In February 1985, a film came to theaters that everyone who had ever browsed a video-store shelf knew: Peter Weir’s Witness. Philadelphia cop John Book, played by Harrison Ford, hides an eight-year-old boy who saw a murder at a train station among the residents of America’s most photogenic community—the Amish of Lancaster. Eight Oscar nominations, two statuettes, a twelve-million-dollar budget, and a place among Kurosawa’s favorite films. Forty years later, people remember the film differently: some for Rachel and John’s slow dance to Sam Cooke’s “What a Wonderful World”; some for the barn.
And yet the most interesting thing about Witness happened off camera. The people the film turned into a global brand refused to appear in it. Entirely: not a single Amish extra, not a single Amish face on screen. They built sets, worked as carpenters on the shoot—and made it a condition that they would not be shown. Producer Edward Feldman wrote about it plainly: “I promised them that if they worked on the film, they would never be photographed by the camera, even as extras.” A film about the Amish, made almost without Amish people—that is the best illustration of how their world works. Not a museum, not the 21st century with the switch turned off, but a functioning system of filters: what we let through, what we hold back, what we throw out. That system has a name—Ordnung. And in the forty years since Witness premiered, science has taken such a close look at this community that the Amish have unexpectedly become one of the planet’s most valuable biomedical resources. Both stories—the one about the film and the one about the people—are really a single story about how a community manages its border with the outside world.
The script for Witness lived a long Hollywood life before it became a film. The screenplay was assembled in reverse over years: Pamela Wallace came up with a novel about an Amish woman who witnesses a murder in Los Angeles; screenwriter William Kelley reworked the premise from an unfilmed episode of Gunsmoke that he himself had written in 1972; together with Earl Wallace, they turned it into a 182-page screenplay titled Called Home—an Amish euphemism for death. Studios rejected it in a loop: Fox passed with the explanation that it did not make “rural movies,” while Paramount was the last to say yes. Feldman dreamed of Peter Weir, but Weir was already preparing The Mosquito Coast and declined; John Badham dismissed the script as “just another cop movie”; David Cronenberg refused with a line for the ages—“I could never be a fan of the Amish.” Then financing for The Mosquito Coast collapsed, Weir became available, and this became his first American film. They shot at a furious pace: principal photography ran from April 27 to June 29, 1984, and the main reason for the rush was a Directors Guild strike hanging over Hollywood. They wrapped principal shooting three days before its announced start—the strike, ironically, never happened. The title Called Home was changed to Witness midway through shooting at the insistence of Paramount’s marketing people, who found a film with the word “home” in its title difficult to sell.
Ford went method in preparation: he spent two weeks with the Philadelphia Police Department’s homicide unit, took part in two raids, saw victims’ bodies—first in photographs, then in person. “The first corpse shocks you,” he recalled later. “The fifth one does, too.” Kelly McGillis, for whom this was only her second film role—five months before shooting, she had been working at a New York coffee shop—spent weeks living in the home of an Amish widow with seven children: milking cows, planting potatoes, learning the Pennsylvania German dialect. In one version of the story, she had to move out after her photograph appeared in a local newspaper, so as not to expose her hosts. Viggo Mortensen, for whom the role of brother Daniel Hochleitner was his debut, had originally been hired to say one word in the funeral scene; Weir liked him so much that he kept him for six weeks of filming.
But the film’s central sequence was not born of dramaturgy. In the script, the barn raising took up a paragraph: Book helps the neighbors, falls off the roof, and is saved by his rival. Weir expanded that paragraph into ten minutes that became the heart of the film—and shot it in a single day. The barn was really built. Ford was put on camera not only as a star, but as a former professional carpenter: in the seventies, he quit acting and earned his living doing woodwork for celebrities—Joan Didion, Talia Shire—and returned to film after accepting a pay cut. Ford demanded technical accuracy in the carpentry and insisted on the proper methods, though he agreed to spruce boards instead of the costly oak Amish builders would have chosen. Ford himself later called the barn one of the largest carpentry projects he had ever done. The structure’s fate was grim: soon after filming, the landowner dismantled the barn and sold the boards.
The landscapes had to be filmed cleverly, too. Actual Amish farmers would not lease land to Paramount; bishops explicitly advised their congregations not to appear in the film or rent out farm buildings. Some local Amish worked behind the scenes as carpenters and technicians, under a written promise that they would not be shown. The extras—the faces on screen, the serious beards—were recruited from neighboring Mennonites, whose interpretation of the faith does not forbid cameras, while makeup artists glued fake beards onto local actors. Hence the bitter irony: the film millions of viewers take to be a portrait of the Amish used Mennonites as extras and Amish behind the scenes—the Amish themselves were kept out of the frame on principle.
Why? The answer is the doorway into the Amish world. The Amish ban on photography is not aesthetic but theological: the Second Commandment forbids “making images”—engravings, depictions, likenesses. For many Amish, that includes photography. But the ban on being photographed is not the main event; it is merely a symptom of a broader principle: the community manages its visibility and its connection to the outside world with caution. Every new channel into the world—the telegraph, telephone, radio, cinema, internet—is a potential hole in the perimeter, and it must either be closed or regulated.
That regulation is called the Ordnung (literally, “order”). It is the unwritten code of a particular church community, governing everything: from the cut of clothing and the shape of hats to whether one may own a car, connect to the electrical grid, sue a neighbor, or run for office. And here is the crucial thing to understand: the Ordnung does not forbid technologies as such. It forbids their consequences. The ban on connecting to the public electrical grid is not a war on electricity itself, but a refusal of dependence on “the world”: as long as wires run to your home from municipal generators, your life is governed by someone else’s schedules, rates, and rules. This is not nostalgia for the 19th century. It is an architectural decision about boundaries.
And the final touch that shatters the stereotype completely: the Amish community votes. The Ordnung is reviewed twice a year before communion. At the pre-communion meeting two weeks beforehand, the community must reach unity: if there are unresolved issues, communion will not take place. The bishop does not dictate rules from above; his real power lies in deciding which questions go to a vote and when. Enough pressure from below—and an issue makes it onto the agenda. Its unwritten nature is not mere archaism here: a rule lives as the community’s shared practice, not as an external contract.
The best part is the mechanism for adopting new technologies. Kevin Kelly, one of Wired’s founders, described the process as an almost perfect release pipeline:
The result of this mechanism is what Kelly called a “50-year lag”: the Amish adopt technologies half a century later than everyone else, but at the point when benefits, prices, and risks are already clear. They refused credit cards, Kelly observed—bishops saw earlier than most how interest rates break farms—and they are among the last people still debating cell phones, ever since Howard Rheingold wrote about it for Wired in 1999. In that time, they have managed to fall in love with solar panels—electricity without wires to the outside world is a brilliant move: the grid remains switched off, while energy flows—master pneumatics (“Amish electricity”: air from a diesel compressor instead of electrical power, pneumatic sewing machines, blenders, and log splitters), grow an entire retrofit industry converting electric appliances to compressed air, work with CNC machines—Kelly describes a $400,000 machine in a Mennonite workshop operated by the owner’s 14-year-old daughter—and plant GMO corn because Bt corn does not break under the corn borer, and broken stalks have to be lifted by hand, which means GMOs preserve the old, debt-free equipment and the family farm. Every one of these decisions looks paradoxical only if you assume their aim is “to be old.” Their aim is to survive as a community, and they choose technologies not for their novelty but for how well they serve that task.
The filter is not perfect, however, and it is not symmetrical. The Amish pipeline can say “yes,” but it has enormous difficulty revoking decisions once adopted: a rule that enters the Ordnung lives on for generations. The cost of such a system is obvious: it is internally conservative, changes slowly, and can crush individual freedom. This is not an idyll. It is a working but highly specific system.
Movie logic suggests that a closed religious community without television ought to be dying out. Reality is exactly the opposite. The Amish are one of the fastest-growing population groups in North America. According to the Young Center for Anabaptist and Pietist Studies at Elizabethtown College, as of June 2026 the Amish population—adults and children—was estimated at 421,200. That is roughly ten thousand more in a year and 32% more over nine years, up from 318,390 in 2017. The population doubles roughly every twenty years. By comparison, Canada’s population doubles over a period of around eighty years.
The mechanics are simple and unfamiliar. A strong attachment to tradition does not limit birth rates here—it multiplies them: a typical Amish family has five or more children. And the critical number people usually forget: 85 percent of Amish teenagers, once grown, voluntarily choose baptism and remain in the community—for a modern voluntary community, that is an extremely high retention rate. No “development” drives people away—people choose to stay.
What happens when 421,200 people descend from several hundred ancestors? Nothing special, if you view it as a success story. But to a geneticist, it looks entirely different—as population quality control, zero mutations from outside, and frequent marriages within a closed pool. This is the founder effect, one of the most vivid examples in human populations: among the founders of a small group, certain genes happen by chance to become very common, and a closed population rolls those frequencies forward across all subsequent generations.
The first high-profile example was Ellis–van Creveld syndrome—a rare form of dwarfism with an additional symptom that never fails with the public: six fingers. Polydactyly, heart defects, abnormalities of the teeth and nails—and an especially high concentration among the Lancaster Amish. (The classic textbook illustration is in the PBS Evolution Library.) The logic of how it arises is simple. In a closed population descended from roughly two hundred founders, a recessive gene that would remain scattered among thousands of unrelated carriers in a larger population can easily meet itself: if both parents are descendants of one founder, the chance of bringing together two copies rises hundreds of times. The Amish simply rolled rare mutations up to visible frequencies.
But then a plot nobody had planned unfolded: instead of merely accepting it, Amish communities in Pennsylvania turned their own genetic burden into a research program. In the late eighties, Harvard-trained pediatrician D. Holmes Morton, then a biochemical genetics fellow at the Children’s Hospital of Philadelphia, diagnosed an Amish child with glutaric acidemia type I—a disease capable of causing severe neurological damage. For families with rare diseases, access to timely diagnosis was especially important. In 1989, Morton, together with his wife Caroline and in partnership with local Amish and Mennonite communities, founded the Clinic for Special Children in Strasburg, Pennsylvania—a “clinic in a cornfield.”
The model was extraordinary. The clinic itself was built in a literal barn style—the first building was raised by the community through volunteer labor, post-and-beam, in 1990—it aims to make diagnosis and treatment affordable, and it conducts DNA diagnostics on site in its own CLIA-certified laboratory. Over 36 years, the clinic has grown into a team of more than 35 staff serving 1,700 active patients; its employees have coauthored more than 130 articles in peer-reviewed journals—from natural-history studies of rare mutations to clinical trials of gene therapy. In 2024, the clinic moved to a new building in Gordonville. Note the architecture of the solution: a community living outside many familiar institutions helped create its own genomic medicine and turned its genetic trap into a model for the whole world. The phrase “clinic in a cornfield” describes the approach well: small, tied to a specific population, and close to the patient.
Then science found in one Amish community something that had previously seemed almost fantastic: a genetic mutation that slows aging. In the Old Order Amish community around Berne, Indiana—the community descends from settlers from the canton of Bern in Switzerland, in the mid-19th century—doctors had known since the 1990s of a family in which children suffered from a strange clotting disorder: a three-year-old girl hit her head and nearly bled to death. Hematologist Amy Shapiro of the Indiana Hemophilia Center built a ten-thousand-person family tree and found twelve patients with a complete absence of the PAI-1 protein, plasminogen activator inhibitor-1. The mutation in the SERPINE1 gene was described in the NEJM in 1992: without PAI-1, blood clots dissolve too quickly—the bleeding does not stop. People with two copies of the mutation have a rare clotting disorder.
But carriers of one copy—43 people out of 177 examined—showed the opposite effect, and this was where the most interesting finding emerged: carriers lived an average of 10 years longer—85 versus 75 (in the paper: 85 ± 10 versus 75 ± 12, P = 0.037); their telomeres were 10% longer (P = 0.007); fasting insulin was 28% lower; none of the carriers had diabetes—0% versus 7% among non-carriers. This was the first documented mutation in human history to protect against several aspects of biological aging at once. Douglas Vaughan of Northwestern University—who had studied the protein for twenty-nine years—once put it almost breathlessly: “We were struck by the consistency of the anti-aging effects across multiple body systems. They not only live longer—they live longer in good health.”
Engineering honesty matters here. This is not an elixir and not a “gene for immortality”—it is a correlation in one isolated extended family, built on family trees and statistics. A sample of 177 people. Average life expectancy among the Amish as a whole is 71 years; carriers of “one copy” add around 10 years to that. The mechanism is thought to work through metabolism—insulin signaling—but full understanding remains far off. And most importantly: the mutation comes at a cost—people with two copies, 7 carriers in the sample, have a clotting disorder. “You don't want too much PAI-1, and you don't want zero,” Vaughan put it in an interview with the Guardian: the optimum lies somewhere in the middle.
Northwestern and Tohoku University in Japan are already doing the next logical thing: if the mutation sells ten years, can the same benefit be bought with a pill? A PAI-1 inhibitor codenamed TM5614 passed Phase I clinical trials in Japan; Phase II involves people with type 2 diabetes and obesity, while Northwestern University was preparing an American trial. In mice, a similar inhibitor produced a fourfold increase in lifespan—the source of widely shared, if fragile, hopes for an “anti-aging pill.” The irony is that one of the main keys to modernity’s biohacking dream turned up precisely in a community that carefully walls itself off from modernity.
Witness became a cultural event and a commercial hit. And it left a rather specific mark on Lancaster. Tourism in the county rose noticeably after the premiere. The National Committee for Amish Religious Freedom called for a boycott of the film, fearing that communities would be “flooded with tourists”: crowds, souvenir hunting, photography, intrusions onto private farms. Amish lawyers said the film portrayed them inaccurately. Pennsylvania Governor Dick Thornburgh agreed not to promote Amish communities as future filming locations. Since then, no major film has been shot in Lancaster County—not because it is prohibited, but because the experience proved too unequal: Hollywood got its hit, the community got crowds with cameras.
The film literally contains the metaphor for this conflict. Rachel tells Book that, to tourists, the Amish are not people but exhibits: they stop, stare, and trespass onto private property. The film, made while observing all the community’s boundaries—no Amish on screen, no shooting on their farms—proved surprisingly sensitive to the problem of the observer. And there is the value of the ending, formulated by negotiator William Ury in Getting to Peace: the crowd of Amish silently surrounding the armed sheriff is not weakness or passivity, but the “third side” in its purest form—a community that stops unauthorized violence through the mere fact of its presence. Kurosawa, incidentally, called Witness one of his favorite films.
The film we remember as the story of a cop among the Amish has, over 40 years, turned out to be an almost painfully precise metaphor for the community itself. The Amish are not people stuck in 1850. They are citizens of the 21st century with a perimeter raised around them: they sift every technology through two rounds of voting, field trials by “alpha geeks,” and one central criterion—“what does this do to our ties?” They receive new things more slowly than anyone else—and are the first to see their real consequences. They refused insurance and grew to 421,200 people. They refused wires—and fell in love with solar panels and pneumatics. By refusing to be filmed, they showed what the camera cannot capture: a working model of slow, communal governance of technology.
And science, largely without their knowledge, did its work: the closed population that refused to be photographed was, for decades, an invaluable subject for geneticists—from research into Ellis–van Creveld syndrome to the mutation that sells ten years of life. There is something of a classic chess problem in this: the strategies of “walling oneself off from the world” and “giving oneself to it entirely” turn out not to be absolute opposites. A community that built a wall around its life accidentally built a window onto the rest of the world’s future. Would Weir make this story today? Unlikely; it is no longer his genre. It is, rather, the genre of science. And, as usual in science, the most interesting result came not from the experiment’s purpose, but from its side effect.
Sources: Wikipedia — Witness (1985 film); AFI Catalog of Feature Films — Witness (1985); Mental Floss — “9 Fascinating Facts About Witness” (February 2025); Kevin Kelly, “Amish Hackers” (The Technium, kk.org); Amish America (Erik Wesner) — “Amish Church Rules: How the Ordnung Guides Daily Life”; Young Center for Anabaptist & Pietist Studies, Elizabethtown College — “Amish Population Profile, 2026”; PBS Evolution Library — “Genetic Drift and the Founder Effect”; Nature Genetics (2000) — “Ellis-van Creveld syndrome and the Amish”; Clinic for Special Children — official website (about-us); Northwestern University Feinberg School of Medicine — “Amish Longevity May Be Due to Genetic Fountain of Youth” (November 15, 2017); Montorsi et al./Vaughan et al., “A null mutation in SERPINE1 protects against biological aging in humans,” Science Advances 3(11), eaao1617 (2017); The Guardian (Ian Sample, November 15, 2017) — “Rare genetic mutation found in Amish community could combat ageing”; Atlas Obscura (Natasha Frost, November 16, 2017) — “The Indiana Amish Are Blessed With a Genetic Fountain of Youth”.