough associated with a class of blood-pressure medicines. Punjabi and Urdu usage of ਭੁਲੇਖਾ, ਭੁੱਲ, ਭੁੱਲਣਾ and ਜ਼ੁਬਾਨੀ, and the shop-bill formula bhūl-chūk lenī-denī, are standard reference knowledge of the language. Vendor representations. None.
Recollection and tradition. The essay asserts no remembered scene. Khalsa College appears only through the public biography; the closed-book examination hall is described as a common arrangement of the period, and no particular paper, hall, examination format or teacher from the author’s own studies is described. The physician with the remembered cough case is hypothetical and is introduced as such; no patient in the essay is real.
Analysis and forecast. The reading of memory as a database with a schema and an index but no dates, sources or change log is the author’s analysis. The suggested habits (distrusting numerical, committee-set and recently changed knowledge; recording corrections with date and source; asking a disconfirming question) are reflection on knowledge practice, not clinical instruction. Readings of Augustine, Plato, James and Darwin are interpretive. 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.
About the Author #
Kanwar Partap Singh Gill, MD, traces his family roots to Khadur Sahib in Punjab’s Majha region and grew up in Amritsar. He attended Spring Dale School and completed pre-medical studies at Khalsa College in 1994–1996. He earned his MBBS from Sri Guru Ram Das Institute of Medical Sciences and Research in Amritsar in 2002.
He completed PGY-1 at the University of Arkansas for Medical Sciences in Little Rock in 2007–2008, and PGY-2 and PGY-3 with UCSF Fresno at Community Regional Medical Center from 2008 through 2010. He has practiced medicine in California since 2010.
His career has crossed four distinct information eras: predominantly paper-based medicine, early and fragmented electronic records, the integrated-EHR era, and the emerging age of generative and agentic artificial intelligence.