Is evolution random? The honest answer is “partly, and the part that is random is not the part people usually argue about.” Evolution runs on at least three processes: mutation, natural selection and genetic drift. One of them is aimed, one is blind, and one is pure luck of the draw. Mixing them up is the source of most arguments about the subject, including the familiar claim that evolution is about as likely as a tornado assembling a jet from a junkyard.

This article takes the three processes one at a time, shows how they fit together, and separates what is well established from what biologists still argue about.

What does “random” mean in evolution?

In everyday speech, random means without pattern. In biology it has a narrower sense: random with respect to need. A mutation is random if its chance of occurring does not depend on whether it would help the organism that carries it.

That distinction matters. Some mutations are far more likely than others, because of how DNA is built and copied. Nobody claims all changes are equally probable. The claim is that the cell does not look ahead, notice that its descendants would benefit from a particular change, and then produce it.

The University of California Museum of Paleontology’s evolution site states the split plainly: the genetic variation that arises through mutation is random, while selection is not, because variants that improve survival and reproduction tend to spread more often than those that do not. Evolution, in other words, has a random step and a non-random step, and the outcome depends on both.

Where does the raw material come from?

Mutation. Copying errors, radiation, chemical damage and mobile genetic elements all change DNA sequences. Sexual reproduction then shuffles existing variants into new combinations each generation.

Darwin did not know any of this. He described variations that “seem to us in our ignorance to arise spontaneously,” a careful way of saying he could see them but could not explain them. The link to genetics came later: the work of Mendel was rediscovered in 1900, and Theodosius Dobzhansky’s 1937 book Genetics and the Origin of Species showed how mutation supplies the raw material that selection works on.

Most new mutations that have any effect are neutral or mildly harmful. Among humans, most of the ones with an effect are slightly deleterious. Helpful ones are rare. That ratio explains why selection is needed at all: without a filter, harmful changes would pile up.

Why is natural selection not random?

Because it is a sorting process with a consistent direction in a given environment. Individuals vary, some of that variation is inherited, and some variants leave more offspring than others. Over generations the successful variants become common.

The textbook case is the peppered moth in Britain. During the Industrial Revolution, soot darkened tree bark, and dark moths gained an advantage. In industrial Manchester nearly all the moths became dark within about fifty years. After the Clean Air Act of 1956 cut pollution, pale moths made a comeback. Researchers later linked the colour switch to the insertion of a transposable element in a gene called cortex.

Look at what happened there. The mutation that produced dark moths was not summoned by the soot. It arose by chance. What was not chance was the result: once soot was in the picture, the dark form was favoured, and it spread in a predictable direction.

Antibiotic resistance follows the same logic. Resistant bacteria exist before the drug is used, or arise by mutation during treatment. The drug then removes competitors, and resistant lines take over. MRSA, the resistant Staphylococcus aureus strain, is the standard example.

What is genetic drift?

Drift is change in the frequency of a gene variant, called an allele, caused by chance alone. It has no direction and does not care whether a variant helps.

A marble analogy shows the idea. Take a jar holding ten red and ten blue marbles. To form the next generation, draw twenty marbles from the jar with replacement. You will almost never get exactly ten of each colour. Repeat the draw from the new jar, and the ratio keeps wandering until one colour disappears.

Wikipedia’s article on genetic drift gives a bacterial version. A colony in which alleles A and B are equally common is cut down to four survivors. The chance that exactly two carry each allele is 6 in 16, or 37.5 percent. The remaining 62.5 percent of outcomes shift the balance. A shift is the likelier result than staying put.

Drift is therefore not a form of selection and is not a force pushing toward anything. It is sampling error, repeated through the generations.

When does chance beat selection?

When populations are small. Drift’s effects shrink as populations grow, because large samples track their source more faithfully. As a rule of thumb from population genetics, drift outweighs selection when the advantage a variant provides is smaller than about one divided by the effective population size. The effective size is a statistical figure that can be much lower than a head count, because of inbreeding and uneven breeding success.

A few cases show drift at work:

  • Greater prairie chickens in Illinois fell from roughly 100 million birds around 1900 to about 50 in the 1990s, and the population lost most of its genetic diversity in the process.
  • The 19th-century hunting of northern elephant seals left them with less genetic variation than their southern relatives, even after numbers recovered.
  • Achromatopsia, a form of total colour blindness, is unusually common on Pingelap atoll in Micronesia, a legacy of a small founding population.
  • In 1744, two Amish migrants to Pennsylvania both carried a recessive allele for Ellis-van Creveld syndrome. In their descendants, a rare condition became common.

The last two are founder effects, a special kind of bottleneck in which a small group splits off and starts a new population. The founders’ alleles are a random sample of the parent group, and the new population inherits whatever quirks that sample contained.

One more point, which surprises people: even a helpful new mutation is at risk. When it first appears, it exists in a single individual. That carrier may simply fail to reproduce, and the advantage disappears with it.

How much of DNA change is chance? The neutral theory

This is where real scientific disagreement begins. In 1968 the geneticist Motoo Kimura proposed that most of the molecular differences between species, and much of the variation within them, come from drift acting on neutral mutations, changes that neither help nor harm. King and Jukes reached a similar view independently in 1969.

The theory did not deny natural selection. Kimura accepted that selection shapes the traits animals actually display. His claim was about the DNA sequence itself, where many changes, such as those at the third position of a codon, often leave a protein unchanged. Evidence cited in support includes the observation that proteins with a minor functional role tend to change faster than those with an essential one.

Tomoko Ohta later refined the picture with the nearly neutral theory. Slightly harmful mutations behave like neutral ones in small populations, where drift overpowers weak selection, but get removed in large ones.

Where does this stand now? Many predictions of the neutral framework have held up, including the steady accumulation of substitutions over time that underlies the molecular clock. Other findings cut against a strict version. Genetic diversity varies far less across species than population sizes would suggest, a puzzle called the paradox of variation, and widespread evidence of positive selection shows that adaptive change leaves its mark on genomes. Biologists still debate the relative weight of drift and selection at the molecular level. That debate concerns proportions, not whether either one exists.

Does that mean evolution is just chance?

No, though chance is certainly involved. A better description is that evolution is a process in which chance supplies options and non-random filtering decides among them.

An analogy, offered with the usual caution that analogies leak: a sculptor who knocks off pieces of marble by tapping at random would produce rubble. But if every tap is followed by a judge who keeps only the results that look more like a horse, the sequence climbs toward horses without any single tap being aimed. The judge in evolution is the environment, and it does not need a plan. It only needs differences in how many offspring each variant leaves.

Chance still leaves its mark in places. Which particular mutations arise, and in what order, shapes which paths are available. Two populations facing the same problem can arrive at different solutions. Drift can also fix a slightly harmful variant in a small group, which is why not every feature of every organism is an adaptation.

None of this makes evolution unpredictable in every respect. Under strong, steady pressure, such as an antibiotic, the direction is easy to forecast. What is hard to forecast is the specific genetic route.

Is “survival of the fittest” the right summary?

It is a poor one. Herbert Spencer coined the phrase, and Darwin adopted it in the fifth edition of On the Origin of Species in 1869 as an alternative to natural selection. Modern biologists mostly avoid it for two reasons.

First, it invites circularity: who survives? The fittest. Who is the fittest? Those who survive. Second, fitness in the technical sense means reproductive success, not strength, speed or even a long life. An animal that lives to a great age without leaving descendants has a fitness of zero.

The phrase also hides drift. Some individuals survive and breed through luck, not merit, and in a small population that luck can reshape the whole group. Selection removes the less fit variants, but it does not guarantee that the survivors are well adapted in any absolute way. Darwin himself called selection “the main but not exclusive means” of modification, and later biologists have taken the qualification seriously.

Common questions

Is evolution just luck?

Not entirely. Mutation and drift involve chance, but selection systematically favours variants that leave more offspring in a given environment. The combination produces outcomes that look designed without being planned.

Do organisms mutate because they need to?

The standard view is that mutations arise without regard to whether they would help. Resistant bacteria are not created by the antibiotic; the drug filters existing variants. Researchers have studied whether stress can raise mutation rates in general, but that is a different claim from mutations being steered toward a useful target.

Is genetic drift a kind of natural selection?

No. Selection depends on heritable differences that affect survival or reproduction. Drift happens regardless of those differences, through chance sampling of which individuals pass on their genes.

Does selection make organisms perfect?

No. It favours whatever works better than the competing variants present, in the current environment, using the variation available. It cannot design from scratch, and it works with history: new features are built from older parts.

Which matters more, drift or selection?

It depends on population size and on how strong the selection is. In small, isolated populations drift can dominate. In large ones selection usually does, at least for traits that affect fitness. At the level of raw DNA sequence the balance is an open research question.

Can chance really produce complex structures?

Chance alone cannot, and no biologist argues that it can. Complex structures arise through many steps of variation followed by selection, each step keeping what works. The key point is that the filtering is cumulative, so progress is retained rather than rebuilt from nothing each time.

What to remember

Evolution is neither pure chance nor pure design. Mutation generates variation with no foresight, selection sorts it with a direction set by the environment, and drift adds statistical noise that matters most when populations are small.

Treat “random” as a statement about the origin of variation, not about the outcome. The established core is that all three processes occur and leave measurable traces. The unsettled part is how much of the change in DNA is neutral drift versus adaptive selection, a question that has been debated since Kimura’s 1968 proposal and has not been closed.

Further reading

A widely read book on how non-random selection acting on random variation can produce complex adaptations. See it on Amazon (affiliate link)

As an Amazon Associate I earn from qualifying purchases. A commission may be earned if you buy through this link, at no extra cost to you.

Sources