A collection of slime-mold sporesJoao Paulo Burini / Getty

Biologists have argued for a long time, long before molecular phylogenetics began complicating matters, about how to define species. The concept dates back to at least Carl Linnaeus, who during the 18th century defined a species in his system of binomial nomenclature as an entity (an aggregation of creatures, but still an entity) that had constancy and essence.

Charles Darwin in the 19th century, with help from Alfred Wallace and others, dismissed that sort of idealism, convincing people that species change, species originate and depart, and species consist of individuals that vary from one another, sharing a certain degree of similarity but no ineradicable common essence.

This post is adapted from Quammen’s new book.

In the 20th century, a clarified definition of species was offered by Ernst Mayr, one of the founding neo-Darwinists. Mayr’s famous 1942 definition was: “Species are groups of actually or potentially interbreeding natural populations which are reproductively isolated from other such groups.” Thanks to revolutionaries like the microbiologist Carl Woese, who defined the “third kingdom” of biological forms known as Archaea, we know enough now to see two problems with that definition.

First problem: It’s inapplicable to bacteria and Archaea, which don’t “interbreed” in anything like the way implied by Mayr. Second problem: How can “reproductively isolated” be an absolute standard if genes are continually transferred horizontally (by viral infection and other mechanisms) and if, furthermore, members of one species sometimes breed with members of another, producing new lineages of hybrid offspring? Answer: Reproductive isolation is a useful and intuitive standard, yes, but not an absolute one.

Take Homo sapiens, the species most dear to our hearts. In the era of DNA sequencing, scientists have recognized that the human genome contains evidence of hybridizing events. Homo neanderthalensis was discovered in 1856, named in 1864, and for many decades considered a discrete species, closely related to us within the hominid family, but distinct. Some experts now consider the Neanderthals to have been a subspecies of Homo sapiens, more properly called Homo sapiens neanderthalensis, but others argue that Homo neanderthalensis is still the right label, representing the group as a full species.

In any case, our lineage diverged from theirs sometime between about 300,000 and 600,000 years ago, maybe more, when pioneers left Africa and colonized Eurasia. From those pioneers would descend several non-African species, including Homo neanderthalensis. Our own lineage, known as “modern humans,” sent another wave of dispersers out of Africa and colonized Europe again, but much later, around 50,000 years ago. Then, for one reason or another, the Neanderthals disappeared.

Paleoanthropologists have long speculated that either our ancestors killed off the Neanderthals, by direct aggression, or forced them to go extinct, by competition, or else absorbed them to some degree, by interbreeding. But there was no conclusive proof. Nowadays, since the recovery and sequencing of Neanderthal DNA by a team including the Swedish biologist Svante Pääbo, analyses indicate that hybrid matings did occur between Neanderthals and modern humans. The human genome, especially as found among non- African peoples descended from those hybrid matings, now contains about 1 to 3 percent Neanderthal DNA.

And it’s not just Neanderthals in our genome. The human lineage diverged from what became the chimpanzee lineage 7 million, or 10 million, or maybe 13 million years ago—nobody knows the timing with any precision. But recent genomic analysis suggests that, sometime well after the big split, hominid ancestors and chimp ancestors came back together for hybrid matings, and that those hybrid matings have left genuine chimpanzee genes (not just close human equivalents) in parts of our genome.

Of course, the presence of chimpanzee genes, or Neanderthal genes, isn’t the half of it. There’s also that viral DNA—including syncytin-2, a gene co-opted from a retrovirus, repurposed to enable human pregnancy. The fact that DNA from endogenous retroviruses constitutes 8 percent of the human genome certainly complicates our sense of Homo sapiens as a species of primate.

It complicates our sense of human individuality even more. So does the recognition that each of us contains, as a necessity for health and digestion and other aspects of our physiology, some dozens of trillions of bacterial cells, representing thousands of different bacterial “species.” And so does the realization that within every one of our human cells reside captured bacteria, long since transmogrified into mitochondria, without which we couldn’t exist.

The concept of an “individual” continues to defy attempts by biologists and philosophers of science to perfectly define it. Some experts have argued that it’s crucial to have such a definition, because the logic of evolution by natural selection—Darwin’s core principle—depends on the differential survival and reproduction of … individuals. If so, what is an individual? Is a single bacterium an individual? The work of the biologist Lynn Margulis, from 1967 onward, vastly advanced and complicated that question, with her theory of endosymbiosis raising the proposition that all complex creatures, including us humans, are chimerical creatures compounded of bacterial and other genomes.

Carl Woese and Nigel Goldenfeld teased at similar questions in their 2007 paper “Biology’s Next Revolution.” Sorin Sonea, who in 1988 argued that all earthly bacteria constitute a single “superorganism,” a single interconnected genetic entity, would say, no, bacteria considered one by one are not individuals. Is a worker ant, incapable of reproducing itself, living its life to maximize the reproductive output of the queen ant, an individual? Or is the ant colony itself an individual? Is it another “superorganism”?

Consider the Portuguese man o’ war, that peculiar relative of jellyfish, floating the ocean surface like a swim bladder with stinging tentacles. An individual? It seems so, but biologists who study these things tell us that, no, a Portuguese man o’ war is not. It too, like an ant hill or a termite community, is a colony of individual creatures (in this case, small multicellular forms known as zooids), aggregated for a common purpose and variously performing specialized functions.

Likewise that very strange thing known as a cellular slime mold, which during one phase of its existence looks and behaves like a garden slug, but at another phase reveals itself to be a fine-tuned team of individual amoebae. When food is scarce, the amoebae aggregate into the slug, unified in their effort to crawl toward better habitat. They raise a stalk atop which sits a fruitlike body, and, when that opens, disperse spores. If the spores land in a place where food particles (bacteria) are available, they awaken as new amoebae.

Likewise again with aspen trees in a grove. They may look like individuals, but, in fact, aspens grow as clonal eruptions from underground rootstock, all interconnected, all sharing the same genome, sometimes including hundreds or thousands of trees across a wide area. The grove is the individual.

By one accounting, the largest organism on Earth may be a single aspen clone composed of many thousands of trees spread across more than a hundred acres in Utah’s Fishlake National Forest. It weighs about 13 million pounds, this aspen individual, and is roughly 80,000 years old.

And then there’s a third challenged categorical: the tree of life. It doesn’t, in fact, look like an oak, or a Lombardy poplar. Even aspens in a grove make an unsuitable metaphor because, though interconnected underground, they don’t reconnect above. Their roots form a network, but their limbs and branches only diverge, growing away from one another, seeking open space in which their leaves may harvest light. They don’t converge, they don’t inosculate, not in the wild. The tree of life is not a true categorical, because the history of life just doesn’t resemble a tree.

Woese, the father of Archaea, knew that, though it wasn’t among his highest priorities to say so. He interested himself in big limbs, not small branches. And of big limbs, in his view, considering the past 4 billion years, there were three: Bacteria, Eukarya, Archaea.

Those three diverged from the last universal common ancestor of all life as we know it— life on Earth, life using one common genetic code, life that began with RNA and then yielded cells and became very complex. Three living domains, from which all else results. A holy trinity. It’s almost religious.

This post is adapted from Quammen’s new book, The Tangled Tree: A Radical New History of Life.

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.