Not long ago in rural Maryland, a mere 50 miles from the bustle of downtown D.C., there was a lab in which a scientist and a few hundred rhesus macaque monkeys, over five decades, turned out gobs and gobs of data to decode what, ultimately, makes us who we are. The monkeys are gone now but their legacy lives on in some of the most remarkable science behind genes and mothering, evolution, and mental health.

Two things make rhesus macaques particularly suited for the job of studying humans: they share 95 percent of our DNA and exhibit personalities, just like we do. Steve Suomi was the lab’s last chief scientist until it closed down in 2008. During his time, he saw generation after generation of monkeys whizz through life, buffeted by biology and circumstance. Suomi had a special interest in a type of macaques he called uptight: shy and fearful as infants, they’d grow into something rather similar to depressed human adults. Suomi wanted to know why. Was it nature or nurture? And what did that mean for us humans? Are some of us wired for suffering? Or is it life that twists us one way or another?

Vulnerability tips into disorder once a certain stress threshold has been passed.

Over the past century, our mental deviances have been blamed on rogue genes or unfeeling parents (also known as refrigerator moms), or some combination of the two. The prevailing view in psychiatry — called the diathesis-stress model — holds that some people have an underlying, usually genetic, predisposition to mental illnesses. On its own, though, being prone isn’t enough. Vulnerability tips into disorder once a certain stress threshold has been passed. For illness to ensue, you need a sufficiently harsh environment — a difficult childhood or adversity later in life — to trigger your vulnerability genes.

But something about the monkeys didn’t quite fit this story. Thanks to advanced gene sequencing technology that became available in the ’90s, Suomi analyzed the DNA of all macaques in his colony and found that some of them carried a genetic mutation, which in humans, has been linked to depression. He looked for the same mutation in the genomes of other primates but found none that had it.

This was odd. “What is it about humans and rhesus monkeys”, Suomi wondered, “that differs from these other primate species?” He mulled it over and then it jumped out at him: It’s our remarkable adaptability. While most chimps and monkeys perish out of their natural habitat, humans and rhesus macaques have spread, survived, and even flourished across an astonishing range of environments, all over the planet. Could it be that the two things that separated us and rhesus monkeys from all other primates were somehow connected? In other words, is it possible that we owe our extraordinary resilience as a species… to our vulnerability genes?

The data suggests that the genes that put us most at risk of mental illness may also be the same genes that help to make us healthier, happier, more resilient, and successful, both as individuals and as a species.

That seemed absurd. It defied both common sense and science canon. But around the same time, a professor of human development at the University of California, Davis, called Jay Belsky, uncovered a curious pattern in the human research that suddenly made Suomi’s speculation entirely plausible. Belsky was poring over studies of gene-environment interaction when he found that, in many cases, subjects with a genetic risk of developing mental disorder had not only the worst outcomes but also the best ones.

Separately from Belsky, other researchers were seeing the same puzzling thing in studies across genetics, personality psychology, and child development. Since then, evidence for the upside of vulnerability genes has only been growing. The data suggests that the genes that put us most at risk of mental illness may also be the same genes that help to make us healthier, happier, more resilient, and successful, both as individuals and as a species. The research, though still new, holds out the beguiling possibility that Leonard Cohen had it right, after all: that sometimes the cracks in us are also how the light gets in.

Reviewing the existing research, Belsky found that people with a genetic vulnerability, who came from a harsh early environment, were at a higher risk of developing mental disorders, just like the diathesis-stress model in psychiatry would predict. But similarly, vulnerable people from a nurturing early environment turned out normal — indeed, sometimes better-off than their good-gene peers. Contrary to what you’d expect, the combination bad genes/good environment often produced greater health and superior psychological functioning than good genes/good environment.

In study after study, this pattern appeared with remarkable consistency, yet somehow researchers had completely missed it. A focus on illness, though well-justified in most cases, had blinded them to the silver linings in their own data. But once Belsky and others pointed out the good side of bad genes, it started cropping up everywhere, in old and new studies alike.

In 2004, for example, researchers from the Netherlands set out to study how genes, and moms, affect kids’ antisocial behaviour. Marian Bakermans-Kranenburg and her colleagues screened over 2,000 families, picking out for further study those with children who scored highest on aggression. The researchers found that kids with insensitive mothers were more aggressive and ran a higher risk of disturbed mental development — but only if they also carried a specific variant of a gene called DRD4. Known as the ADHD gene, the bully gene, and the slut gene, DRD4 helps to process dopamine in the brain and one of its mutations has been linked to violence, hyperactivity, and novelty-seeking. It was kids with this mutation, raised by harsh or aloof mothers, that turned out the most antisocial of all kids in the study.

So far, this is standard diathesis-stress theory: risky gene plus bad environment equals disorder. The surprise came in the second part of the experiment, when Bakermans-Kranenburg and her team split the families into two groups. The first group were followed around by camera-wearing researchers, who then used the video clips to teach the moms how to read their children’s behaviour and use empathy to quell tantrums. The second group of moms served as a control and received no video feedback or parenting advice.

The kids in the two groups started out similarly aggressive but ended up dramatically different. Two years after the experiment, the research team followed up with the families to measure the impact of the video intervention. Here’s what they found: The kids in the intervention group who carried the risky variant of the DRD4 gene, and whose moms had improved their parenting skills, were 37 percent less aggressive. Meanwhile, the kids in the control group, who carried no genetic risk, had reduced their problem behaviors by a mere 8 percent — and that’s when their moms, too, had become more sensitive and empathetic (only without the intervention). And even this paltry improvement, the researchers suggest, may have happened anyway simply as a result of growing up.

In another oft-cited study, Ariel Knafo, a psychologist at Hebrew University, brought 167 three-year-olds to his lab. There the youngsters met a female researcher disguised as a sort of damsel in distress: she “accidentally” dropped her pen, then “accidentally” hurt her knee, then, “accidentally” again, lost her finger doll. This was, of course, a ruse designed to test what psychologists call pro-social behavior, meaning the children’s willingness to help. As you may expect by now, the kids who carried the risky variant of the DRD4 gene turned out the least pro-social of all, provided that they were also raised by strict, insensitive mothers. And the most pro-social kids? They, too, carried the genetic risk for aggression but had, instead, nurturing mothers. The positive parenting had somehow tipped these kids’ antisocial predisposition into its polar opposite — kindness.

But it’s not only at-risk kids who seem to respond so strongly to nurturing environments. Shelley Taylor, a psychologist at UCLA, looked at 118 young men and women to see if they would react differently to stress based on which variant of the 5-HTTLPR gene they carried. The 5-HTTLPR gene regulates serotonin levels in the brain and comes it three variants — or alleles, in genetics speak. There’s the long/long, the long/short and the short/short allele, the latter of which, research suggests, poses the highest risk of depression and the former protects against it.

Here’s what happened. Young adults with the short/short allele and frequent high stress over the previous six months had, indeed, the most depressive symptoms. On the other hand, young adults with the same risky allele but low amounts of recent stress had the fewest depressive symptoms — two times fewer, in fact, compared to their genetically protected peers. Just like in Knafo’s experiment, where the kindest kids carried aggressive genes, here, too, the most resilient young adults were those with the depressive genes. Something about their environment — be it loving parents or low stress — seemed to turn their genetic vulnerability into protection against the very disorders it made them susceptible to.

This striking conclusion and the growing evidence behind it has spurred a new theory that is forcing us to profoundly rethink mental health and illness. Spanning evolutionary psychology, child development, neuroscience, and genetics, this theory goes under different names, among them Biological Sensitivity to Context and Differential Susceptibility Theory. I prefer the more poetic rendition coined by science journalist David Dobbs: the Orchid Hypothesis.

The name was inspired by the pioneering work of Bruce Ellis and Thomas Boyce, who, in 1995, proposed that some children are like dandelions and others are like orchids. Dandelions, the innocuous weeds that crop up everywhere, can feel at home in all terrain and climate — from the side of pavements to waste dumps, from mountain slopes to burned forests. Not orchids. Take them out of their native rainforests and they perish. But plant them in the right soil, give them proper care, and you’ll have a thing of unutterable, mesmerizing beauty. So irresistible is their appeal, in fact, that over the centuries orchids have fueled an obsession among collectors called orchidelirium. Lives have been risked, blood spilled and billions of dollars squandered on the rarest of species.

Just like dandelions, some children turn out fine no matter what circumstances they find themselves in. In the Bakermans-Kranenburg study, kids who didn’t carry the risky variant of the DRD4 gene were all but unaffected by the quality of parenting they received. There was little difference between those with good moms and those with bad moms: their good genes seemed to protect them from the negative aspects of their environment.

We might call these kids normal or well-adjusted. But according to the Orchid Hypothesis, that’s not quite accurate. Dandelion kids are simply more impervious, more insensitive, and unresponsive to their environment and, it turns out, not just to its negative effects but to the positive ones, too.

By contrast, Orchid children are extremely susceptible in both directions. They respond to good and bad environments alike, suffering the most in the latter but also making the most out of the former. Theirs, according to Dobbs, is not so much vulnerability to negative experience as sensitivity to “all experience.” Where dandelions are hardy, orchids are malleable; while dandelions resist the influence of the outside world, orchids yield to it. Their genes, in this sense, are not vulnerable but extremely plastic, responding to the environment and amplifying its effects — “for better and for worse”, as Jay Belsky likes to put it.

One interesting implication of this new research is what Belsky calls “the dark side of resilience.” Resilient people, we think, have this highly honed ability to bounce back from setbacks which is key to their health and happiness. And so we scramble to build resilient classrooms and resilient companies that will thrive in an uncertain world. But what if resilience is not a mental superpower that anyone can develop? What if, instead, it is more like insensitivity to adversity, a case of Dandelionism? The danger, according to Jay Belsky, is that resilience training won’t work for the Orchids, who can’t override a hardwired propensity to feel stress more intensely just by picking up a few emotional regulation techniques. And if we see resilience as a learnable skill, we are also bound to see its failure to work for some as a personal flaw: they simply didn’t put in the effort.

On the other hand, Orchids’ very sensitivity may present unexpected treatment advantages and suggest interventions that will actually work for them. It’s not hard to imagine that, being more susceptible to good environments, people most at risk of suffering mental distress will also most respond to therapy. Indeed, a 2012 study of Romanian orphanages found that it was only kids with the risky variant of the 5-HTTLPR gene — the short/short allele — that benefited when placed in high quality foster care. Although the research on positive interventions is still in its early stages, the initial evidence seems promising. If it gets confirmed, it can profoundly change things for many people who won’t even seek help fearing a verdict that they may be well beyond repair.