We put five of the most commonly asked questions about evolution to some of the leading experts in the field. Here are their answers.

1. How did life begin?

Four billion years ago, the Earth’s interior was superheated by nuclear and gravitational energy and its exterior battered by asteroids – not, you would think, a promising nursery for life. Yet here it began. Most researchers trying to explain this extraordinary event have taken the “vivocentric” approach, retracing the steps from today’s life forms back to the origin of organic building materials. I believe this approach is doomed to fail because it neglects the primary cause of life and disregards the geochemistry of the early Earth.

It is 50 years since the University of Chicago’s Stanley Miller created amino acids, the building blocks of proteins, by heating methane, hydrogen and ammonia in an enclosed glass apparatus and adding a spark of electricity. His lab experiments were taken as evidence that life could have started on a scorching Earth struck by lightning and ultraviolet radiation. But these days not many people think that proteins came first. The orthodox view now is that life began in an RNA world in which RNA not only transmitted information but also acted as primitive enzymes, catalysing life’s reactions using organic compounds from the primordial oceanic soup.

An aqueous soup, however, would not have provided the concentration of organic molecules needed for life. No matter. Theorists have come up with all sorts of ideas to explain this problem away. Some argue that life must have started on dry land, perhaps in a pond subject to cycles of evaporation. Or perhaps the ocean froze, concentrating the supposed soup in a residual fluid. Other theories hold that metabolism began in two dimensions on clay or pyrite surfaces until lipids created at the same site somehow organised themselves into cell membranes. If this seems unlikely, then there are always those ubiquitous organic molecules that inhabit space. They might have landed on Earth on meteorites, collecting on the ocean surface as a slick before clumping together to form vesicles in which organic reactions could take place.

I’m not convinced by any of this. My view is that the origin of life is not a biological but a geological issue. Instead of trying to trace back the evolutionary tree to its origins, we should start from the bottom up, considering the geochemical environment of early Earth.

In our universe, structure is built with material to hand, in processes that degrade energy from one level to another with an associated increase in entropy. So in our hunt for life’s origins we should be asking where on the early Earth material and energy might have come together to produce life-like structures, and what thermodynamic and chemical reactions would have been involved, with what opportunities for the disposal of heat and waste. In essence, what we are looking for is a self-regulating electrochemical vehicle that runs off a few hundred millivolts, uses redox reactions to maintain its system, reproduces spontaneously and emits waste.

The early Earth offers only two candidates for life’s first construction and test site: in mounds of mineral precipitates either at acidic springs on ocean ridges, or at alkaline seepages elsewhere on the ocean floor. Both types of spring bring material and energy continuously to bear on a cooler carbonic ocean, both are out of harm’s way, and both are home to living organisms even to this day. But there are several reasons why alkaline seepages are the more likely hatchery of life.

First, temperatures at these seepages are a bearable 75 °C, whereas organic molecules could not survive let alone be generated in ancient acidic springs at over 350 °C. In addition, alkaline seepages favour the solubility of organic molecules, so you have a hydrothermal fluid containing building materials. And where alkaline seepages meet acidic ocean water there is more energy available because protons from the sea augment the energy supplied by electrons in the seepage, producing about half a volt in total – quite enough to allow metabolism.

If life did originate at ancient alkaline seepages, what might it have looked like? I believe that it took the form of immobile iron sulphide compartments – tiny bubbles at the boundary between hot seepage sites and the cooler ocean. Semipermeable, semiconducting and capable of catalysing reactions, iron sulphide membranes were the precursor to organic membranes. And, crucially, they would have brought molecular building blocks together, providing the right environment for life’s chemical reactions to take place.

Inside iron sulphide compartments, hydrogen, ammonia and cyanide bubbling up from below would be strongly out of equilibrium with the carbon dioxide driven through the membrane from the ocean. The energy needed for these to react together would have come from an electron gradient building up spontaneously across the iron sulphide membrane, allowing the production of sugars, ribonucleic acids and amino acids. To join these building blocks together giving the familiar molecules of life would need further energy – enter polyphosphates also originating in the ocean. These would polymerise the nucleic acids and amino acids into RNA and peptides. And protons streaming across the membrane would continually recharge the polyphosphates.

The newly polymerised RNA chains would in turn help build simple proteins by folding to create different shaped clefts each of which would accept a specific amino acid. This process would bring together strings of amino acids to create peptides, as suggested a decade ago by Anthony Mellersh (żìĂš¶ÌÊÓÆ”, 2 October 1993, p 13). And, since protopeptides are useful for copying RNAs and as settings for the active metal centres of enzymes, you have a positive catalytic feedback loop in which living matter could form and evolve.

What’s more, if iron sulphide compartments did bring together organic building blocks to create life on Earth, then these inanimate bubbles could play the same role on any other wet, rocky, sunlit planet in the universe. So, the emergence of life would be an early feature on all such worlds, surviving as long as there was liquid water.

Michael Russell studies the origins of life at the Scottish Universities Environmental Research Centre in Glasgow.

2. How do mutations lead to evolution?

Genetic mutations are the raw material of evolution. But what sorts of mutations are important? Traditionally, biologists have emphasised changes in genes – the DNA sequences that code for proteins. The assumption is that such mutations occasionally result in proteins with slightly different sequences of amino acids that give an individual a survival advantage and lead to evolution through natural selection. However, many gene sequences show hardly any change over millions of years, except in accumulating neutral substitutions that do not alter the amino acid code. Can this slow rate of change really account for the evolution of morphology and behaviour?

An alternative view has recently emerged, promoted mainly by developmental biologists and evolutionary geneticists like myself. We have come to realise the important role played in adaptive evolution by mutations in regions of DNA that regulate gene expression.

One of the most striking and least expected discoveries of the past decade was the finding of common developmental genetic pathways across a diverse range of animal groups. The classic example is the Hox genes that lay down the basic body plan. They were discovered in fruit flies, but turn out to be just as important in fish, frogs and humans. Despite the diversity of body form, the Hox gene sequences are almost identical across the species. Even more astounding, several important Hox genes can be swapped between distant species, with little obvious effect on development. It seems that evolution has conserved the genes and instead altered their interaction and regulation to give rise to animals as diverse as fruit flies and humans.

One of the main systems controlling gene expression is cis-regulation – the binding of proteins known as transcription factors to DNA in so-called “promoter regions” of genes. Each promoter has multiple binding sites that lie close to the start of the gene. Transcription factor binding turns genes on or off and so controls the timing of gene expression during development. In addition, transcription factor binding permits an organism to switch between different forms of the same gene, producing slightly different proteins, for example, or to switch from embryonic to adult, or from male to female forms. It is now clear that cis-regulation is crucial in development and in the differentiation of cells into different tissues with specialised roles.

Unfortunately, the evolution of cis-regulation has been hard to study. The main problem is that changes in promoter sequences have no clear association with altered gene expression. There are many confounding factors. First, although individual transcription factors typically bind to a particular DNA sequence (between 5 and 12 base pairs in length), there is plenty of slop in recognition, and the core binding sequence used by any given transcription factor varies between different species. Second, many different transcription factors bind to different sites within the same promoter region. The complex interactions between these transcription factors not only turn genes on and off, they also control the quantitative expression of genes. Finally, many sites within the promoter region appear to have no specific binding function, and are just neutral spacers. As a result, single base mutations, or even deletions and insertions, have unpredictable effects on gene expression. This is in marked contrast to mutations in regions of DNA that code for proteins, where single base changes cause highly predictable changes in the amino acid sequence of the resulting proteins.

Despite these difficulties, some changes in gene regulation have been linked to morphological modifications. A good example is abdominal pigmentation in the fruit fly Drosophila melanogaster. Unlike females, males have dark pigments at the tips of their abdomens which are thought to be important in mate choice. The ancestral condition is no pigmentation, controlled by expression of the bab gene, which suppresses production of abdominal pigments. But Sean Carroll’s group at the University of Wisconsin has found that in D. melanogaster, bab is suppressed by the transcription factor Abdominal-B, which allows pigments to accumulate in males. Females, however, remain colourless because the effects of Abdominal-B are counteracted by another transcription factor, DSX-F, a protein only produced in females, which stimulates bab.

The regulation of bab reveals that it is no longer sufficient to conceive of evolution in terms of single genes. A mutation in one gene inevitably causes a cascade of effects through the gene regulatory network. Recent work by Junhyong Kim at Yale University and Marty Kreitman’s group at the University of Chicago shows just how far reaching such networks can be. They found that in closely related species of Drosophila there is rapid evolutionary change in the promoters of genes such as hairy (hh) and even-skipped (eve), both associated with early pattern formation. These changes, which include loss and gain of binding sites, alter the interaction of hh and eve genes with a variety of transcription factors such as hunchback, KrĂŒppel, giant and knirps. And in turn, they alter how hh and eve interact with other genes.

Until recently we could say very little about how genes interact. Now it is becoming clear that we can unravel gene networks, and this will greatly enhance our ability to understand how mutations result in adaptive changes in form and function. But a word of caution. In all science, there is a tendency to exaggerate the new. Beware the zealots who claim that evolutionary change in gene networks and regulatory pathways are the true driving forces of morphological evolution. These are wild extrapolations. There is no doubt that mutations within genes and the evolution of new genes also play a part in evolution. We now have some remarkable accounts of functional change in gene sequences. One example is the rapid and convergent evolution of the enzyme lysozyme that has independently allowed foregut fermentation in leaf eaters ranging from monkeys to cows and birds. Recent work has also established the importance of gene duplication as a source of novel genes. There is no need to push one explanation to the exclusion of others.

Andrew Pomiankowski is an evolutionary geneticist at University College London.

3. How are new species formed?

Not long ago, we thought we knew how species formed. We believed that the process almost always started with complete isolation of populations. It often occurred after a population had gone through a severe “genetic bottleneck”, as might happen after a pregnant female was swept off to a remote island and her offspring mated with each other.

The beauty of this so-called “founder effect” model was that it could be tested in the lab. In reality, it just didn’t hold up. Despite evolutionary biologists’ best efforts, nobody has even got close to creating a new species from a founder population. What’s more, as far as we know, no new species has formed as a result of humans releasing small numbers of organisms into alien environments.

These days, the focus has changed. Biologists still think that most speciation is “allopatric” – the result of geographical isolation – but the ideas have shifted away from chance and small populations. Biologists are looking instead at all kinds of weird ways that species can change rapidly. The main forces at work are ecological selection (where new species form as the result of adaptations to changing environmental conditions) and sexual selection (in which changing sexual traits and preferences for such traits lead to divergence in populations). The big questions revolve around the relative importance of these two forces.

One of the most dramatic illustrations of the power of ecological selection is “parallel speciation”, where essentially the same species arise independently in different places in response to a similar environment. The best example of this is among stickleback fish in Canadian lakes. Several lakes contain two different species of stickleback, one bottom-feeding and the other plankton-eating. Analysis of mitochondrial DNA (mtDNA) reveals that sticklebacks living in the same lake are more closely related to each other than to their counterparts in different lakes. In other words, they probably arose by parallel speciation.

These findings also indicate “sympatric speciation” – speciation without geographical isolation. Biologists who believe a species almost never splits into two without first being physically isolated have hotly contested the idea of sympatric speciation. But those who support sympatric speciation have seized on the stickleback findings, as well as mtDNA studies of several other species, which seem to support the idea. Recent research, however, shows that mtDNA doesn’t always give the full picture. Analysis of stickleback nuclear DNA indicates that dissimilar morphs might have invaded the same lake independently and then occasionally interbred, exchanging mtDNA and making them look more closely related than they really are.

Sympatric speciation remains contentious, but other research suggests how it might occur. The evidence comes from a group of fish that have undergone the most extravagant burst of speciation we know of. I have spent much of my professional life studying the cichlids of Africa’s great lakes. Between them, Lakes Malawi, Victoria and Tanganyika contain around 1700 species of cichlids, many of which have evolved since the last ice age, a mere 12,500 years ago. One puzzle about cichlids was explaining the evolution of over 500 species in Lake Victoria that live together without physical barriers to prevent interbreeding. Sexual selection seems to be the key, with males varying in colour and picky females showing distinct colour preferences. In this way, populations of fishes that looked remarkably similar in every other way might have become reproductively isolated, with sexual selection leading eventually to the emergence of new species.

This particular form of sexual selection relies on females being able to distinguish between differently coloured males. But as pollution clouds the waters of Africa’s great lakes, cichlids are losing this ability. In the murky waters, hybridisation is becoming increasingly common, and because cichlid species are evolutionarily close, they often produce viable hybrid offspring. Surprisingly, some biologists now think that hybridisation might actually be a creative process, churning out new species, and it has probably happened naturally in Lake Victoria many times in the past. I am beginning to suspect that hybridisation may be a significant factor in some of the evolutionary explosions we call adaptive radiations.

In theory, we can test whether species are the product of parallel evolution, sexual selection or hybridisation by looking for “speciation genes” – genes that are responsible for preventing interbreeding. As more and more genomes are sequenced, biologists eagerly anticipate finding such genes. Also, there is a big push to look at differences in the way genes are expressed. These are nice ideas, but I don’t think we know enough about which genes are involved in speciation to give us a realistic chance of finding them by such methods. We would do better to use careful Mendel-style crossing experiments to find out if speciation can really be caused by a single gene or a pair of genes, like a male courtship signal and a female preference for that signal. I think most people would bet against this being common. But then, most studies are carried out on relatively old pairs of species that are likely to have evolved lots of other differences following speciation. So, we need to look at species that have recently diverged and can still be crossed in the lab to give fertile hybrids. My old friends the cichlids look like the perfect candidates.

George Turner is an evolutionary biologist and behavioural ecologist at the University of Hull, UK.

4. Is evolution predictable?

The late Stephen Jay Gould once famously suggested a thought experiment in which the tape of life is rewound and replayed. Would the repetition bear any resemblance to the original? Gould’s answer was that it would not: each play of the tape would produce a different outcome. His answer stemmed from the fact that evolution proceeds from a continuous interplay of both random and selective forces. The existence of a random element (mutation, recombination and migration) and a stochastic component (daily chance events that determine the survival of individuals and the probability of finding a mate) suggest that evolution cannot be repeatable, predictable or even follow rules.

Yet, as Darwin so clearly saw, working hand in hand with contingency is natural selection, a most potent force that systematically sorts among variant types, favouring characteristics in organisms that give them a better chance of surviving. Indeed, Darwin’s theory of natural selection makes a prediction – that organisms will adapt to their environment.

Importantly, though, any predictions based on Darwin’s theory will be probabilistic: they require us to know the odds against particular events happening. The trouble is, we rarely do. But all is not lost. Although today’s evolutionary biologists do not anticipate “laws” analogous to those in the physical sciences, as Darwin and other 19th-century biologists did, there is mounting evidence for the existence of certain fundamental rules of evolution. Our growing understanding of the mechanism of evolutionary change is providing tantalising hints that certain outcomes may be more likely than others.

The evidence stems from research on topics as diverse as language and learning theory, evolutionary and developmental genetics, biochemical systems theory and metabolic network analysis. What these all have in common is their focus on establishing the basic design principles of complex systems. It is becoming clear that such systems are often assembled from combinations of a few simple modules. The loose linkage that typically exists between modules allows a huge number of possible combinations. It also ensures that different combinations of modules have a high probability of generating biologically viable scenarios. In evolutionary terms, this suggests that even though there may be a limited number of successful solutions to a particular evolutionary challenge, there may be many ways of achieving the same end.

Fortunately, there is a way of testing some of these ideas. My colleagues and I study populations of the bacterium Pseudomonas fluorescens that rapidly diversify by mutation and selection into distinct types or “morphs” when you grow them in test tubes of nutrient broth. These experiments in test-tube evolution allow us to replay life’s tape, albeit on a small scale, as often as we like.

We have found that when we seed our mini-worlds with genetically identical microbes and the population size is large (around a billion cells per millilitre), each “replay” results in highly similar patterns of evolutionary change. After just one week, P. fluorescens evolves into two new morphs that we call “wrinkly” and “fuzzy” spreaders. But this doesn’t happen if we limit the mutation supply rate, reducing it by more than two orders of magnitude. Evolution only repeats itself if certain phenotypic innovations have a high probability of arising and are strongly favoured by selection.

By looking more closely at wrinkly spreaders, we have found that there are several different pathways to this particular morph. Although all wrinkly spreaders have a similar appearance and tendency to clump together on the surface, there is substantial variation. Genetic analysis reveals that there are many routes to becoming a wrinkly spreader, but that most are the result of simple point mutations in one of the components of a pathway that regulates the expression of adhesive factors. And the design of this pathway? It turns out to be modular with loose linkages between components: precisely the kind of genetic system we predicted would accommodate the evolution of new phenotypes, and possibly even facilitate it.

If bacterial colonies that start out as identical clones evolve down different routes to reach a similar end point, what about colonies that start off as distinct? Can they converge on the same solution to an environmental challenge, or will their evolution be constrained by their genes? To test these ideas we have deleted genes encoding critical components of wrinkly spreaders and then allowed these “defective” colonies to evolve. In all instances wrinkly spreaders eventually emerge by co-opting alternative genetic systems and structural components to bring about the necessary change. So, in the face of similar selective conditions, different lineages can find similar solutions to the same problems. Replay life’s tape, then, and while Homo sapiens may not evolve there is a high probability that introspective bipedal organisms with binocular vision will.

Given historical contingency, it is impossible to come up with a definitive answer to Gould’s thought experiment. But we can start to make predictions about the course of evolution based on our growing understanding of the fundamental architecture of biological systems and the way in which these work together under natural selection. We are already starting to predict how organisms will adapt to fit their environment. In the future we should be able to make quantitative predictions about the kinds of changes that will occur and the particular pathways leading to them

Paul Rainey is an experimental evolutionary biologist who divides his time between the University of Auckland, New Zealand, and the University of Oxford, UK.

5. What’s God got to do with it?

Perhaps your sentiments lie with Karl Popper. Religion, argued the great philosopher of science, lies in the realms of metaphysics and is not open to scientific enquiry. That is the line most biologists take to justify sidestepping the issue. But there is no denying that religion and gods are a core part of human behaviour. That’s why I and a growing number of biologists think we must offer some insight into the questions of why religion exists and at what point in human evolution it began.

Humans exhibit one feature that is very odd by animal standards, namely our extraordinary willingness to accept the will of the community and even to die for it. This level of altruism is the key to our success, allowing us to exploit cooperative solutions to the problems of individual survival and reproduction. For these to work, however, the individual has to be prepared to trade immediate personal interests for long-term gains. And high levels of group conformity expose us to the risk of free riders – those who take the benefits of sociality but will not pay the costs.

Of course, we can and do control free riders with policing and appeals to decency. But in the end, both strategies carry only so much weight: who cares if you don’t like what I do if I gain enough by doing it. Religion offers a significant advance because the threat of intervention by forces beyond our control – whether now or after death – offers a level of penalty that far exceeds anything the civil estate can manage. But it only works if people believe in the existence of a shared supernatural world.

That’s where our species’ special talent for mind reading comes in. This phenomenon is best known in the form of “theory of mind”: the ability to understand that someone else has a mind driven by belief-states. This might be represented in the sentence: “I believe that you suppose that there is a supernatural being who understands that you and I want to aspire to behave decently.” It is this kind of thinking that enables us to go beyond holding personal supernatural beliefs to organising religion as a shared, social phenomenon.

So, our brains allow us to create gods and religions. But is this ability simply an accidental product of the evolution of big brains, or is it an adaptation? My own studies show that in primates, including humans, the volume of the neocortex – and especially the frontal lobe – is directly correlated with group size and social skills. In other words, the evolution of brain size has been driven by the need to provide the computational capacity to support the social skills needed to maintain stability among large groups. And in the case of humans, these social adaptations include religion.

By recognising that religion requires a large amount of mental power, we can also start to ask when it might have evolved. Plotting theory of mind abilities as a function of brain size in our fossil lineage suggests that the complexity required to support religion is likely to have arisen very late in our evolutionary history. It could not have happened before the appearance of Homo sapiens half a million years ago, and possibly not until anatomically modern humans appeared 200,000 years ago. That tallies with evidence for the evolution of language, another prerequisite for religion.

Of course, religion isn’t all stick and no carrot. While religious sanctions help enforce conformity, religious experiences make us feel part of the group. Once again, evolution seems to have furnished us with mental mechanisms that make this possible. In recent years neuroscience has revealed the so-called God-spot – part of the brain’s left parietal lobe, responsible for our sense of spatial self – an area that shuts down when individuals experience ecstatic states (żìĂš¶ÌÊÓÆ”, 21 April 2001, p 24). As well as being linked with a sense of “oneness with the universe”, it also creates the blinding flash of light associated with trances and religious experiences.

But perhaps the most powerful device for reinforcing commitment to the group must be endorphins. These brain chemicals are released when the body is under stress. It’s surely no coincidence that most religions involve practices such as flagellation or long periods spent singing or dancing, which trigger a flood of endorphins whose opiate-like effects make us feel relaxed and at peace with those we share the experience with.

So, gods are created by big brains to prevent free riders benefiting from cooperative society without paying the costs. But religious experience can also be seen in a more positive light, as a way to help reinforce the group’s effectiveness as a bulwark against the vagaries of the natural world.

Robin Dunbar studies evolutionary psychology and behavioral ecology at the University of Liverpool, UK.

RICHARD DAWKINS lives in awe of the wonder of science. His faith runs deep.
Not only in the power of scientific explanation, but also in science as a deep
aesthetic experience to rival the finest music or poetry. With typical verve and
lucidity, Dawkins rides headlong into the ragtag army of critics who claim that
science’s tedious and plodding message robs nature of her beauty and
inspiration. While he easily dispels these shallow views. His wider aim is to
create a work of poetic science. Unfortunately, in this he is less successful.
Unweaving the Rainbow works as a savage critique of populist nonsense,
but fails to unify good science and good art.

The title of his book, Unweaving the Rainbow, comes from John
Keats’s famous lament on Newton’s dissection of the rainbow into light of
different wave lengths. For Keats and many of his contemporaries, Newton’s
science was a cold wind that reduced nature to a “dull catalogue of common
things”, a conquering of mystery, beauty and romance. This is echoed in more
recent writing, both in the popular press and, more surprisingly, within science
itself. But can it be that ignorant enjoyment surpasses informed
understanding?

And indeed Dawkins answers the despair of Keats with a delightful exposition
of Newtonian optics. Almost 300 years later, Newton’s elegant prism experiments
can’t help but take one’s breath away. Light split into a rainbow of colours by
one prism is brought back into white light when refracted by a second, whereas
one part of the spectrum retains its colour when passed through a second prism.
These experiments convinced Newton that prisms don’t colour light, rather they
separate light into its component parts or wavelengths.

Dawkins uses these ideas to lead us on to deeper insights about how spherical
raindrops break up and bend light so that we end up seeing a rainbow. There is
still plenty of mystery: why is a rainbow steady even though rain is falling,
how come we can never reach a rainbow’s end? But now we can enjoy these puzzles
more because they are illuminated by an understanding of the natural forces in
play, not surrounded with poetic obscurity.

Dawkins also lays into what he sees as perversions of science. Spoon benders,
astrologists and even The X-files get a drubbing. These are easy
targets. Dawkins’s serious intent is to refill science with wonder. He sees
wonder as a natural attribute of childhood (allowing children to absorb
knowledge from others) that needs to be balanced by the cultivation of healthy
scepticism. Pseudoscience makes headway by hijacking our sense of wonder into
superstition and obfuscation, leaving adults open to “unhealthy and
reprehensible gullibility”. This is all a bit simplistic and preachy.

So Dawkins wants more good poetic science (that is, what he does). He doles
out excellent chapters on the evolution of ears and cooperation, biological
cycles and the runaway expansion of the human brain. But in his espousal of
science as poetry, Dawkins adopts a highly judgmental attitude towards others.
This results in an extraordinary attack on the writing of the American
palaeontologist Stephen Jay Gould. His work is dismissed as bad poetic science,
using intoxicating rhetoric to make colourful yet empty explanations. Gould’s
punctuationism and belief in contingency are written off as so-much muddled
thinking. Is this tirade personal? Dawkins says no, but it certainly looks like
it.

Where I felt Unweaving the Rainbow fell short was in its discussion
of how poetry and art connect with science. Dawkins thinks art is good at
generating “awed wonder”. He wants this to be part of science as well. Well, of
course, it should be. Wonder can’t be the only thing to draw the two cultures
together. The best science is probably little different from the best art, but
on this Dawkins is silent.

WHY is it that I am still in love with my wife, have had an effortless
weekend playing with my nephew and niece—yet have just had another blazing
row with my brother? Fresh from the runaway success of Sperm Wars,
Robin Baker and partner Elizabeth Oram claim in Baby Wars that all can
be explained by evolutionary biology. So my messy family relationships
apparently make sense once my actions and reactions are seen as the servants of
genes fashioned by unremitting natural selection.

In one sense the main message of Baby Wars is banal. Our nearest
common ancestors, chimps and gorillas, differ from us not only in morphology but
also in their social organisation. There can be no doubt that human anatomy
evolved through natural selection. This logic must also apply to our
interactions with others: the complex behaviour seen in courting, parenthood and
relations between family members has evolved. Does this mean, though, that
natural selection can illuminate all aspects of contemporary family life?

At first sight the answer must be no. Natural selection can’t have
specifically adapted us to exploit single parent benefit, bottle-feed our babies
or use modern contraception. But Baker and Oram’s thrust is to convince us that
natural selection has equipped us with general adaptations, and that these
underlie modern sexual and family behaviour. So single parenthood results from
male uncertainty over paternity and the reproductive success men can gain by
switching partners. Bottle-feeding allows a mother to regain her sexual appetite
earlier than if she breast-feeds. Baker and Oram claim that what we actually do
enhances our reproductive success and so makes evolutionary sense.

Baker and Oram even try to bring child abuse, pregnancy sickness and parental
favouritism into the evolutionary fold. Their central metaphor is war: between
siblings, mother and father, parents and children, even grandparents. This seems
an odd word to use in describing what are essentially cooperative relationships.
Baby Wars has little to say about the biology of guilt and shame, or
how sharing relationships are maintained or needs are communicated and sympathy
given. These capabilities appear to have been forgotten in the emphasis on
exploitation and conflict—but they are surely just as central to family
life.

Underlying Baby Wars is a strong genetic determinism. Fortunately,
the authors largely avoid the old and sterile nature-nurture debate. Instead
they offer us a simplistic adaptationism. Genes dictate why a baby’s cry is
stressful, when it pays to commit child abuse, why mothers suffer post-natal
depression and who commits incest. Genes also account for variation between
individuals and cultures as natural selection has programmed into people the
ability to monitor their situation, subconsciously judge the pros and cons, and
choose the best answers. Well: maybe. But such a view is as yet mainly
guesswork. Baby Warsdoes not provide any true demonstration of adaptive
value in present or past environments.

What is crucially missing from Baby Wars is evidence. For example,
sex during pregnancy seems odd because there is little or no chance that it
results in conception. So why does it happen? Baker and Oram propose that
female-initiated sex during pregnancy is an adaptation for her to check whether
he remains interested and faithful. Male-initiated sex, they say, can sometimes
be more aggressive during pregnancy than at other times—especially when
men have lower certainty of paternity and hope to increase the chances of
miscarriage. But no experiments or observations are given to support either of
these propositions.

Each section of the book is preceded by “scenes illustrating different
aspects of parenthood and family life”. These have the feel of the real-life
dramas daily reported in tabloid and broadsheet newspapers—but they are
entirely made-up. What follows is little better—more speculative thinking
covered by a thin veneer of data. Perhaps Baker and Oram’s hypotheses are
correct, but without evidence, it is hard to differentiate good ideas from
wishful speculation.

COULD biochemistry be Darwin’s Achilles heel? When a rhodopsin molecule in
your retina is hit by a photon of light, it causes an ion channel to open, a
neuron to fire, and thus allows you to see. According to Michael Behe, in
Darwin’s Black Box: The Biochemical Challenge to Evolution, this molecular
cascade is beyond the reach of Darwinism. Modern biochemistry has revealed a
cellular world of precisely tailored molecules and staggering complexity that
throws an enormous monkey wrench into the plausibility of gradual evolution. I
am, however, totally unconvinced by Behe’s case.

Darwin knew all about the problem of “organs of extreme perfection and
complication”. He devoted a brilliant chapter to them in On the
Origin of Species. Gradual evolution of a complex organ such as the human
eye seemed to be impossible in the 19th century. But Darwin showed that many
modern animals have eyes that are much simpler in structure. He gave a plausible
account of how complex eyes gradually evolved from simple light-sensitive
pigmented cells, through a series of functional intermediates. His explanation
of complexity is one of his great triumphs.

Limited by the science of his day, Darwin could not delve much beneath the
anatomical structures visible under a light microscope. Since the 1950s a deeper
understanding of the molecular basis of life has been possible fuelled by
increasing knowledge of the workings of DNA, molecular biology and better
instrumentation. Is anything different within the Lilliputian workings of the
cell? Behe’s claim is that biologists have been quietly overlooking the true
difficulties posed by our newer understanding of the molecular basis of
life.

Take, for example, the structure of the cilium. A cilium is a hair-like whip
used by primitive unicellular organisms for swimming. In more advanced
organisms, cilia have a variety of uses, from causing liquids to move over
stationary cells to helping propel sperm. Electron microscopy has uncovered the
common structure of the cilium, a ring of microtubules that run along its
length. Each microtubule is made up of tubulin molecules, stacked into rods
which are joined together by dyenin motor proteins and nexin linker proteins.
These three molecules work together to cause the paddle-like motion of the
cilium.

Isolated cilia still beat when provided with energy. If dyenin is removed,
however, the cilium becomes stiff and inert; it cannot move without its motor.
But if you remove the nexin protein, something rather unexpected happens. The
rod-like tubulin stacks slide past each other and the whole cilium unravels.
Nexin linkers are needed to convert the sliding force of the tubulin stacks into
the bending motion of the cilium.

Behe describes the cilium as irreducibly complex: each of its three molecular
parts is essential to its function, and if any one is removed, it fails. This
leads to Behe’s bigger claim that irreducible complexity appears throughout the
biochemistry of the cell—in blood clotting, the movement of molecules
across cell membranes, the immune system, and in fact almost everywhere
biochemists have looked in detail. What at first glance look like simple systems
turn out on closer inspection to be highly interdependent complexes that lose
all function once one part of the whole is removed.

If Behe is right, a direct gradual evolutionary route—such as that of
the complex eye—seems unlikely. The final function cannot have dictated
the evolution of the parts. In this case, the standard Darwinian explanation is
to show that the intermediate stages had other functions. For example, our jaw
was once a leg. And this looks like an entirely plausible explanation for cilia.
Microtubules are used in a variety of structural roles in the cell, and motor
proteins use microtubules to move themselves and other molecules around the
cell. It seems a small step to link pre-existing microtubules and motor proteins
to produce a bending rod-like structure and a proto-cilium. But Behe despairs of
such guesses. He dismisses them as the mere ramblings of a fertile imagination.
What, he asks, were the critical events and how exactly did they happen?

At this point I find myself partly agreeing with Behe. If biochemistry is to
fall within evolutionary biology, we need detailed case histories. Behe’s trawl
of the scientific literature on cilia found only three major attempts to
understand their evolution. Each is interesting, covering issues such as the
possible origin of cilia as independent symbiotic bacteria, the use of cilia in
phototaxis and mechanical difficulties in the evolution of cilia.

Evolutionary thinking is pushing at the frontiers of knowledge. But the
general lack of interest in biochemical evolution that Behe reveals is typical.
Only in the area of DNA and protein sequence analysis has evolution been taken
seriously. Why haven’t biochemists tried harder to understand how specific
complex systems evolved?

I think there are two reasons that explain the dearth in understanding. The
first is the sheer difficulty of the problem. Consider cilia, for instance.
Cilia are found throughout the unicellular and multicellular
eukaryotes—cells with nuclei containing genetic material carried on
chromosomes. So their origin must have been very early, around the time
eukaryotes became distinct from the bacteria, whose genetic material is
distributed between plasmids and nuclei. Deducing the causes of evolutionary
events so long ago is never going to be easy. Cilia might be related to the
bacterial flagellum, a similar whip-like propulsion devise. But flagella have a
very different construction and rotate to help bacteria to swim, rather than
show the paddling motion of cilia.

And the comparative biochemistry of cilia is also almost nonexistent. Nothing
is known about variation in cilia design across different groups of organisms,
although it undoubtedly exists. Nobody has attempted to relate this variation to
differences in function or to reveal hints of the proto-cilium. Similar problems
bedevil many attempts to uncover biochemical evolution.

But there is a bigger problem. Most biochemists have only a meagre
understanding of, or interest in, evolution. As Behe points out, for the
thousand-plus scholarly articles on the biochemistry of cilia, he could find
only a handful that seriously addressed evolution. This indifference is
universal. Pick up any biochemistry textbook, and you will find perhaps two or
three references to evolution. Turn to one of these and you will be lucky to
find anything better than “evolution selects the fittest molecules for their
biological function”.

Behe is good at exposing the paucity of evolutionary thought in the field of
biochemistry. But in Darwin’s Black Box, he reveals that he is also
part of the problem, falling back on the old, limp idea of “design”. He takes
irreducible complexity as a statement of fact, rather than an admission of
ignorance, claiming that the “purposeful arrangement” of biochemical parts must
be the result of an intelligent designer. So what we have here is just the
latest, and no doubt not the last, attempt to put God back into nature. But it
is an old blind alley. To understand molecular design, we need a biochemical
account of evolution.

So I think that Darwin’s Black Box is a missed opportunity. You can
read it to tell you what is wrong with biochemistry. Behe is also very good at
making biochemistry easy to understand. But don’t be fooled by his claim that
molecular systems are irreducibly complex, or that a supernatural designer is
needed. Biochemistry is yet another area of biology still awaiting its Darwinian
revolution.

Dawkins has already had big hits with The Selfish Gene and The Blind Watchmaker. God also sells. This week River Out of Eden has joined them, rising to the top of the science bestseller list. Dawkins’s success is well deserved; he writes with great fluidity and clarity about the big evolutionary issues and has not shirked the arguments of his adversaries. Dawkins’s armchair reasoning sometimes has the air of an old scholastic musing upon the number of genes dancing on a pinhead. But he frequently introduces a novel twist to make matters clear. I was also surprised how often I found myself chuckling away. River Out of Eden reveals Dawkins to have a lighter, witty side as well.

The central metaphor of Dawkins’s new book is a river of genes, flowing througb geological time rather than space. Each river is a population of genes held together by sexual interbreeding. The river may fork in time. Each branch slowly drifts apart to give rise to a new species. Some branches dry up in extinction. Others split repeatedly, generating the 30 million branches that exist today.

The notion of a River of Eden has a majestic ring with its borrowed Biblical authority. Its illumination of biology is, however, somewhat limited. After all, rivers flow downhill, joining together, becoming larger and deeper, the reverse of the effect of diversifying speciation. And why do species divide? The river metaphor hardly illuminates this old conundrum. I still prefer Darwin’s tree of life that has evolution going the right way.

Dawkins’ next scriptural play is with mitochondrial Eve and her, probably younger, playmate Y-chromosome Adam. Dawkins makes scientific ancestor hunting simple. DNA allows the reconstruction of the historical relations between separate species. But deciding ancestry within species, in particular within humanity, is much harder. Genes within species are mixed continually together by sexual union. Each gene tells a different story – this one came from my mother’s father’s father, the next my father’s mother’s mother. And the problem gets worse the further back one goes.

Only when sex is absent can ancestry be easily reconstructed. This is why we can classify species using ordinary genes – species have been separate and usually sexless for thousands, if not millions, of years. Uncovering relationships within species is possible using mitochondrial genes, because mitochondria do not have sex. We get all our mitochondria from our mothers, never from our fathers. So all the genes within your mitochondria are your mother’s mother’s mother’s 
 back ad infinitum.

This allows us to locate mitochondrial Eve, the last common female ancestor of mankind, both in time and space. All Europeans, Asians, Australians and some Africans sbare a female common ancestor who is surprisingly recent. She lived only a few tens of thousands of years ago. Other African lineages branch earlier, but still remarkably recently, around 200 000 years ago. This points to Africa as the location of mitochondrial Eve. Dawkins is surely right that mitochondrial Eve is far more fascinating than her mythical namesake from Eden.

Dawkins next takes us through some familiar themes. He deals with the problem of intermediate forms, good design without a designer and the power of self-replication. Dawkins asks whether nature reveals the hand of God. The clear answer is no. In fact, says Dawkins, removing God from nature is often the first step towards understanding.

Take, for example, the common attack on evolution by the denial of intermediates in the pathway to complex organs such as the eye. The force of this argument lies in personal incredulity. How could natural selection have favoured half an eye? Even if it could, would not the huge number of small steps take an inordinate length of time to accumulate? These objections can be overcome by building simple evolutionary models. The most recent computer study by the Swedish scientists Dan Nilsson and Susanne Pelger shows just how readily and quickly a flat epidermal surface can evolve gradually into a high-acuity camera-lens eye. Far from being an insurmountable problem needing the hand of God, the eye just needs a bit of imaginative computer programming.

Other oddities in nature at first glance seem to violate good biological design. What is the use of an even ratio between the sexes in species such as the elephant seal where 4 per cent of males account for almost 90 per cent of all the copulations. A sex ratio biased towards females would be far more economically efficient. Another puzzle is why selection has left us open to the depredations of age. Why is senescence a universal? Should not natural selection have done better and produced super long-lived species?

Conundrums such as these melt away once you see that nature is maximising the survival of DNA, not the welfare of the population. Most male elephant seals may end up bachelors, but occasionally they are destined to be prodigious harem holders. On average a male leaves as many offspring as a female elephant seal when the sex ratio is even, so there is no pressure for natural selection to produce a surplus of females. We can understand ageing as a natural outcome of an asymmetry in selection. All our ancestors were young, rather fewer made it to be old. Genes that have beneficial effects in youth, but a deleterious effects later in life, are favoured much more strongly than genes promoting a trouble-free old age at the cost of glorious youth. As Dawkins puts it, selfish gain underlines God’s utility function rather than the greater good of all.

The eye, sex and death – these are themes touched on or delved into deeply in Dawkins’ previous books. But I found them well worth encountering again. River Out of Eden is strong on argument, but I would have enjoyed more examples, more of the riches of anomalies and exceptions. In many species the sex ratio is uneven, for example, the parasitoid wasp Nasonia, which has genes producing female biased broods, others causing male-biased broods, and still others trying to suppress these deviations. Dawkins’s approach from his armchair is to deduce the utility function of God, but the details are needed to complete the story.

Dawkins strays further from his well-versed arguments to challenge the reading of nature as a struggle between good and evil. Nature is neither kind nor unkind, it is wholly indifferent to suffering. We may find inspiring the lengths to which parents may go to defend their young or horrifying the pain and fear of those pursued to death. But none of this is good or evil. It is simply the result of genes being favoured by selection.

Dawkins’s forthright advocacy of the Darwinian view exposes the shallow thinking of its many critics. The simple logic of selection on self-replicating entities remains an immensely powerful explanation. Dawkins takes this evolutionary message beyond its usual biological bounds. Darwinism has too often been portrayed as a grim and pessimistic creed. River Out of Eden shows there is little reason for this. Dawkins has lodged Darwinism on the high ground of imagination from where we can truly wonder at life.

Though very popular with the public, Stephen Jay Gould has a divided
reputation amongst scientists. Gould’s ‘extensions’ of Darwinism – punctuated
equilibrium, hierarchical selection theory, a greater role for constraint,
chance and unpredictability – have equally inspired and infuriated evolutionary
biologists.

I felt similar ambivalence about Gould’s sixth collection of essays.
At best, Gould is wonderfully entertaining. An unusual entree, followed
by a not-too-believable metaphor from Mozart, Michelangelo or Monty Python,
illuminating a byway of natural history. But too often I had to stifle a
yawn at some of Gould’s overfamiliar and rather unconvincing themes.

Gould’s title essay embodies the good and bad. It is a stark reminder
of the perils of taking generalities for granted; in biology there always
is an exception. The leading feature unifying terrestrial vertebrates (amphibia,
reptiles, birds and mammals) is that they are four-legged and five-fingered.
Digits have been reduced, fused and lost. Hummingbirds have only three toes
and horses only one. But embryology reveals that all this flexibility originates
from a primitively five-digit limb.

The belief in five digits as an ancestral character has even extended
to fossil reconstructions of Ichthyostega, one of the earliest terrestrial
vertebrates from the Devonian (about 390 to 340 million years ago). As Gould
recounts, this neat view crumbled with the finding in October 1990 of a
complete Ichthyostega hind limb. It had seven toes! Other deviant Devonian
species have since been discovered: Tulpeton bore six digits and Acanthostaga
seven. Five can now be seen as a secondary stabilisation rather than a canonical
feature of terrestrial vertebrates.

What then is so special about five? This is where my irritation starts.
Gould claims that five is simply a historical contingency. Successful lineages
just happened to have five toes. They could equally well have had six, seven
or eight (as could we) if evolution was re-run. This seems highly implausible.
It seems far more likely that variation in early terrestrial vertebrates
was honed down to five because this was an adequate number for scraping
around on land. The paradox remains why five became so strongly fixed that
we can still see it in bird and horse development. Historical contingency
seems no explanation at all.

Other essays in Eight Little Piggies cover familiar Gould territory.
The body of the book concerns ‘Grand patterns of evolution’ and ‘Revising
and extending Darwin’. Gould has a palaeontologist’s vision of evolution.
His timescale is millions of years. From this perspective the rise and fall
of groups come to the fore, rather than individual adaptations and design.
Gould uses this to debunk the idea of progress, particularly in the lineage
of Homo sapiens. We often naively think of an evolutionary ladder from monkey
to ape to man. Gould points out that the pattern of evolution is more like
a bush than a ladder. The ape-human branch of the bush reached its peak
of diversity in the Miocene, some 20 million years ago. Since then it has
suffered inexorable decline while the monkeys have expanded greatly. The
few apes left hang on in increasingly restricted habitats (orang-utans in
Asia, gorillas in Africa) and are stuck with odd specialisations (knuckle
walking, large brains). Far from dominating the primate lineage, man and
the apes are the last hangers-on in the monkey’s domain.

The only new ground that Gould breaks in this collection is a set of
essays on man’s environmental destruction and extinction. The theme is most
emotively expressed about Partula snails on the idyllic Polynesian island
of Moorea, which Gould visited in 1991. Partula’s trouble began with the
introduction in 1967 of Achatina, the giant edible tree snail. Escargots
were happily consumed (Moorea is a French colony) but Achatina spread so
wildly that it quickly became a major agricultural pest. To control Achatina
the colonial authorities decided to make a second introduction (as if one
failure was not enough). This time they turned to the ‘cannibal’ snail Euglandia,
a natural snail predator.

Almost predictably, Euglandia ignored Achatina. Instead, it rapaciously
consumed the unwary Partula. The effect was devastating. Within nine years,
all seven Moorean Partula species were completely gone. This loss might
well have gone uncharted but for the work of Henry Edward Crampton. From
the turn of the century to 1956, Crampton made a series of heroic expeditions,
studying the distribution and variation of Partula on Moorea and the surrounding
islands. His followers were still hard at work in the 1970s when Euglandia
struck. All this work has been undone by Partula’s untimely demise. For
Gould this makes Partula’s loss all the greater.

What is most depressing about this story is its predictability. Introductions
of Euglandia had already devastated a number of other island snail faunas.
Warnings were given but ignored. Perhaps the one hope is that popularisers
like Gould will finally get the message across.

Andrew Pomiankowski works at the Galton Laboratory, University College
London.