Institutional legibility explains why organizations treat people as numbers. Payroll, scheduling, and hiring all depend on workers who fit fixed categories and can swap in for one another. People keep changing anyway. They learn, tire, doubt, and grow between one week and the next, and the file records none of it.
This essay argues that organizations need both views at once. Flat records keep operations running and keep favoritism out of decisions. Seeing each person as an ongoing event lets managers respond to what someone does now, not what a file said last year.
The essay also turns the argument on itself. “Institution” names thousands of separate human choices. Readers flatten themselves too, each time they answer with a job title.
The question has a 2026 edge, since nearly two-thirds (63%) of workers expect AI to make the workplace feel less human this year (Source: Trade and Industry). The essay leaves that tension open and shows how some organizations count people without flattening them.
Two Needs At Once
Quick Summary: Organizations depend on interchangeable workers to run payroll and schedules. Real employees change from week to week in ways no file records. So every institution needs flat records and singular people at the same time, and those two needs work against each other.
Institutions run on flattened people. Payroll, shift schedules, and hiring pipelines all depend on one worker swapping in for another without anyone asking what changed in either of them since last week. A field on a form has to fit whoever fills it in.
The people keep changing anyway. Picture a nurse on Tuesday’s roster. She slept four hours, found a better way to calm one anxious patient on Monday, and started wondering whether she wants this job a year from now. Her file says “RN, Level II.”
So any institution needs two things that work against each other. It needs people flat enough to count. It also needs each person seen as singular, as something still happening, because a living person does the work and the file only records it.
Those two needs don’t settle. This essay won’t pretend they do.
How Legibility Taught Officials To See People As Entries
Quick Summary: Governments and companies built tools for reading people from a distance. James C. Scott calls this legibility, and Max Weber and Frederick Taylor built the same logic into bureaucracy and factory management. Words like headcount and human resources carry that habit into today’s offices.
James C. Scott named this way of seeing in Seeing Like a State (1998). He called it legibility. According to Scott, premodern states knew little about their subjects’ wealth, land, identity, or location. So officials built tools to read a population from a distance, including permanent surnames, standard weights and measures, land surveys, and population registers. Each tool swapped a local detail that kept shifting for an entry a clerk far away could read.
Max Weber described the same move inside the office. In the ideal type of bureaucracy he sketched in Economy and Society, published after he died in the early 1920s, officials follow calculable rules “without regard for persons.” Weber counted that impersonality among bureaucracy’s technical advantages. An official who skips over who you are also skips over whose nephew you are.
Frederick Winslow Taylor brought it to the factory floor. In The Principles of Scientific Management (1911), he wrote, “In the past the man has been first; in the future the system must be first.”
Then listen to the vocabulary that followed. Headcount. Human resources. Full-time equivalent, the unit that lets two half-time people add up to one. Each term lets a manager total people without meeting any of them.
I did this for a living, too. I’ve spent about fifteen years in paid search and paid social, sorting people into audience segments. “In-market for running shoes” describes a person for roughly as long as it takes them to buy the shoes.
What Changes When Software Does The Counting
Quick Summary: Workplace software now does much of the counting that headcount and job titles started. Many workers expect that shift to make work feel less human. Some already notice it in their own writing, as AI-drafted messages stop sounding like them.
Software didn’t start the flattening. It just counts faster than any clerk. A scheduling system or a dashboard of productivity scores does what Scott’s officials did with a register, without a person reading each line.
Workers expect the change. In a survey of 1,003 employed U.S. adults, conducted by Resume Now using Pollfish in October 2025, 63% said AI would make the workplace feel less human in 2026. 42% cite dehumanization of work as one of the biggest workforce issues linked to AI.
Leaders see some of it too. In Globalization Partners’ 2026 AI at Work report, 82% admit AI has lowered the value placed on human employees.
globalization-partners
The flattening also reaches people’s own sentences. In a survey of 921 U.S. workers conducted by Omni Calculator in June 2026, a quarter of workers who use AI for workplace messages say their writing no longer sounds like them. Shane Saunderson, an assistant professor at McMaster University’s DeGroote School of Business, described these tools as “statistically averaging machines”.
An average voice belongs to nobody in particular. That makes it the newest field on the form.
What Scientific Forestry And The Identifiable Victim Effect Left Out
Quick Summary: Flattening drops whatever the count can’t hold. Scott’s forestry case shows a managed system declining after it dropped the details its tables ignored. Research on charitable giving suggests people respond more to one named person than to statistics, though a later replication challenges that finding.
Scott opens his book with forestry. In the 18th and 19th centuries, German foresters wanted woods they could count. Mathematicians defined a standard tree, the Normalbaum, and built yield tables from it. Foresters then cleared the underbrush, cut the number of species, often down to one, and planted in straight rows across large tracts.
The first rotation, about 80 years of growth, went well. Norway spruce became the reliable crop. By the second rotation, production had dropped. Scott puts the loss at 20 to 30 percent, and he notes that Waldsterben, “forest death,” entered German to describe the worst cases.
The yield table counted wood. Scott’s account points to what it left out: the underbrush, the dead wood, the mix of species, the soil.
Why One Name Can Outweigh A Count
Quick Summary: In the original research, people gave more to one named person than to a set of statistics. A 2007 charity study found that adding numbers to a single child’s story lowered donations. A later replication failed, so the essay treats the finding as an open question.
The same split shows up when people give money. In a 2007 study, Deborah Small, George Loewenstein and Paul Slovic asked participants to give to Save the Children. Some read about Rokia, a 7-year-old girl in Mali facing severe hunger. Others read statistics about hunger across Africa. Rokia drew more money. A third group got Rokia and the statistics together, and that group gave less than the Rokia-only group. As Small put it, “It’s easy to override people’s feelings by giving them statistical information.”
Hold that one loosely. A later preregistered replication of two of the studies, run with hypothetical donations by Maximilian Maier, Yik Chun Wong and Gilad Feldman, found no support for the effect. The forestry numbers hold up better than the donation numbers. Both stories still involve the same choice: a count or a name.
How Flattened Workers Start Flattening Others
Quick Summary: Flattening reaches past the office and into service work. Burnout research names one of its dimensions depersonalization, a cold, distant response toward the people a worker serves. Healthcare staff who refer to a patient by diagnosis and bed number repeat the same move on a smaller scale.
Workers who are treated as interchangeable can start treating others that way. I can’t prove that link with one study. Burnout research does give the second half a name.
In 1981, Maslach and Susan Jackson published an instrument to assess burnout: the Maslach Burnout Inventory (MBI). Its three scales measure emotional exhaustion, depersonalization, and reduced personal accomplishment. Maslach describes depersonalization as treating clients, students, customers, or colleagues in a distant and/or cynical way.
waybackmachine
A researcher writing in the British Psychological Society’s magazine, The Psychologist, described how staff in her interviews treated people as objects to protect themselves. Her example came from healthcare, where staff treat patients as objects, for example ‘the infarction in bed two’.
bps
Look at what that phrase does. It repeats Scott’s move at the scale of one bed. A diagnosis and a location take the place of a person, the same way “RN, Level II” took the place of the nurse at the start of this essay.
So the flat view doesn’t stop at the HR file. A tired worker can turn it on the patient in front of them.
Why Blind Auditions Make The Case For Impersonal Rules
Quick Summary: Flattening also protects people. Blind auditions and impersonal rules keep bias about names, ages, or appearance out of hiring decisions. The essay treats flat systems as necessary and argues for pairing them with attention to each person.
Flattening protects people, too. Many American orchestras started auditioning musicians behind a screen in the 1960s and ’70s, so the jury could hear the playing without seeing the player. Some orchestras also used carpet to hide the difference between men’s and women’s footsteps, Bloomberg columnist Faye Flam reported.
In 2000, economists Claudia Goldin and Cecilia Rouse found that the screen raised the chance a woman would advance and get hired. They also flagged large standard errors in some estimates, and statistician Andrew Gelman later called the results “not very impressive at all.” Either way, the screens did the job the orchestras built them for. They hid the singular person on purpose.
A boss who “really sees” each person can also see an accent, an age, a last name (the same thing Scott’s officials made permanent). A rule that skips over the person keeps that kind of seeing out of the decision. So the flat version has real uses. Dropping it would leave plenty of people more exposed.
The Title Gets It Wrong
Quick Summary: The essay’s own title makes the error it criticizes. An “institution” decides nothing on its own, because specific people make separate choices that look like one will. Readers do the same to themselves when a job title stands in for the work they actually do.
This essay keeps saying “institutions need.” Institutions don’t need anything. The title makes the same move the essay criticizes.
“Institution” names a few thousand people making separate calls on separate days. Those calls happen to add up to something that looks like one will. A forest office can’t want straight rows. People inside it ordered them, and other people planted them.
You flatten yourself, too. Someone at a party asks what you do, and you answer with a job title. That title stays true on days you did none of the work. It would still read as true five minutes from now if you did nothing at all.
So the people who flatten and the people who get flattened turn out to be the same. One week they write the entry. The next week they appear in one.
How Buurtzorg Counts Hours Without Flattening Nurses
Quick Summary: Some organizations count people without flattening them. Buurtzorg, a Dutch home-care provider, still tracks hours but lets small nurse teams decide how to care for each patient. The TV series Severance shows the opposite extreme: workers with no history, and what they want once they notice.
By the 1990s, Dutch home care had merged into large regional organizations. Care got divided into standardized “products,” each tied to a task and a time allotment.
Jos de Blok, a nurse who had also worked as a manager in that system, co-founded Buurtzorg (“neighborhood care”) in 2006. Its nurses work in self-managing teams of no more than 12. Teams that grow past 12 split in two. Each nurse meets the patient and family to work out what that household needs, and the teams plan their own work. By 2015, the organization employed 8,000 nurses in 700 teams.
Buurtzorgdidn’tt stop counting. Teams have to spend at least 62 percent of their time on billable care, and Dutch insurers pay for home care by the hour. The difference comes from who does the counting and what it serves. The nurse still logs the hours. The nurse, not the hour count, decides what happens on the visit.
Severance shows the other end. Apple TV+ premiered the series on February 18, 2022. At the fictional Lumon Industries, a surgical procedure splits workers’ memories between work and home, so the “innie” at the desk remembers nothing of the “outie’s” life. From a scheduling view, that makes an ideal hire. Nobody brings last night’s argument into the office.
Then Helly, the new hire, tries to quit almost at once. Over the nine-episode first season, the innies keep trying to learn about the lives they can’t remember. The way I watch it, the writers imagine a fully flat worker and then show that worker wanting the first thing a person wants. A history.
The Clerk Has A File Too
Quick Summary: The people who keep records change as much as the people in them. Kashmir Shaivism describes reality as an ongoing vibration with no finished object behind it. The essay closes by asking readers who writes their one-line description, the person they became this week or the frozen entry.
Kashmir Shaivism describes reality as spanda, a vibration that keeps going with no finished object behind it. Under that view, a person keeps happening. So does the clerk writing the file, the manager reading it, and the forester counting trees.
An institution needs flat records to function and singular people to do the work. Those singular people include the ones keeping the records.
So the next time a form asks you to describe yourself in one line, who writes the line? The person who changed this week, or the entry that stayed the same?
Questions and Responses
James C. Scott used legibility for the tools states built to read populations from a distance, such as permanent surnames, standard measures, and land surveys.
Payroll, scheduling, and hiring depend on one worker swapping in for another. Fixed entries let managers total people without meeting them.
Yes. Screened orchestra auditions hid each musician from the jury, and Goldin and Rouse found the screen raised women’s chances of advancing and getting hired.
Buurtzorg nurses work in self-managing neighborhood teams of no more than 12. Teams that grow past 12 split in two.
The Apple TV+ series imagines a procedure that splits workers’ memories between work and home, leaving the at-work self with no outside history.


Leave a Reply
You must be logged in to post a comment.