life

Life is a characteristic of matter that distinguishes entities capable of extracting energy from their environment to sustain replication, growth, metabolism, and evolution. All known life is carbon-based, relies on liquid water as a solvent, and encodes genetic information in DNA or RNA molecules. The earliest evidence of life on Earth dates to approximately 3.5–4 billion years ago through abiogenesis — the natural emergence of living organisms from non-living chemistry. Modern life is classified into three domains: Bacteria, Archaea, and Eukarya, each representing a fundamental divergence in cellular architecture and biochemistry. Charles Darwin's theory of natural selection provides the unifying mechanism by which populations adapt over generations, driven by heritable variation and differential reproductive success. From single-celled microbes to complex multicellular organisms, life exhibits an extraordinary diversity of forms shaped by billions of years of evolutionary history.

Life is the condition that distinguishes organisms from inorganic matter, defined by the capacity to grow, reproduce, maintain homeostasis, and respond to stimuli through the extraction and transformation of energy from the environment. All known life is carbon-based, relies on liquid water as a solvent, and encodes hereditary information in nucleic acids — DNA and RNA — which direct the synthesis of proteins that carry out cellular functions. The origin of life from non-living chemistry, known as abiogenesis, remains one of science's deepest open questions, with leading hypotheses centering on hydrothermal vents and RNA-world scenarios approximately 3.5–4 billion years ago. Living organisms are classified into three domains — Bacteria, Archaea, and Eukarya — spanning an extraordinary range from single-celled extremophiles to complex multicellular organisms. Charles Darwin's theory of evolution by natural selection provides the unifying framework for understanding how populations adapt and diversify over time, driven by heritable variation and differential reproductive success. Biology, the scientific study of life, integrates disciplines from molecular genetics to ecology to investigate these processes across every scale of organization.

Life is the quality that distinguishes matter capable of self-sustaining biological processes — including growth, reproduction, metabolism, and response to stimuli — from inert matter. All known life on Earth is carbon-based, depends on liquid water, and uses DNA or RNA as its genetic material to encode hereditary information. Life originated approximately 3.5 to 4 billion years ago through a process called abiogenesis, and has since diversified into three domains: Bacteria, Archaea, and Eukarya. Through evolution by natural selection, as described by Charles Darwin, organisms adapt to their environments over successive generations. The scientific study of life in all its forms is known as biology.

You are, right now, a self-sustaining chemical system capable of Darwinian evolution [1]. That is NASA's working definition of life, and the awkward secret is that nobody — not the biologists, not the astrobiologists, not the philosophers — actually agrees on what life IS [2]. We are the thing we cannot define.

So what is life, and why can't scientists pin it down?

Life resists definition because Earth has only ever shown us one example of it, and you cannot generalize from a single data point [2].

The most quoted attempt comes from NASA: life is "a self-sustaining chemical system capable of Darwinian evolution" [1]. That sentence was hammered out by the NASA Exobiology Working Group at Carl Sagan's prompting, and it was never meant to be philosophically airtight — it was a tool for designing probes that sniff for life on other worlds [1]. It tells a Mars rover what to look for. It does not tell a philosopher what life is.

Other frameworks do exist. Some define life by metabolism — the business of eating and excreting to stay ordered. Others define it genetically, by the presence of heritable information. Chilean biologists Humberto Maturana and Francisco Varela proposed autopoiesis: a living system is one that continuously produces the very components that maintain its own boundary — a cell building the membrane that keeps it a cell [2]. Erwin Schrödinger framed life thermodynamically: living things locally reverse the universe's slide into disorder, importing order (food, sunlight) and exporting entropy (heat, waste) to stay organized against the second law's pressure [2]. None of these frameworks wins outright [2]. A virus metabolizes nothing but evolves furiously. A flame metabolizes beautifully but inherits nothing. Fire is not alive; a dormant seed is. The boundary is annoyingly fuzzy.

Britannica's working list is more practical and less elegant: living things show order, sensitivity, reproduction, adaptation, growth, regulation, homeostasis, and energy processing, and they are always built from one or more cells [10]. That is a checklist, not a definition — but when you are staring at something weird in a microscope, a checklist is what you use.

Imagine trying to define "music" if you had only ever heard one song. You would not know which features were essential (rhythm? pitch?) and which were just quirks of that one tune. That is biology's situation with life [2]. Every organism we have ever studied — bacterium, blue whale, baobab — descends from a single common ancestor, and the fingerprints of that shared descent are everywhere once you look. All terrestrial life uses the same triplet genetic code, in which three DNA or RNA letters spell out one amino acid, and the translation table is nearly identical from gut bacteria to giant squid [2][6]. All life builds its proteins from the same twenty amino acids, out of the hundreds that abiotic chemistry can produce. All life uses left-handed (L-) amino acids and right-handed (D-) sugars — a chirality choice that is arbitrary at the chemistry level but universal in biology. All life runs translation on ribosomes whose core RNA structure is conserved across Bacteria, Archaea, and Eukarya [6]. These are not independent strokes of convergent genius; they are inherited quirks from LUCA [3]. We have N = 1. Finding even one independent origin elsewhere — a microbe on Enceladus using a different code, or right-handed amino acids, or a non-ribosomal replicator — would immediately break the N=1 trap, because comparing two independent solutions would finally let biologists distinguish the essential features of life from the accidents of our particular lineage [2][8]. That is why astrobiology matters so much more than its budget suggests [8].

How did life start on Earth?

Life began here roughly four billion years ago, and the gap between "sterile rock" and "fully functioning cell" may have been shorter than the time since the dinosaurs died [3].

The LUCA paper of 2024 — that's the Last Universal Common Ancestor, the microbe every living thing on Earth descends from — placed LUCA at about 4.2 billion years ago, within a window of 4.09 to 4.33 Ga [3]. The shocking part is that LUCA was not simple. Its genome was already at least 2.5 megabases, encoding around 2,600 proteins, which is prokaryote-grade complexity [3]. LUCA was already eating, already dividing, already part of an ecosystem with neighbors [3]. So the real origin of life happened some unknown time before — and the gap may have been somewhere between 100 and 400 million years [3].

Where? Stanley Miller's 1953 experiment showed that the easy part is the chemistry: spark electricity through methane, ammonia, hydrogen and water vapor and within days you get glycine and a soup of other amino acids [5]. The hard part is getting them to organize.

One leading answer is alkaline hydrothermal vents on the young seafloor [4]. Their porous mineral chimneys acted as ready-made cellular compartments, and — this is the elegant bit — the natural pH gradient across their walls looks uncannily like the proton-motive force that every cell alive today uses to make ATP [4]. Life may have borrowed the vent's plumbing before it built its own. This is why the vents beat Darwin's old "warm little pond" guess: a tide pool gives you neither a sustained energy gradient nor a ready-made compartment, whereas a vent hands you both for free [4].

Miller and Urey's flask was a triumph of simplicity: seal gases, add sparks, wait [5]. But making amino acids is not making life. Amino acids do not spontaneously string themselves into functional proteins, and proteins alone cannot copy themselves — you need information and catalysis in the same molecule, or you need two molecules that already know how to cooperate. The modern origin-of-life puzzle sits in three stubborn gaps that Miller-Urey never touched [2][4]. First is the chicken-and-egg of replication and catalysis: the leading bridge is the RNA world hypothesis, in which early RNA molecules served double duty as genetic tape and as ribozymes — catalytic RNA that could copy itself and a few simple reactions, before DNA and proteins took over the specialist roles [2]. Second is compartmentalization: a naked replicator in open water gets diluted into irrelevance, so life needed a boundary. Candidates range from self-assembling lipid vesicles in a warm puddle to the porous mineral chambers offered by alkaline vents, which come pre-compartmentalized with semi-permeable walls [4]. Third, and most often ignored, is energy. Miller-Urey used lightning as a one-shot kick, but a living system needs a continuous, dissipating energy flow to stay organized against entropy. That is exactly the gap the hydrothermal-vent hypothesis is designed to fill: the pH gradient across a vent wall is a free, sustained, proton-motive energy source — the same kind of gradient every cell alive today uses to make ATP [4]. Miller-Urey gave us the bricks; the vents may have given us the scaffolding and the power supply.

What do all living things have in common?

Strip away the scales, feathers, bark and fur and every living thing on Earth runs on the same eight-item list: order, sensitivity, reproduction, adaptation, growth, regulation, homeostasis, and energy processing [10].

All of it happens inside cells [10]. The cell is biology's atom — the smallest thing that is unambiguously alive. A single bacterium is a cell; a redwood is trillions of them cooperating. Everything in between — metabolism, heredity, response to the environment — happens inside or between cells.

And every cell, no matter how weird, needs the same short shopping list: liquid water as a solvent, a usable energy source, and the bioessential elements carbon, hydrogen, nitrogen, oxygen, phosphorus and sulfur [8]. That austerity is why astrobiologists get excited about Europa and Enceladus — they tick the water box.

How is all of life related?

Every organism on Earth fits into one of three great domains: Bacteria, Archaea, and Eukarya [6].

That three-way split was not obvious. Until 1990, biology taught a prokaryote-versus-eukaryote binary. Then Carl Woese, Otto Kandler and Mark Wheelis compared 16S ribosomal RNA sequences across hundreds of microbes and discovered that what people had been lumping together as "bacteria" was actually two wildly different kingdoms [6]. Archaea look like bacteria under a microscope but are molecularly about as distant from them as we are [6]. We share a domain with mushrooms and mold; we do not share one with E. coli.

The other plot twist in the tree of life is that its branches occasionally fuse. Your mitochondria — the little power stations in every one of your cells — are descended from a free-living alphaproteobacterium that got swallowed and stayed [9]. Plant chloroplasts have the same origin story with a cyanobacterium [9]. Evolution is not only gradual mutation; it is also the occasional corporate merger [9].

Where else could life exist?

Life on Earth is tougher than your high-school textbook let on — it thrives from about -20°C up to 130°C, which dramatically widens the map of where else it might live [7].

Extremophiles — microbes that laugh at boiling hot springs, Dead Sea brine, nuclear-reactor coolant and the crushing pressure of the deep ocean — have been found in essentially every environment we have bothered to sample [7]. Theoretical limits stretch the envelope even further, to roughly -40°C and 150°C [7]. That matters because it turns previously "dead" worlds back into candidates: the subsurface oceans of Europa, Enceladus and possibly Titan all fall inside that envelope [7][8]. The old "habitable zone" of planets at Earth-like distances from their star has quietly expanded to include icy moons warmed from below [8]. Worth distinguishing, though: Europa and Enceladus hide liquid-water oceans beneath ice, which is the familiar Earth-style habitability mode, whereas Titan offers cryogenic lakes of liquid methane and ethane — a non-water solvent at around -180°C that would require a fundamentally different biochemistry, if life is possible there at all [7][8].

The catch is that finding life elsewhere will be harder than finding signatures that merely look like life. Abiotic chemistry can mimic biosignatures, so any detection has to survive the false-positive gauntlet [8].

Most things in the universe fall apart. Stars burn out, mountains erode, coffee cools. And then there is this one strange exception: a 3.5-billion-year-old chain of molecular copies that has never once been broken, passing a description of itself from parent to child across roughly a hundred trillion generations. Break the chain anywhere — even once — and you are not here reading this. That chain is what we mean by life, and it is weirder than the textbook sentence makes it sound.

What does life actually require?

NASA's working definition cuts surprisingly deep: life is "a self-sustaining chemical system capable of Darwinian evolution" [1]. Notice what the definition refuses to mention — no cells, no DNA, no water, no carbon. A salt crystal is self-sustaining too; what the crystal cannot do is evolve, because it has no heritable, imperfectly-copied description of itself [1].

The sharpest version of this insight came from a mathematician, not a biologist. In the 1940s, John von Neumann asked what the minimal architecture for self-replication would look like, and proved you need three parts: a universal constructor that builds things, a universal copier that duplicates descriptions, and a description tape that tells the constructor what to build [2]. The twist is that the tape is used twice — once interpreted (to build the machine) and once copied uninterpreted (to hand to the offspring) [2]. A decade before Watson and Crick, von Neumann had predicted exactly what DNA would have to do: translation and replication, from the same molecule, in two different modes [2].

Richard Dawkins took this one step further in 1976 and flipped the usual picture on its head [3]. The organism, he argued, is not the thing evolution cares about. The replicator is the gene; you are a temporary "survival machine" it built to carry copies of itself into the next round [3]. The word "selfish" is a metaphor for propagation, not for consciousness — genes do not want anything, but the ones that act as if they do are the ones still around [3].

How does life blur its own boundary?

Ask a biologist whether a virus is alive and watch them squirm. Giant viruses make it worse. Pandoraviruses carry genomes larger than some bacteria, and the majority of their genes — originally estimated at 93%, revised to around 70% with later annotation — have no known counterpart anywhere else on the tree of life [7]. They still cannot translate their own proteins — no ribosomes — so most researchers keep them out of the life club, but some have proposed an entire fourth domain to house them [7].

Go smaller and it gets stranger. Viroids are naked circles of RNA a few hundred bases long, with no protein coat at all; they infect plants by hijacking host enzymes to copy themselves [8]. Prions are even more minimal — infectious proteins, containing zero nucleic acid, that spread by forcing normal proteins to misfold into the prion shape [8]. Both replicate. Both evolve. Neither has a cell or a metabolism [8]. The inability to draw a clean line here is not a failure of classification; it is the universe telling us "alive" is a gradient.

Two Earth edge cases sharpen the point. The potato spindle tuber viroid (PSTVd) is a naked circle of ~359 RNA bases — no protein coat, no genes coding for enzymes, nothing but the tape itself — and it still infects, replicates, and mutates inside plant cells [8]. It is arguably the smallest autonomous replicator we know, and a cell-detector would walk right past it. Prion-like inheritance pushes the other way: in yeast, the Sup35 protein can flip into the [PSI+] conformation, which is then propagated to daughter cells as a heritable trait encoded in protein shape rather than nucleic acid [8]. That is non-pathological, stable, heritable information with no DNA or RNA involved.

The RNA-world implication is uncomfortable and liberating at once. The weird edge cases of today — naked RNA circles, self-templating proteins, genome-bloated viruses with no homologs [7] — probably rhyme with what most of the replicating matter on early Earth looked like before the first proper cell. If life is a gradient, its origin was almost certainly gradient too, and our definitions have to stretch to meet it [1].

How did complexity actually happen?

In 1995 John Maynard Smith and Eors Szathmary noticed that the history of life is not a smooth ramp of improvements but a short staircase of roughly eight huge reorganizations [4]. Free replicators corralled themselves into compartmentalized populations. Separate replicators bundled into chromosomes. RNA handed off translation to DNA-plus-protein. Prokaryotic cells swallowed other prokaryotes and became eukaryotes. Asexual clones adopted sexual reproduction. Single cells glued themselves together as multicellular bodies. Solitary animals became colonies, then eusocial societies. Primate calls became language [4].

Every transition has the same signature: a change in how information is stored, and a loss of independence for the smaller units [4]. Your mitochondria used to be free-living bacteria; now they cannot reproduce outside your cells. The chromosomes in your nucleus used to be separate replicators competing with each other; now they only ride together. Complexity, on this view, is the accumulated cost of tiny things agreeing to stop running their own lives.

Can we build life from scratch?

In 2016 a team at the J. Craig Venter Institute finished a twenty-year project called JCVI-syn3.0 — a bacterium with a chemically synthesized genome of just 531,560 base pairs and 473 genes, the smallest any freely-growing cell has ever had [5]. The embarrassing part: 149 of those 473 genes, essential for life, have no known biological function [5]. We built the minimal cell and still cannot read a third of its parts list.

Then, in 2023, another team pushed on a different axis: they synthesized and sequenced DNA containing up to twelve chemical letters, six matched pairs instead of the familiar two [6]. The work demonstrates that Watson-Crick-style pairing can accommodate a radically expanded alphabet far beyond ACGT [6]. Nothing about ACGT is sacred; it is one solution, not the solution. Somewhere between a minimal cell we do not understand and a 12-letter alphabet that works fine, the honest answer to "what is life made of?" is: we are still figuring out how much of biology is necessary and how much is just what Earth happened to do.

How has life rewritten its planet?

Around 2.4 billion years ago, the atmosphere changed color. Cyanobacteria had already invented oxygenic photosynthesis earlier in the Archean, but by that point enough of their waste product — O2 — had accumulated to poison the anaerobic world, rust the oceans into banded iron formations, and trigger what we now call the Great Oxidation Event [9]. Molecular phylogenies and fossil microbial mats both point back to cyanobacteria as the culprits; their descendants later got absorbed into plant cells as chloroplasts [9].

James Lovelock and Lynn Margulis looked at this and asked a heretical question in 1974: what if the biosphere is not just living on the planet but regulating it [10]? The Gaia hypothesis argues that life maintains the atmosphere, temperature, and ocean chemistry through negative-feedback loops — and points to the simultaneous presence of oxygen and methane, a combination wildly out of chemical equilibrium, as evidence that something is actively holding the system there [10]. You do not have to accept strong Gaia to see the shape of the claim: life is not a passenger on Earth. After 3.5 billion years, it is a geological force.

Margulis later reformulated Gaia in the language of autopoiesis — systems that produce and maintain their own components and their own boundary [10]. A cell is the minimal autopoietic unit; the biosphere, in her reading, is a scaled-up one, a self-producing, self-bounded entity whose "membrane" is the atmosphere and whose metabolism is the sum of biogeochemical cycles.

Run this back through von Neumann and the picture sharpens [2]. No single genome on Earth carries the description of the whole biosphere — there is no planetary tape. But the set of coexisting genomes, taken together, functions as a distributed description that the ecosystem interprets (by building organisms) and copies (by reproducing them) across generations [9]. Gaia, on this reading, is not mystical at all; it is what a planet-scale von Neumann machine looks like from the inside, three billion years into its run.