The California gnatcatcher, Polioptila californica, is a little gray bird. Males are identifiable by their black cap (absent in winter), females by the slight brownish tinge of their plumage. Distinguishing the California gnatcatcher from the fifteen other species of gnatcatcher is easy, the ornithologist Jonathan Atwood told me—“if you’re a gnatcatcher freak.” All the others are little gray birds, too.

Scientists have divided Polioptila californica into several subspecies. The northernmost of these, the coastal California gnatcatcher (Polioptila californica californica) lives along the western coast of the Baja Peninsula, from El Rosario, Mexico, north to Long Beach—a range that includes some of the choicest undeveloped real estate in southern California. When the U.S. Fish and Wildlife Service listed the subspecies as threatened under the Endangered Species Act in 1993, the region’s ranchers and developers were furious. The listing would cost them more than $900 million, according to Fish and Wildlife’s own estimate.

In 2010, a group of these ranchers and developers, represented by a libertarian law firm called the Pacific Legal Foundation, began a long campaign to delist the bird. Eventually, after a series of failures, the firm decided to target a certain vagueness at the center of the ESA. While the law is clear in its intent—to protect endangered and threatened species and subspecies, along with the ecosystems they depend on—the act is somewhat less clear in its definitions of those categories. In a 2017 petition, Pacific Legal sought clarification. What, it demanded to know, is a species?

“For the man in the street,” the ecologist Arthur Shapiro wrote, “species are species, self-evident bits of nature.” For scientists, things are fuzzier. They have proposed dozens of different ways to define species, all of which work quite well when comparing organisms that are obviously unalike. But when comparing organisms that are physically or genetically similar, determining where one species ends and another begins gets tricky.

Taxonomy, wrote the biologist Ernst Mayr, is biology’s oldest branch, but its modern incarnation can be traced to Carl Linnaeus’s Systema Naturae, first published in 1735. Linnaeus proposed a descending taxonomic scale—kingdom, class, order, genus, and species—along with the binomial system of identification still in use today. For Linnaeus, species were discrete and unchanging, created and maintained by God. With the advent of evolutionary theory, however, these boundaries began to break down. While Jean-Baptiste Lamarck is best known for his misguided ideas about the mechanisms of evolution, the French naturalist was right about the problem of species. Old species gradually give rise to new ones, Lamarck wrote in his 1809 Philosophie zoologique, with no clean break. The taxonomic rankings used to organize the world’s life-forms were thus a rigid framework built around a fluid reality—“purely artificial aids which we have to use in the arrangement and division of the various observed natural productions. . . . Nature has made nothing of the kind.”

Half a century later, Darwin seemed even more impatient with the question. “I look at the term species,” he wrote,

as one arbitrarily given for the sake of convenience to a set of individuals closely resembling each other, and that it does not essentially differ from the term variety, which is given to less distinct and more fluctuating forms.

In the ensuing decades, some scientists followed this position to its extreme, arguing that species were not only fluid but fictitious. “ ‘Species’ have only a subjective existence,” wrote the paleontologist Benjamin Burma. “They are merely convenient labels for arbitrary groupings and have only a minimum of biological meaning.” The botanist Charles Bessey was equally blunt: “They are mental concepts and nothing more.”

Even so, species had their defenders. Some have focused on organisms’ shared ancestry, others on their shared habitat, others on shared genes. But the definition that remains the most widely used was proposed by Mayr in 1942. It is known as the biological species concept, and posits that a species is a collection of populations of organisms that interbreed or could potentially do so.

In the Endangered Species Act of 1973, Congress echoed this definition. According to the law, species include

any subspecies of fish or wildlife or plants, and any distinct population segment of any species of vertebrate fish or wildlife which interbreeds when mature.

Mayr’s biological species concept worked well enough as scientific shorthand, but it faltered as a diagnostic tool. Taxonomists say that a small but significant proportion of organisms—perhaps as much as 20 percent—are edge cases, where even professionals sometimes disagree on classification. The question of whether the organisms interbreed sounds simple, but in reality this doesn’t provide a practical or objective answer to whether a given organism belongs to a particular species. Even organisms that are indisputably considered separate species often interbreed, especially plants. Deciding where one species ends and the next begins is a matter of interpretation. As the taxonomist Scott Thomson told me, “You’ll have your data; that data is objective. But where you draw those lines that say that something is a species or a subspecies—that’s a subjective judgment.”

This ambiguity escaped the notice of lawmakers. Beyond the small group of staffers who drafted the bill, Congress had an unsophisticated view of species, said Holly Doremus, a professor of environmental law at the University of California at Berkeley. “I don’t think Congress understood in 1973 how mushy this was,” Doremus told me. Legislative discussion focused on easily recognizable icons such as grizzly bears, bald eagles, and wolves. “The arbitrariness is a difficult academic problem,” said Damien Schiff, the lawyer who drafted the Pacific Legal petition. “But incorporated into statute, it becomes a civics problem.”

The California gnatcatcher, according to Atwood, has a “rather tortured” taxonomic history. The bird entered the scientific record in 1881, when, on the basis of a slight difference in the gray of its feathers, an ornithologist distinguished it from the other gnatcatchers of the southwestern United States and northwestern Mexico. It remained a species until 1926, when a zoologist demoted it to a subspecies of the black-tailed gnatcatcher, Polioptila melanura. In the Eighties, Atwood, then pursuing his PhD at UCLA, took up the question again. After examining dozens of museum specimens and recording hours of birdcalls in the field, Atwood concluded that the demotion had been a mistake. His proposal was accepted by the American Ornithologists’ Union—“the taxonomic gods,” as he called them—and with that, the bird once more became its own species, P. californica.

Soon after, Atwood told me, he ran into Patrick Mock, a former classmate from UCLA, at a meeting of the ornithologists’ union. Mock suggested that Atwood petition the U.S. Fish and Wildlife Service to list the California gnatcatcher under the Endangered Species Act. This would not only protect the bird, but also its rare habitat—the coastal sage scrub. Mock couldn’t file the petition himself; the consulting firm he worked for at the time was often hired by developers who threatened that ecosystem. So he asked Atwood, then working at the nonprofit Manomet Bird Observatory, to do so.

Listing the entire species was out of the question. California gnatcatchers are common (“South of El Rosario, you could swing a dead cat and hit one,” one scientist told me), and it wasn’t entirely clear how many subspecies there might be. Over the years, scientists had suggested as many as four, with varying ranges and descriptions. Atwood decided to try listing the northernmost of these proposed subspecies, P. californica californica. This rare bird could be distinguished from other California gnatcatchers, Atwood wrote, by the width of its bill, the length of its middle toe, and other, more arcane morphological measures. In 1990, the Manomet Bird Observatory and the Natural Resources Defense Council filed their listing petition. Three years later, over the protests of ranchers and developers, Fish and Wildlife listed the coastal California gnatcatcher as a threatened subspecies.

The bird (and its habitat) was safe from developers—but not from taxonomists. In the late Nineties, a group of scientists led by the biologist Robert Zink revisited the California gnatcatchers with a new tool: genetics. Technological advances had made it possible for scientists to quickly compare organisms or entire populations by looking at short segments of their DNA. Whereas taxonomists had earlier been limited to what they could see (or, in the case of vocal creatures such as birds and frogs, what they could hear), now they could look directly at organisms’ genetic blueprints.

DNA sequencing lent itself to what is known as the phylogenetic species concept. This view regards species as part of an evolutionary tree whose limbs, branches, and twigs are defined by common ancestry. Where the biological species concept makes determinations based on whether organisms can interbreed, the phylogenetic view relies on histories of evolutionary descent. In Zink’s estimation, this definition makes the taxonomic rank of subspecies more or less superfluous. “The problem is that the subspecies label has been used idiosyncratically,” Zink told me. “Some subspecies are barely discernible. Some are really different.”

In their 2000 report, Zink and his colleagues confirmed that California gnatcatchers formed a species, as Atwood had determined. But the group found no evidence of P. californica californica. “Coastal sage scrub populations of California Gnatcatchers are not genetically distinct from populations in Baja California,” they wrote. The coastal California gnatcatcher—the subspecies protected by the Endangered Species Act—did not exist.

Pacific Legal relied on Zink’s independent assessment when it filed a petition to delist the coastal California gnatcatcher in 2010. The ESA, they argued, was protecting a nonexistent bird. But Fish and Wildlife denied the petition, writing that by analyzing mitochondrial DNA and not nuclear DNA, Zink had insufficiently supported his findings. Two years later, Zink began conducting a second study, this time using a broader set of genetic markers. Again, Zink announced that he had found no evidence of a distinct coastal subspecies. Pacific Legal filed a second petition.

Fish and Wildlife hired an outside consultant to convene a panel of six ornithologists, geneticists, and taxonomists. They met for two days in Carlsbad, California, to determine their ruling. In the resulting 269-page report, they wrote that Pacific Legal had focused too much on genetics at the expense of the subspecies’ physical and ecological characteristics, and had failed to demonstrate any existing genetic distinctions to begin with. Once again, Fish and Wildlife dismissed the petition.

The firm’s lawyers began thinking about a different approach. The coastal California gnatcatcher wasn’t the only organism protected by the ESA whose existence was debatable. Legal battles over the Preble’s meadow jumping mouse and the southwestern willow flycatcher had also centered on uncertain taxonomic boundaries. “We realized there’s a larger problem,” Schiff told me. In 2017, Pacific Legal filed a new petition demanding that the government clearly define both species and subspecies.

The law firm offered its own suggestions. A species, it argued, is

a group of actually or potentially interbreeding populations that are reproductively isolated from other such groups to the extent that the rate of fertile hybridization is less than 1% per generation.

A subspecies, meanwhile, is “a population for which at least 75% of its distribution lies outside the distribution of any other population within the species,” and is distinct on the basis of at least two traits, whether genetic, physical, or ecological. The former follows Mayr’s biological species concept while adding the hard cutoff of 1 percent; the latter is a definition proposed by the ornithologist Dean Amadon in 1949. Schiff said he is open to other definitions, as long as they include a quantifiable edge, some way to consistently determine what constitutes a species and what doesn’t. “Even if the definitions are not that great,” he told me, “they’re going to be improvements over the status quo.”

Few believe that Pacific Legal has the interests of any little gray birds at heart. The firm—which has ties to the petroleum industry and right-wing outfits such as the Charles Koch Foundation and the Adolph Coors Foundation—has a long history of litigation against the ESA. According to the Guardian, it has fought, often successfully, to delist or reduce protections for many species, including the black-capped vireo, the greater sage-grouse, the golden parakeet, the valley elderberry longhorn beetle, the Kuenzler hedgehog cactus, the Tobusch fishhook cactus, the lesser long-nosed bat, the wood stork, the West Indian manatee, and gypsum wild-buckwheat. Critics say that adopting the proposed definitions would only make it easier for groups like Pacific Legal to prevent new creatures from being added to the endangered-species list and strip protections from those already listed.

But although the scientists, legal experts, and conservationists I spoke to almost all disagreed with the particular definitions Pacific Legal offered, some were sympathetic to the idea that the ESA could benefit from greater precision. “I actually agree with PLF on the issue of clarifying what species and subspecies mean,” said Ya-Wei Li, the director for biodiversity at the Environmental Policy Innovation Center. He told me that unclear definitions of species and subspecies make ESA listings vulnerable to never-ending petitions and lawsuits, as Pacific Legal has spent decades demonstrating: “When there is uncertainty, people tend to argue.”

Imprecise definitions can also contribute to a bigger problem. As Doremus explained, by mandating that listing decisions be made according to “the best available scientific data,” Congress disguised a political and philosophical decision as a purely scientific one. Which organisms get listed is not only a question of data, but also one of money, time, and value, she said. It is unavoidably terrible to calculate the worth of any creature’s very existence, but in the face of mass extinction, and with finite resources, it is impossible to save everything. Some choices we make implicitly—the ESA offers no protections to fungi or microorganisms, for instance, although human activities have endangered many of their members. Other choices we avoid entirely—for example, should species that are the only member of their genus, family, and order be given more weight than species that share a genus with many others? “We’re making value choices about what we think is important and what isn’t,” Doremus said. “We should be making those choices more openly.”

Codifying definitions of species and subspecies could be a start. Pacific Legal’s petition has found few supporters in the conservation community, but it’s possible to imagine definitions that would help protect, rather than hurt, threatened or endangered populations. These definitions might serve as translations between the worlds of taxonomy and law, announcing what we want to save and why, and ensuring that the organisms we spend our time and money on satisfy those criteria. Our understanding of species—the way we conceive of them, count them, and conserve them—has always involved an element of choice. As Lamarck argued, a species, independent of its existence in nature, is a human tool. When it comes to the Endangered Species Act, it isn’t hard to imagine that the tool could be made to better fit its task.