Improbable Destinies: How Predictable Is Evolution? 
by Jonathan Losos.
Allen Lane, 364 pp., £20, August 2017, 978 0 241 20192 3
Show More
Show More

The issue​ of evolutionary inevitability was brought sharply into focus by the late Stephen Jay Gould in his book Wonderful Life (1989). Gould discussed the bizarre fossils uncovered by the Cambridge palaeontologist Simon Conway Morris in an outcrop of rock in the Canadian Rockies, known as the Burgess Shale. The shale was formed 511 million years ago, in the period when animal life was first emerging. Buried within it Conway Morris found the fossils of extraordinary creatures with no modern equivalents, animals such as Hallucigenia, a long tube with a blob of a head at one end, an upturned tail at the other, seven pairs of pointed, stilt-like legs and seven matching tubes projecting above its back. Another creature had five eyes and a long claw-tipped hose on the front of its head. The subsequent disappearance of such body plans, Gould argued, suggests that chance events shaped the direction of evolution, and that if we were to rewind the tape of life to its beginning and let it run again, it is extremely unlikely that anything like humans would come about. Conway Morris, as much a predestinarian Christian as Gould was a Marxist, strongly disagreed. In The Crucible of Creation (1998) he attacked Gould, ‘biting the hand that once fed him’ as Richard Fortey put it in his review, in a way that made ‘a shoal of piranha seem decorous’.* The range of evolutionary options is tightly constrained, he insisted, and wherever there is life, on earth or any other planet, human-like creatures are likely to emerge.

For palaeontologists such as Gould and Conway Morris, re-running the tape of life could only ever be a thought experiment. Darwin believed that evolutionary change through natural selection was glacially slow, unobservable within a human lifespan, and hence not amenable to experimental testing. That became the conventional wisdom among evolutionists. But no longer. Jonathan Losos, an evolutionary biologist, takes the dispute between Gould and Conway Morris as the starting point for his richly detailed account of pioneering research in experimental evolution. Recent studies, based on field observations as well as direct experimental manipulations, have shown that under certain circumstances evolutionary change can be fast enough to observe in action – witness the rapid emergence of antibiotic-resistant strains of pathogenic bacteria, or the decline in effectiveness of many common agricultural pesticides. Such phenomena make it feasible to test alternative possibilities. If Conway Morris is right, you could vary the genetic profile you started out with, but the process of evolution would always result in life forms which, when faced with similar problems, would converge on similar solutions. If Gould is right, you can start the process from identical conditions every time, but accidental, contingent variations resulting from random mutation, epigenetic pressure or tiny environmental differences, will cause life forms to diverge steadily.

Of course Gould would have accepted that the possibilities available to life forms are not unbounded. Evolution is a tinkerer; it can only build on what is already there, and any change must be physically and chemically possible. There are physical constraints that prevent pigs – or humans – developing wings and flying. And versatile an element as carbon is – compared, say, to silicon, which is often suggested as a possible basis for life forms on other planets – there are limits to the types of molecule that can be built from it and hence participate in living processes. There are also, some have argued, deep ‘laws of form’ based on mathematical principles which, independent of evolution, shape organisms – from the repeated pattern of stripes on a zebra’s back to the whorls of seeds on the head of a sunflower. More straightforwardly, if you are a fast-swimming fish, natural selection is going to favour a streamlined body shape. If you are going to find food by hunting, a bigger brain, powerful jaws and sharper senses will serve your needs better than if you live by munching grass – although herbivores need to increase their chances of escaping the hunter by evolving faster reflexes.

Observing evolution in progress once required many decades of patient, non-interventionist field studies. Over a period of forty years beginning in 1973, the Princeton biologists Peter and Rosemary Grant studied the seed-eating ground finches on a small island in the Galapagos, noting the ways in which year by year changes in the weather pattern, from heavy rainfall to drought, drove changes in body and beak size in the population. Drought affects the supply of seeds and favours the survival of birds with larger beaks, strong enough to crack the big seeds that smaller-beaked birds can’t manage. Larger beak size is heritable; within a couple of generations, the average beak size in the island population had increased by up to 5 per cent.

The Grants’ project required a rare dedication to observing the evolutionary consequences of naturally occurring changes in the environment, but it didn’t directly address the issue of determinism. For that purpose, deliberate manipulation of the environment is a better bet. Consider the guppy, a favourite of the domestic fish tank owner, with its brightly coloured, spotted males and somewhat less flashy females. The females favour the brightest spotted males, which, by way of a process called sexual selection, is said to drive the males towards more and more exotic colouring. Guppies are good experimental subjects, as they breed fast. The ancestors and present-day cousins of fish tank guppies live in shallow South American and Caribbean ponds. In some ponds, the male guppies lack the brightly coloured spots of their domestic relatives; in others the colours and spots are there. Why the difference? It turns out that in some ponds the guppies coexist rather uncomfortably with predators, such as the pike cichlid, which feast on the highly visible brightly coloured males; in others, there are no voracious cichlids and the males can sport their colours. Transfer dun-coloured guppies to cichlid-free pools and within a few generations, the pools fill with colour-spotted males. Vice versa, if cichlids are put into a previously predator-free pool, coloured males disappear, leaving dun descendants. Is this evidence in favour of Conway Morris? Not quite. When later researchers repeated the transfer experiment, putting dun guppies in a predator-free pool, instead of evolving spots the males became brightly iridescent. Whether this outcome was the result of chance differences in the environment between the two experiments, or of selection favouring a different route to presumably the same end (more sexually attractive males), it tips the scales towards Gould.

Such field studies require a degree of patience – not to mention physical fitness, as Losos graphically depicts in his accounts of hiking into the jungle interior of Trinidad in the search for guppy pools – that may deter many evolutionists. Fortunately, it is possible to approach the question of determinism in the lab, as long as you exercise some ingenuity – and choose the right organism. The Nobel Prize-winning biochemist Hans Krebs once told me that for every problem a biologist works on, God has provided the ideal organism, and in this case the ideal choice has been a laboratory biochemical workhorse, the bacterium Escherichia coli. E. coli breeds very fast, its single cell dividing into two sibling offspring every twenty minutes. Take 12 identical samples of E. coli, grow them in separate but identical flasks, and every day take a new sample from each of the flasks and grow it on in turn. Repeat, generation after generation, and check whether the populations diverge. In effect, instead of rewinding the evolutionary tape and running it again, the 12 flasks can be seen as 12 tapes of life running in parallel.

This was the experiment started in 1988 by the microbiologist Rich Lenski at Irvine, California. It is still running today. To put the bacteria under stress, and hence encourage evolutionary change, the medium in which they were grown was short on their normal foodstuff, glucose. Six years and ten thousand generations later, differences began to emerge; in a few of the E. coli populations the cells responded to the food shortage by becoming larger and growing more rapidly. This looks like support for Gould, but by the 50,000th generation, the differences had largely vanished, which could be taken as evidence of convergence à la Conway Morris. But in 2003 Lenski and his team noticed something odd happening in a small number of flasks. The E. coli in these flasks had responded to the challenge of glucose depletion by undergoing a completely novel mutation enabling them to feed on citrate, another standard component of the medium in which they were growing. Mutations – changes in the DNA sequences that code for proteins or regulate their functions – are predominantly the result of cosmic radiation. Most are deleterious, so it would seem the E. coli in these flasks just got lucky.

The answers you get from experiments like this one seem to depend on how they are designed and how long you are prepared to wait. The physical and chemical properties of the universe underpin all living processes, but these processes, and the life forms they generate, transcend the tidy laws that govern physics and chemistry. Biology is a historical science. The Grants’ study shows the way in which environmental pressures can drive evolutionary change. The guppy experiment shows this too, and also that the evolutionary process permits different solutions – coloured spots v. iridescence – to the same problem, namely, if you are male, making yourself more attractive to females. The E. coli experiment shows that originally identical populations can diverge.

Natural selection works at the level of the phenotype – the observable characteristics of an organism – not directly on the genes. In the wild most populations are not clones but genetically heterogeneous, and it may be that the selection pressure imposed by environmental changes, as with the Grants’ finches, favours genetic outcomes already present in potential in the population; release the pressure – for instance by alleviating the Galapagos drought – and the average beak size may revert. In other cases, as in Lenski’s E. coli, it would seem that a genuinely new mutation occurred. Irrespective of the molecular mechanisms involved, there are many ways in which evolving organisms find solutions to environmental challenges. In response to the massive application of agricultural pesticides many different insect species have evolved resistance. This is convergent evolution as predicted by Conway Morris. It is often the same gene, in mosquitoes, aphids, fleas and many other insects, which is active in enabling the biochemical changes that confer resistance. But in other cases bugs have found a quite different genetic route to the same outcome, just as various species of fish, in evolving to swim faster, have varied their body shape, fin pattern and skin surface characteristics.

Losos relishes the details of these and the many other evolutionary experiments he discusses, some natural, some lab-based. Coming finally down on the side of Gould, he sees the contrast between the unique animal species found in Australia (kangaroo, platypus) and New Zealand (kiwi) before the colonists arrived with those in the other main land masses of Eurasia as an indication that contingency led to alternative evolutionary paths. As for humans, we are, he argues, an outlier, a unique oddity, as improbable as Hallucigenia; no primate relative has evolved similar characteristics, and the chance of human-like life on other planets seems remote.

Nonetheless, the idea that evolution leads inexorably to intelligent human-like beings has a strong purchase on popular culture; witness Star Trek and any number of little green men from Mars, or for that matter the propensity of designers to equip robots with human-like features. Unlike the early Darwinians, evolutionists are increasingly chary of stressing human superiority. They no longer draw a tree of life with humans at the top, but rather a bush at the end of whose many branches are placed all current (and hence ‘equally evolved’) species, whether ‘intelligent’ or not. As humans, we tend to regard large brains and intelligent behaviour as the high point of evolution, an apogee towards which life has been striving ever since the appearance of the first single-celled organism. But intelligence isn’t the only measure of success, and anyway, even on brain size and complexity, dolphins beat us. Choose other criteria, and we fail miserably. On the basis of biomass or numbers of individual organisms, ants match and bacteria outnumber us. If it’s longevity you want, better be a giant redwood.

That said, there is a well-mapped lineage leading to the branch on which humans sit. It runs from the earliest mammals as they emerged from obscurity with the death of the dinosaurs, through the primates to us. Throughout this history there is a general – though by no means absolute – tendency to increased brain size, more flexible behaviour, an enlarged sense of self, and autobiographical memory. Bigger brains require larger skulls, more easily balanced by being placed on the top of the body, favouring an upright posture with no need for a tail as a counterbalance. So the evolution of humanoid creatures could perhaps have been predicted, and their features may well be shared by life forms on other planets.

This was the conclusion reached by the Canadian palaeontologist Dale Russell. If the dinosaurs hadn’t been wiped out by an asteroid hitting the earth (and/or by climate change, an increasingly popular alternative explanation), humans might not have emerged, but, he proposed, an upright, human-sized, large-brained, two-legged dinosauroid would probably have evolved, perhaps from the relatively larger-brained dinosaur Troodon, a relative of Jurassic Park’s formidable Velociraptor. But before predestinarians set too much store by Russell’s inevitable – albeit green, scaly or perhaps feathered – humanoid, they should recognise that ours is not the only route to superior intelligence. There is increasing evidence that, as well as mammals, several species of birds, notably ravens and jays, have a sense of self, of memory, of past and future. And it isn’t a requirement that brains be encased in heads. Octopuses, which share many of these abilities, have a distributed nervous system, much of it located along their eight arms, each capable of independent action if severed from the main body. We can barely begin to enter the minds of birds, whose brains are not so dissimilar in structure to ours; the octopus remains utterly alien. Even if Conway Morris were right about convergence, and intelligence were an evolutionary end point, becoming human would not be the only way of achieving it.

Evolution itself does not predict; natural selection can act only on present contingencies. But with the arrival of Homo sapiens, evolution has produced a creature that can not only predict the future, but also endeavour to change it, whether by social and political struggle, or by technological innovation. Improbable Destinies ends with some cautious reflections on the human condition but shies away from current debates about the ethics, potential and hazards of rapidly advancing gene-editing techniques. It is no longer just the wild prediction of futurists that humans will soon be superseded by cyborgs of our own creation: scientific journals now entertain the prospect that new technologies will enable humans to direct their, and other species’, evolution. Yet, even if they are successful, success is transient. If an asteroid were to hit the earth, the deranged president of a nuclear power launch a war, or gene editing result in environmental catastrophe, the outcome might be the extinction of our species. Indeed, the speed of anthropogenic destruction of the conditions of human existence may achieve this end not with a bang but a whimper. In that case, as the evolutionary biologist Lynn Margulis once pointed out, slime moulds would inherit the planet.

Send Letters To:

The Editor
London Review of Books,
28 Little Russell Street
London, WC1A 2HN

Please include name, address, and a telephone number.


Vol. 40 No. 9 · 10 May 2018

Steven Rose mentions Hallucigenia, the extraordinary animal which wobbled along on spiky legs as if on some Paleozoic version of extreme stilettos (LRB, 22 March). He omits to mention that Stephen Jay Gould’s account in Wonderful Life was based on a fundamental mistake: Hallucigenia had been reconstructed upside down. In reality the ‘legs’ were spikes, presumably defensive, mounted on its back, and the ‘matching tubes’ were its legs. Visualised in this way, it turned out to be not ‘something with no modern equivalent’ but just another arthropod, though an unusual one. Gould based much of his contingency argument in Wonderful Life on the existence of creatures like Hallucigenia. Subsequent work has shown that this evidence does not in itself support his conclusions – though that doesn’t mean the conclusions themselves were wrong.

Rory Allen

Vol. 40 No. 7 · 5 April 2018

It shouldn’t in any way detract from the contribution made by Simon Conway Morris to point out that he didn’t, as Steven Rose puts it, ‘uncover’ the ‘bizarre fossils’ in the Burgess Shale (LRB, 22 March). That was the work of Charles Doolittle Walcott. Conway Morris’s doctoral supervisor, Harry Whittington, ‘discovered’ the fossils later in a drawer, probably in the Smithsonian, and passed them on to his student for analysis.

Jim Provan
Aberystwyth University

send letters to

The Editor
London Review of Books
28 Little Russell Street
London, WC1A 2HN

Please include name, address and a telephone number

Read anywhere with the London Review of Books app, available now from the App Store for Apple devices, Google Play for Android devices and Amazon for your Kindle Fire.

Sign up to our newsletter

For highlights from the latest issue, our archive and the blog, as well as news, events and exclusive promotions.

Newsletter Preferences