The women who made America’s microchips and the children who paid for it

Mark Flores sits at the kitchen table drawing birthday cakes. At 44, he loves to draw circles, a skill he’s mastered over the past decade of his life. His thick black hair is neatly combed like the Superman cartoon on his T-shirt. Grasping thick Crayola markers, he lines up small circles in rows within larger oblong shapes. Mark has accomplished much more than his mother, Yvette, was told he ever could when he was born — when doctors said he wouldn’t be able to interact with people because of his intellectual disabilities.

Instead, Mark greets most people with a big, toothy smile, stoops over to give them a hug if they’re willing, and is quick to answer most questions with an enthusiastic “yeah.”

When Mark coughs at the table, Yvette asks him if he needs water. “You don’t have to, Mark, you can say no,” she says, her soft brown eyes behind black cat-eye glasses. “We’re learning ‘no,’” she says to me as an aside. Like her son, she’s quick to flash a smile. Her dark hair falls loosely around her shoulders, streaked with silver against her face.

Yvette was working at a factory in the early days of Silicon Valley when she got pregnant with Mark. An employment center placed her there in 1975. She was 18 and had recently gotten her GED. She just needed a job, she says. Today, she is one of hundreds of former Silicon Valley workers who say they unknowingly handled harmful substances on the job and later filed suit against their employers.

Manufacturing semiconductors, the chips needed for all our consumer electronics today, was reshaping Yvette’s community. The industry employed many women and young people from Hispanic and Asian immigrant families who’d previously worked in canneries that were closing up shop as Americans started importing more fresh fruit. The factories offered a new kind of assembly line work that you didn’t need a degree or much training to land.

But a lack of appropriate safety measures left workers vulnerable to a slurry of chemicals that posed dire health risks. Over the years, many of the workers had miscarriages, including Yvette. Some gave birth to children with disabilities; Mark was one of those kids.

Yvette has painstakingly built a safe haven for Mark in their home filled with toys and holiday decorations year-round. An inflatable Santa greets guests in the living room, even though it’s March when I visit. There’s a giant heart-shaped balloon from Valentine’s Day, deflating but still floating above Mark’s head. Plush Chuck E. Cheese characters are seated around their home, dressed up in old clothes complete with mismatched shoes peeking out below too-big pants. Mark’s drawings are tucked beneath many of the stuffed animals.

The Verge sought out legal documents and old media reports to find Mark and some of his peers — a generation of kids, now adults, whose parents say they paid the price for lax workplace standards in an industry that uses an ever-evolving cocktail of chemicals to make computer chips.

For some time, miscarriages associated with the chip manufacturing industry were well documented in research papers. The risks discovered by scientists prompted companies to pledge to stop using certain chemicals — at least in the United States. After that, though, the industry did an about-face on transparency with researchers. As personnel data dried up, it became harder to show adverse reproductive health outcomes among employees on paper, even as many parents said their kids were living with the consequences.

US-based research into these health risks slowed down as semiconductor manufacturing moved to Asia in the 1980s. But with the new push to revive chip manufacturing in the US, these concerns are bubbling back up to the surface. In recent years, as diplomatic relations with China have deteriorated, and with the AI arms race continuing to accelerate, the Biden administration poured billions of dollars into subsidizing a domestic chip industry. Now, unions and health and environmental advocates are once again fighting to protect workers to make sure that Silicon Valley’s toxic history doesn’t repeat itself.

“Now looking back I mean, it just gets me so so upset. I understand they have to make money,” Yvette tells The Verge, “but not at the cost of another person.”

Like many of her colleagues in the early days of Silicon Valley, Yvette didn’t know what kinds of risks were posed by the substances she was handling. In some manufacturing facilities — called fabs — workers donned masks and head-to-toe smocks called “bunny suits” before entering “clean rooms.” Those protective measures weren’t meant to shield workers from the industrial solvents, adhesives, and other chemicals in the room; they were for the chips, which were so sensitive they needed to be shielded from airborne microbes.

Nothing had been designed to protect the people making them.

By the mid-1980s, there was increasing talk of miscarriages among fab workers, the majority of whom were women. No link had yet been established between semiconductor chip manufacturing and miscarriages; perhaps for that reason, companies freely cooperated with academic researchers, confident that they would ultimately dispel concerns about working conditions. But the series of studies that followed would reach grave conclusions.

Scientists at the University of Massachusetts Amherst surveyed a small group of workers at a single fab in Massachusetts. Then, the University of California, Davis launched a larger study from a sample of about 50,000 employees at 14 different semiconductor companies representing around a fifth of semiconductor workers in the US at the time. IBM, believing it took more precautions than its peers, decided to fund its own separate study with researchers from Johns Hopkins University.

Ultimately, these three separate industry-backed studies found higher than expected rates of miscarriages among women who worked on semiconductor production lines — as much as twice as high as those who didn’t work there.

The large study by University of California researchers was published in the American Journal of Industrial Medicine in 1995. The authors first conducted a historical analysis by interviewing workers who had gotten pregnant while working for a semiconductor company between 1986 and 1989. They found a 43 percent higher risk of miscarriage among women who worked in chip fabrication rooms compared to employees in other departments (often office workers) without the same exposure to chemicals.

Researchers also carried out a second analysis of current employees, conducting interviews and collecting urine samples to monitor pregnancies and miscarriages. In that cohort, they found a 25 percent increase in miscarriage risk among those who worked in fab rooms. The figures were adjusted for age, ethnicity, and other potentially confounding factors like whether each person smoked cigarettes. A majority of the miscarriages in that cohort were subclinical, meaning the woman didn’t know about the pregnancy before losing it.

“The bottom line is that both studies showed a small but significant risk of increased miscarriage associated with fabrication work,” says Marc Schenker, a coauthor of that research and a distinguished professor emeritus of public health sciences and medicine at the UC Davis School of Medicine.

The highest risk for miscarriage was among women who worked with certain ethylene glycol ethers (EGEs) used in photoresist chemicals that coat semiconductors, which the industry subsequently pledged to phase out in the 1990s.

“The single biggest takeaway from my point of view, is that low concentrations, what we think of as very safe concentrations, can actually have a serious impact and reproductive effects,” says Katharine Hammond, an associate dean at the University of California, Berkeley’s graduate School of Public Health and another one of the authors of the industry-backed miscarriage research.

And EGEs are just one group of chemicals in a constantly evolving mix — not only in semiconductor clean rooms, but in manufacturing facilities across the supply chain for computer chips. A 1999 article in the journal Environmental Health Perspectives names 10 “chemicals of concern in the semiconductor industry” that include known carcinogens and other substances tied to reproductive health problems, including lead and arsenic. IBM alone faced some 250 lawsuits filed by employees who developed cancer, suffered miscarriages, or whose children were born with intellectual disabilities or birth defects.

Semiconductor manufacturing was beginning to move to Asia more steadily by the ’90s, with a majority located in South Korea, Taiwan, Japan, and China by the 2010s. Since then, it’s been harder to suss out the potential reproductive health risks workers face. Many in Asia were still exposed to EGEs through at least 2015, a Bloomberg investigation found.

But later studies on workers’ reproductive health have been a mixed bag. One study in South Korea in 2015 attempted to estimate miscarriages using claim data from the National Health Insurance. Looking at instances of women who sought healthcare for a miscarriage, the researchers found higher rates among microelectronics workers compared to control groups of bank employees, the larger working population, and “economically inactive women.” The authors also said, however, that more studies “based on primary data collection and careful surveillance are required to confirm these results.”

In 2017, a separate study relying on questionnaires from workers at two semiconductor manufacturing plants in South Korea didn’t find a statistically significant difference in miscarriages among workers on the production floor compared to clerical workers. Workers in one production area, however, showed higher risk prior to 2008. That suggested that changes in working conditions, like more automation, might have reduced risk. A 2019 review of prior English-language research on semiconductor work and miscarriage subsequently confirmed a statistically significant association between fab work and miscarriage from early research through the mid-1990s, but “due to the lack of studies … could not evaluate spontaneous abortion risk based on current working environment.”

There are limitations to each of these studies that make it difficult to determine what the reproductive health risks are for semiconductor workers today. But there are more clear outcomes from research conducted on miscarriages in the early days of Silicon Valley, when companies were open to sharing employee data. A generation of workers unknowingly handled hazardous substances on the job without adequate protections and paid the price.

At the kitchen table, Yvette borrows one of her son’s markers to illustrate the job she had when she was pregnant with him.

Between 1975 and 1980, she worked on an assembly line in Santa Clara County, California, for Spectra-Physics, a company that billed itself as a leading supplier of lasers for semiconductor manufacturing. Lasers would, for instance, be used to cut individual chips from larger silicon wafers.

But first, workers like Yvette would have to put those lasers together. With a bubblegum pink marker, she draws a long cylindrical shape with an angled edge. Her job was gluing a slanted glass bevel to a tube, then hitting the joint with a blowtorch to cure it.

She wore a mask, one thin enough that she could still smell through it as she stirred and sprayed the green adhesive they called frit. She’d dump any unused chemicals down the sink in the small room she shared with one other worker cooking beryllium in an oven behind her. She could feel the blast of heat each time he opened the oven door.

“There was so much going on there,” Yvette recalls. “I had to walk out of there a lot of times because it got so stuffy.”

One day at work in 1978, Yvette began bleeding in the restroom of Spectra-Physics. The experience is mostly a blur to her today, but there are things she still distinctly remembers — the intense, painful cramping; another woman inside the restroom telling her she should go to the hospital; and finally, a doctor at the hospital telling her that she had had a miscarriage.

It’s still difficult for Yvette to talk about what happened. Her voice falls to a quiet monotone as she tells this part of her story. She went back to work after what happened in the bathroom, still unaware of what risks she faced on the job. She didn’t suspect at the time that anything in her workplace could have led to the miscarriage. In 1979, when she was 22, she got pregnant with Mark. She kept working at Spectra-Physics until Mark was born with a severe intellectual disability, when she left her job to become his primary caregiver.

She’s a tenacious advocate for her son. When he graduated from a special education school at 23, she borrowed a cap and gown from a photography studio so that he could attend the celebration in the traditional garb, even though it wasn’t part of the ceremony.

“I wanted him to experience the cap and gown, which he did,” she says. The tassel was “a real pretty blue.”

Things got harder after Mark’s father died in a motorcycle accident in 2007 at the age of 48 — not much older than Mark is now. He had trouble understanding what happened, why his dad wasn’t around anymore. He acted out, sometimes flipping chairs over or suddenly grabbing people. Yvette struggled at first to find help for her son, looking for therapy and day care programs that would take him. “I tried calling out, ‘help me,’ you know, to a lot of agencies,” she says. “I’m not the only one who can have this happen to you.”

It wasn’t until she heard an ad on the radio a couple of years later that she started to understand how many other people might be facing something similar. The ad was from a law firm trying to reach former electronics industry workers who had children with birth defects.

“I kept hearing it,” she says. “And I’m going, that sounds like everything that’s happened to me … it was checkmarks in my head going yup, yup,” she says matter of factly.

Yvette called the number from the ad and filed suit against Spectra-Physics in 2010. During the discovery process, another employee came forward to share a document revealing what was in the green “frit” Yvette handled each day at work: the biggest ingredient was lead.

Lead dust likes to travel. It’ll all too readily stick to skin and clothes, following a person all the way home after work. It doesn’t easily wash away with regular soap — there are special deleading soaps for a reason. If a pregnant person breathes or ingests lead, it can move through the placenta into a developing fetus. In fact, the fetus is likely to have the same level of lead in its blood as the adult carrying it.

“It’s a very potent developmental hazard, so it very much harms a developing brain,” says Carissa Rocheleau, PhD, a reproductive occupational epidemiologist in the National Institute for Occupational Safety and Health Division of Field Studies and Engineering.

That could have a wide range of effects, Rocheleau says, potentially impacting behavior, cognitive abilities, motor skills, and speech. Research has also linked high levels of lead in the blood to increased risk of miscarriage.

The impact lead has on the body is well understood — certainly much better understood relative to the many other substances found in semiconductor fabs that use an ever-changing cocktail of chemicals. What many of these substances can do to fetal development is an open question.

LeAnn Severson got a job at a semiconductor plant for Electronic Arrays in Mountain View, California, the same year Mark was born. She worked the evening swing shift sorting wafers — the large silicon discs from which individual computer chips are cut. She met her soon-to-be husband Keith on the job; he was a wafer sorter, too.

During each shift, they placed wafers under a microscope for a machine to test for defective chips. If it spotted one, it would mark the chip with a dot of red ink. To be really sure, the machine would need to test it again. So wafer sorters like the Seversons would take the potentially defective disk, wash the ink off with a solvent in a sink, and then place the wafer back under the microscope.

LeAnn remembers squirting solvent onto the wafers using plastic bottles with bent straws. When the squirt bottles ran empty, they would refill them from large brown bottles sitting on the floor. The bottles were labeled methyl ethyl ketone (MEK), a name that was hard for her to forget.

“Back in the ’60s, my dad would always talk about filling the car with ethyl gasoline or something like that. And this kind of stuck in my head, the methyl ethyl, it kind of rhymes. And so I remembered it,” she says. She didn’t think the chemical was any harsher than nail polish remover; MEK has a sickly sweet smell similar to acetone.

Gloves were optional on the job, LeAnn says. “If you were wearing fingernail polish, it would come off. We would just continue to get a paper towel and take off the rest of our polish, you know, with that same solution because it smelled like fingernail polish remover. No one told us not to.”

There’s been limited research on what constant exposure to MEK does to the body — it’s not the kind of thing that can be easily studied in humans. It’s been linked to developmental effects in fetuses in studies with mice and rats. But results from animal studies often don’t hold true for humans, especially when it comes to reproductive outcomes. A suite of other industrial solvents and chemical products were used on the job, according to the lawsuit LeAnn Severson filed against her former employer — any of those could have affected her body as well.

The Severson’s son, Darryl, was born in 1980 with microcephaly, a condition in which a baby is born with a smaller head than usual. Today, he loves watching Sesame Street and going to the library to pick out books with his mom. He reads at about a fifth-grade level. “He’s an adult child,” LeAnn says.

LeAnn didn’t draw the connection between Darryl’s condition and the chemicals she worked with until the late ’80s, when she saw a PBS news special on birth defects in Silicon Valley. She wrote a handwritten letter to the state’s health department, which she says subsequently referred her to the same occupational health attorney that represented Yvette.

Both women settled out of court with their employers, which helped each of them care for their sons full time. IBM also reached settlements in at least two closely watched suits filed by other parents who worked at its semiconductor plant in East Fishkill, New York. One child was born with microcephaly. Another was born with a rare disorder called Hallermann-Streiff syndrome that causes face and skull malformations. IBM avoided what likely would have been high-profile trials by settling, and denied wrongdoing.

“I mean, everybody trusts their workplace,” she says. “They think they’re going to be safe when they walk in.”

Amanda Hawes has represented people who gave birth to children with intellectual disabilities or birth defects after working in California’s early electronics industry since the 1970s. Hawes moved to San Jose, California, before it was known as the heart of Silicon Valley. Back then, it was called the “Valley of Heart’s Delight” because of its orchards lush with peaches and plums — fruit that was used in the region’s many canneries. (Yvette Flores still remembers her mom coming home from her cannery job late at night, smelling of peaches.)

Hawes started off representing cannery workers fighting for lost wages. As the tech industry started to crowd out agriculture, many of the same employees started working for semiconductor fabs instead of canneries — and coming to Hawes with new health issues they traced back to chemicals at work. Hawes and other advocates have spent decades trying to figure out just how many people have been affected. It’s a hidden epidemic, she believes, in part because semiconductor companies stopped cooperating with outside researchers after those early studies into miscarriages.

In 1997, California’s Department of Health Services crafted a proposal to develop a database to study birth defects, cancer rates, and deaths among electronics workers. The state had been collecting data from birth certificates in the San Francisco area since 1983 through its Birth Defects Monitoring Program. They already had names and birth information; they just needed job titles from employers to study whether there were any links between birth defects in children and their parents’ occupations. The initiative even garnered a $100,000 funding commitment from the Environmental Protection Agency, The Wall Street Journal reported the following year.

This time around, industry leaders decided it wasn’t a good idea to participate. “I might as well take a gun and shoot myself,” Intel’s director of environmental affairs at the time reportedly said during a January 1998 EPA meeting where the proposal was ultimately blocked. Agreeing to share personnel records for the study “would be like giving discovery to plaintiffs’ lawyers,” The Wall Street Journal reported him saying.

Even if the chip industry had cooperated, there was a fairly large hurdle, just from a purely academic perspective. Ascertaining any connections between birth defects and toxins at work would have been a much taller order than linking miscarriages and toxins. Defects affect around 3 percent of births in the US, while roughly 20 percent of pregnancies end in miscarriage. Because birth defects are uncommon, studies would need to include a larger pool of people. The California Department of Health Services proposal hoped to gather data from at least 100,000 workers for its proposed database, compared to the large UC Davis study on miscarriages based on around half that number of employees.

There are also many different kinds of birth defects that may each warrant their own study. “Something can be a risk factor for breast cancer, but not stomach cancer, right? And the same is true with birth defects,” says Rocheleau. In other words, different triggers can lead to different kinds of health outcomes, whether that’s varying types of cancer or birth defects.

Many factors throughout each stage of fetal development and even before pregnancy might lead to a birth defect. Does a chemical damage DNA or alter the physical development of a fetus? And if so, where and how? Which parent was exposed to the chemical, and what impact does that have?

Miscarriages themselves also complicate research into birth defects. The fetus has to survive long enough to be diagnosed, and the most severe birth defects might end with a miscarriage, according to Rocheleau. In that sense, you can think of birth defects and miscarriages as different points along the same spectrum of possible outcomes from chemical exposure. Some pregnancies might end. Others could come to full term, even though the fetus is vulnerable to the effects of that chemical.

“Sometimes people tend to think of the placenta as being a barrier, but that’s really not true for a lot of small molecules, like most chemicals are,” Rocheleau says. “It’s maybe more accurate to think of it as an expressway for nutrient delivery and it can deliver toxicants to the fetus as well.”

A fetus doesn’t have all of its defense systems up yet; its organs are still a work in progress. A single cell in a fetus may be destined to become an entire organ system — and damage to that cell, of course, could have widespread effects on the development of that system. The liver, which filters out toxins, isn’t histologically complete until two weeks before birth. And the toxin doesn’t even need to reach the fetus to affect it — there are a number of indirect risks to a fetus simply because its parent is sick. If a toxic substance triggers health issues in the fully grown adult who is carrying it, the adult’s health issues can have additional effects on fetal development.

That’s all to say, it’s complicated. A pair of studies on birth defects among semiconductor workers in Taiwan published in 2007 and 2008 also found diverging outcomes based on sex. The researchers found a heightened risk of death from birth defects among the children of male workers. They didn’t see the same elevated risk in female workers, however — they instead documented an increased risk of cancer among the children of female workers.

From the front seats of their car, Amanda Hawes and her husband, Ted Smith, survey the campus of nondescript office buildings and vacant parking spaces. While Hawes represented workers, Smith started the nonprofit Silicon Valley Toxics Coalition in the 1980s with other residents concerned about chemicals leaching into the soil and groundwater.

“It’s a ghost town,” Smith says. It’s all empty now, following a wave of layoffs in Silicon Valley over the past year. In the parking lot, Ted is looking for a site where he remembers an old carbon treatment system running to clean up contaminated water. Now, it’s just a suburban landscape.

Much of it, however, sits atop land so contaminated that it was placed on a national priority list for cleanup. They’re called Superfund sites, and Santa Clara has more of them than any other county in the US. It’s a legacy of the early days of Silicon Valley, when companies simply flushed chemicals down sinks or stored them in leaking underground tanks.

Across the US, new fights are starting to crop up in communities near new semiconductor manufacturing sites. Now in their 70s, the couple is an integral part of CHIPS Communities United, a national coalition that’s backed by Sierra Club, the Communications Workers of America union, and other labor groups.

The rush to develop more powerful chips for AI is raising the stakes. But without stronger protections for workers, says Hawes. “We’re getting ready to repeat the same stuff.”

The 2021 CHIPS and Science Act, spearheaded by the Biden administration, set aside $52.7 billion in funding for domestic semiconductor manufacturing. That included $200 million specifically for semiconductor workforce development. The administration launched two semiconductor “workforce hubs” in the last two years in Ohio and Arizona. In Ohio, community colleges are working with Intel to create curriculums for semiconductor technician roles. In Arizona, Intel, TSMC, and NXP Semiconductors announced new apprenticeship programs with support from state and local governments. And in April, the Biden administration announced a third workforce hub in upstate New York. The Biden administration said that its incentives attracted nearly $450 billion more in private investments for semiconductors and electronics manufacturing. All in all, it’s expected to create more than 145,000 jobs in construction and manufacturing if the subsidies survive the current Trump administration.

More than $30 billion of CHIPS Act funding has already been doled out to companies building new factories across the US. Meanwhile, health, safety, and environmental concerns have started to crop up with some projects. A truck driver died in an explosion in May, reportedly while transporting a pressurized tank of sulfuric acid at a site where TSMC is building a new manufacturing facility.

The CHIPS Communities United coalition wants more transparency around what kinds of chemicals semiconductor manufacturers are using in the next generation of fabs. Last summer, the group asked the Department of Commerce to conduct more thorough environmental reviews of at least three federally funded projects to build or expand fabs. That included TSMC and Intel’s plans in Arizona and a Micron project in Idaho. The draft assessments didn’t include enough information on what kinds of substances workers might be exposed to or adequate details on what protections would be in place, they contend in comments they submitted to the Commerce Department and shared with The Verge.

TSMC didn’t respond to requests for comment. Intel didn’t comment on the Arizona facility. With regard to the industry’s historical use of hazardous chemicals and comments that Intel’s director of environmental affairs reportedly made in 1998, spokesperson Patricia Oliverio said in an emailed statement that “the health and safety of our workforce is our top priority. Intel uses advanced engineering controls and best practices based on regulatory and industry standards to protect our workforce.”

In October 2023, CHIPS Communities United published a letter to semiconductor industry execs asking them to sign legally binding community benefit agreements when they build new fabs. They asked companies to replace chemicals that can cause cancer, miscarriages, birth defects, and fetal brain damage. The letter urged companies to be more proactive, rather than merely following the Occupational Safety and Health Administration’s exposure limits for chemicals — most of which haven’t been updated since the 1970s after years of industry pushback and Republican efforts to kneecap OSHA.

The agency itself says its exposure limits “are outdated and inadequate for ensuring protection of worker health” and recommends that employers adhere to more recent guidelines developed by the state of California, the NIOSH, and the nonprofit American Conference of Governmental Industrial Hygienists.

Semiconductor manufacturers say that they’ve limited the exposure of workers to different chemicals, in part by automating much more of the manufacturing process.

“Just as chips themselves have improved dramatically over the last several decades, so too has semiconductor worker health and safety, thanks to advancements in fabrication techniques and automation, the adoption of enhanced safety standards and engineering controls, and improvements in the quality of personal protective equipment,” Dan Rosso, vice president of communications at the Semiconductor Industry Association, said in a statement to The Verge.

Even so, health and labor advocates remain worried. “It still takes people to actually make these things run even though they’re highly automated,” says Ted Smith. “And so there still is the opportunity for exposure — and it’s never to one chemical, it’s mixtures of chemicals.”

The chemical cocktail is constantly changing in workplaces, and often at a faster rate than scientists can study the potential risks at play. On top of that, companies are often reluctant to share what chemicals they’re using, protecting them as trade secrets. The semiconductor industry’s reassurances are, simply, difficult to verify.

In the meantime, Silicon Valley, where it all started, is still cleaning up the mess left behind by the first generation of semiconductor fabs.

The place where Yvette Flores worked when she got pregnant with Mark is also a Superfund site. But unlike the cluster of empty office buildings in Santa Clara County, this site is buzzing with activity on Sundays. Instead of a laser manufacturing plant, there’s a church that serves a primarily Korean American community, complete with a colorful playground and basketball court in the back. Picnic tables and lawn chairs sit on Astroturf behind a white picket fence labeled the “seniors’ lounge.” The church’s website advertises a Sunday school program named Love Land for “our littlest ones” and Joy Land for toddlers to kindergarteners.

“Thousands of pounds” of volatile organic compounds have been removed from the groundwater and soil at this Superfund site over the past three decades, according to the EPA’s website. More assessments are needed to determine whether the possibility of vapors rising from lingering contaminants into indoor spaces is still a concern, it says. Part of the ongoing cleanup plan, as stated by the EPA, is to prevent homes and schools from being built there. Neither the EPA nor church leadership answered questions from The Verge about why the church was located there.

MKS Instruments, which owns the brand Spectra-Physics now, said in an email that “Spectra-Physics no longer exists as a legal entity, and we have not owned or leased any of the properties that comprise the superfund site since 2010. We cooperate closely with EPA and the San Francisco Regional Water Quality Control Board on the ongoing site cleanup measures.”

Yvette learned about the church on the site for the first time when I mentioned it to her. She was alarmed — an hour later, she tells me that it’s still bothering her.

We’re back at her kitchen table after taking Mark to eat some pizza.​​ The thought of the church — the Sunday school, the playground — is lingering in her mind.

“That’s so sad,” she says, drawing out the word sad into two syllables.

She’s soon distracted by cats that have wandered into the backyard, part of a menagerie of animals in the neighborhood that she and Mark watch through the sliding glass door across from the kitchen each day.

“As long as I have breath, me and you, huh?” Yvette says, standing up from the table to give her son a hug.

“Me and you,” Mark says.