Evolution is not impossible. It has been directly observed in laboratories and in the wild, and the arguments claiming it cannot work rest on misunderstandings of biology, physics, and probability. These arguments tend to resurface in similar forms, so it’s worth examining each one against what the evidence actually shows.
The “Too Complex” Argument
The most popular version of this claim focuses on structures like the bacterial flagellum, a tiny spinning motor that bacteria use to swim. The argument goes: this motor has dozens of protein parts, and if you remove any one of them, it stops working. Therefore it could never have evolved piece by piece, because each intermediate step would be useless. This concept is called “irreducible complexity.”
The problem with this reasoning is that it assumes each part always did the same job. In reality, many flagellum components have clear relatives elsewhere in biology that serve completely different functions. The flagellum’s power source, an enzyme called FliI, is structurally similar to a widespread enzyme that cells use to produce energy (ATP synthase). The motor proteins that spin the flagellum are related to proteins in bacterial secretion systems that pump molecules across membranes. And the flagellum shares well-established structural similarities with the type III secretion system, a needle-like apparatus bacteria use to inject proteins into other cells. The two structures descended from a common ancestor.
A 2007 study in the Proceedings of the National Academy of Sciences traced the evolutionary history of the flagellum’s core genes and found something striking: they originated from one another through a series of gene duplications. A single precursor gene duplicated and diversified repeatedly, building the structure outward. Even more telling, the evolutionary relationships between these genes mirror the physical arrangement of their proteins in the finished flagellum. The parts closest to the base evolved first, and the outer components came later. The order in which the flagellum assembles itself today recapitulates its evolutionary history. Far from being an all-or-nothing structure, it was built stepwise from simpler systems that already had functions of their own.
The Thermodynamics Argument
Another common claim is that evolution violates the second law of thermodynamics. This law states that disorder (entropy) in a closed system always increases over time. Life, with its extraordinary organization, seems to move in the opposite direction. How can simple organisms become more complex if the universe trends toward disorder?
The key word is “closed.” The second law applies to systems with no energy flowing in or out. Earth is not a closed system. It receives an enormous, continuous flood of energy from the sun. That solar energy drives weather, ocean currents, photosynthesis, and the entire food web that sustains life. When you account for the entropy increase of sunlight being absorbed and re-radiated as heat, the total entropy of the Earth-sun system increases, exactly as thermodynamics predicts. Local decreases in entropy, like a plant building sugar from carbon dioxide or an embryo developing from a single cell, are perfectly consistent with the second law as long as they’re powered by an external energy source. And they are.
This isn’t a loophole. It’s how thermodynamics has always worked. Ice forming in your freezer reduces local entropy too, and nobody argues that refrigerators violate physics. They’re powered by electricity. Living systems are powered by sunlight, either directly through photosynthesis or indirectly through the food chain.
The “No New Information” Argument
A persistent claim holds that mutations can only destroy genetic information, never create it. If evolution requires organisms to gain new capabilities, and mutations only break things, then evolution cannot produce the diversity of life we see.
This misunderstands what actually happens in genomes. Gene duplication, where a stretch of DNA gets copied so that an organism carries two versions of the same gene, is among the most frequent types of mutation in all living things. It was first documented over a century ago. Once a gene is duplicated, one copy can continue doing its original job while the other is free to accumulate changes. Over time, that spare copy can take on an entirely new function. This is not theoretical. It has been observed repeatedly in laboratory experiments and confirmed across genomes from bacteria to humans.
Whole-genome duplication, where the entire set of chromosomes gets copied, is even more dramatic. It’s common in plants and has occurred in the ancestry of all vertebrates, including humans. These events instantly double the amount of raw material available for evolution to work with. On smaller scales, structural variations like duplications, insertions, and rearrangements occur at higher rates than single-letter changes in DNA. In E. coli experiments running about 500 generations, researchers observed that structural variations appeared sooner and more frequently than point mutations, though point mutations often provided higher fitness gains when they did arise. The interplay between these different types of mutation gives evolution multiple paths to adaptation simultaneously.
The Probability Argument
Some calculations claim to show that the odds of life arising by chance are astronomically small, sometimes compared to a tornado assembling a Boeing 747 from a junkyard. These calculations typically estimate the probability of a specific modern protein forming all at once, randomly, from scratch. The resulting number is indeed absurdly small, but the calculation doesn’t describe anything evolution actually does.
Evolution is not random assembly. It works through selection acting on variation over time. Each generation, organisms with traits better suited to their environment are more likely to survive and reproduce. Beneficial changes accumulate incrementally. No step requires assembling a complex protein from nothing. Each step only requires a small improvement on what already exists.
It’s also important to distinguish between the origin of life (abiogenesis) and evolution. Evolution explains how living things change over time once self-replicating systems exist. How those first self-replicating molecules arose is a separate question with its own active field of research. Conflating the two, and then declaring evolution “impossible” because we haven’t fully explained abiogenesis, is a category error.
What the Fossil Record Shows
Critics sometimes point to gaps in the fossil record as evidence against evolution. But the record is far richer than this argument suggests, and some of the most detailed sequences come from transitions that were once considered the most challenging to explain.
The transition from fish to land-dwelling animals is documented by a series of fossils showing gradual changes in skull shape, limb structure, and gill cover. Eusthenopteron, a lobe-finned fish from about 385 million years ago, had large skull bones and eyes positioned toward the front, with a full set of gill-covering bones. Panderichthys, which came later, had a flattened skull, eyes shifted toward the top of the head, and had already lost its dorsal and anal fins. Tiktaalik, discovered in 2004, bridges the gap further: its skull is flat like a land animal’s, its pectoral fins contain bones arranged like a wrist and fingers (while still retaining reduced fin rays), and it had lost the bony connection between its skull and shoulder girdle, freeing its head to move independently. Acanthostega had true limbs with digits but retained many fish-like features. Each fossil fills in more of the picture, and their features change in exactly the order predicted by evolutionary theory.
The Cambrian explosion, roughly 540 million years ago, is sometimes presented as evidence that complex life appeared “suddenly” without predecessors. But multicellular animals first appear in the fossil record almost 600 million years ago, tens of millions of years before the Cambrian. Known as the Ediacarans, these soft-bodied organisms included sponges and other creatures unlike anything alive today. The Cambrian wasn’t a moment of creation from nothing. It was an acceleration in diversification that built on lineages already in place.
Genetic Evidence for Common Ancestry
Perhaps the strongest evidence that evolution has occurred comes not from fossils but from DNA. Humans have 46 chromosomes. Chimpanzees, gorillas, and orangutans have 48. If humans and other great apes share a common ancestor, one explanation would be that two ancestral chromosomes fused together in the human lineage. And that is exactly what the genetic evidence shows.
Human chromosome 2 contains two head-to-head arrays of telomere repeats (the sequences normally found at chromosome tips) buried in its interior, right where a fusion point would be. It also has a deactivated second centromere, the structure chromosomes normally have only one of. The sequences flanking this fusion site are duplicated at the tips of other human chromosomes, exactly where you’d expect them if they were once at chromosome ends themselves. The fusion site sits at band 2q13-2q14.1, and the surrounding 600,000+ base pairs of DNA have been mapped in detail, confirming the fusion origin.
Endogenous retroviruses provide another line of evidence. These are remnants of ancient viral infections that inserted themselves into an ancestor’s DNA and were passed down to descendants. About 8% of the human genome is composed of these retroviral sequences. When researchers compare the locations of these viral insertions across species, closely related species share the same retroviruses at the same positions in their genomes. Old World monkeys possess a few copies of a retrovirus called HERV-K that is absent in New World monkeys, indicating the virus entered the lineage after those two groups diverged but before Old World monkeys and apes split apart. These shared insertions are essentially molecular timestamps. Because each viral insertion is an independent event at a specific location among billions of base pairs, the odds of the same virus landing in the same spot independently in two species are vanishingly small. Shared insertions at identical locations are direct evidence of shared ancestry.
Speciation Has Been Observed Directly
Evolution’s critics sometimes argue that while small changes within a species are possible, one species becoming two has never been witnessed. This is incorrect. Speciation through polyploidy (whole-genome duplication) in plants has been documented many times and is not controversial even among skeptics. But speciation has also been observed in animals. In northern Israel, at a site researchers call “Evolution Canyon” on Mount Carmel, scientists have documented incipient speciation occurring in spiny mice of the genus Acomys. Populations living on opposite slopes of the canyon, separated by only about 200 meters but facing dramatically different microclimates, are diverging genetically and reproductively. Similar patterns of speciation without geographic isolation have been described at the same site in bacteria, fungi, wild barley, fruit flies, and beetles. In the eastern Upper Galilee Mountains, blind subterranean mole rats also show signs of splitting into separate species despite living in the same geographic area.
In laboratory settings, bacterial evolution experiments running hundreds of generations have directly tracked the appearance, competition, and fixation of beneficial mutations in real time. These experiments show natural selection doing precisely what evolutionary theory predicts: favoring changes that improve survival and reproduction in a given environment, with those changes accumulating and compounding over generations.