Return to essays top page
It seemed to me, looking at websites which tried to explain how we know that evolution is real, that many of them were mired in technicalities and confused by needless detail. It really isn't necessary to go into minutiae, any more than it is necessary to examine the precise design tolerances of a spark plug in order to have a fair idea how the internal combustion engine works, and see that it works, right there in front of us. Hence, this essay.
Evolution is heritable changes taking place in the genetic makeup of a population of living things, resulting in a progressive alteration in the average form of individuals in that population.
The commonly-heard term "survival of the fittest" is a misnomer - "superior reproductive success of the fittest" is closer to the mark. What usually drives evolution is that a change in the environment, or a social pressure such as mate-selection, means that a particular variant turns out to do better than its relatives, has more offspring through whom to spread its genes, and so comes to dominate the population. In some cases heritable changes in the average genetic makeup of a group can also happen by chance. Heritable changes caused by deliberate human intervention (e.g. when developing a specific breed of domestic animal) are usually excluded from the definition but the process is essentially the same as natural evolution by preferential mate-selection.
The underlying mechanism and tool of evolution is genetic variation, both the re-arrangement of pre-existing genes, and the acquisition of new ones. If you want to understand in detail how this works, see the appendix at the bottom of this page.
We know that differing colour genes - for example - can be advantageous in different situations, and that selection pressure in the environment can change the average colour, and colour genes, of a population, because since the Victorian era we have watched it happen to the pepper moth, where the predominant colour of the species darkens or lightens according to how sooty or pale its environment is. This is probably because moths which match the prevailing background are less likely to be spotted and eaten by predators, and certainly because changes in the environment exert some sort of pressure on the moth population. We can watch bacteria change and mutate and develop heritable resistance to drugs from year to year and even day to day: the increasing resistance that bacteria are showing to antibiotics is evolution in action, happening in a few decades.
We can watch rat populations becoming genetically resistant to poisons, as the most susceptible individuals die and the least susceptible spread their genes through the group. Pure deductive reasoning ought to have told us that this would happen, even before we saw it happening. So long as susceptibility or resistance to a lethal poison has a genetic component, it's obvious that if a population is exposed to that poison, the animals with genes which make them susceptible will tend to die early, thus producing few or no offspring, and the ones with genes for resistance to poisoning will be less affected, have a healthy number of offspring and spread their genes, resulting in an average change in the genetic makeup of the population. It can't not happen. And if evolution can and indeed must happen when driven by an extreme selection pressure such as exposure to poison, it can and must happen due to exposure to milder selection pressures (only more slowly).
Creationists like to cite the fact that Darwin himself called evolution a "theory" and said that some new theory might come along and overturn his idea. But this was because Mendel's work on genes and the mechanism of inheritance had not yet been published when On the Origin of Species came out, so Darwin had only a very vague idea of how the changes wrought by selection pressure might be passed along. Once you know that most of an organisms's characteristics are controlled by its genes, and that those genes are handed on from generation to generation, evolution due to selection pressure becomes self-evident. Any change in the environment will tend to favour individuals whose genes best fit them for that changed environment, who will then go on to have greater reproductive success and an increased genetic presence in subsequent generations. The changes in form which we see in the fossil record are only the outward sign of changes in the average genetic make-up of populations, due to selection.
We also see man-made selection pressure being used to generate new breeds of domestic animal, by preferentially breeding from individuals with desired characteristics. Although we do not normally consider this man-made selection to be evolution, it shows beyond a doubt that selective breeding can change the genetic makeup of a population, and it is self-evident that different climates and environments will favour different characteristics - longer or shorter hair, thicker or thinner build etc.. Why, since we know beyond a shadow of a doubt that man-made selection pressure can change the appearance of a population, would natural selection pressure not be able to do so?
We also know beyond any doubt that mutations causing harmless, potentially useful changes such as altered colour do occur, because we can see them happening. In some individuals, mutations occur in the developing embryo, after conception, resulting in what's called a mosaic organism, patched with two different cell lines. We can see the cell lines differing by only a single gene, so we know they both spring from the same original, single fertilised egg-cell (otherwise they would have many more differences): the anomalous gene is a new mutation. We can see these mutations often having harmless, potentially useful effects, because they often do things like causing odd patches of differently-coloured or textured fur. We also know, and can see happening, that if one of the mutated blocks of cells includes a gonad (ovary or testicle, in mammals), then the affected organism is able to produce gametes (egg-cells or sperm) which carry the new mutation on to its offspring.
Since we can see mutations which affect things like colour taking place before our eyes, and we can see selection pressure (both natural and artificial) influencing e.g. the average colour of a species, as well as resistance to poisons etc., it necessarily follows that selection pressure can cause a new mutation to spread through a species, changing it relative to what it was before the mutation arose.
I have seen a Creationist argue that the rate of mutation is too low to account for the genetic changes which need to have taken place to account for all the species we see, but this is a misunderstanding. There is believed to be a slow, fairly steady accumulation of minor, neutral mutations in proteins which many species share, such as haemoglobin, due to transcription errors and background radiation, but mutations, specially non-neutral ones, can happen much faster than that for a variety of reasons. Some genetic locii (such as the one associated with achondroplasia) are fracture-points which mutate at the drop of a hat, mutation rates can be increased drastically by prolonged sunbathing and by exposure to mutagenic toxins, and genetic material can be carried from one species to another by viruses.
Even the rate of the kind of minor mutations which are used in the "molecular clock" varies with the generation-time of the species; its population size (with small populations leading to changes in the genetic makeup op a population due to random sampling effects, called "genetic drift"); its metabolic rate; changes in overall selection pressure; and evolutionary changes in the function of the protein being studied, which may cause formerly neutral changes to become significant and be specifically selected for or against. Nor need it take very many mutations to create radical changes of form: a single change of gene can have a big effect, especially if it interferes with embryological development.
This of course is the answer to the question of how evolution can produce forms which are wildly different. It could be said that at the most basic, metabolic level it doesn't and hasn't - we still share some remarkable chemical and genetic similarities with bacteria. But quite small changes in gene can result in major reorganisations of physical form - especially if the change occurs in one of the "Hox" genes which control the order of embryological development.
We can also see from first principles that changes to the average genetic makeup of a small population due to random genetic drift, or evolution-by-chance, must happen.
To understand how genetic drift works, you have to bear in mind that in any given population some individuals - often, most individuals - will die without ever having offspring, thus removing their genes from the gene pool.
Let's imagine that we have a sack containing an even number of coloured balls, half of them red and half white. These represent individual organisms which carry one of two alternative colour genes.
Now assume that half of those organisms go on to breed, and those that breed have two offspring like themselves, maintaining a constant population. Without looking, take out half the balls from the bag and throw them away (they are the ones that didn't breed), then duplicate each of the balls you still have left in the sack. Repeat several times.
If you started off with five thousand red balls and five thousand white balls, the odds are good that even after several "generations", you will still have roughly five thousand red balls and five thousand white balls, give or take a few. At the other extreme, if you started with only one red and one white ball and discarded one, then inevitably all your following generations will be all red or all white.
In between, though, you have a situation where the smaller the initial population, the more likely you are to lose one of the colours (genes) due to a random sampling error. If you have two red and two white balls, for example, and you keep two and discard two at random, you might end up with one red and one white ball, carrying both "genes" through to the next "generation" - but each time you choose from a mixed foursome there's a 50:50 chance that you will discard both the red balls, or both the white, and the next generation will be all one colour. One of your colour genes has been lost from the population, purely due to random chance.
In real terms that means that if you have, for example, a population of only a couple of hundred of some rare finch, it may happen just by chance that all the individual finches who have a particular gene get eaten by hawks before they can reproduce, even though the gene was not as such a getting-eaten-by-hawks gene; and so the gene is lost and the average genetic makeup of the population changes.
By observing man-made selection-pressure in action, we can see how natural selection can produce the tangle of different species we see, both in the present day and in the fossil record. If humans set out to create a new breed of sheep or dog or hen from an existing one, it rarely if ever happens that the whole of the pre-existing breed is converted into a new one. Rather, a group of individuals from the existing breed are separated out from the main population - in this case, by human intervention, but in nature by things like a river changing its course or a rock-fall blocking a pass - and selected for a specific characteristic.
Again, in this case the selection is being applied by human intervention, but except that it is more intense and directed, this is not different from a climate-change at one end of an animal's range favouring individuals with thicker coats and stockier build; or the arrival of a new, sharp-sighted predator eliminating individuals with poor camouflage from those areas which the predator has reached.
So, a new breed arises, but in nearly all cases the old breed it came from is still there. Sometimes both the old and new breeds continue side-by-side, each suited for slightly different situations and uses. Sometimes other new breeds are created from the old one - or from the new one itself. Sometimes the old breed eventually dies out, leaving the newer breed or breeds to hold the stage. Sometimes a new form turns out to be a failed experiment and it's the new breed which dies out, leaving the old one still thriving.
To understand how one group can be lost naturally, while another related group survives, you only have to think about family dynamics. It often happens that one branch of a family dies out because even though Great Aunt Mabel had five children, by chance or by war none of her children had children. Extinction is the same sort of thing, happening on a larger scale.
Of course, when we humans create a new breed by artificial selection, the new group can still interbreed with the old one, at least in theory. In order to create a new species - according to the old definition of a species as a group whose members can interbreed and produce fertile offspring with each other but not with individuals outside the group - the new breed has to cease to be able to reproduce successfully with the one it came from, or with other new breeds which may arise from the same parent breed. Once this happens, the new species will diverge from the old one much more noticeably, because there is no longer any possibility of the two groups mingling their genes and averaging out.
We can see speciation too happening before our eyes. Even in domestic breeds, it is usually only possible for very large and very small breeds to cross-mate with human help. If we were to put a group of Rottweilers and a group of Chihuahuas on an island together, with no dogs of intermediate size and no stepladders, they would probably function as if they were separate species. Unable to mingle their genes, over time the two groups would become more and more different. This is a situation we see in nature, where differences of habitat and mate-selection have caused leopards and lions to grow apart and appear so unalike that we consider them to be separate species, and yet under certain circumstances they are still able to interbreed and produce fertile offspring (fertile female offspring, at any rate).
The ship rat, Rattus rattus, seems to be particularly prone to chromosome mutations (where the genes remain the same but the order in which they are strung together to form chromosomes changes, or where chunks of chromosomes are duplicated or missing or have become separated and free-floating) which result in populations which cannot breed with each other, or can do so only very inefficiently, even though they look like the same species, share all the same genes and probably were the same species very recently (see A new population of Rattus rattus with 38 chromosomes in north-western India and Studies on rodent chromosomes. VI. Co-existence of Rattus rattus with 38 and 42 chromosomes in south-western India). All these groups will have the same amount of genetic material, but it's been assembled slightly differently.
That these groups with their different chromosome counts seldom if ever interbreed successfully is shown by the fact that samples of the populations found almost no individuals with an intermediate chromosome count. We know that chromosome mutations can just happen - we see them just happening all the time - so we can reasonably assume that that's how it came about that some populations of ship rats in India have 42 chromosomes while others have 38 (two slightly different ways of having 38 chromosomes, too, resulting in two discrete populations with 38 plus another with 42).
By re-assorting their chromosomes in this way, these rats are basically forming new species, new circumscribed breeding groups, pretty much on the spot, before developing different characteristics. We can also see species forming the other way, starting by developing differences between populations, and then gradually becoming unable to interbreed.
There is a natural occurrence called a cline, where the range of a species stretches across a progressively changing habitat - from a valley floor to halfway up a mountain, for example - and the individuals in that species also change to fit the changing habitat. For example, hedgehogs tend to be larger the colder the area they live in, because the colder it is in winter the more bulk they need to get through hibernation, and the larger you are the smaller your surface area to volume ratio is, so the less heat you lose.
Sometimes, a group of organisms living along a cline form what is called a ring species. What this means is that there is a progressive change in form from a point at one end of the cline to a point at the other, and the two extreme ends cannot breed with each other.
If we call the population at one end of the cline A and the one at the other end Z, there are a whole range of intermediate forms in between which interbreed with the groups either side of them. Group A breeds with B; B breeds with both A and C; C breeds with B and D and so on: but A and Z cannot breed with each other even if they are neighbours (if they are, it's called a closed ring species). We can think of a ring species as being one species because there is this continuum of cross-breeding, yet if some natural disaster were to eliminate the groups in the middle to leave only A and Z, there would be no doubt that they were now two quite separate species.
Fossilisation is a rare event. We could predict that this was so, even without looking at the fossil record at all, simply by observing what usually happens to dead organisms: and what usually happens is that they get eaten, or moulder away, or dissolve into the soil around them. Fossilisation requires that a dead organism lands up somewhere dry enough to mummify it, preferably including burying it in sand; or wet enough to bury it in deep anaerobic mud before it has a chance to decay; or in some specialised preservative medium such as ice, tar, amber or peat. Finding fossils is an even rarer event than making them: the great majority of those that ever formed, must still be in the ground.
Also, some organisms and bits of organisms fossilise better than others. Teeth are often preserved; soft tissue, hardly ever. The bigger and denser the bits, the better they last. Something like an oak tree or a rhinoceros has a much better chance of leaving at least some fossil traces than a jellyfish or even a mouse, yet small and soft and often boneless creatures make up most of life.
Looking at the fossil record, therefore, is like trying to assess the wildlife of modern Africa based on what casual tourists happen to spot. Anything which was both large and common will probably be represented but most of the small, the soft and the rare will never be seen. You would probably get wildebeest, zebras, hyaenas, maybe also elephants, rhinos, hippos, crocodiles and at least a few of the big cats and the primates: but your chances of getting an okapi or an aardvark or a Cape hunting dog would be nearly zero.
It's not surprising that the fossil record of the human family tree is a little patchy, because early humans and their ancestors seem never to have been very numerous, and to have been changing quite fast, and not to have lived in conditions which were very favourable to fossilisation: hence, the odds of any given type of human ancestor being fossilised were not great. It's not surprising if some are missing from the record, although the analysis of Australopithecus sediba in 2011 seems to have filled in the human line back to a genuine "missing link" between humans and the other Great Apes. Whether future discoveries bear this out or not, to prove that successive evolution of new forms occurs it is enough to be able to prove it even once: and we have an exceptionally complete record of the evolution of the horse family.
Dating fossils is more complex than is usually assumed. Although we believe that most fossils are ancient, there are some areas today where freak mineral deposits can cause dead organisms to fossilise in about twenty years. An animal burying the remains of its prey can push the carcase of an organism down into a lower layer of mud than the one it lived in; and a deep footprint may leave traces in an older level than the one the maker walked on, even if the top layers of the footprint have later been eroded away. A burrowing animal can drag fossils up to a higher layer, or itself die down a hole and fossilise far below the layer it lived on. We used to be sure we knew that heavy sauropods swam in inland lakes, pushing off from the bottom with the tips of their toes, because there exist fossilised sauropod tracks which show the toe-tips only: now we think that the animal put its whole foot down in mud or sand on dry land, with the toes digging deeper than the sole, the whole print dried out and hardened and then the wind or other forces scoured away the shallower parts of the track, leaving only the toe-prints, pushed down into a slightly deeper layer than the one the animal actually walked across.
However, a high proportion of fossils are in sedimentary rock - that is, they are the remains of organisms which died in water, drifted to the bottom and then were slowly covered by silt and the shells of dead micro-organisms. We have a fair idea how long it takes a layer of sediment to form because we can watch it happening today, and we can see that the organisms fossilised into the layers are themselves layered, showing a succession of different ecosystems in which new creatures appear and some of the old ones fade out. Sometimes we can see mass die-offs in which a high proportion of pre-existing species disappear and then are slowly replaced by new forms with similar functions.
Although there's a degree of guesswork involved in dating anything beyond the limits of tree-ring counts, we know that there have been many thousands of successive ecosystems, each probably taking many thousands of years to develop and die away. We know that there are individual organisms much older than Mediaeval churchmen thought the whole world was, which gives us some idea of how long ecosystems persist. Bristlecone pines exist which can be dated to five thousand years ago; the underlying root system of the still-living Old Tjikko spruce in Norway has been carbon-dated at nine and a half thousand years; there are creosote-bush rings estimated at well over eleven and a half thousand years. There are clone-colonies of Quaking Aspen in the USA - that is, clumps of trees which all have the same genetic makeup and are really just large shoots arising from an ancient, shared root system - whose massive size and very slow rate of growth suggests that they may be close to a million years old. And these are the children of one or at most two of the many successive ecosystems we can see in the fossil record.
So, although there's a lot of guesswork involved in working out exactly how old the fossil record is, unless you are assuming the presence of a Creator who deliberately made an artificial world to mislead us, in the manner of the world-builders in Terry Pratchett's book Strata, it's safe to say that the earliest layers of the fossil record are very, very ancient indeed. We can see that humans came along towards the top of the fossil record, certainly many millions of years along and, so far as palaeontologists and geologists have been able to estimate, actually some billions of years after life first began.
We can also see that the layers which make up the fossil record are more complex than the layers of creation described in Genesis. It is not the case that all plants appeared, then all reptiles and so on: rather, there were early plants, and then there were reptiles, but new plants continued to develop after the arrival of the reptiles, and so on up the list. New bacteria continue to happen probably every day or even every hour, despite the fact that bacteria are nearly the oldest living things there are: the old things are still living and growing and generating brand-new varieties, right alongside the new things.
It used to be thought that there were anomalies in the fossil record because, for example, most of the evidence suggested that birds had evolved from therapod dinosaurs, yet birds possessed a particular finger which therapods had lost before birds evolved. It was thought that a feature once lost could not be regained, other than by re-evolving it.
The discovery of Hox (homoeotic complex) genes, however, showed how such reversals can happen. More information on Hox genes can be found in the appendix below, but basically they control the order in which groups of other genes are expressed during embryological devlopment. A single Hox mutation can multiply the segments of an already segmented organism, making it longer, or cause a fly to grow legs where its antennae should be. Simple changes in Hox genes can switch other genes off altogether, with the result that the code for a lost feature may be carried invisibly in an organism's genes for thousands of generations, waiting for another simple mutation to switch it on again.
At its most extreme an induced mutation in a gene called Hoxa2 can cause a mouse embryo to develop with extra, reptilian-style bones in its lower jaw, even though so far as we know mammals diverged from reptiles, and lost these extra jaw-bones, hundreds of millions of years ago. If mice, or even mammals as a whole, had been planned in detail by a Creator without having to evolve from reptiles, it's difficult to see why they would have been designed with a hidden, optional reptilian feature which was not expressed in nature and therefore did the mouse no good, and yet could be uncovered by man - unless that Creator set out deliberately to deceive, or was using Object-Oriented Programming and made a bit of a fist of it.
That does not mean that the Genesis story has no value, or that science and religion are enemies. The best evidence that the Judaeo-Christian version of the Creation myth is any more valid than anyone else's Creation myth is the fact that if you squint at it a bit, it does actually fit the fossil record remarkably well, so long as we assume that "day" is being used in the sense in which one says "the day of the covered wagon" - a period of time of arbitrary length during which something was especially prevalent and noteworthy.
To deny evolution is, from a religious point of view, perverse. I don't think even the most determined Creationist has ever suggested that the different layers of the fossil record represent single days, so if the successive layers of groups of life-forms described in Genesis are referring to the successive emergences of fauna in the fossil record, which they match pretty well, then the "days" in Genesis have to be geological eras. If the creation event described in Genesis is assumed not to refer to the layers of the fossil record, it cuts it away from the only evidence it has that it's anything more than a teaching-myth based around an early attempt at science-fiction.
Genesis describes a long phase during which we see a half-formed planet coalescing out of the dust of the void, the sun igniting, the planet rotating and land masses rising. It describes the advent of plants on dry land - the mention of grasses and fruit trees is an anomaly because they were late arrivals, but the idea of plants of some kind making it onto land early on is good, and you can read it as "ground cover and large plants with edible seed-pods" - and lights in the firmament, which might correspond with the period in which sight first evolved, or with a clearing of the skies.
Then you have the advent of reptiles (this should be amphibians followed by reptiles, but the original author of the text might have lumped them together, as many people do) emerging from the waters, and flying things in the air - the fossil record shows the arrival on land of amphibians and arthropods, including large flying insects, followed by early reptiles. The inclusion of whales at the same point as early amphibians/reptiles in Genesis is an anomaly but "large swimming vertebrates" would certainly make sense: we know that huge sharks existed around that time, and most of our ancestors didn't know a whale from a whale shark. Then we have the development of land-dwelling reptiles and beasts (which do belong together because mammal-like reptiles came along very early), eventually followed by the arrival of humans.
[I'm also toying with the idea that the night-and-morning pairs which Genesis places between each successive "day" of Creation coincide with major extinctions and renewals. You could certainly say that the one after reptiles climbed out of the ocean and before the development of quadruped beasts was the Great Dying towards the end of the Permian period, and the one after the quadrupeds and before man was the extinction event at the end of the Cretaceous.]
That the account in Genesis corresponds so well to later discoveries about the formation of the Earth and the emergence of life from the water argues either divine inspiration, clairvoyance or very astute early fossil hunters.
There is a minor problem with the "astute fossil hunters" option, in that we know the Ancient Greeks collected fossils, but they misunderstood them: they thought they were the bones of monsters from their mythology, and that the surprisingly fresh-looking Protoceratops skeletons which litter the Gobi were the remains of recently dead gryphons. If their fairly near neighbours the Jews had worked out that fossils were the remains of ancient ecosystems and that they came in stratified layers of different life-forms, you'd think the Greeks, with their interest in the subject, would have heard about it and discussed it in their own writings. That they do not seem to have done so could be said to argue in favour of divine inspiration or clairvoyance - and some people think clairvoyance is a form of divine inspiration anyway.
Although the fact that Genesis matches the fossil record so well at least raises the possibility of genuine spirtual knowledge, we have no actual evidence that the world has any sort of Creator, as opposed to just happening (this is a separate issue from whether there is a god, or gods, since the emergence of a powerful spiritual entity could be one of the things which just happened). The idea that because our world suits the development of life so perfectly it must have been engineered to do so is a mathematical fallacy. We are living on a world which is suitable for life because if it wasn't suitable for life, we wouldn't be here to be thinking about it: and we have no idea how many failed worlds there are, or how common life is.
If it were the case that there were no other planets even vaguely suited to life, and no other life, and this world of ours was unlike all other worlds, and unlike them in specific ways which made it both unique and able to generate life, then the hypothesis that the world must have been created to support life would have merit. But in fact, even in our own solar system we can see other worlds (Europa, Titan, Ganymede, Callisto, Mars....) which potentially might harbour life, at least at a very simple level, and even here on Earth, we can see simple organisms thriving in extreme, highly toxic (to humans) conditions such as around underwater volcanoes, suggesting that life is probably a common phenomenon which can occur in a wide range of conditions. Some scientists even believe that life may have evolved on Earth more than once and that there may still be "weird life" microbes existing in odd corners.
As for Intelligent Design at the species level, you have to ask what kind of Intelligent Designer would design mammals in general, and humans specifically, with back-to-front retinas and testicles which don't work at body-temperature (unless, for some reason, you're a hedgehog) and have to be hung outside in a bag to keep them cool, and in the specific case of humans sinuses which don't drain properly unless you go down on all fours, and a semi-redundant appendix which serves little purpose except occasionally to kill the owner.
In 2009, great advances were made in the creation of life in the laboratory. A team in Manchester managed to get RNA bases to form from raw chemicals exposed to conditions similar to those which geologists believed prevailed on the early Earth; and a team in Maryland claimed to have generated a semi-artificial new species of bacterium by inserting a whole man-made genotype into a natural bacterial coat. Given the vast length of time available for RNA strands to form and break and re-form in the waters of the early Earth, there seems no doubt that the most primitive initial life-forms could have come about by chance, like a coherent word emerging from Alphabetti Spaghetti.
On the other hand, absence of evidence is not necessarily evidence of absence. If a forest fire rages and destroys the evidence of its own origin, it's not necessary to assume that it was started by a human being - there are plenty of other ways a fire can start - but we also can't rule out that it was started by a human. It might well have been. Although it is not necessary to assume the existence of a Creator in order to explain everything we have seen so far, that isn't proof that there isn't one.
We can say that there's no obvious sign of any divine intervention, so if such an intervention took place, it must have been in the form of tweaks and adjustments to an independently running system. This ties in with the normal Jewish belief that miracles are not usually spectacular unnatural events, but natural events which have been arranged by G_d to happen at the most appropriate and useful moment. Such an intervention, were it to occur, would register not as a strikingly anomalous, remarkable event but as a blip in statistical probability.
There is one feature of life on Earth which could be taken as evidence of very early spiritual prompting, if you wanted to do so, although that isn't the only possible explanation. In the case of minor mutations there are chemical buffers involved in embryological development which can often steer the organism towards normal development, even if it has some odd genes. These buffers tend to fail if the developing embryo is subjected to serious environmental stress, with the result that minor mutations which have been carried in a population for many generations may only start to affect the developed organism when that population is exposed to environmental changes which make it a good idea to try out some new variants to see if they fit the altered situation.
It is difficult to see how this feature could have evolved in the usual way, by slow increments, because it's a method which only comes into play under extreme circumstances where getting it wrong could be fatal. It would be like evolving an ability to survive being shot in the head: if you don't get it spot-on right the first time, it doesn't matter if you are slightly better at it, because you won't be alive to pass on your slightly improved genes. If this characteristic, this sudden introduction of variation just when variation is most needed, just happened by chance to be there the first time living organisms were put under that potentially fatal stress, that's a suspiciously heavy coincidence.
However, these buffers go all the way back to bacteria, so the explanation is probably that in the very early days of life on Earth there were several independent occurrences of simple life, either successively or simultaneously, and only a line which happened by chance to have this buffering system fully-formed survived. There could have been a thousand "weird life" lines of proto-bacteria coalescing out of the primordial soup before one came along which had what it took to survive, so it still isn't neccessary to assume Divine Intervention - but that's not the same as saying you can prove it didn't happen, and if you want to look for highly suspicious statistical probability-bucking this is the place to go to.
I myself do in fact believe that the universe has a spiritual as well as a physical layer, that the spiritual layer is very malleable to mind and that the physical layer is slightly malleable to mind, so that it is possible for things on the spiritual level to exert at least a slight influence over the outcome of events in the physical layer. This is based on my own experience of things such as psychic healing, and the fact that the validity of psychic healing has been confirmed pretty definitely in the laboratory.
That we cannot measure the spiritual level with instruments yet is not proof that it does not exist. If it turns out that we will never be able to detect it with instruments that probably means it's not there to be detected, but before we find out how to see something, there is always a period when the thing exists but we cannot measure it. Pluto existed before we had telescopes to see it, and X-rays were there before we had instruments to measure them - unless you believe in a completely fluid and malleable, spiritual world in which these things only popped into being because we started wanting them to be there.
Nor is the fact that we are edging closer and closer to being able to generate life in the laboratory absolutely from scratch any proof that there is no such thing as a soul. If you think of living organisms as a collaboration between spirit and substance, like a hermit crab inhabiting a found shell or radio waves being turned into audible words through the medium of a radio set, there's no reason why a putative life-force shouldn't be able to form such a collaboration with a sufficiently detailed man-made construct, as easily as with a naturally occurring or divinely created one.
Proteins are themselves made up of strings of smaller chemicals called amino acids, and the DNA/RNA code tells these amino acids how to line up to make a particular protein. It takes three bases in the DNA chain to code for a particular amino acid, and these three bases are called a codon. It takes a string of amino acids - and therefore of codons - to code for a particular protein, and this string of codons resulting in one complete type of protein is called a gene. Genes themselves are strung together in long thick coils of DNA called chromosomes, and a living organism will normally have several chromosomes of different shapes, at least one complete set in every cell nucleus. The precise arrangement of the chromosomes in the cells of a given organism is called its karyotype.
You can think of bases as letters, codons as syllables and genes as words. A chromosome is a sentence or maybe a paragraph, and the whole genotype for a living organism is a complete story. The karyotype is the way the story is laid out and bound.
In sexual reproduction, each parent passes on one complete set of chromosomes to the offspring via its gametes (egg cells, sperm, pollen etc.). Two gametes, one from each parent, fuse to produce an embryo, an offspring, and most of the cells in that offspring thereafter contain two sets of chromosomes, one from each parent. That also means that the organism has two of every gene, because for any given gene it has one copy on a chromosome derived from the mother, and one copy on the matching chromosome derived from the father.
[The exception is the sex chromosomes, the ones which determine gender. In mammals such as humans or cats or whales, for example, as well as pairs of all the different non-sexual or "autosomal" chromosomes a female has two X chromosomes, but a male has one X and one Y. Egg cells - female gametes - always have an X chromosome, since eggs come from the mother and the mother never has a Y chromosome. Sperm cells - male gametes - may contain either an X or a Y and this determines the gender of the resultant embryo. Y chromosomes do not contain all the same genes as X chromosomes - they are missing a lot of the genes which are on the X, plus they have a few of their very own - so in males a lot of the genes on the sex chromosomes are unpaired.]
When the offspring in turn makes its own gametes, prior to reproducing, each gamete contains one complete set of chromosomes - that is, one of each chromosome, not two as in non-sexual cells - but they are a random mix of genes from either parent. It isn't just a matter of saying "This grain of pollen contains chromosomes A and D derived from the father of the plant and chromosomes B and C from the mother", either. A process called meiosis ensures that as the gamete is being produced, chromosomes break, reassort and reform so that instead of the gamete receiving chromosome A from the father of the organism, chromosome B from the mother and so on, each chromosome which goes into the gamete contains a mixture of DNA from both parents.
On average, every sperm or egg cell which your body makes gets roughly half its genes from your mother and half from your father, although each individual cell will contain a slightly different mix. One gamete cell may have an eye gene from your mother, a hair gene and a bone gene from your father, while the next one will have eye and bone genes from your father and a hair gene from your mother, and so on.
This process of sexual reproduction, of mixing genes from two parent organisms to produce an offspring which contains a mix of genes from both, is the main source of genetic variation. Organisms which reproduce asexually, where the offspring is simply a copy of its mother and there is no genetic contribution from a father, have no source of variation except mutation, and so they adapt only very slowly.
Mutation can produce new genes which weren't in the species before. Mutations can occur during the normal cell division and duplication which enables the organism to grow in size or maintain itself - these are called "somatic mutations" and can result in anything from a white spot in the fur to a cancer - but only changes which affect the genetic material of the gametes themselves are preserved and passed on.
There are two major kinds of mutation - genetic and chromosomal. Genetic mutations are usually transcription errors which occur during cell division. They are equivalent to "typing errors", resulting in changes in the bases - the "letters" - which make up a gene "word". These in turn subtly alter the protein for which that gene is the template. Genetic mutations - such as the "blue" fur colour in mammals - can also sometimes occur because a virus has introduced a few new bases, or a whole gene, into the host's DNA.
[Viruses are tiny parasitic quasi-organisms which live by inserting their own genetic material into a cell and tricking the cell into reproducing it. Occasionally they accidentally pick up a bit of the host's genetic material and then carry it with them to their next victim; or they may start life as pieces of the host species' genetic material which broke off and went freelance, and carry with them some genes from their original "parent" which are not directly relevant to the virus itself.]
Some genetic mutations are harmful, others neutral or even useful. If you think of them as typing errors, for example, most typing errors result in bad text, but "the singing waves" and "the swinging waves" or "the long trail" and "the lone trail" are equally sensible. In one of WH Auden's poems, for example, a typing error changed "the poets have names for the sea" to "the ports have names for the sea" and he decided to keep the change, commenting that "as so often before, the mistake seems better than the original idea".
Even mutations which produce large bodily changes are not necessarily fatal. For some reason the FGFR3 (fibroblast growth-factor receptor 3) gene represents some sort of weak spot which is very prone to mutation, especially in the sperm-cells of fathers who are getting on in years and have done a lot of sunbathing. Depending on the exact nature of the mutation, changes in this gene generate three different types of dwarfism - thanatophoric dysplasia, achondroplasia and hypochondroplasia.
Thanatophoric dysplasia is fatal at or soon after birth. Achondroplasia and hypochondroplasia are fatal if you have two copies of the gene, but one copy produces a healthy individual with an odd shape. These are dominant genes - that is, if you have even one copy of this gene, it will usually be possible to see by looking at you that you have it, because it will affect your development - so anybody who has one of these conditions and doesn't have a parent who is a dwarf, must have got it by a fresh mutation. Because the conditions normally arise through a new mutation every time, they vary slightly, but in its most extreme form achondroplasia results in an individual with a normal-sized torso and head but very short limbs; a sway back; a domed forehead; a flattened bridge to the nose; short, square hands and feet and sometimes with the first and second fingers, and the third and fourth, fused together in pairs. There is a slightly increased risk of heart disease but individuals with this condition are otherwise healthy and of fully normal intelligence. Hypochondroplasia produces similar symptoms to achondroplasia but much milder, so that it isn't immediately obvious whether the person has hypochondroplasia or is just plain short.
Many genetic mutations result in a failure to produce a given protein, or production of a protein which is altered so that it doesn't quite do its usual job. Such genes are usually recessive. What this means is that if the organism has one original gene, and one mutated gene, the original gene produces so much of whatever protein it codes for that it doesn't matter that the mutated gene doesn't produce any; the organism's phenotype is the same either way. Only if the organism has two of the mutated gene does it fail to produce the protein in question, resulting in a visible change.
Hence, mutations which result in recessive genes can become established in the population even if they are harmful, because they have no effect unless an individual happens to inherit the same recessive gene from both parents. And many will be useful in some circumstances even if not in others, and will be selected for and become more common when those circumstances arise. For example, when early humans were living in the bright sunlight of Africa, long before the invention of sunscreen, they nearly all had dark skin and the recessive gene for fair skin was harmful if you had two of it, resulting in potentially lethal sunburn. But once humans began to spread into the dark, soggy bits of northern Europe fair skin was selected for and became almost universal - even though it was and is still recessive to dark skin - because prior to the invention of vitamin supplements, dark-skinned infants in northern Europe had difficulty absorbing enough sunlight to manufacture vitamin D, and so were liable to develop rickets.
Even if a mutation is dominant, there are chemical buffers built into cells which in many cases ensure that a normal protein continues to be produced, even if the gene which codes for it is a bit dodgy. These buffers tend to fail if the organism is exposed to extreme environmental stress, such as a rise in temperature, so that genetic variations which may have been carried invisibly for thousands of generations suddenly start to produce variations in the developed organism, offering a range of types from which selection pressure can sort out those who best fit the altered circumstances. [This seems to me more likely to be a fortuitous side-effect of how the buffers work than something which has evolved, since the opportunities for this feature to be exposed to selection would be rare: but it's so massively convenient that if you wanted to look for evolutionary events which could be interpreted as having an element of divine intervention, this would be a good candidate.]
Chromosomal mutations are changes in the karyotype, the chromosome count and shape of the organism. They are of two kinds: those where part of a chromosome is duplicated or missing, so that the organism either has three copies of some of its genes, called trisomy, or only one, called monosomy; and those where the amount of genetic material is normal but the chromosomes have been broken and put back together wrong, so that they are a different shape from the chromosomes of the parents. If genetic mutations are like typing errors, chromosomal mutations are like binding a book with the pages in the wrong order, or with some of them duplicated.
Chromosomal mutations in which the number of genes is correct but the chromosomes have been split or jumbled shouldn't do any damage to the offspring which receives those odd chromosomes, but they may make it infertile, because when it goes on to produce its own gametes, the pairs of chromosomes it received from its parents probably won't be able to match up and exchange genetic material properly and its gametes will be abnormal. If two individuals with the same mutation mate, however, they may produce fertile offspring who all have a new karyotype and can only breed with each other or with their parents, producing a new mini-species (circumscribed breeding group) on which selection pressure can act.
Trisomy is usually harmful in animals, unless the duplication is very small (plants are better able to survive even major duplications because their body-plan is both simpler and more flexible). Down's Syndrome in humans, for example is caused by a partial duplication of chromosome 21. If the duplication is large, an affected child may have severe learning difficulties, major deformities of heart and gut and a protruding tongue, as well as changes to the facial features and a greatly reduced lifespan; if the duplication is very small it may result in nothing worse than unusually-shaped creases in the palms of the hands, and possibly slight difficulties with concentration.
Most of the really dreadful deformities which sometimes occur in animals are due either to trisomy, or to teratogens - chemicals in the environment which cause the developing embryo to suffer from deformities which are not genetic and not inherited. There is a syndrome called Holoprosencephaly, for example, which in its most extreme form results in a stillborn or dying infant with no facial features apart from a single eye in the middle of the forehead, and a kind of nozzle above the eye where the root of the nose should be: exact causes are uncertain but it is thought that it can be caused either by duplication of chromosome 13, or by some environmental toxin. But these sort of really harmful chromosomal duplications, and also any seriously harmful gene mutations which are not recessive, have no effect on the communal gene pool or on evolution, because affected individuals die either in the womb or soon after birth without passing their genes on. Their only long-term effect is to waste some of their parents' reproductive capability.
Mutation is a lottery in which the relative chances of getting a winner are fairly small, but the number of players is so large that winners nevertheless come by quite frequently. And since the losers - those individuals with seriously harmful genes - usually die without reproducing, it's usually only the winning mutations which get handed on in large numbers.
Harmful recessive mutations such as cystic fibrosis and muscular dystrophy persist in the population because they do no harm in individuals with only one affected gene, but they remain rare. As soon as they start to become common you get carriers mating together, with on average a quarter of their offspring being born with two affected genes. These individuals are sick and will probably die without passing their genes on.
At the same time, these non-viable offspring have used a quarter of the reproductive potential of their parents who are carriers of the recessive gene, and used it up on a biological dead-end. In a population in which a recessive gene is fairly widespread, there is always going to be a risk that a carrier will mate with another carrier and produce a proportion of affected offspring, whereas an individual who is not a carrier has no risk of producing affected offspring (except by a rare new mutation): so carriers of seriously harmful genes will tend on average to have fewer healthy, breeding offspring and so fewer descendants than those without the harmful gene.
There are two exceptions to this. One is genes such as Huntington's Disease which are definitely harmful, but which don't usually manifest until after the victim has had, and raised, their children, so that prior to the introduction of genetic testing there was no selection pressure to weed these genes out. The other is genes which are useful if you have a single copy, but harmful if you have two. To have two genes for sickle-cell anaemia is extremely harmful, resulting in crippling pain and weakness, but the gene remains common in damp areas of Africa because to have one sickle-cell gene, combined with one normal gene, conveys a life-saving resistance to malaria.
Mutations are not entirely random - there are fracture-points in our genetic material which cause certain gene or chromosome mutations to recur. Achondroplasia, for example, keeps on happening, even though it's usually due to a mint-new mutation. Certain trisomies such as Down's Syndrome also keep happening. This explains how chromosomal mutation can lead to the formation of a new species.
If you have more than one individual in the same population at the same time, with the same (because recurring) chromosomal abnormality, and those individuals mate, you instantly have a small group who can breed with each other but probably not (or not very successfully) with those around them. Once the population is genetically isolated it will drift away from the main group, because its genes are no longer being mingled with and averaged out with that main group.
Of course, you will still sometimes get individuals cropping up in the parent group who have that same chromosome mutation and can breed with the new group. But they will contribute only a small percentage of the group's genetic material, and as it drifts further away and becomes more and more different, interbreeding even with individuals with the right chromosome count will become more difficult.
Other factors to bear in mind when considering the genetics of evolution are Hox genes, and the existence of separate mitochondrial and chloroplast DNA.
Hox (homoeotic complex) genes are a whole class of genes containing specialised clusters of bases, called homeoboxes, which bind to proteins and thus can switch other genes on or off. Homeoboxes are pretty much the same whatever organism they occur in but the rest of the Hox gene varies according to the function of the gene, that is, to where and when the homeobox is to be applied. It's like having a simple toggle electrical switch which can be inserted into many different points in a circuit, with differing results.
One of the things Hox genes do is control the head-to-tail order in which genes are expressed as an embryo develops: they initiate e.g. a whole sequence of genes for forming a limb, in the manner of a program calling sub-routines. Mutation of a Hox gene can result in aberrations such as an insect with legs growing where its antennae should be, but they can also have useful effects. It may well have been a change in a Hox gene which created the first millipede, for example, or the first vertebrate ribcage, by causing a body segment to be repeated multiple times.
Hox genes explain the existence of so-called throwbacks, a.k.a. atavism. In some cases a new genetic variation is achieved when a Hox mutation switches off a gene or group of genes - switching off the genes which gave the earliest birds teeth and long bony tails, for example, instead of lightweight beaks and pygostyles - but those other genes are still present in the organism's genome. Because they are not actually being acted on, there is no selection pressure to keep them healthy, and over a very long time they will accumulate so many random mutations that they cease to work: but it takes a long time. Artificially switching off the Hoxa2 Hox gene, for example, can cause a mouse embryo to develop with some ancient reptilian characteristics in its jawbones, dating back, so far as we know, a couple of hundred million years.
The discovery of Hox genes explains how it could happen that the fossil evidence strongly indicates that birds evolved from therapod dinosaurs, and yet early birds had a well-developed finger which their presumed dinosaur ancestors had lost. They didn't have to laboriously re-evolve that finger: it will have been switch off by a Hox mutation, and then switched back on by another one.
Hox genes also explain the evolution of body features which are useful in their finished form, but where it's hard to see how they could have been favoured by natural selection in a half-developed form. Extra limbs, for example, do not have to be laboriously developed or lost - just switched on or off, inserted and shuffled around by simple Hox mutations. The ones that turn out well, do well, and the ones that don't turn out well, die, leaving no mark on the population.
Finally, the cells of most eukaryote organisms (that is, organisms whose cells have a clearly defined nucleus containing genetic material and bounded by a cell membrane, which is basically all life-forms except bacteria, viruses and some very odd, ancient single-celled things called archaea) contain ancient, highly-adapted symbiotic bacteria, called mitochondria, which enable them to utilise oxygen to break down food and to power their metabolisms. Plants and algae in addition contain primitive bacteria called chloroplasts which carry on most of the business of photosynthesis. Without these ancient indwellers, animals could not be animals nor plants, plants. They have their own individual genetic material and reproduce asexually, so they change over time only as a result of mutation, not genetic re-assortment: this means that if you have a rough idea of how fast their DNA mutates you can use them as a sort of clock. If you have two groups of organisms you think are related, the more the DNA in their mitochondria or chloroplasts differs between the two groups, the longer ago they split off from each other.
In animals, sperm cells are too small to contain mitochondria, so mitochondrial DNA is inherited only through the maternal line, in the egg cell. Most types of plants have the same issue with chloroplasts and pollen, which is the plant equivalent of sperm.
You may hear scientists speak of an Adam and an Eve from whom all modern humans are descended, but this is their idea of a joke: it does not mean that this Adam and this Eve were the sole progenitors of mankind, quite apart from the fact that "Eve" lived about 143,000 years ago and "Adam" 59,000 years ago.
What it means is that out of the tens or hundreds of thousands of women alive 143,000 years ago, most of whom probably contributed to modern humanity, only Eve's descendants went on to produce at least one daughter in every generation who had a daughter, who had a daughter and so on in an unbroken female line. All the other women's multiple lines of descent went through at least one generation in which none of the women in that generation had daughters, and so their mitochndrial DNA was lost. The mitochondria of Eve are therefore the sole parents of the mitochondria of all humanity, but she is not the sole mother of humanity - just the only woman of her time to have given rise to an unbroken female line.
Clearly, since everyone has a mother, who had a mother, who had a mother there is an unbroken female line stretching back beyond "Eve" all the way back to the first ancestral species to evolve gender. She is the most recent woman whom all known humans alive today are known to have as a common ancester (even the men, because their mothers were descended from Eve), but clearly we are all also descended from Eve's mother, grandmother, great-grandmother etc.. There may well be other women of her time, or even earlier, from whom all humanity can also claim descent but in Eve's case we can prove it, because she left us her mitochindria.
Similarly, the Y-chromosome "Adam" of about 59,000 years ago was the only man, out of all the men alive at that time, whose descendants include an unbroken line of sons. All the others produced at least one generation of all daughters and their Y-chromosomes were lost. My mother's mother and my father's mother are equally my grandmothers, but only my mother's mother's mother's female line gives me mitochondria, and if I were a bloke, only my father's father's father's male line would give me a Y-chromasome.
As with women, there's a chain of fathers of fathers of fathers all the way back to the dawn of gender itself, but "Adam" is the most recent man from whom all humans alive today can be proven to be descended (including the women, because their fathers were descended from Adam), because he left us his Y-chromosome.