I have been fascinated with living things since childhood. Growing up in northern California, I spent a lot of time playing outdoors among plants and animals. Some of my friends and I would sneak up on bees as they pollinated flowers and trap them in Ziploc bags so we could get a close look at their obsidian eyes and golden hairs before returning the insects to their daily routines. Sometimes I would make crude bows and arrows from bushes in my backyard, using stripped bark for string and leaves for fletchings. On family trips to the beach I learned how to quickly dig crustaceans and arthropods out of their hiding spots by watching for bubbles in the sand as the most recent wave retreated. And I vividly recall an elementary school field trip to a grove of eucalyptus trees in Santa Cruz, where thousands of migrating monarch butterflies had stopped to rest. They clung to branches in great brown globs, resembling dead leaves—until one stirred and revealed the fiery orange inside of its wings.
Moments like that—along with a number of David Attenborough television specials—intensified my enthrallment with the planet’s creatures. Whereas my younger brother was obsessed with his K’Nex set—meticulously building elaborate roller coasters—I wanted to understand how our cat, well, worked. How did she see the world? Why did she purr? What were fur and claws and whiskers made of? One Christmas I asked for an encyclopedia of animals. After ripping the wrapping paper off a massive book that probably weighed half as much as I did, I sat near the tree reading for hours. Not too surprising, then, that I ended up writing about nature and science for a living.
Recently, however, I had an epiphany that has forced me to rethink why I love living things so much and reexamine what life is, really. For as long as people have studied life they have struggled to define it. Even today, scientists have no satisfactory or universally accepted definition of life. While pondering this problem, I remembered my brother’s devotion to K’Nex roller coasters and my curiosity about the family cat. Why do we think of the former as inanimate and the latter as alive? In the end, aren’t they both machines? Granted, a cat is an incredibly complex machine capable of amazing behaviors that a K’Nex set could probably never mimic. But on the most fundamental level, what is the difference between an inanimate machine and a living one? Do people, cats, plants and other creatures belong in one category and K’Nex, computers, stars and rocks in another? My conclusion: No. In fact, I decided, life does not actually exist.
Allow me to elaborate.
Formal attempts to precisely define life date to at least the time of ancient Greek philosophers. Aristotle believed that, unlike the inanimate, all living things have one of three kinds of souls: vegetative souls, animal souls and rational souls, the last of which belonged exclusively to humans. Greek anatomist Galen proposed a similar, organ-based system of “vital spirits” in the lungs, blood and nervous system. In the 17th century, German chemist George Erns Stahl and other researchers began to describe a doctrine that would eventually become known as vitalism. Vitalists maintained that “living organisms are fundamentally different from non-living entities because they contain some non-physical element or are governed by different principles than are inanimate things” and that organic matter (molecules that contained carbon and hydrogen and were produced by living things) could not arise from inorganic matter (molecules lacking carbon that resulted primarily from geological processes). Subsequent experiments revealed vitalism to be completely untrue—the inorganic can be converted into the organic both inside and outside the lab.
Instead of imbuing organisms with “some non-physical element,” other scientists attempted to identify a specific set of physical properties that differentiated the living from the nonliving. Today, in lieu of a succinct definition of life, Campbell and many other widely used biology textbooks include a rather bloated list of such distinguishing characteristics, for instance: order (the fact that many organisms are made from either a single cell with different compartments and organelles or highly structured groups of cells); growth and development (changing size and shape in a predictable manner); homeostasis (maintaining an internal environment that differs from an external one, such as the way cells regulate their pH levels and salt concentrations); metabolism (expending energy to grow and to delay decay); reacting to stimuli (changing behavior in response to light, temperature, chemicals or other aspects of the environment); reproduction (cloning or mating to produce new organisms and transfer genetic information from one generation to the next); and evolution (the change in the genetic makeup of a population over time).
It’s almost too easy to shred the logic of such lists. No one has ever managed to compile a set of physical properties that unites all living things and excludes everything we label inanimate. There are always exceptions. Most people do not consider crystals to be alive, for example, yet they are highly organized and they grow. Fire, too, consumes energy and gets bigger. In contrast, bacteria, tardigrades and even some crustaceans can enter long periods of dormancy during which they are not growing, metabolizing or changing at all, yet are not technically dead. How do we categorize a single leaf that has fallen from a tree? Most people would agree that, when attached to a tree, a leaf is alive: its many cells work tirelessly to turn sunlight, carbon dioxide and water into food, among other duties. When a leaf detaches from a tree, its cells do not instantly cease their activities. Does it die on the way to the ground; or when it hits the ground; or when all its individual cells finally expire? If you pluck a leaf from a plant and keep its cells nourished and happy inside a lab, is that life?
Such dilemmas plague just about every proposed feature of life. Responding to the environment is not a talent limited to living organisms—we have designed countless machines that do just that. Even reproduction does not define a living thing. Many an individual animal cannot reproduce on its own. So are two cats alive because they can create new cats together, but a single cat is not alive because it cannot propagate its genes by itself? Consider, also, the unusual case of turritopsis nutricula, the immortal jellyfish, which can indefinitely alternate between its adult form and its juvenile stage. A jelly vacillating in this way is not producing offspring, cloning itself or even aging in the typical fashion—yet most people would concede it remains alive.
But what about evolution? The ability to store information in molecules like DNA and RNA, to pass on this information to one’s offspring and to adapt to a changing environment by altering genetic information—surely these talents are unique to living things. Many biologists have focused on evolution as life’s key distinguishing feature. In the early 1990s, Gerald Joyce of the Scripps Research Institute was a member of an advisory panel to John Rummel, manager of NASA’s exobiology program at the time. During discussions about how best to find life on other worlds, Joyce and his fellow panelists came up with a widely cited working definition of life: a self-sustaining system capable of Darwinian evolution. It’s lucid, concise and comprehensive. But does it work?
Let's examine how this definition handles viruses, which have complicated the quest to define life more than any other entity. Viruses are essentially strands of DNA or RNA packaged inside a protein shell; they do not have cells or a metabolism, but they do have genes and they can evolve. Joyce explains, however, that in order to be a “self-sustaining system,” an organism must contain all the information necessary to reproduce and to undergo Darwinian evolution. Because of this constraint, he argues that viruses do not satisfy the working definition. After all, a virus must invade and hijack a cell in order to make copies of itself. "The viral genome only evolves in the context of the host cell," Joyce said in a recent interview.
When you really think about it, though, NASA's working definition of life is not able to accommodate the ambiguity of viruses better than any other proposed definition. A parasitic worm living inside a person’s intestines—widely regarded as a detestable but very real form of life—has all the genetic information it needs to reproduce, but it would never be able to do so without cells and molecules in the human gut from which it steals the energy it needs to survive. Likewise, a virus has all the genetic information required to replicate itself, but does not have all the requisite cellular machinery. Claiming that the worm’s situation is categorically different from that of the virus is a tenuous argument. Both the worm and virus reproduce and evolve only "in the context" of their hosts. In fact, the virus is a much more efficient reproducer than the worm. Whereas the virus gets right down to business and needs only a few proteins inside a cell’s nucleus to initiate replication on a massive scale, the parasitic worm’s reproduction requires use of an entire organ in another animal and will be successful only if the worm survives long enough to feed, grow and lay eggs. So if we use NASA's working definition to banish viruses from the realm of life, we must further exclude all manner of much larger parasites including worms, fungi and plants.
Defining life as a self-sustaining system capable of Darwinian evolution also forces us to admit that certain computer programs are alive. Genetic algorithms, for instance, imitate natural selection to arrive at the optimal solution to a problem: they are bit arrays that code traits, evolve, compete with one another to reproduce and even exchange information. Similarly, software platforms like Avida create "digital organisms" that "are made up of digital bits that can mutate in much the same way DNA mutates." In other words they, too, evolve. "Avida is not a simulation of evolution; it is an instance of it," Robert Pennock of Michigan State University told Carl Zimmer in Discover. “All the core parts of the Darwinian process are there. These things replicate, they mutate, they are competing with one another. The very process of natural selection is happening there. If that’s central to the definition of life, then these things count.”
I would argue that Joyce’s own lab delivered another devastating blow to NASA's working definition of life. He and many other scientists favor an origin of life story known as the RNA world hypothesis. All life on our planet depends on DNA and RNA. In modern living organisms, DNA stores the information necessary to build the proteins and molecular machines that together form a bustling cell. At first, scientists thought only proteins known as enzymes could catalyze the chemical reactions necessary to construct this cellular machinery. In the 1980s, however, Thomas Cech and Sidney Altman discovered that, in collaboration with various protein enzymes, many different kinds of RNA enzymes—or ribozymes—read the information coded in DNA and build the different parts of a cell piece by piece. The RNA world hypothesis posits that the earliest organisms on the planet relied solely on RNA to perform all these tasks—to both store and use genetic information—without the help of DNA or an entourage of protein enzymes.
Here’s how it might have happened: Nearly four billion years ago, in Earth’s primordial soup, free-floating nucleotides—the building blocks of RNA and DNA—linked into longer and longer chains, eventually producing ribozymes that were big enough and complex enough to make new copies of themselves and thus had a much greater chance of surviving than RNAs that could not reproduce. Simple self-assembling membranes enveloped these early ribozymes, forming the first cells. In addition to making more RNA, ribozymes may have joined nucleotides into chains of DNA; nucleotides may have spontaneously formed DNA as well. Either way, DNA replaced RNA as the main information-storing molecule because it was more stable. And proteins took on many catalytic roles because they were so versatile and diverse. But the cells of modern organisms still contain what are likely remnants of the original RNA world. The ribosome, for example—a bundle of RNA and proteins that builds proteins one amino acid at a time—is a ribozyme. There's also a group of viruses that use RNA as their primary genetic material
To test the RNA world hypothesis, Joyce and other researchers have tried to create the types of self-replicating ribozymes that may have once existed in the planet’s primordial soup. In the mid-2000s, Joyce and Tracey Lincoln constructed trillions of random free-floating RNA sequences in the lab, similar to the early RNAs that may have competed with one another billions of years ago, and isolated sequences that, by chance, were capable of bonding two other pieces of RNA. By pitting these sequences against one another, the pair eventually produced two ribozymes that could replicate one another ad infinitum as long as they were supplied with sufficient nucleotides. Not only can these naked RNA molecules reproduce, they can also mutate and evolve. The ribozymes have altered small segments of their genetic code to adapt to fluctuating environmental conditions, for example.
“They meet the working definition of life,” Joyce says. “It’s self-sustaining Darwinian evolution.” But he hesitates to say that the ribozymes are truly alive. Before he goes all Dr. Frankenstein, he wants to see his creation innovate a completely new behavior, not just modify something it can already do. “I think what’s missing is that it needs to be inventive, needs to come up with new solutions,” he says.
But I don’t think Joyce is giving the ribozymes enough credit. Evolution is a change in genes over time; one does not need to witness pigs sprouting wings or RNAs assembling into the letters of the alphabet to see evolution at work. The advent of blue eye color between 6,000 and 10,000 years ago—simply another variation of iris pigments—is just as legitimate an example of evolution as the first feathered dinosaurs. If we define life as a “self-sustaining system capable of Darwinian evolution,” I cannot see any legitimate reason to deny self-replicating ribozymes or viruses the moniker of life. But I do see a reason to ditch this working definition and all other definitions of life altogether.
Why is defining life so frustratingly difficult? Why have scientists and philosophers failed for centuries to find a specific physical property or set of properties that clearly separates the living from the inanimate? Because such a property does not exist. Life is a concept that we invented. On the most fundamental level, all matter that exists is an arrangement of atoms and their constituent particles. These arrangements fall onto an immense spectrum of complexity, from a single hydrogen atom to something as intricate as a brain. In trying to define life, we have drawn a line at an arbitrary level of complexity and declared that everything above that border is alive and everything below it is not. In truth, this division does not exist outside the mind. There is no threshold at which a collection of atoms suddenly becomes alive, no categorical distinction between the living and inanimate, no Frankensteinian spark. We have failed to define life because there was never anything to define in the first place.
I nervously explained these ideas to Joyce on the phone, anticipating that he would laugh and tell me they were absurd. After all, this is someone who helped NASA define life. But Joyce said the argument that life is a concept is “perfect.” He agrees that the mission to define life is, in some ways, futile. The working definition was really just a linguistic convenience. “We were trying to help NASA find extraterrestrial life,” he says. “We couldn’t use the word ‘life’ in every paragraph and not define it.”
Carol Cleland, a philosopher at the University of Colorado Boulder who has spent years researching attempts to deliniate life, also thinks that the instinct to precisely define life is misguided—but she is not yet ready to deny life's physical reality. “It’s just as premature to reach the conclusion that there is no intrinsic nature to life as it is to define life,” she says. “I think the best attitude is to treat what are normally taken as the definitive criteria of life as tentative criteria.”
What we really need, Cleland has written, is "a well-confirmed, adequately general theory of life." She draws an analogy to chemists in the sixteenth century. Before scientists understood that air, dirt, acids and all chemical substances were made of molecules, they struggled to define water. They could list its properties—it was wet, transparent, tasteless, freezable and it could dissolve many other substances—but they could not precisely characterize it until researchers discovered that water is two hydrogen atoms bonded to an oxygen atom. Whether salty, muddy, dyed, liquid or frozen, water is always H20; it may have other elements mixed in, but the tripartite molecules that make what we call water water are always there. Nitric acid may resemble water, but it is not water because the two substances have different molecular structures. Creating the equivalent of molecular theory for life, Cleland says, will require a larger sample size. She argues that, so far, we have only one example of what life is—the DNA and RNA-based life on Earth. Imagine trying to create a theory about mammals by observing only zebras. That’s the situation we find ourselves in when trying to identify what makes life life, Cleland concludes.
I disagree. Discovering examples of alien life on other planets would undoubtedly expand our understanding of how the things we call living organisms work and how they evolved in the first place, but such discoveries would probably not help us formulate a revolutionary new theory of life. Sixteenth century chemists could not pinpoint what distinguished water from other substances because they did not understand its fundamental nature: they did not know that every substance was made of a specific arrangement of molecules. In contrast, modern scientists know exactly what the creatures on our planet are made of—cells, proteins, DNA and RNA. What differentiates molecules of water, rocks, and silverware from cats, people and other living things is not "life," but complexity. Scientists already have sufficient knowledge to explain why what we have dubbed organisms can in general do things that most of what we call inanimate cannot—to explain how bacteria make new copies of themselves and quickly adapt to their environment, and why rocks do not—without proclaiming that life is this and non-life that and never the twain shall meet.
Recognizing life as a concept in no way robs what we call life of its splendor. It's not that there's no material difference between living things and the inanimate; rather, we will never find some clean dividing line between the two because the notion of life and non-life as distinct categories is just that—a notion, not a reality. Everything about living creatures that fascinated me as a boy are equally wondrous to me now, even with my new understanding of life. I think what truly unites the things we say are alive is not any property intrinsic to those things themselves; rather, it is our perception of them, our love of them and—frankly—our hubris and narcissism.
First, we announced that everything on Earth could be separated into two groups—the animate and inanimate—and it is no secret which one we think is superior. Then, not only did we place ourselves in the first group, we further insisted on measuring all other life forms on the planet against ourselves. The more similar something is to us—the more it appears to move, talk, feel, think—the more alive it is to us, even though the particular set of attributes that makes a human a human is clearly not the only way (or, in evolutionary terms, even the most successful way) to go about being a 'living thing.'
Truthfully, that which we call life is impossible without and inseparable from what we regard as inanimate. If we could somehow see the underlying reality of our planet—to comprehend its structure on every scale simultaneously, from the microscopic to the macroscopic—we would see the world in innumerable grains of sand, a giant quivering sphere of atoms. Just as one can mold thousands of practically identical grains of sand on a beach into castles, mermaids or whatever one can imagine, the innumerable atoms that make up everything on the planet continually congregate and disassemble themselves, creating a ceaselessly shifting kaleidoscope of matter. Some of those flocks of particles would be what we have named mountains, oceans and clouds; others trees, fish and birds. Some would be relatively inert; others would be changing at inconceivable speed in bafflingly complex ways. Some would be roller coasters and others cats.