Two thousand cotton briefs buried on purpose in gardens, parks, and fields across Switzerland. That was the experiment. Eighty days later they dug them up to measure how much fabric remained. The logic is so direct it sounds like a joke: the faster the cotton decomposes, the more active the microbial and invertebrate life in that soil. A shredded pair of underwear is good news. A nearly intact pair is an alarm bell.
The project is called "Proof by Underwear" (in French, La preuve par le slip). Its most uncomfortable result wasn't the national average, but the comparison between land uses. Private gardens showed the fastest decomposition, with 58% of the fabric gone in two months. Agricultural and forest grasslands came in at about half that figure. Lawn areas—those perfectly mowed green rectangles we associate with care and order—had the lowest biological activity of all the categories measured.
Healthy soil is soil where organic matter disappears quickly because there's enough life to break it down. Depleted soil is soil where that same matter stays intact because there's no one to process it. The lawn, which looks like lush plant life, turned out to be the setting with the least life beneath the surface. Aesthetics had been fooling science for years.
Marie Jahoda documented something similar, though in a completely different register, when she studied Marienthal in the 1930s. An Austrian town was left without work after its textile factory closed. What she found, along with Paul Lazarsfeld and Hans Zeisel, wasn't just material poverty. It was a gradual disintegration of invisible structures: time lost its rhythm, community bonds loosened, people walked more slowly without noticing. None of it was visible to the naked eye at first. It took measurement, as with the underwear, to notice that something essential was eroding beneath the apparent surface.
Why does a well-manicured lawn have less biological life than an apparently neglected garden? The practices needed to maintain that uniformity explain most of the answer. Constant mowing, artificial irrigation, herbicides and synthetic fertilizers that eliminate so-called weeds also reduce the diversity of roots, fungi, and microorganisms that sustain the decomposition cycle. A private garden usually incorporates compost, uncollected leaf litter, different plants with roots exploring different depths. That apparent disorder translates, in biological terms, into richness.
This finding contradicts a deeply held intuition, especially in urban and suburban contexts across North America and Europe. We believe well-maintained soil is the kind that looks most uniform and controlled. Researchers have been saying the opposite for decades. Functional soil tends to be heterogeneous, not flat.
What does this mean for agriculture and the management of public spaces? The visual metrics we use to judge how well a piece of land is cared for—even greenery, absence of weeds, order—may be measuring exactly the opposite of real biological health. Swiss public parks, managed with aesthetic and low-maintenance criteria, ended up reflecting that same pattern. Acceptable appearance, poor biological activity. This isn't a problem exclusive to Switzerland. It's a problem with how we design and value green space in general.
Here it's worth being honest about the experiment's limits. Eighty days is a short window. Cotton decomposition depends on temperature, humidity, soil type, and season. The project tried to control for those variables, though some noise always remains. Nor is it a precision laboratory instrument. It's a public awareness tool designed to involve hundreds of volunteers in data collection. That doesn't diminish its value—it changes its function. We're not looking at a peer-reviewed article with margins of error reported to three decimal places. We're looking at citizen science with a double goal: generating real data and generating public awareness. Two things that almost never come together.
Turning an abstract problem—soil degradation, the loss of microbial biodiversity—into something tangible that people can bury, dig up, and see with their own eyes. Reports on organic matter loss measured in carbon percentages rarely move anyone. Almost anyone understands a pair of underwear that comes out of the ground in tatters versus one that comes out nearly intact.
The Generosity in the Doorway argues that sustainable human systems need feedback that makes real progress visible, not artificial indicators disconnected from what actually matters. The perfect lawn is, in a sense, an artificial indicator. It communicates order and care without communicating biological health. It's a vanity metric applied to landscape. And like every vanity metric, it produces behaviors that optimize the metric instead of optimizing what the metric is supposed to represent.
Real alternatives exist. The no-mow movement has gained ground in the UK and the Netherlands, where entire municipalities have stopped mowing strips of public parks to let diverse vegetation grow. The results for pollinator insect biodiversity have been consistent. More wild flora, more microbial activity, more life overall. There's no need to turn every park into an uncontrolled jungle. What's needed is accepting that functional nature doesn't always look like a green carpet.
The Swiss project didn't emerge out of institutional nowhere either. It was driven by organic farming and soil biodiversity organizations, with support from universities and the collaboration of thousands of volunteers who buried and dug up their own underwear. This model of decentralized coordination, where distributed participants generate data that no single institution could collect on its own, has relatives in other fields. I think of community air quality monitoring projects or bird observation networks like eBird. Citizen science doesn't replace institutional science. It complements it with a geographic scale that no lab could afford on its own.
Does everyone need to bury their own underwear in the garden? Probably not. The central lesson holds without going to that extreme. It's worth questioning why we associate visual order with ecological health when the empirical evidence points in the opposite direction. I'm not certain how much this kind of experiment can scale outside contexts with funding and a participation culture like Switzerland's. It's as much a social experiment as a biological one. And that type of experiment depends on conditions that don't always replicate the same way in other countries.
Soil degradation isn't an abstract topic reserved for FAO reports on global erosion. It happens under every garden, every park, every strip of lawn we mow every week without asking ourselves what we're killing underneath. Archaeological science has shown something similar about entire civilizations: land-use patterns predict agricultural collapses long before they become visible on the surface. The Sumerians salinized their land without knowing it until it was too late. We have, for the first time, the chance to measure it beforehand.
Perfect lawn. Life fading away underneath.
How willing are we to change the metrics by which we judge a landscape before the evidence beneath our feet becomes irreversible?
