The Magic of Reality

by Richard Dawkins

  Just as some species are more similar than others and are placed in the same family, so there are also families of languages. Spanish, Italian, Portuguese, French and many European languages and dialects such as Romansch, Galician, Occitan and Catalan are all pretty similar to each other; together they’re called ‘Romance’ languages. The name actually comes from their common origin in Latin, the language of Rome, not from any association with romance, but let’s use an expression of love as our example. Depending on which country you are in, you might declare your feelings in one of the following ways: ‘Ti amo’, ‘Amote’, ‘T’aimi’ or ‘Je t’aime’. In Latin it would be ‘Te amo’ – exactly like modern Spanish.

  To swear your love to someone in Kenya, Tanzania or Uganda you could say, in Swahili, ‘Nakupenda’. A bit further south, in Mozambique, Zambia, or Malawi where I was brought up, you might say, in the Chinyanja language, ‘Ndimakukonda’. In other so-called Bantu languages in southern Africa you might say ‘Ndinokuda’, ‘Ndiyakuthanda’ or, to a Zulu, ‘Ngiyakuthanda’. This Bantu family of languages is quite distinct from the Romance family of languages, and both are distinct from the Germanic family which includes Dutch, German and the Scandinavian languages. See how we use the word ‘family’ for languages, just as we do for species (the cat family, the dog family) and also, of course, for our own families (the Jones family, the Robinson family, the Dawkins family).

  It isn’t hard to work out how families of related languages arise over the centuries. Listen to the way you and your friends speak to each other, and compare it to the way your grandparents speak. Their speech is only slightly different and you can easily understand them, but they are only two generations away. Now imagine talking, not to your grandparents but to your 25-greats-grandparents. If you happen to be English, that might take you back to the late fourteenth century – the lifetime of the poet Geoffrey Chaucer, who wrote descriptions like this:

  He was a lord ful fat and in good poynt;

  His eyen stepe, and rollynge in his heed,

  That stemed as a forneys of a leed;

  His bootes souple, his hors in greet estaat.

  Now certeinly he was a fair prelaat;

  He was nat pale as a forpyned goost.

  A fat swan loved he best of any roost.

  His palfrey was as broun as is a berye.

  Well, it is recognizably English, isn’t it? But I bet you’d have a hard time understanding it if you heard it spoken. And if it was any more different you’d probably consider it a separate language, as different as Spanish is from Italian.

  So, the language in any one place changes century by century. We could say it ‘drifts’ into something different. Now add the fact that people speaking the same language in different places don’t often have the opportunity to hear each other (or at least they didn’t before telephones and radios were invented); and the fact that language drifts in different directions in different places. This applies to the way it is spoken as well as to the words themselves: think how different English sounds in a Scottish, Welsh, Geordie, Cornish, Australian or American accent. And Scottish people can easily distinguish an Edinburgh accent from a Glasgow accent or a Hebridean accent. Over time, both the way the language is spoken and the words used become characteristic of a region; when two ways of speaking a language have drifted sufficiently far apart, we call them different ‘dialects’.

  After enough centuries of drift, different regional dialects eventually become so different that people in one region can no longer understand people in another. At this point we call them separate languages. That is what happened when German and Dutch drifted, in separate directions, from a now extinct ancestral language. It is what happened when French, Italian, Spanish and Portuguese independently drifted away from Latin in separate parts of Europe. You can draw a family tree of languages, with ‘cousins’ like French, Portuguese and Italian on neighbouring ‘branches’ and ancestors like Latin further down the tree – just as Darwin did with species.

  Like languages, species change over time and over distance. Before we look at why this happens, we need to see how they do it. For species, the equivalent of words is DNA – the genetic information every living thing carries inside it that determines how it is made, as we saw in Chapter 2. When individuals reproduce sexually, they mix their DNA. And when members of one local population migrate into another local population and introduce their genes into it by mating with individuals of the population they have just joined, we call this ‘gene flow’.

  The equivalent of, say, Italian and French drifting apart is that the DNA of two separated populations of a species becomes less and less alike over time. Their DNA becomes less and less able to work together to make babies. Horses and donkeys can mate with each other, but horse DNA has drifted so far from donkey DNA that the two can no longer understand each other. Or rather, they can mix well enough – the two ‘DNA dialects’ can understand each other well enough – to make a living creature, a mule, but not well enough to make one that can reproduce itself: mules, as we saw earlier, are sterile.

  An important difference between species and languages is that languages can pick up ‘loan words’ from other languages. Long after it developed as a separate language from Romance, Germanic and Celtic sources, for example, English picked up ‘shampoo’ from Hindi, ‘iceberg’ from Norwegian, ‘bungalow’ from Bengali and ‘anorak’ from Inuit. Animal species, by contrast, never (or almost never) exchange DNA ever again, once they have drifted far enough apart to have stopped breeding together. Bacteria are another story: they do exchange genes, but there isn’t enough space in this book to go into that. In the rest of this chapter, assume that we are talking about animals.

  Islands and isolation: the power of separation

  So the DNA of species, like the words of languages, drifts apart when separated. Why might this happen? What might start the separation? An obvious possibility is the sea. Populations on separate islands don’t meet each other – not often, anyway – so their two sets of genes have the opportunity to drift away from one another. This makes islands extremely important in the origins of new species. But we can think of an island as more than just a piece of land surrounded by water. To a frog, an oasis is an ‘island’ where it can live, surrounded by a desert where it can’t. To a fish, a lake is an island. Islands matter, both for species and for languages, because the population of an island is cut off from contact with other populations (preventing gene flow in the case of species, just as it prevents language drift) and so is free to begin to evolve in its own direction.

  The next important point is that the population of an island need not be totally isolated for ever: genes can occasionally cross the barrier surrounding it, whether this be water or uninhabitable land.

  On 4 October 1995 a mat of logs and uprooted trees was blown onto a beach on the Caribbean island of Anguilla. On the mat were 15 green iguanas, alive after what must have been a perilous journey from another island, probably Guadeloupe, 160 miles away. Two hurricanes, called Luis and Marilyn, had roared through the Caribbean during the previous month, uprooting trees and flinging them into the sea. It seems that one of these hurricanes must have torn down the trees in which the iguanas were climbing (they love sitting up in trees, as I have seen in Panama) and blown them out to sea. Eventually reaching Anguilla, the iguanas crawled off their unorthodox means of transport onto the beach and began a new life, feeding and reproducing and passing on their DNA, on a brand new island home.

  We know this happened because the iguanas were seen arriving on Anguilla by local fishermen. Centuries earlier, although nobody was there to witness it, something similar is almost certainly what brought the iguanas’ ancestors to Guadeloupe in the first place. And something like the same story almost certainly accounts for the presence of iguanas on the Galapagos islands, which is where we turn for the next step in our story.

  The Galapagos islands are historically important because they probably inspired Charles Darwin
’s first thoughts on evolution when, as a member of the expedition on HMS Beagle, he visited them in 1835. They are a collection of volcanic islands in the Pacific Ocean near the equator, about 600 miles west of South America. They are all young (just a few million years old), formed by volcanoes punching up from the bottom of the sea. This means that all the species of animals and plants on the islands must have arrived from elsewhere – presumably the mainland of South America – and recently, by evolutionary standards. Once arrived, species could make the shorter crossings from island to island, sufficiently often to reach all the islands (maybe once or twice every century or so) but sufficiently seldom that they were able to evolve separately – ‘drift apart’ as we have been saying in this chapter – during the intervals between the rare crossings.

  Nobody knows when the first iguanas arrived in the Galapagos. They probably rafted across from the mainland just like the ones that arrived in Anguilla in 1995. Nowadays the nearest island to the mainland is San Cristobal (Darwin knew it by the English name of Chatham), but millions of years ago there were other islands too, which have now sunk beneath the sea. The iguanas could have arrived first on one of the now sunken islands, and then crossed to other islands, including those still above water today.

  Once there, they had the opportunity to flourish in a new place, just like the ones that arrived in Anguilla in 1995. The first iguanas on Galapagos would have evolved to become different from their cousins on the mainland, partly by just ‘drifting’ (like languages) and partly because natural selection would have favoured new survival skills: a relatively barren volcanic island is a very different place from the South American mainland.

  The distances between the different islands are much smaller than the distance from any of them to the mainland. So accidental sea crossings between islands would be relatively common: perhaps once per century rather than once per millennium. And iguanas would have started turning up on most or all of the islands eventually. Island-hoppings would have been rare enough to allow some evolutionary drifting apart on the different islands, between ‘contaminations’ of the genes by subsequent island-hoppings: rare enough to allow the different groups of iguanas to evolve so much that when they eventually met again they could no longer breed together. The result is that there are now three distinct species of land iguana on Galapagos, which are no longer capable of cross-breeding. Conolophus pallidus is found only on the island of Santa Fe. Conolophus subcristatus lives on several islands including Fernandina, Isabela and Santa Cruz (each island population possibly on its way to becoming a separate species). Conolophus marthae is confined to the northernmost of the chain of five volcanoes on the big island of Isabela.

  That raises another interesting point, by the way. You remember we said that a lake or an oasis could count as an island, even though neither consists of land surrounded by water? Well, the same goes for each of the five volcanoes on Isabela. Each volcano in the chain is surrounded by a zone of rich vegetation, which is a kind of oasis, separated from the next volcano by a desert. Most of the Galapagos islands have only a single large volcano, but Isabela has five. If the sea level rises (perhaps because of global warming) Isabela could become five islands separated by sea. As it is, you can think of each volcano as a kind of island within an island. That’s how it would seem to an animal like a land iguana (or a giant tortoise), which needs to feed on the vegetation found only on the slopes around the volcanoes.

  Any kind of isolation by a geographical barrier which can be crossed sometimes but not too often leads to evolutionary branching. (Actually, it doesn’t have to be a geographical barrier. There are other possibilities, especially in insects, but for simplicity’s sake I won’t go into them here.) And once the divided populations have drifted far enough apart that they can no longer breed together, the geographical barrier is no longer necessary. The two species can go their separate evolutionary ways without contaminating each other’s DNA ever again. It is mainly separations of this kind that were originally responsible for all the new species that have ever arisen on this planet: even, as we shall see, the original separation of the ancestors of, say, snails from the ancestors of all vertebrates including us.

  At some point in the history of iguanas on Galapagos, a branching occurred which was to lead to a very peculiar new species. On one of the islands – we don’t know which – a local population of land iguanas completely changed their way of life. Instead of eating land plants on the slopes of volcanoes, they went to the shore and took to feeding on seaweed. Natural selection then favoured those individuals that became skilled swimmers, until nowadays their descendants habitually dive to graze on underwater seaweeds. They are called marine iguanas and, unlike land iguanas, they are found nowhere but Galapagos.

  They have lots of strange features that equip them for life in the sea and this makes them really rather different from the land iguanas of Galapagos and everywhere else in the world. They have certainly evolved from land iguanas, but they are not especially close cousins of today’s land iguanas of Galapagos, so it is possible that they evolved from an earlier, now extinct genus, which colonized the islands from the mainland long before the present Conolophus. There are different races of marine iguanas, but not different species, on the different islands. One day these different island races will probably be found to have drifted apart far enough to be called different species of the marine iguana genus.

  It’s a similar story for giant tortoises, for lava lizards, for the strange flightless cormorants, for mockingbirds, for finches, and for many other animals and plants of Galapagos. And the same kind of thing happens all over the world. Galapagos is just a particularly clear example. Islands (including lakes, oases and mountains) manufacture new species. A river can do the same thing. If it is difficult for an animal to cross a river, the genes in populations on either side of the river can drift apart, just as one language can drift to become two dialects, which can later drift to become two languages. Mountain ranges can play the same role of separation. So can just plain distance. Mice in Spain may be connected by a chain of interbreeding mice all across the Asian continent to China. But it takes so long for a gene to travel from mouse to mouse across that vast distance that they might as well be on separate islands. And mouse evolution in Spain and China might drift in different directions.

  The three species of Galapagos land iguana have had only a few thousand years to drift apart in their evolution. After enough hundreds of millions of years have passed, the descendants of a single ancestral species can be as different as, say, a cockroach is from a crocodile. In fact it is literally true that once upon a time there was a great-great-great- (lots of greats) grandparent of cockroaches (and lots of other animals including snails and crabs) which was also the grand ancestor (let’s use the word ‘grancestor’) of crocodiles (not to mention all the other vertebrates). But you’d have to go back a very very long way, maybe more than a billion years, before you found a grancestor as grand and ancient as that. That is much too long ago for us even to begin to guess what the original barrier was that separated them in the first place. Whatever it was, it must have been in the sea, because in those far-off days no animals lived on land. Maybe the grancestor species could only live on coral reefs, and two populations found themselves on a pair of coral reefs separated by inhospitable deep water.

  As we saw in the previous chapter, you’d only have to go back six million years to find the most recent shared grancestor of all humans and chimpanzees. That’s recent enough for us to guess at a possible geographical barrier that might have occasioned the original split. It’s been suggested that it was the Great Rift Valley in Africa, with humans evolving on the east side and chimpanzees on the west. Later, the chimp ancestral line split into common chimpanzees and pygmy chimpanzees or bonobos: it’s been suggested that the barrier in that case was the Congo river. As we saw in the previous chapter, the shared grancestor of all surviving mammals lived about 185 million years ago. Since then, its descendant
s have branched and branched and branched again, producing all the thousands of species of mammals we see today, including 231 species of carnivores (dogs, cats, weasels, bears etc.), 2,000 species of rodents, 88 species of whales and dolphins, 196 species of cloven-hoofed animals (cows, antelopes, pigs, deer, sheep), 16 species in the horse family (horses, zebras, tapirs and rhinos), 87 rabbits and hares, 977 species of bats, 68 species of kangaroos, 18 species of apes (including humans), and lots and lots of species that have gone extinct along the way (including quite a few extinct humans, known only from fossils).

  Stirring, selection and survival

  I want to round off the chapter by telling the story again in slightly different language. I’ve already briefly mentioned gene flow; scientists also talk of something called the gene pool, and I now want to spell out more fully what that means. Of course there can’t literally be a pool of genes. The word ‘pool’ suggests a liquid, in which genes might be stirred around. But genes are found only in the cells of living bodies. So what does it mean to talk of a gene pool?

  In every generation, sexual reproduction sees to it that genes are shuffled. You were born with the shuffled genes of your father and your mother, which means the shuffled genes of your four grandparents. The same applies to every individual in the population over the long, long reach of evolutionary time: thousands of years, tens of thousands, hundreds of thousands of years. During that time, this process of sexual shuffling sees to it that the genes within the whole population are so thoroughly shuffled, indeed stirred, that it makes sense to talk of a great, swirling pool of genes: the ‘gene pool’.

  You remember our definition of a species as a group of animals or plants that can breed with each other? Now you can see why this definition matters. If two animals are members of the same species in the same population, that means their genes are being stirred about in the same gene pool. If two animals are members of different species they cannot be members of the same gene pool because their DNA cannot mix in sexual reproduction, even if they live in the same country and meet each other frequently. If populations of the same species are geographically separated, their gene pools have the opportunity to drift apart – so far apart, eventually, that if they happen to meet again they can no longer breed together. Now that their gene pools have moved beyond mixing they have become different species and can go on moving further apart for millions of years to the point where they might become as different from one another as humans are from cockroaches.