The first thing to say is that Natural Selection is not the same thing as evolution. The word "evolution" simply means "gradual change". It means the same thing in biology too, except that the timescale is usually measured in thousands or millions of years—way beyond the timescales we are used to in everyday life. Biological evolution simply means that the genome of a lineage is constantly changing—that every descendant is a tiny bit different from its ancestors (and from each other), and that these differences pile up, bit by tiny bit, as the generations go by. After a few million generations of this "descent with modification" as Darwin called it, the accumulated differences are often so great that descendants no longer resemble their distant ancestors at all.

Darwin wasn’t the first person to think about evolution—the Greeks beat him to that more than two thousand years ago—but he was the first to explain how evolution works. His Big Idea—Natural Selection—is how evolution works.^[1]

Unfortunately, descriptions of Natural Selection are often rather dry and technical. If you've ever taken a peek into a biology text, you probably found something like this:

Where there is phenotypic [physical, structural] variation in a population, and this variation is heritable, and there is competition for resources, then by virtue of their inherited qualities some individuals will fare better than others and leave more viable descendents (on average). Accordingly, the overall genetic composition of the population will gradually change, as genes that increase fitness are preserved and become more frequent, while those that decrease fitness are weeded out. This gradual genetic change is what constitutes visible evolution...

All well and good, but this doesn't really convey the idea of evolution—the simplicity of the principle, or the wild complexity of the actuality—not least because it assumes you're already familiar with the language of the field, if not the concepts themselves. The philosopher Daniel Dennett (author of the brilliant Darwin's Dangerous Idea) even offers us this uncharacteristically gristly mouthful:

[natural selection is] the process in which reproducing entities must compete for finite resources and thereby engage in a tournament of blind trial and error from which improvements automatically emerge... (Daniel Dennett, "Show me the Science", NYT Aug 28, 2005) 

And he's right, bless him, but if you're unfamiliar with Nash equilibriums, game theory, and iterated prisoners' dilemmas, this kind of explanation doesn't really help very much.

I think we need something simpler and more straightforward.

Let's start with Darwin, and what he noticed.

Actually, before that, I should briefly mention the beginning of life on Earth.

No one knows how life got started on Earth. Darwin didn't, and we still don't. Darwin was famously tight-lipped on the matter; the only thing he said (as far as we know) was in a private letter to a friend in which he speculated on what might have happened in "some warm little pond" millions of years ago. Darwin tended to be extremely cautious in his claims, not least because he was acutely aware that little was known about geology or biology at the time. He didn't know, for instance, how reproduction worked—he didn't even know how old the Earth is—but he knew that if the world turned out to be young (as some of his colleagues at the time believed—perhaps only a few thousand years old), his theory would crumble.

What we know, in the 21st century, is that our planet has been orbiting the Sun for some four billion years or so, and that life has been thriving here for at least 3.5 billion years—although for most of that interminable time the only "life" to be found was slime and bacteria and algae. We'll come to the evidence for this shortly; for now, I simply want to stress just how long 3.5 billion years is. Three thousand five hundred million years. It's a very long time indeed. The geologists call this "deep time". It's an absolute eternity. I mention this because from the human perspective evolution happens so slowly we can't see it directly. Evolution needs plenty of time. If only a human lifespan was a million years, we'd have no trouble seeing it at all.

OK, back to what Darwin noticed:

If we look at the biological world—by which I mean all the animals, all the plants, all the fungi, and all the microbes (in other words anything with DNA in it)—there are a few features that are so ubiquitous, so glaringly obvious, that we usually overlook them. One of the most remarkable, and what Darwin set out to explain, is the sheer number of species we share the planet with. Why are there so many? Where did they come from? And what of the enormous number of species we see in the fossil record that are no longer here?

When Carl Linnaeus started classifying plants and animals back in the eighteenth century, he thought there might be as many as ten thousand species on Earth. Current estimates put the number somewhere in the tens of millions.^[2] One of the reasons for such a huge discrepancy in these estimates is because we now know that there are many more small things (which are hard to count) than there are big things (which are easier to count). Many of the big things were named early, while many small things went unseen and uncounted (because much of the deep jungle and ocean were unexplored), or because they were too small to be visible (until scientists had good enough microscopes, in the 19th Century).

We now know that there are about 30,000 species of fish, 10,000 species of bird, 4,500 species of mammal, and 7,000 kinds of reptile. But there are an astonishing 10,000,000 (ten million) kinds of insect (estimates vary a bit on this)—of which at least 250,000 are distinct species of beetle! Hundreds of new species are still described every year, and this is just in the animal kingdom! There are also hundreds of thousands of species of flowering plants, a similar number of fungi, and goodness knows how many millions of species of microbe (Ref and Ref). If variety is the spice of life, then it looks as if evolution is the mother of all Vindaloos.

OK, so there are lots of species alive now, and there were lots more species—gazillions of them—alive in the past (and the past, as we shall see, goes back a long way—disappearing off into "deep time"). Darwin reasoned that if everything alive today has descended from ancestors, who in turn descended from ancestors, but these older ancestors differed in structure from their contemporary descendants, then there must have been some change in the meantime. He ruled out sudden, drastic change in a lineage, in favour of tiny, generation by generation changes. This generation-by-generation change is evolution.

Darwin didn't originally set out to explain this rather racy idea, but he was familiar with it from an early age. His grandfather, Erasmus, had written a book on the subject 15 years before Charles was born. The solution finally came to Charles a year after he got back from his famous voyage on the Beagle (1831-1836), though he kept it to himself for nearly 20 years.

Darwin showed that for biological evolution to happen, four ingredients were necessary: 

1 There Must Be Variation

From humming birds to jellyfish to the mites that live in the ears of moths, the amount of difference between species is both obvious and breathtaking. But this isn’t quite the kind of "variation" that Darwin had in mind when he was wondering how evolution might work. He noted that there are differences within species too—even between closely related individuals—and it was these much more subtle but heritable differences (i.e. differences that can be passed on to offspring) that interested him, and are what biologists mean when they talk about “variation”.

Variation can be very hard to see sometimes. Indeed, organisms often resemble each other to such an extent that it’s almost impossible to tell individuals apart. I’m thinking penguins, but the same can be said of poppies, porcupines, parrots, and even people too. Nevertheless, no two living things are exactly alike—not even if they start out with the same genes, as “identical” twins do. There are always some differences, even if they are incredibly slight. As we shall see, sometimes a tiny difference can make all the difference.

2 There Must Be Heredity

Living things reproduce, and when they do, they usually pass on at least some of their differences to their descendents, which is why offspring tend to resemble their parents more closely than they resemble other individuals. But along with some of their differences, individuals usually pass on 100% of their similarity too: penguin eggs hatch into penguins, not parrots (and certainly not porcupines!). This is because every individual in a population possesses the same basic set of genes—the species’ genome—that is reliably inherited by every individual. Darwin could see that there is heredity, obviously, just as there is variation; what he didn't know was why, or even how, it worked. Darwin was a worrier and this kept him awake at night.

3 There Must be Population Constraints

As Thomas Malthus pointed out in his famous Essay on the Principle of Population (published in 1798, eleven years before Darwin was born), a population can only get so large. To be precise, while the size of a population is affected by various things such as climate and disease, it is ultimately constrained by the available food supply. Only so much grass (and thus so many rabbits) can grow on a hillside; only so many fish (and thus so many otters) can live in a stream. There are limits to what an environment can sustain, and life, ever-hungry, spends most of its time bumping up against these limits. Malthus showed that while a population tends to increase geometrically (2, 4, 8, 16, 32, etc), the food supply at the bottom of the food chain can only increase arithmetically (1, 2, 3, 4, etc), and even then to a maximum. As a result, a population tends to hover near the maximum size allowed by its food supply, and there are consequently many, many, many more seeds, shoots, spores, runners, eggs and offspring produced each season than can possibly survive. Most end up as lunch.

4 There is Competition for Resources

Which seeds, shoots, spores, runners, eggs and offspring end up as lunch? The answer, as Darwin reasoned, are the unlucky, the unwell, and the unfit.^[3] Those that remain constitute a population of survivors, some of which may simply be lucky, but most of which are fitter than those who perish (see note^[3]). The fittest individuals are those best equipped to cope with what Darwin called the "conditions of life"—what we now think of as "the environment". Fit individuals are faster, fatter, thinner—whatever it takes—to survive, reproduce, and if necessary, raise young. Fit individuals have better immune systems, lay more eggs, are better parents—whatever it takes—to become a great-grandparent, an ancestor. Most crucially, fit individuals are able to out-compete their peers—thanks to some inherited traits which they then in turn bequeath (at least in part) to some of their offspring. Fit individuals leave more descendants than unfit individuals.

When resources are scarce or conditions harsh—and they very often are—some individuals are bound to do better than others. Darwin called this “the struggle for existence”—a phrase his friend Herbert Spencer famously turned into “the survival of the fittest.” (If you haven't yet, please take a look at note^[3])

If only Darwin had known about genes. They explain why every individual is unique, and they explain how inheritance works. He might have slept better. And if we—the public—understood genetics a little better today, perhaps evolution would cease to be such a controversial idea.

Putting it together

The Four Elements of Natural Selection:

1. Difference—no two individuals are perfectly alike, because no two individuals' genomes are perfectly alike (variation). If all individuals were exactly the same, it wouldn't matter which ones lived or died. No variation means nothing for natural selection to act on.
2. Differences tend to get passed on to offspring (heredity). Individuals must be able to inherit their parents' traits—good and bad. No heredity, and again, there can be no natural selection.
3. Population constraints (Malthus—never enough food to go round—most perish). Getting the hang of this yet? If no one ever goes hungry or gets crowded out (or eaten!), not only would the world be amazingly crowded, but again, there could be no natural selection.
4. Competition—among members of a species—for resources (food, territory, mates, etc.). This isn't really competition in the normal human sense—plants and animals aren't "trying to win"—they are simply doing what they do. But because there is genetic variation (and thus some differences between individuals in size, shape, abilities, etc), there are winners and losers in terms of survival and reproduction. Perhaps it would be better to say that "winning" in this competition simply means leaving more descendents than your peers. If seed A sprouts sooner than seed B right next to it, A might be able to hog the available sunlight while B withers. Or if one bird-of-paradise, let's call him Bruce, happens to have brighter, longer feathers than his neighbour, Nigel, then lucky Bruce might well win more copulating opportunities. (On the other hand, being so bright he might get eaten by a sharp-eyed predator before he gets any mating opportunities—now that's what I call "losing"!).

Sorry to keep repeating myself, but here's the thing: some individuals are a tad better able to cope in their world than others—thanks to their inherited differences. These individuals are thus a tad more likely to successfully reproduce than others, and thus over the generations their "good" genes crop up more frequently in the population as a whole. These are the "fittest" individuals; they leave more descendants than their less fit peers (whose genes will become slightly less frequent with every generation). (More on fitness here)

And what happens when every member of a population has these "good" genes? Is every member equally fit? No; these genes become part of the species' genome and the competition shifts to other genes, other traits. The environment is constantly changing also—predators evolve too, as do parasites and diseases, and there will be climate change, chemical change, etc—so evolution cannot stop, even if it could. Every individual is at least a tiny bit different from all the others, remember?

So, call me a rebel, but I think there's a fifth element too: The environment—the thing that does the "selecting". Just as plant and animal breeders "select" the qualities they want when breeding new varieties, so the environment blindly wipes out some and leaves others. Of course "the environment" (which includes all sorts of things like predators and viruses and the social rules of your pack) doesn't actually sit down and "choose" any particular traits—it is a mindless and contingent process—but out of necessity we use the language of desire and will, even though this can be confusing. We talk as if the environment was an agent that could choose. It can't.

To repeat then, the environment is more than temperature, weather, acidity and so on (what Darwin called the "conditions of life"). It is also (and primarily in many cases) an organism's peers, predators, prey, and so on. Indeed, for many species, including us, the environment is mainly political; it consists of relationships and contracts and what others think and feel. Indeed, for most life a central part of "the environment" is the frequency of particular genes in the world. Chloroplasts..

There is also the crucial matter of time—mind-bogglingly vast spans of time. Darwin knew that unless the world was hundreds of millions of years old (at least), his theory was doomed. (More on geologic, or "Deep" Time here)

Evolution has been happening ever since living things started reproducing, passing on their traits, and creating unique new individuals in the process—some of which leave descendants who are in turn slightly different, and so on. If we accept this, then all we need are the mechanisms of inheritance. The physiology of sex, and genetics.

How good is the evidence? As we shall see, it's pretty solid—among the most reliable in science in fact. Click here for more on this (coming soon).

Remember those little differences between individuals? The ones that tend to get passed on? The tiny differences that make all the difference? Well, they also accumulate: thousands of genetic changes—random additions, deletions, and mutations—pile up over the centuries until sooner or later some of them result in some behaviour or structure that is either beneficial or (usually) disastrous to the individuals bearing those genes. Most often this results in some new feature that is decidedly not helpful to its owner, and the gene dies along with the body it helped build. On the other hand, the changes might not be accumulating fast enough to cope with the changes in the environment, in which case extinction also beckons (you can see why most species go extinct). If the lineage survives, we can either see or infer these changes from the organism's phenotype, either now or in the fossil record.

Any differences in phenotype that get passed on, can only be passed on via an organism's genes. Thus we get to "genotype leads to phenotype"—what is sometimes known as the "central dogma" of biology. This is the idea that dolphin genes don't beget eagles. (But see also epigenetics)

Most individuals don’t make it (to the “goal” of successful reproduction). A combination of bad luck and constitutional disadvantage mean that very often only a fortunate few (those lucky enough to possess certain genes—"the fittest") from each generation live long enough to breed.

The “Economy of Life”, as Darwin called it, is for most organisms a miserable business in which the break-even point is staying alive, and profit comes with reproduction. Needless to say, in this brutal economy there are no rules or regulations or small claims courts.

The more individuals there are in a species, the more ways there are of being slightly different.

Who knows what the future will hold—and therefore what traits will be helpful or harmful for your descendants? Our best weather forecasters can't see more than a day or two into the future, so what hope for a frog, or a flea, or a fern? A thick fur coat in unexpectedly hot weather may doom its owner to heat exhaustion, but when it snows that coat may save its life. What's a critter to do? Nature's solution is to hedge all bets: some variation in the genes for coat thickness means that at least some individuals will survive—hot or cold—so long as the environment doesn't change too much, or too fast. And as almost every facet of an individual's structure (its phenotype) is variable to some degree (because every individual's genome is different), some individuals are inevitably favoured, while others are punished, by the ever-changing environment.

Evolution is about life gradually changing, but this isn't just any old change. And it certainly isn't "random"—at least not in the sense meant by critics of evolutionary theory. The randomness happens at the molecular level—with the handful of point mutations that occur among the billions and billions of faithful copies made during development and meiosis. These mutations are a bit like the random spelling and punctuation errors—typos—that creep in during the repeated copying of a lengthy book. As with genes, most errors are harmless and easily correctable, and only very occasionally would the sense of a passage be changed.

Having said that, some phenotypic changes aren't inherited in the way that antibodies or the ability to fly are. The colour of feathers or fur, for example, which are often vital in securing a mate, sometimes depend on the presence of certain chemicals in the diet. These epigenetic factors—phenotypic changes that are caused by environmental variability—play an important role in the story of evolution, but this is not to say that penguins could ever lay eggs that hatch into parrots, no matter what epigenetic factors might apply.

Darwin: "...This review, however, and Harvey's letter have convinced me that I must be a very bad explainer. [They don't] really understand what I mean by Natural Selection. I am inclined to give up the attempt as hopeless. Those who do not understand, it seems, cannot be made to understand." (emphasis added Ref)

When Darwin set off on his famous voyage, his job was to keep the Captain company, and his official expertise, such as it was, was in the brand new science of geology. His interest in rocks was fortunate however, as one of the most important observations Darwin made on the voyage was that the Earth is an ancient body whose surface is slowly and constantly changing.

Notes

1 Strictly speaking, natural selection isn’t the only process that leads to change in a lineage. But as we shall see, it is responsible for any change that is adaptive—i.e. change that confers reproductive advantage over individuals that do not possess the relevant gene.

2 And this is just the number alive today—perhaps a billion species have come and gone since life on Earth began. This has lead some experts, such as Craig Venter, to wonder whether we should abandon the task of counting species altogether and concentrate instead on the number of genes in the world—many of which are common to many species.

3 A quick note on the word fitness: in bio-speak fitness is meant in the sense that a key “fits” a lock. Being “fit” in the sense of “fighting fit” or “in good physical condition” really has nothing to do with it. Only keys with the right shape will fit a particular lock, and likewise only organisms with the right qualities can flourish in a particular environment—no matter how “fit and strong” they may be. Thus, a fit individual is one that is better adapted (to survive and reproduce) to its world than most of its peers; an unfit one is less well adapted. Fitness is also a statistical measure. The Biggest, baddest, meanest predator ever might catch cold and die one day, or fall off a cliff, or get run over. In other words, having a slight edge is no guarantee at all of reproductive success. Nevertheless, if, in some particular circumstance, possessing gene X is more likely to keep you alive than not possessing gene X, then among a whole population we would expect to see most survivors possessing that gene.

More on fitness here.