How Your Family Tree Could Catch a Killer
On Thanksgiving morning, 1987, Rick Bart, a homicide detective in Snohomish County, Washington, got word that a pheasant hunter had discovered a body in a field beneath High Bridge, an overpass spanning the Snoqualmie River. Bart was preparing to spend the day with his family, but he went anyway. He was one of only two homicide detectives in Snohomish—a jurisdiction, just north of Seattle, that covers more than two thousand square miles. He was familiar with the crime scene. It was near the Monroe Honor Farm, where inmates milked cows to provide dairy to the state prison system. The bridge was secluded enough to be private, but accessible by a country road. Teen-agers drank there.
When Bart arrived, morning fog was clinging to trees along the riverbank. The body was partially shrouded by a blue blanket. Lifting it revealed signs of a brutal death. The man’s head had been struck with a rock. A clump of hair, ripped from his scalp, was in the grass. A ligature, made from plastic twine and two red dog chokers, was around his neck. An autopsy later revealed that he had been gagged with a tissue and a pack of Camel Lights.
“We had no I.D.—didn’t know who he was,” Bart recalled. “We didn’t know when he was put there, at all. There was nothing.”
The following day, he got a call from a detective in nearby Skagit County, who thought the body belonged to Jay Cook, a twenty-year-old from British Columbia. The detective told Bart what he knew. Several days earlier, Cook’s father had asked Jay to drive the family’s van to Seattle to pick up some furnace parts for his business. Jay had brought his girlfriend, Tanya Van Cuylenborg, an aspiring photographer. Tanya had an open smile, and friends called her “sweetie.” Jay had a boyish, elfin face and wispy brown hair.
Regarding the errand as a chance at adventure, the two decided to sleep in the van, and packed it with foam mattresses and provisions. They drove down on November 18th, taking the Coho ferry from Canada to Washington. Then they vanished.
After days with no news, Tanya’s father filed a missing-persons report with the Royal Canadian Mounted Police. Then he hired a plane to look for the van himself. When the search was unsuccessful, he drove hundreds of miles, interviewing people at restaurants and convenience stores.
While he was on the road, detectives in Skagit County discovered Tanya’s body near a creek. She had been raped, and shot in the back of the head. The following day, the van was found sixteen miles north of the creek, parked close to a tavern and a Greyhound bus terminal. Behind the tavern, police found keys to the van, a camera lens cap, a wallet with Tanya’s I.D., surgical gloves, zip ties, and a box containing .380-calibre bullets, which matched those found at her crime scene.
After Bart heard all this, he rushed back to High Bridge and collected zip ties that he had noticed on the ground near Jay Cook’s body. The killer, it was clear, had intercepted the couple with a “murder kit” in hand.
Bart wondered if he was pursuing a serial offender—perhaps an inmate who had once been at the prison farm. About two weeks later, taunting letters began to appear at the Cook and Van Cuylenborg homes, from someone claiming to be the killer. (“I had a gun in Jay’s back. Tanya was pleading.”) The letters were written by hand, and postmarked from all over: Seattle, Los Angeles, New York.
The detectives desperately ran down every lead. A forensic analysis of the correspondence offered nothing conclusive. All the suspects affiliated with the prison had alibis. Jay and Tanya’s movements could only partially be constructed. “We couldn’t even put them for sure in Seattle,” Bart said. “It all added up that this was a cold case.”
VIDEO FROM THE NEW YORKER
How a Notorious D.E.A. Informant Busted Criminals
In 1989, the detectives opened their files to “Unsolved Mysteries,” and hundreds of tips poured in. Some were easy to dismiss—one was a long missive from a psychic—but many seemed worth pursuing. All were dead ends. A lens belonging to a camera that Tanya had with her on the trip turned up at a pawnshop in Portland. It was also a dead end. The detectives learned that a serial killer had lived near the spot where the van was ditched; they searched a crammed storage locker that he used for weapons and other items. Another dead end.
In the nineteen-nineties, policing was revolutionized by forensic DNA analysis, which could identify criminals from biological evidence. The F.B.I. created a national DNA database of convicted offenders, and another for missing persons and for samples taken at crime scenes. Together, they would aid in the investigation of more than half a million crimes. But, when Snohomish detectives uploaded their crime-scene DNA samples, they received no hits. Their suspect was not a convicted felon, and his DNA had not been found at another crime scene. It was possible that he was dead.
In 1999, Rick Bart became sheriff. “One of the first things I did was to create a cold-case squad,” he told me. “I wanted this case solved. It just haunted me.” In 2005, the investigation fell into the hands of a detective named Jim Scharf. By then, its file had grown to fill a dozen binders, containing the names of more than a hundred possible suspects. Scharf had a reputation for absorbing volumes of detail and pursuing every tip. Delving into the material, he established that the letter writer was not the killer; he was a disturbed itinerant, who travelled around on buses and trains. But, after more than a decade on the case, Scharf was still no closer to finding the killer.
In July, 2016, Scharf’s captain learned about a company in Virginia, Parabon NanoLabs, that had created a tool called Snapshot. From a DNA sample, the company claimed, it could derive information about physical traits: hair color, eye color, complexion, possibly even facial structure. “I want you to find a case where we can use this tool,” the captain told Scharf, who decided to try it on the Cook and Van Cuylenborg investigation. A DNA sample was shipped to Parabon, and a report came back noting that the killer was of northwestern-European descent and likely had reddish-blond hair, green or hazel eyes, and a light complexion.
Scharf, along with a partner, returned to the case binders and found four men with comparable traits. Two were deceased. Two were living. None, it appeared, was the killer. The detectives published a computer-generated rendering of the Snapshot, but the only result was a new wave of useless tips.
Having hit another obstacle, Scharf began to consider a new approach: forensic genealogy. For years, genealogists had been using ever-larger private DNA databases to compare genetic information among populations, allowing them to chart family networks more completely. What if those same tools could be used with the DNA that detectives had gathered back in 1987?
In April, 2018, Scharf permitted Parabon to pass the killer’s DNA profile to one of the world’s leading genetic genealogists, CeCe Moore. The company’s C.E.O. told him that she would likely identify the killer within a week. Scharf was skeptical, but three days later Parabon reported that Moore had a name: William Earl Talbott II, a truck driver who lived not far from High Bridge. It had taken her two hours on a Saturday to figure this out.
Scharf thought of all the names in the voluminous case file. There had been no William Earl Talbott. After thirty years of detective work, after hundreds of tips and leads, the man’s name had never come up. Still, he had officers tail Talbott, hoping to retrieve something that had his DNA on it. When a coffee cup fell from Talbott’s truck one afternoon, they rushed it to a crime lab to compare his genetics with the killer’s. Scharf waited anxiously in the lab, until a technician emerged, and said, “Jim, it’s him.” Scharf’s eyes filled with tears, and, raising a fist, he called out, “We got him!”
By then, Rick Bart had retired. From a beach in front of his house, he could see where Jay’s and Tanya’s families had lived in Canada—a daily reminder of the brutal deaths. When Scharf called with the news, he was looking across the water. He burst out laughing. “For thirty years, we couldn’t do a damn thing,” he told me. “And here comes this lady who says, ‘Hey, I know who did it!’ ”
CeCe Moore lives on a hill in coastal California, about an hour’s drive south of Los Angeles. This summer, when I arrived at her home, her husband, Lennart Martinson, a film producer, greeted me and walked me inside. Moore was sitting on a puffy leather sofa, facing a laptop on a stand. The air-conditioning was cranked up, and she had a fuzzy purple blanket over her bare feet. On one side of the room, a skylight illuminated a print of a mermaid. On the other, a Teddy bear sat by windows offering a stupendous view: eucalyptus and palm trees, and the endless Pacific.
Moore has a warm but intense manner. On a black cord around her neck, she wears a pendant of the Finnish flag—she is a quarter Finnish—and a tiny square of metal engraved with the word Sisu. It means something like “grit” or “daring”—adrenaline-fuelled doggedness. Holding the bit of metal, Moore told me, “Sisu is me.”
Although she is in her fifties, Moore pulls all-nighters with the frequency of a college freshman. While working on a difficult case, she will let phone calls and e-mails go unanswered for hours as she stares at archival data—hunting for patterns in families she has never met, while her own family quietly places food and coffee beside her. “I can hardly see the screen many times, but I keep going,” she told me. During one of our conversations this summer, she sounded tired, and I asked if she had slept. “I was up late, and saw that a building in Florida collapsed,” she said. She had noticed the news online, at 3 a.m., and turned on CNN. “That kind of woke me up, so I kept working with that in the background.”
Moore is typically juggling at least two full-time jobs. Last year, ABC aired “The Genetic Detective,” a prime-time show based on her work. She is also a genealogist for “Finding Your Roots,” a PBS show hosted by Henry Louis Gates, Jr., through which she oversees what is likely the world’s largest collection of celebrity DNA. She is a co-founder of the Institute for Genetic Genealogy, and runs DNA Detectives—a Facebook group, with a hundred and seventy thousand members, in which volunteers help people find their biological parents and unravel other family mysteries.
Since working on the murders in Snohomish County, Moore has also become an avid crime solver, one of several prominent genealogists—nearly all women—who have combined the study of ancestry with genetics to forge a powerful new policing tool. Moore leads a team of three at Parabon. They have helped resolve more than a hundred and fifty criminal investigations since 2018—averaging about one a week. No other group using genetic genealogy, not even one within the F.B.I., has documented more successes.
Most of Moore’s cases were long cold, and more than one police detective told me that the technique that she helped pioneer was a kind of forensic wizardry, which might one day rival the fingerprint. A few years ago, Moore was brought into an investigation in Utah, where a man had raped a seventy-nine-year-old woman in her home. Local detectives had worked every lead but got nowhere. Days after Moore took the case, she sent them the names of four brothers, explaining that the rapist had to be among them. When the officers questioned the oldest brother, he immediately confessed. “It was mind-blowing,” one officer said at the time. “It seemed like magic what she was able to do.”
Crime-solving genetic genealogist is not a profession that one chooses by picking up a leaflet at a career fair. Moore fell into it—partly by accident, and partly by helping to invent the field.
She grew up, with two older sisters and a younger brother, in Rancho Bernardo, a planned community on the outskirts of San Diego. Her parents—a senior manager at J. C. Penney and a homemaker—were deeply religious, and did not expect education to play an important role in their children’s lives. But Moore was academically inclined. Her teachers drew up an independent curriculum that she could pursue while the rest of the class focussed on conventional lessons. She tested into Mensa, but didn’t entirely fit in at school, or at her congregation, or at home. “We had a tree next to my house that I would climb up in, and read my books, to be alone,” she told me.
Neither of Moore’s sisters had gone to college, and she assumed that she wouldn’t, either. “I started having teachers say to me, ‘You’re joking, right?’ ” she told me. So she bought a Pee-Chee folder and wrote “$30,000” at the top—the scholarship money that she would need to attend the University of Southern California.
She liked science, journalism, and law, but a music teacher, impressed by her singing, encouraged her to study music. “She had an influence over me—more than anybody outside of my parents,” Moore said. “She told me that I had to stop with calculus, stop with this and that, just focus on singing. She thought that I could make it.”
U.S.C.’s music school was a world-class destination for aspiring classical musicians. Moore was admitted, but quickly discovered that she had little interest in studying opera. She wanted to be in musicals. When she got a part in one, staged by the theatre department, her instructors were aghast, fearing that her participation would ruin her training. “They said, ‘You have to choose,’ basically,” she said. “ ‘It’s the program or the musical.’ ” She transferred to the theatre department.
As a senior, she lived in a friend’s home, in Irvine, acting in a community production and commuting an hour to campus. That year, her friend committed suicide. Distraught, Moore struggled to complete her one remaining course. The university told her that she could walk at graduation and finish the work afterward, but she was already acting, and decided that there was no point in getting a degree.
For an actress, Moore was introverted—more comfortable reading a book than jumping on a table and launching into soliloquies. But she was relentlessly focussed, and memorizing lines came easily. (“I used to have a photographic memory,” she told me. “I joke now that I used up all the film.”) She spent hours at the gym, training her body. She was likewise disciplined about organizing the jumble of gigs that aspiring actors must negotiate; once, she lined up fifty days of work in a row. She landed roles in the theatre and small parts on TV and in movies. (During an oceanside scene in Francis Ford Coppola’s “The Rainmaker,” she can be seen in the background, a bikini-clad beachgoer.) In between, she made ends meet with infomercials and convention work. She had to forgo a chance to appear on “The Young and the Restless” because she was at a toy fair, modelling as Barbie.
On September 11, 2001, Moore had three auditions scheduled for the day, but after the terror attacks her gigs were all cancelled. With no work, she threw herself into an old neglected project: building a family tree. Virtually every genealogical quest, Moore learned, begins with a psychological mirage. What appears to be ego-driven—a desire to map relationships that affirm one’s centrality in the world—at some point reveals itself to be about others, people we can no longer see, hear, or perhaps even name.
Moore’s family, like everyone’s, had its uncharted branches and enigmas. She knew, for instance, that, after her mother’s Finnish grandparents emigrated, they had mysteriously cut off communication with their relatives. “They would never talk about their families—not their parents, their siblings, not anyone,” Moore said. “That intrigued me.” Her father’s heritage was a quarter Norwegian, and two of his cousins had travelled to Norway to collect genealogical details about the family. Moore dug further, poring over church records, many of them in Old Norwegian, a language that she taught herself to navigate.
“I would kind of come and go from genealogy,” she told me. “But one thing that I consistently kept doing was read about DNA testing.” A vanguard of genealogists was seeking to bring genetics into the field; they were often dismissed by peers who regarded archives as paramount, but Moore loved the science. At the time, the DNA tests available were too expensive for her. But within a few years the technology would evolve, and genetics would fill her entire waking life.
The human cell is a masterpiece of data compression. Its nucleus, just a few microns wide, contains six feet of DNA: helical molecules that string together some three billion pairs of nucleotides, each represented by an initial—A, C, G, and T—the programming language of our genetic code. These strands are divided into coiled chromosomes. Two of them—labelled either X or Y—determine our biological sex. The remaining twenty-two pairs, known as autosomal DNA, are encoded with information about our traits: bone structure, eye color, skin color, the stuff of ourselves.
Genealogists grew interested in genetics at the turn of the millennium, when it became possible to analyze bits of information from the Y chromosome—known as Y-DNA—on a commercial scale. Because the Y chromosome is passed from father to son with little mutation, and because surnames historically were passed down the same way, it seemed worth exploring whether the confluence could be useful to researchers. In the late nineties, Bryan Sykes, an Oxford geneticist, persuaded forty-eight men who shared his surname to take Y-DNA tests. “Sykes” comes from a Middle English word meaning “spring” or “stream,” and the name was thought to have arisen separately among unrelated families that lived near various sources of water. But the genetics suggested that the men descended from a single ancestral line. “If this pattern is reproduced with other surnames, it may have important forensic and genealogical applications,” Sykes concluded. Theoretically, researchers could use Y-DNA to establish the pedigree of a man with an unknown identity. Sykes made a similar case for mt-DNA, which is passed down on the maternal line, in a book titled “The Seven Daughters of Eve.”
Sykes was a popularizer with a knack for flamboyance. He once declared that an accountant in Florida was a descendant of Genghis Khan. The claim was quickly disproved, but it remained evident that Y-DNA and mt-DNA had genuine applications in tracing ancestry. In Utah, the Sorenson Molecular Genealogy Foundation began collecting genetic samples, hoping that they would reveal linkages across humankind. A company called FamilyTreeDNA started selling mail-in Y-DNA tests to consumers, to build a database that offered clues to genealogical puzzles.
Moore was intrigued by Sykes’s work, and as the costs of the technology dropped she had her father take a Y-DNA test and her mother an mt-DNA test. Her acting gigs had returned, but genealogy remained a comforting focus, especially as she was hit by personal difficulties. She had fallen in love with a medical researcher, become pregnant, then fallen out of love. Moore worked as long as her pregnancy allowed, then directed clients to friends who could fill in—collegial gestures that grew into a business called Commercial Casting. After running into Lennart Martinson on the set of a commercial, she developed a close friendship with him. They soon became a couple, and also business partners, merging her casting agency with his film-production company.
In 2009, Moore persuaded executives at FamilyTreeDNA to hire her to make a commercial. During one shoot, a genealogist showed her the Web site of a competitor, 23andMe. The company had been developing technology that allowed users to access their autosomal DNA for genealogy, by tracking tiny genetic mutations. These mutations, called single-nucleotide polymorphisms, or SNPs, combine in unique patterns that are passed from one generation to the next: a child will share fifty per cent of them with each parent, about a quarter with each grandparent, 12.5 per cent with each great-grandparent, and so on.
23andMe had created a simple online dashboard that compared users’ SNPs and made rudimentary estimates about their relatedness—say, whether they were first or second cousins. After about six generations, the mutations would become too scarce to offer insight, but Moore was still floored. “This opened up the inner branches of the family tree for genetic exploration,” she told me. “I knew that was huge. I just knew it in my deepest place.”
Soon afterward, Moore called the genealogist who had shown her the site, Katherine Borges. “This is what I want to do with my life,” she told her. “How can I get involved?” Borges ran the International Society of Genetic Genealogy, which had a Web forum for “newbies” who were curious about DNA. She told Moore that she could take it over. “Just start answering people’s questions,” she said. “Read as much as you can, and become an expert.”
The peer-reviewed literature was scant, but a small group of citizen scientists was working to fill the gaps. Moore experimented with 23andMe’s technology by systematically testing her own family, to compare the results with relationships that she had vetted. “I was finding interesting data,” she told me. “Second cousins are supposed to share 3.125 per cent of their DNA on average, but some of my second cousins shared almost six per cent. Others shared one per cent.” She became fluent in terms like “haplogroup” (an ancestral society that shares SNP patterns) and “centimorgan” (a unit for measuring DNA segments). Moore was soon able to identify, for instance, that a cluster of SNPs on her own seventh chromosome indicated an ultra-distant Jewish ancestor. She became active on genealogy forums, and created blogs where she reported on her findings, and adopted the role of a promoter, noting when new companies offered DNA testing, and which offered sales.
By then, Moore had ceded her business responsibilities to Martinson. “I dropped everything,” she told me. “I am sure I’ve done more genetic genealogy than anyone in the world—probably by far—because from that time on I did it a ridiculous number of hours. I’m so obsessive-compulsive.” The tools were limited, and the databases still small, but the technology’s power was revealing itself. An increasing number of people were taking DNA tests, many of them at Moore’s encouragement, and some were learning that their paternity was not what they had thought it was. Because Moore had put herself forward as an accessible expert, they often came to her, and she helped them solve the riddles of their parentage. “I would just dive in,” she said. “I wouldn’t sleep sometimes, and just work on somebody’s case non-stop.”
The sun was glinting off the ocean outside Moore’s window. On her computer, she pointed out an open tab, for GEDmatch. “It’s just this very basic-looking account,” she said. It looked like it had been designed in 1997.
GEDmatch was the brainchild of Curtis Rogers, a former marketing executive who had spent the sixties and seventies in Hong Kong and the Philippines, representing brands like Quaker Oats and Mennen. In the eighties, he moved to Florida and ran candy stores, but by the early two-thousands he was retired and devoting himself to genealogy.
Rogers built GEDmatch with John Olson, a transportation engineer in Texas, whose day job involved devising systems for optimal traffic flow. Their original intent was to support software that could compare family trees—a difficult problem, since many trees include thousands of names. Soon, the site also allowed for segment-by-segment comparisons of autosomal DNA. GEDmatch was free and open—a nonprofit, commercially agnostic place for serious genealogy. Unlike 23andMe, it provided detailed results. People were encouraged to extract their DNA profiles, or “kits,” from private testing companies and upload them to the platform.
Moore began uploading profiles in 2011, and now manages ninety-four personal kits on GEDmatch—her family members’ and her own. (She tests herself often, tracking improvements in the technology.) As we sat at her computer, she initiated a comparison between her profile and that of one of her sisters. The screen filled up with horizontal bands, each representing one of the twenty-two chromosomal pairs. Vertical stripes—green, yellow, and red—ran across them, like a bar code. Red stripes indicated segments where the two siblings shared no DNA. Yellow indicated where they shared DNA from one parent. Green was where they inherited identical DNA from both.
Moore pointed to a chromosome with a green segment that was a hundred and eighty-five centimorgans—a long stretch of shared DNA. “So, there’s 27,803 SNPs in a row,” she said. “You’re not going to have fully identical segments with most people. You could have them with double first cousins—two brothers marry two sisters. Even then, it would be nowhere near this amount.”
“Do you know this so well that you could just scroll through these colors and say ‘sister’?” I asked.
“Oh, absolutely,” she said. Then she stopped and reviewed the stripes again. “The only other thing that looks like this is a ‘three-quarter sibling’ ”—a term that she and her collaborators invented. “When a father has children with two sisters, or a woman has children with two brothers, their offspring are half siblings, plus first cousins with each other. Instead of sharing fifty per cent of their DNA, the children will share 37.5 per cent. It’s something we’ve actually seen quite a bit of.”
Moore called up another profile; this time the color coding showed large bands of identical SNPs inherited from both parents. In such cases, GEDmatch issues users a warning: contact CeCe Moore. Years ago, she began volunteering to examine such data, to determine if the results indicate incest or a genetic anomaly. In cases of incest, Moore tries to identify the relatives. She also founded a private support group for people wrestling with the news, but the work was overwhelming, and she recently turned over some responsibilities to an assistant. “I was getting multiple e-mails a week from people who had first-degree relatives for parents,” she told me. “It’s the worst thing a person can find out from direct-to-consumer testing, other than a relative is a serial killer.”
Genetic anomalies can also be devastating. Once, a parent approached Moore with horrifying news. Her children, conceived by sperm donation, had been born with significant disabilities; a DNA test suggested that they had chromosomal abnormalities consistent with an embryo produced by sperm from an elderly man—a person who clearly had not been her selected donor.
Moore knew the clinic. It was affiliated with the University of Utah. In 2012, she and another parent had figured out that the facility had employed a felon—a former professor who had kidnapped a woman for an “experiment” designed to compel her to love him. The clinic had served some fifteen hundred couples during his time there; following an official investigation, the university conceded that it did not know how many children he had fathered through tampering. “He was playing God,” Moore said. “He was mixing up vials.”
All genealogy is a search for human continuity. When researchers cannot trace someone’s lineage beyond a certain ancestor, they say that they have hit a “brick wall.” There are always brick walls, but the farther back in time one hits them the less painful it tends to be.
For adoptees, who live right up against their brick walls, the proximity can be heartrending, a primordial loss. Genealogists known as “search angels” have long volunteered to help them break through. Many are adoptees themselves, or are relatives of adoptees, and understand firsthand the importance—psychological, and perhaps physical—of finding birth parents. Unlike conventional genealogy, search-angel work is not a race to reclaim memory. It is a revolutionary act, a hack of privacy laws.
By 2011, Moore was working as a search angel, noting on one of her blogs, “I am, and have been for some time, committed to helping adoptees utilize their DNA results to learn more about their ancestry, especially in light of the unjust laws on the books of so many states blocking adoptees from their inherent right to know.” Eventually, she found her way to like-minded volunteers in a Yahoo discussion group, and joined in an effort to develop an elegant, powerful technique for identifying people. They called it the Methodology.
The first step was to establish a DNA profile for the adoptee in a database like GEDmatch, to look for partial genetic matches with other users. The people linked with those matches were not always easy to identify; some users logged on without any personal information or, worse, under aliases. But, when the genealogists succeeded, they could trace back family trees until they identified common ancestors. Then they would reverse the process: starting from the common ancestors, they would build a complete tree of all the descendants, knowing that the adoptee’s parents had to be among them. The amount of DNA that the adoptee shared with matches in the database was a key clue to where he or she belonged in the larger tree; personal details, like birth dates and geography, could also provide clues.
Among the search angels working on the Methodology, Moore had the deepest experience with genetic genealogy. She was not adopted, but she was personally invested in the work. Growing up, she had often heard family members talk about an aunt of hers who had a son stolen from her while she was sedated during labor; the theft, the family was sure, had been orchestrated by her husband at the time, with help from a doctor. “They took the child, and told her that it died,” Moore explained. “They never let her see the child or bury it.” The story implied a wild conspiracy, and Moore was skeptical at first. But, after encountering similar cases as a search angel, she came to take the scenario seriously.
“I have been trying to solve this my whole life,” Moore told me. We were hunched over her computer, reviewing the people who shared her DNA. Scrolling through the list, she stopped at a young man named Erik, who had appeared among her matches on Ancestry earlier this year. He shared about four per cent of her DNA—indicating a second cousin—but Moore did not recognize him. Curious, she accessed an Ancestry account that she maintains for her mother, and another that belongs to her aunt’s daughter. Each woman shared eight per cent of her DNA with Erik—double the amount that Moore did. “I said, ‘Who is this?’ ” she recalled.
Moore built Erik’s extended tree; finding no connections with her own, she decided to reach out to him. “I am shocked to see how closely you are related to my family,” she wrote. “Is there an adoption in your family?” She asked to examine his profile, and he agreed.
After grouping Erik’s relatives into distinct “genetic networks,” Moore traced them to common ancestors: Martin and Julia Timm, who lived in Minnesota in the nineteenth century. With days of painstaking work, she filled in their descendants, until she noticed that one of the Timms’ great-granddaughters had married a man who was born on November 6, 1950, the same day as her aunt’s stolen son—and in the same town. She found an obituary for him, from 2018, and a photo. “I was like, ‘Oh, my God,’ ” Moore told me. “He was full siblings with my first cousin that I’m closest to. And they look alike! They have the exact same red hair.”
Moore had found her missing cousin, but another mystery remained: how were he and Erik linked? The obituary noted that her cousin was survived by two daughters and by a son named Ed. Delving into their biographies, she figured out that Ed had served at a military base near where Erik’s mother lived. Erik was the result of a drunken liaison. Neither man knew of the other.
Excited, Moore called her aunt to explain that the family’s suspicions were true. “I found him,” she said. “He is dead, unfortunately—he died pretty recently.” But, Moore explained, she had located his descendants. Erik, recently married, had just become a father.
Moore expected the news to be earth-shaking. “My aunt found out she has got three more grandchildren,” she told me. “She’s got great-grandchildren. A great-great-grandchild! I was hoping everyone would get to meet.” But her aunt was ninety-one, and the pandemic was raging. “So, um,” Moore said. “My aunt got covid. She got over covid, but never got better, and died.”
Moore flushed. Her hand was quivering. “I feel guilty,” she said. “I found these people. I told them, ‘Hey, your grandmother is alive.’ She never even reached out, because she was so—” She quieted. “Now I feel very bad. I’m the one who figured it out and told everybody.” Moore had hoped that her research would heal a family wound. Instead, she feared it had only added to the sense of loss. She looked past me, out to the Pacific. “It’s so complicated,” she said.
Genetic genealogy, it turned out, could function as an all-purpose de-anonymizer. As long as it targeted the secrets of the living, it would likely become entangled, in some way, with police work. Moore’s effort to track down her lost cousin had apparently identified a crime: the theft of a child. The same possibility existed for people who had been abandoned as babies; their unidentified mothers were often the subjects of criminal investigations. Some adoptees who had followed a genetic trail to their biological parents ended up learning that their mothers had been raped.
“If one of my loved ones was murdered and I had access to that DNA sample, you better believe that I would be using our databases to try to figure out who was guilty,” Moore wrote in 2010. “Wouldn’t you?” But, as she grew into a public figure—one who encouraged people to take DNA tests—she developed a more cautious attitude. People who handed over their genetic data to private companies, or to GEDmatch, never consented to their use by police. “I was very concerned that if I went behind the scenes, and worked with law enforcement, that it would look like a betrayal,” she told me.
Early attempts by police officers to use genetic genealogy had triggered controversy. In 2011, a physicist and former nasa contractor named Colleen Fitzpatrick worked with detectives in Washington State to help identify the killer of a high-school girl. Using Y-DNA testing, she concluded that the suspect was a descendant of Robert Fuller, a colonist who had lived in Salem, Massachusetts, in 1630. The suspect, she told detectives, could also be a man with the surname Fuller. The tip led police to the girl’s neighbor, a family friend, who was totally innocent.
Not long afterward, police in Idaho uploaded a killer’s Y-DNA to the Sorenson database—the archive in Utah, which by then had been acquired by Ancestry. A partial match led them to Michael Usry, a filmmaker in New Orleans who had made a movie, “Murderabilia,” that seemed to echo the crime. Usry also turned out to be innocent. “I am probably still rocked by this,” he told me recently. “Anybody who has read a science-fiction book has an idea of what it could mean.” After the episode, Ancestry shut down access to the entire Sorenson database, which had grown to include a hundred thousand profiles—many belonging to dead people, who could no longer be tested. Moore was horrified. “For us, that’s like burning libraries,” she told me.
With the advent of autosomal-DNA testing, detectives began to surreptitiously rummage through those repositories, too. In 2014, a police department in Florida uploaded a DNA profile from a rapist to GEDmatch, but failed to identify him. It was only a matter of time before people skilled in genealogy would try the same procedure. A year later, a detective in California teamed up with Barbara Rae-Venter, a retired patent lawyer who knew the Methodology, to work on a case. Decades before, a drifter had kidnapped an infant girl and renamed her Lisa; he kept her captive for several years, before abandoning her in an R.V. park, in 1986. Although Lisa had grown into adulthood, she still didn’t know what her given name was, or where she was born; the drifter had zigzagged across the country and perhaps even into Canada. He went by multiple aliases—and was later convicted, as “Curtis Kimball,” of murdering and dismembering a woman. He died in prison, but the detective, convinced that he had more victims, still hoped to piece together the details.
After Lisa took a DNA test at Ancestry, Rae-Venter and a team of volunteers helped establish her profile on all the major databases, including GEDmatch and 23andMe. Each had different users, offering different possible matches. The team identified a pair of common ancestors four generations back, only to learn that the couple had fourteen children, twelve of whom could have been Lisa’s distant ancestors. After more than a year—and twenty thousand hours of research and analysis—the genealogists figured out that she was Dawn Beaudin, from New Hampshire. The drifter had apparently kidnapped her after killing her mother. In 2016, Rae-Venter and her team began working to identify him, too.
Moore was aware that Rae-Venter was taking on criminal casework, but she remained uncomfortable doing so herself. She had asked executives at 23andMe and at Ancestry if they would allow genealogists to use their databases to identify killers or rapists; they flatly rejected the idea. At a talk before law-enforcement officials, she urged the policing community to build its own database for forensic genealogy, to avoid the moral and legal quandaries that private databases posed. None did.
In 2017, Moore attended the International Symposium on Human Identification. “I would like to work more with the forensics community,” she said there. “I am a little bit more hesitant to identify a killer, much as I want those solved.” But she signalled that she was ready to help identify Jane and John Does—a step that other genealogists were also exploring. By then, she was already in contact with Parabon, which had relationships with detectives who were wrestling with unsolved Doe cases. “Identifying deceased people for their families, so they can get some relief—that’s something very much along the lines of what I do now, just kind of reversing it a little bit,” she said. “So that’s really where I think my focus will be.”
The offices of Parabon NanoLabs are on a tree-lined street in Reston, Virginia. Amid the myriad federal contractors and agencies in the area, it occupies a curious niche. The company was founded in 1999 by Steve Armentrout, a computer scientist who specialized in machine learning, and by his wife, Paula. They hoped to build a cloud-computing service, but, as Amazon, Microsoft, and Google began to dominate the field, they pivoted to focus on the intersection of machine learning and biotech. One early contract was with the Department of Defense, which wanted to know if traces of DNA left on improvised explosive devices in Iraq and elsewhere could be used to identify the people behind them. Was it possible to construct a person’s phenotype—all of the observable traits—from genetic clues?
Armentrout regarded this as a computational problem well suited to machine learning. He hired a young geneticist from Harvard to help with the biology. Within a year, the Defense Department was pouring more than a million dollars into the project, and expanding its scope. The genetic profiles that Parabon used to build the phenotypes could also help determine how two people were related. In places like Iraq, where clan affiliations are strong, such comparisons could potentially identify combatants. The tool could also help identify the remains of soldiers killed in action. Parabon called the new product Kinship Inference. Significantly, it was designed to work even when DNA samples were degraded.
To train a machine-learning algorithm to assess degrees of kinship, Armentrout and his staff had to feed it genetic profiles of people whose relationships were already known. At first, this seemed like costly information to harvest. But they realized that genetic genealogists had already done it. Canvassing genealogy conferences, the company found its way to Moore, who agreed to promote its project. Soon, Parabon was inundated with data.
As Moore got to know Armentrout, she told him that she wanted to work on Doe cases, and agreed to an unpaid trial run. “I needed to convince them it was viable,” she told me. Parabon reached out to GEDmatch, and after weeks of discussion obtained permission.
In 2018, Moore took on her first two cases, from a police department in Texas. They proved to be unexpectedly challenging. One involved a Louisiana woman with roots in Acadia, the early French settlement in eastern Canada. “It’s a population that has stuck together and intermarried for centuries,” Moore told me. As she worked on the woman’s tree, she kept running into the same surnames. Worse, because of intermarriage, the quantities of shared DNA were exaggerated. A match that looked like a second cousin, for instance, represented something far less. “I was finding these big segments—thirty to forty centimorgans—that came from before the expulsion of the Acadians from French Canada, before 1755,” she said.
That April, Moore was working on the cases when she awoke to a startling news flash: the Golden State Killer, a serial rapist and murderer who had terrorized the state in the seventies and eighties, had been arrested, after he was identified by a task force that included California detectives and the F.B.I. His real name was Joseph James DeAngelo, Jr. He had once been a cop.
Moore was certain that genetic genealogy was behind the breakthrough. Aware that Barbara Rae-Venter had been quietly working with law enforcement, she asked her if she had been involved. Rae-Venter confirmed that she had. The research had been conducted through GEDmatch and other databases. Moore called Curtis Rogers, the co-founder of GEDmatch, to tell him that genetic genealogy had crossed a Rubicon: his site had been used to catch a killer.
Rogers had been running his nonprofit out of a small house in Florida, which doubled as his wife’s painting studio. When the authorities made the role of GEDmatch public, the property was swarmed with television crews, and Rogers was caught by conflicting imperatives. The spectre of police officers prowling through his site risked eroding public trust; he knew that he could denounce the intrusion and prevent it from happening again. But he also believed that GEDmatch was in a unique position to perform a societal good. The site had helped capture a man who had killed at least thirteen people and raped as many as fifty women. He told Moore that she, too, could use the database to help pursue violent offenders. “You need to be doing this,” he said.
By the end of the week, Rogers had posted a notice on the site, warning users that their profiles might be accessed for “non-genealogical” purposes, and that if they objected they should remove their data. He also redrafted the terms of service, noting that detectives would be issued special research accounts, to be used only for cases of homicide or sexual assault, or for identifying human remains. Moore called Steve Armentrout, at Parabon, to say that she had changed her mind. “I feel like I can work suspect cases now,” she said. “The cat’s out of the bag.”
A week or so earlier, Jim Scharf, the detective in Snohomish County, had asked Parabon to retrieve his suspect’s DNA profile, so that Rae-Venter could take on the case. The company had been slow to respond. But now Armentrout recognized a business opportunity. Parabon had a repository of DNA profiles of criminal suspects, provided by detectives who had purchased Snapshots; these could be used with GEDmatch to solve cases. Armentrout called Scharf back and said that Parabon could perform the genealogy instead of Rae-Venter—and at no charge.
For Moore, identifying William Earl Talbott II, the truck driver, turned out to be straightforward. On GEDmatch, she found two users who shared a significant amount of his DNA—“genetic witnesses,” as they are now sometimes called. One was Chelsea Rustad, a second cousin living in Tacoma. Moore traced Rustad’s ancestry to her great-grandparents, then strove to reconstruct their lives. They had lived in North Dakota, but the wife in the couple, Janna, had died in Seattle. Searching there, Moore discovered that Janna had a granddaughter who had married a man named Talbott. The name struck her, because the other genetic witness she found had a great-grandmother, Ada Marie, who had also married a man named Talbott. Homing in on the convergence, she figured out that Ada Marie’s son had married Janna’s granddaughter. They were the suspect’s parents.
Moore spent the rest of the weekend double-checking her work, uncomfortably aware that she was the only person, other than the killer, who knew who had committed the crime. When the case went to trial, she was able to observe the verdict. Talbott stood, a hoary behemoth of a man. (Scharf told me that he couldn’t get cuffs around his wrists, because they were so thick.) “When they said he was convicted, he collapsed,” Moore recalled. “His female attorney grabbed him, and he said, ‘I didn’t do it.’ I saw it, and I thought, Oh, my God, can I be wrong? Then I thought, No, no, no. His semen was on her pants. Talbott’s DNA was on the zip ties. There is no other explanation.” Tanya Van Cuylenborg’s parents had by then passed away, but her brother was at the trial. Moore said, “It looked, physically, like a burden was lifted off of his shoulders.”
Moore began using GEDmatch to work through a lineup of horrific cold cases. On May 5th, she identified the killer of Terri Lynn Hollis, an eleven-year-old who was murdered in California in 1972. The officers investigating her death had conducted two thousand interviews, over half a century, to no avail. On May 15th, she identified the killer of a teacher who was raped and murdered at her home in Pennsylvania in 1992. On May 30th, she identified the murderer of a twelve-year-old in Washington whose body was dumped in a gulch in 1986. Three days later, she identified a man who had kidnapped, raped, and killed an eight-year-old girl in Indiana in 1988.
She continued this way in the weeks ahead—as if she had discovered a master key to investigative cryptographs made up of imperfect memories, bad evidence, and evasive wrongdoing. Some of the men she had identified were deceased. Some were aging and free; they had apparently been one-time offenders—contrary to the conventional belief that a successful rapist-murderer will likely become a serial rapist-murderer. Paul Holes, who worked the Golden State Killer case, told me that genetic genealogy was revealing a new criminal profile: the rapist or murderer who never “escalates.”
Moore was gaining momentum, but so was a fractious debate, prompted by the Golden State Killer’s arrest. Even in the best of circumstances, the nature of DNA made the question of consent particularly thorny. As one commenter on a genealogy blog pointed out, “When YOU give consent, you are also giving consent for fifty percent of your mother’s and fifty percent of your father’s DNA, too.”
Judy Russell, a blogger known as the Legal Genealogist, noted that, in addition to the problems of consent, police searches were being conducted without judicial oversight. “I think of the DNA results—the links that allow us to reconnect our families—as delicate and priceless vases on glass shelves,” she wrote. “Right now, there’s a bull loose in that china shop.”
In 2019, police in Centerville, Utah, asked Parabon for help with an investigation: someone had broken into a church where an elderly organist was practicing, choked her until she passed out, and then fled. Steve Armentrout told the officers that the crime did not meet GEDmatch’s new terms of service—it was neither a homicide nor a sexual assault—and so the company could not assist them. One of the officers, fearing that lives were at stake, went to Curt Rogers and requested an exception. “The detective said, ‘This guy is out there, and I think he is going to do it again,’ ” Rogers told me. “So I said, ‘O.K., let’s try it this one time.’ ”
Moore’s team quickly identified the strangler. But when news surfaced that GEDmatch had again been involved it caused an even greater uproar. Rogers’s unilateral decision to redefine the policy only stoked fears that private genetic data were being managed by fiat. “In 2012, I called it a ‘DNA geek’s dream site,’ ” Judy Russell noted. “Now that dream has turned into more of a nightmare.”
Striving to navigate the complicated ethics, Rogers rushed to make two key changes. He broadened the site’s terms of service to permit police searches for a wider range of violent crimes. But he also decided that users would, by default, be opted out of those searches, until they explicitly gave permission. GEDmatch by then contained more than a million kits. For anyone who was doing police work on the site, it was now effectively empty.
“To go to zero on this—oh, it was very hard,” Rogers told me.
At the time, Moore was in Idaho Falls, having just helped resolve the investigation that had ensnared Michael Usry, the filmmaker from New Orleans. Not only had she managed to identify the killer; she had helped to exonerate a man, Chris Tapp, who had been wrongfully convicted of the murder. She flew home feeling triumphant.
“I woke up the next morning with zero matches,” she told me. “I went from the highest high to the lowest low.” She had a backlog of half-complete cases; families had been waiting, in some instances for decades, for a resolution. “There were the loved ones of victims, and women who had been raped, who were writing to me,” she told me. She called Rogers. The two had a tearful conversation, not knowing how they could proceed, or even if they could.
The chaos at GEDmatch underscored a fundamental problem with genetic genealogy and policing: there were no rules; anyone could do it, in just about any way. Inside the F.B.I., there was a movement to formally adopt the technique, and with that came attempts to clarify some of the uncertainties. In Los Angeles, Steve Kramer, an F.B.I. lawyer who had helped lead the Golden State Killer case, joined with an agent to prove to the Bureau’s leadership that it was not a fluke. They set a goal to solve twelve cold cases using genetic genealogy, and in 2019 they flew to Washington to present their work to the F.B.I.’s deputy director, David Bowdich. “This is the Lord’s work,” Bowdich told them. “The F.B.I. should own this.”
The Department of Justice, meanwhile, began to consider a legal framework for the new tool. In September, 2019, it issued provisional guidelines, indicating that genetic genealogy could be used only for violent crimes, or for cases that presented a clear threat to public safety or national security. Federal agents were instructed not to upload DNA profiles to consumer repositories covertly, or against the terms of service, and were urged not to trick suspects’ relatives into providing DNA samples. Most important, genetic genealogy had to be treated like a tip, and could not serve as the only basis for an arrest.
There were other significant limitations, but the guidelines remained only advisory, and held little sway at the state level, where many violent crimes are tried. Perhaps for this reason, states began to take notice, too. In 2019, Barry Scheck, a co-founder of the Innocence Project, worked with a legislator in Maryland to develop a bill that would codify and expand the guidelines. “For us to allow private companies to be engaging in this kind of incredibly private surveillance without government oversight, I think, is crazy,” Scheck told me. It is not unusual for law-enforcement officers to get help from outside contractors. But genetic genealogists were being brought into the most sensitive aspects of the investigative process: generating key evidence, selecting suspects. And, while government agencies have strict controls on such data, companies like Parabon faced few legal limits on how they might monetize information that they gathered.
Earlier this year, the bill became law. Moore had advised legislators as they crafted the statute, but, after it passed, she reacted with a police officer’s protectiveness: “If it means that these cases don’t get solved because you’ve added too much of a burden, is that a good thing?” The law requires genealogists to be credentialled, but there are no agreed-upon credentials; flaws like this bothered her. But, when we spoke several weeks later, she seemed confident that the law’s unworkable elements would fall away. “Things will get clarified in time,” she said.
By then, nearly half a million GEDmatch users had opted to allow police to use their kits to identify violent criminals. Some of those people, undoubtedly, were new to the site. More and more people were taking DNA tests. In 2014, barely two hundred thousand people had been tested across all platforms. By 2018, the total was approaching twenty million. Researchers calculated that sixty per cent of all Americans with European ancestry could be identified from their DNA. Before long, they speculated, the number will approach a hundred per cent.
Moore’s furious output resumed. After getting a covid vaccination, she suffered for days from fatigue and migraines. Nonetheless, she spent her recovery resolving two cold cases. One had been with her for years. “The detective is going to retire, so I have been putting a ton of pro-bono hours into it,” she told me. At the same time, Moore volunteered to help a retired N.F.L. player search for his birth parents. “I got an e-mail from someone who heard a rumor that his parents were half siblings, so I have been working a little bit on that, too,” she said.
Her deadlines at Parabon were piling up, but each new request was a plea that she could not ignore. “I know that if I spend a few hours we’ll have the answer,” she told me. The older the mystery, the more pressure Moore felt to solve it quickly—fearing that opportunities for healing would be lost, as people passed away. “I just identified a John Doe, and I think his mom died in May—and he was lost in 1979,” she told me. “He just missed her. After the first couple of times that happened, it was so devastating. That’s why I don’t want to go out and see a movie, or do something recreational. I could be helping to get these answers for somebody before it’s too late.”
One evening, at Moore’s house, she spoke about a lingering mystery that she still hoped to resolve. It was for George R. R. Martin, the celebrated fantasy author whose books inspired the HBO series “Game of Thrones.” He is now seventy-three.
Moore had first encountered the case years earlier, through “Finding Your Roots.” She began working on the show in 2013, after Henry Louis Gates, Jr., heard her speak in Burbank and hired her on the spot. At first, his producers were skeptical, but within a few episodes Moore had established herself as a force. “We have five geneticists who vet her work,” Gates told me. “There were a couple of things she found that were so astonishing to me—I was, like, ‘We’re going to triple-check this,’ and each of the geneticists said, ‘No, CeCe is absolutely right.’ ”
George R. R. Martin had come on the show hoping to learn more about the family of his father, Raymond. He knew a lot about Raymond’s mother, Grace Jones, who had grown up in Bayonne, New Jersey. But he knew little about his father—Luigi Mazucola, an Italian immigrant who took on the name Louie Martin. Grace and Louie were married in 1915, but they separated after Raymond was born. Louie married a younger woman; for the family that he left behind, he became persona non grata.
Martin believed that Louie had left after having an affair. As Moore delved into the family’s genealogy, though, she learned that the opposite was more likely true. The genetics indicated that Raymond’s father was not Louie but another man, an unknown Ashkenazi Jew.
For Martin, the news was wrenching. “It’s uprooting my world here!” he told Gates on the set. “It doesn’t make any sense! So I am descended from mystery?” After the taping, Martin followed the show’s production crew to a local restaurant, wanting to talk more about what they knew. In the years that followed, he and his sisters strove to solve the mystery, to no avail.
It upset Moore that her work, intended to give people a sense of ancestral belonging, had left Martin with only disconnection. She continued to work the case. Initially, there was just one lead to pursue—a Jewish man, named Scott Ross, who shared three per cent of Martin’s DNA. But there are more than a quarter of a million Americans with the surname Ross. She built trees for dozens of Scott Rosses, hoping to locate one who could plausibly share DNA with Martin. Years passed. A combination of deduction and intuition led her to a family in New Jersey, but she could not complete their tree. Uncertain that they were the right family, she refrained from reaching out.
As I sat with Moore, she opened up Martin’s DNA profile for the first time in months. Up popped a new genetic match: another Jewish man, Corey Roberts, who also shared about three per cent of Martin’s DNA. It appeared that the two men and Ross all shared a set of great-grandparents. But how?
Roberts had built a rudimentary family tree, and Moore quickly vetted and expanded it, identifying all of his great-grandparents. But none of them seemed to connect to the Rosses. So Moore returned to the Ross family’s remaining brick wall. To break through, she needed a marriage certificate from the New York City municipal archives. Back in New York, I was able to apply for it, and weeks later a copy printed on pale-blue card stock arrived by mail. On a line at the bottom, pounded into the original document by a government typewriter, was the name of a woman who tied the families together. Her surname was Perlmutter.
I sent a copy to Moore. “WhooHoo!” she wrote. “Here we go.” Within minutes, she was assembling a detailed genealogical portrait of Chaim Yossel Perlmutter. He was born in 1896 in Poritzk, a shtetl in what is now northwestern Ukraine. Two years later, his mother brought him and his two older brothers to Holland, where they boarded the S.S. Rotterdam for Ellis Island. Her husband had made the journey before them, and found work as a tailor. The family lived in Bayonne, just blocks from where Martin’s grandmother Grace had lived. Chaim Yossel Americanized his name to Joseph, but in Bayonne people called him Patty. He worked at a local oil refinery, and became friends with Louie Martin, who worked there, too. When the First World War broke out, both men registered for the draft. Joseph fought as an infantryman in the forests of the Ardennes. He returned to Bayonne in the spring of 1919, and several months later Raymond was conceived.
A few weeks ago, Moore invited Martin to a Zoom chat, so that she could present her findings. At first, he seemed irritated by problems with the connection, but as she spoke his mood softened, and after two hours it was clear that he was moved. Moore hinted at a Shakespearean romance. Joseph and Grace seemed to have fallen in love—but, by the time she separated from her husband, Joseph had married someone else. “They might have just missed each other,” she said.
The two had remained friends, however. Moore shared photos of Joseph and Grace in middle age. In one, they are grinning, their bodies close, heads nearly touching; she is holding his hand. “It’s a very affectionate picture,” Martin said. “Clearly, there is something going on!” In another photo, from 1942, Raymond, a young man in a suit, sits at a cocktail table between Joseph and Grace, his arms around both his biological parents. The resemblance between father and son was striking.
“They look like a happy family,” Moore said.
Martin agreed, but then he wondered if his father ever knew. He explained that Grace’s mother—“a stern matriarch”—would have judged her daughter harshly for having an affair. “There might have been a very good reason that Grace never told anyone—if indeed she never did tell,” he said.
Eventually, Joseph and his wife bought a tavern and moved into the apartment above it. Joseph struggled with drinking, but he ran the tavern as a locus for community: a place for wedding receptions, or Friday-night oysters and beer.
Joseph died before turning fifty, from liver disease. He and his wife never had children, but a niece and a nephew remembered him as a kindhearted uncle. Moore ended her presentation with a photo of his gravestone—a beautifully carved, well-tended monument. A Hebrew inscription began, “Here is buried an upright man, our teacher, the master.”
“Wow,” Martin said.
“How do you feel?” Moore asked.
“I am glad you discovered I am related to him,” he said.
Martin could only glimpse the other side of the brick wall. He asked about connecting with the Roberts and Ross clans, and wondered if there were any living Perlmutters who had known Joseph. There were relatives to meet, questions to ask. “I’m a storyteller,” he said. “I want to know all the details.” ♦
New Yorker Favorites
- How we became infected by chain e-mail.
- Twelve classic movies to watch with your kids.
- The secret lives of fungi.
- The photographer who claimed to capture the ghost of Abraham Lincoln.
- Why are Americans still uncomfortable with atheism?
- The enduring romance of the night train.
- Sign up for our daily newsletter to receive the best stories from The New Yorker.
No comments:
Post a Comment