Modern biology​ , at its conception in the 17th century, inherited one unshakeable belief, two mysteries and an unfortunate error of timing. The belief was in the immutability of species, that each species has essential, unalterable characteristics, which can be traced back at least as far as Aristotle. The mysteries were, first, over what it is about life that distinguishes it from death, and second, the process by which a fully developed organism, be it chicken or human, emerges from a fertilised egg. The first mystery was solved, tautologically, by answering that creatures were animate rather than inanimate because they were infused with the breath of life. The second mystery, the chicken and egg problem, was a matter for dispute: was the final adult form of the organism in some way present in miniature in the egg or sperm (preformationism), or did it develop by stages from an original formless mass (epigenesis)? These questions may have been reformulated over the centuries, but they are still at the heart of the life sciences.

The unfortunate error of timing, which made the questions harder to answer, was that biology developed as a science later than physics. Physics – above all Newtonian physics – had become established as the ideal modern science, and biologists sought to understand living processes through analogy with physical and mechanical systems: the heart as a pump; the brain and nerves first as hydraulic systems, later as telephone exchanges and these days as computers. The hydraulic metaphor was pioneered by Descartes, writing early in the 17th century. He regarded living organisms as mere machines, though humans, uniquely, possessed a soul, which communicated with the body by way of a mini-organ at the centre of the brain, the pineal gland. A hundred years later, the physician, philosopher and self-declared ‘mechanical materialist’ Julien Offray de la Mettrie dismissed Descartes’s dualistic waystation in his manifesto L’Homme machine. He argued that mental processes were no more than manifestations of the workings of the brain, a heretical view at the time but today shared by many neuroscientists.

The conflict between the mechanists and a diminishing group of vitalists rattled on through the 19th century. Darwin’s theory of natural selection put paid to any remaining belief in the immutability of species. His was a thoroughly materialist explanation of evolution based on three simple propositions: like begets like, with minor differences; all creatures produce more offspring than can survive into adulthood; and the ‘fittest’ – those best adapted to their environment – are most likely to survive and reproduce in their turn. Thus species will gradually be transformed – evolve – over time. This is natural selection. However, he wasn’t able to explain how such minor differences in fitness could be passed on from a carrier to its offspring. (He speculated about a variety of possible mechanisms, including the idea that each body organ contained minute particles – gemmules – that circulated in the bloodstream before concentrating in the reproductive system.) This gap in the theory was seized on by Darwin’s critics, some of whom were opposed to the very idea of evolution, others to Darwin’s gradualism, preferring to believe that the formation of a new species could only occur as the result of a large leap.

Working at the same time as Darwin but unknown to him, the Moravian monk Gregor Mendel discovered the mechanism that he could not. Mendel showed in breeding studies that the colour and shape of sweet-peas was transmitted from one generation to the next by what he called ‘hidden determinants’. The full implications of his work weren’t appreciated at the time, but around 1900 his results were rediscovered and repeated by several scientists, whose experiments with plant breeding turned up cases in which sudden changes in some characteristic appeared and could be transmitted to the subsequent generation. They called these changes mutations, and Mendel’s hidden determinants were renamed genes. For several decades, dramatic Mendelian mutation, rather than gradual Darwinian natural selection, was the favoured explanation for evolutionary change. Preformationism was reborn in the guise of genetics, in which development was understood as little more than the carrying out of a molecular program embedded in the genes.

A subterranean current continued to oppose this sort of reductionism. For romantics, quantification emptied the world of its observable features: colour, scent, feel, shape and pattern. It was absurd to believe that the swirling motion of a flock of starlings in flight, or the fast co-ordinated turns of a shoal of herring, could be reduced to the motion of each individual bird or fish. Newton was the archetypal enemy, captured in Blake’s image of the scientist measuring the world with a pair of callipers (misunderstood in Paolozzi’s massive reimagining of Blake’s Newton as hero in the courtyard of the British Library).

In the 1930s, a group of young biologists in Cambridge associated with the embryologist Joseph Needham formed the Theoretical Biology Club (TBC), calling themselves ‘organicists’ in an attempt to transcend the tired opposition between mechanism and vitalism. At the International Congress on the History of Science and Technology in London in 1931, they were galvanised by the dramatic intervention of a delegation from the Soviet Union led by Nikolai Bukharin, a favourite of Lenin but soon to be purged. The Soviets rejected the Whiggish view of the history of science as a progressive, disinterested search for truth, insisting instead that science – even its greatest and most theoretical achievements – was driven by the political economy of the time. From this perspective, the dominance of mechanical materialism could be understood as a function of the needs of capital in the rapidly industrialising 19th century. Gradgrindism, as they might have said, was no accident. When, a few years later, an English translation of Engels’s Dialectics of Nature appeared with an introduction by the geneticist J.B.S. Haldane, a close associate at Cambridge, the group felt that at last they had the theoretical tools they needed.

Life could not be reduced to mere molecules, they argued, but neither was some non-materialist vital principle required to explain it. The material world consists in a multitude of entities and processes of various levels of complexity. Each level is governed by a set of organising principles dependent on, but irreducible to, those that govern lower levels. The properties of water cannot be deduced from the properties of hydrogen and oxygen; the behaviour of the basic unit of life, the cell, is not simply an aggregation of the properties of its constituent proteins, lipids and so on. At each level of complexity, from molecule to cell to organism to ecosystem and society, new properties and organising relationships emerge, and to each belongs its proper science. Above all, the TBC insisted, the living world is self-organising and dynamic: it should be understood not so much as an assemblage of things but of dialectically interacting processes. Biology, unlike physics, is a historical science, Needham argued: the present status of any living system can be understood only with reference to the immediate (developmental) and longer term (evolutionary) past that brought it about. Life forms themselves are not static but are continuously regenerated, persisting even as every molecule in the body is broken down and resynthesised thousands – sometimes millions – of times an hour.

The TBC strove to unite three biological sciences that had become divorced earlier in the century: evolution, genetics and embryology (soon renamed ‘development’). Haldane was one of a trio of mathematically minded biologists – the other two were Ronald Fisher and Sewall Wright – who brought Mendelian genetics and natural selection together in what became known as the Modern or neo-Darwinian Synthesis, a comprehensive theory of evolution which persisted throughout much of the rest of the 20th century. In the Modern Synthesis, small chance mutations in genes provide the heritable variations in an organism’s fitness on which natural selection acts. Rapid mutational jumps were out; gradualism was in. Living organisms were now considered mere vehicles for genetic transmission, and evolution itself became redefined as ‘a change in gene frequency within a population’.

Bringing genetics together with development was a tougher proposition. Genetics had since Mendel been a science of differences, seeking to explain why some peas are yellow and wrinkled, others green and round, or why one person has blue eyes, another brown. Development, though, was a science of similarities, asking for instance why humans, in their trajectory from fertilised egg to adult, are generally bilaterally symmetrical, each with two eyes, two arms terminating in five-fingered hands. An attempt to unify them was made by another of the Cambridge group, the polymath biologist C.H. (Hal) Waddington, who in the early 1940s coined the term epigenetics to refer to the study of the ‘causal interactions between genes and their products which bring the phenotype into being’. (Phenotype is a Humpty-Dumptyish word, but can be roughly taken to mean any observable feature of a living organism, at any level from the molecular to the cellular to the entire organism and its behaviour. Richard Dawkins extended its definition by asserting that the dam a beaver builds is part of its phenotype.)

Epigenetics seeks to explain how, starting from an identical set of genes, the contingencies of development can lead to different outcomes. To illustrate this, Waddington imagined what he called an ‘epigenetic landscape’ of rolling hills and valleys. Place a ball at the top of the hill and give it a little push. Which valley it rolls down depends on chance fluctuations; some valleys may converge on the same endpoint, others on different ones. Waddington called this process ‘canalisation’, though the material basis for the metaphor was, at the time, unknowable. He imagined the hills and valleys as held in place by strings stretching from nodes (genes) located below the surface landscape.

He also went further, proposing that if a strongly canalised phenotypic change was repeated generation after generation, some random mutation would eventually catch up with it and it would be assimilated into the genome. He demonstrated that this was possible by exposing developing fruit fly embryos to ether, which induces them to develop a second thorax. After some twenty generations (it takes a fruit fly about seven days to develop from a fertilised egg to an adult ready to mate, so experiments using them are fast and easy), the flies developed the second thorax without exposure to the ether – the epigenetically induced bithorax had become fixed in the fly’s genome. To many of his contemporaries, it appeared as if Waddington was arguing for a version of the ultimate evolutionary heresy, Lamarckism – the inheritance of acquired characteristics. It was easy for them to dismiss Waddington’s results as the artificial product of extreme laboratory conditions, irrelevant to the real world.

The TBC sought funding from the Rockefeller Foundation to set up a theoretical biology institute in Cambridge, but Rockefeller turned the proposal down in favour of a major investment in biochemistry, which presaged the later triumphs of molecular genetics. By now, many of the group’s members had been drafted into war work. Needham was posted to China, where he began the work on the history of Chinese science for which he is now best known. Waddington worked on operations research for the air force. In 1947 he left for Edinburgh, where he remained for the rest of his career, but despite his continued advocacy of the theory, epigenetics faded from view.

With​ the discovery of the structure of DNA by Francis Crick and James Watson in the 1950s, there was a renewed conviction among biologists – especially the physicists and engineers turned biologists like Crick – that what was needed was a ruthless reductionism. It was immediately recognised that DNA’s helical structure provided the chemical form of a program – a code made up of the molecule’s four subunits or ‘bases’, adenine, cytosine, guanine and thymine, represented by the letters A, C, G and T – that could direct an organism’s development, and also a copying mechanism by means of which information could be transferred from generation to generation. Life, it seemed, was computable. The triumph of reductionism seemed so secure that by the 1990s ambitious molecular biologists were able to persuade their funders, public and private, to embark on the massive project of sequencing the entire three billion As, Cs, Gs and Ts that spell out the human genome. The information the sequence provided would, they claimed, transform our understanding of medicine, and in so doing give a powerful boost to a languishing economy.

As the project got underway, the sequencers conducted a poll. How many genes – that is, mini-sequences of A, C, G and T coding for specific proteins – would they discover embedded in the human genome? The betting suggested around a hundred thousand, roughly the same as the number of different proteins in the human body. When it came to it, the chastened researchers reported that the actual number of genes was just over twenty thousand, about the same number as in a millimetre-long nematode worm. Twenty thousand genes to direct the development of the human embryo from fertilised egg to newborn baby, to code for the hundred thousand proteins, to determine the fates of the 37 trillion cells in the human body.

The numbers made a nonsense of the idea that there is a ‘gene for’ any particular human characteristic, from eye colour to IQ to sexual orientation, and has confounded the hope that sequencing the genome would generate a cornucopia of precision-tailored treatments for complex diseases. The problem lies in the common misconception of genes as ‘master molecules’ directing the operation of the cells in which they reside. In fact DNA is a rather inert molecule, as it has to be if it is to serve as a code. It is the cells that do the work. Cellular enzymes read, edit, cut and paste, transcribe and translate segments of DNA – the literary metaphor, universally employed by molecular biologists, isn’t accidental; they think of DNA as the language in which the Book of Life is written – in a scheduled flow during the development of the foetus, according to whether the cells are destined to become liver or brain, blood or bone. No gene works in isolation but as part of a collaboration. Many genes may be required to produce a single phenotype – more than fifty main gene variants have been shown to affect the chances that someone will contract coronary heart disease, for example – and a particular gene may influence many different phenotypic traits, depending on which organ’s cells it is active in. It is during this period of rapid growth that living organisms are at their most plastic, responding to environmental challenges by modifying anatomical, biochemical, physiological or behavioural phenotypic traits. This is epigenetic canalisation.

For molecular biologists, the task has been to discover the mechanisms by which external causes switch genes on and off. This has meant coming to terms with the significance of the fact that DNA is not a naked molecule but is protectively wrapped in a cling-film of proteins – histones – portions of which have to be peeled away before any particular length of DNA can be read; environmental factors affect the peeling process, and therefore the selection of genes to be read. A second important finding has been that during development segments of DNA become ‘marked’; a small molecular chunk, a methyl group (CH3), is attached to one of the DNA bases (generally C, cytosine). The presence of the methyl group prevents the DNA from being read – that is, it switches the gene off. Removing the methyl switches the gene on again. As the field of epigenetics develops, many more such mechanisms are likely to be discovered.

The environment in which a developing embryo is immersed is not unchanging; in mammals the hormonal status or diet of the pregnant female will affect the embryo and foetus, which responds adaptively to environmental challenges as methyl groups are added to or removed from specific regions of its DNA, thus controlling the direction of its development down one or another of the valleys in Waddington’s epigenetic landscape. What’s more, there is growing evidence that methyl marks placed on DNA during development persist and can be transferred to the next generation during reproduction, along with their phenotypic effects. Such transgenerational phenomena, though not their molecular mechanisms, have been known for decades, demonstrated experimentally in animals and observed in humans. One of the best-known studies followed the effects of the Rotterdam famine during the Second World War. During the winter of 1944 the retreating Germans imposed a blockade of food and fuel across western Holland, affecting some 4.5 million people. Children born to women who conceived or were pregnant during the famine period were found to be more susceptible to health problems such as diabetes, obesity and cardiovascular disease than their contemporaries born in the liberated eastern parts of the country. More surprising, at least to orthodox geneticists, is that similar susceptibilities have been found in their children and even their grandchildren.

The data from such ‘natural experiments’ is complex and any conclusions drawn from it must necessarily be fragile, but similar, sometimes quite subtle effects have been found in laboratory experiments. One involved feeding pregnant or lactating rabbits food containing strongly flavoured juniper berries. Given a choice, their offspring, as adults, also favoured a juniper berry diet, and so did the next generation down. The transmission of food preference could be a consequence of the epigenetic marking of DNA, or it could, as Eva Jablonka and Marion Lamb say in their book Evolution in Four Dimensions (2006), be behavioural transmission. Evolution is a multidimensional process, they argue, combining genetic, epigenetic and behavioural transmission across generations. To this they add symbolic transmission, through what they call animal traditions or, in humans, language and culture. A diminishing band of geneticists remain sceptical, arguing that unless transgenerational effects are constantly reinforced, they are gradually diluted and will eventually disappear, rather than being assimilated into the genome. Epigeneticists respond with the bold claim that an epigenetic trait is, as one recent definition has it, a ‘stably heritable phenotype resulting from changes in a chromosome without alterations in the DNA sequence’.

Whatever further insights it may bring, epigenetics has already achieved one of the goals that eluded the biologists of the 1930s: the bringing together of genetics and development, now no longer understood as contrasting sciences of differences and similarities, but as part of a greater unity, the Extended Evolutionary Synthesis (EES). The EES, which was presaged by Waddington, challenges the Neo-Darwinian picture of living creatures as ‘lumbering robots’, in Dawkins’s phrase, whose sole function, crushed as they are between the millstones of genes and environment, is to survive long enough to transmit the genes they carry to the next generation. Chickens, one might say, are merely an egg’s way of making more eggs.

In the EES, by contrast, selection operates not just on the individual adult organism but, through epigenetic processes, across the entire life cycle, and at multiple levels – genes, genomes, organisms, populations and even entire ecosystems. Co-operative interactions within and between species become important – not just competition. In the EES, as for the dialectical biologists of the 1930s, organisms do not merely accept the environment into which they are born, but work to seek out a more favourable one (the term for this is ‘niche construction’) and, having found it, they transform it, just as the beaver does by building a dam. In this way, the EES recognises that processes other than natural selection contribute to evolutionary change. This would have been no surprise to Darwin, who repeatedly emphasised that he saw natural selection as the major, but by no means the only evolutionary mechanism. Yet the EES remains contentious. The announcement this year that the John Templeton Foundation, a Christian funding agency, has awarded an $8 million grant to a multinational team of researchers, led by the evolutionary biologist Kevin Laland, to work on the EES has caused a rumble of dissent from hardline Neo-Darwinists.

Epigenetic mechanisms have been adduced to explain differences between genetically identical twins and in the aetiology of such diseases and disorders as cancer, hypertension, obesity and addiction. New prefixes appear: nutri-epigenetics, for example, attributes many disorders to epigenetic marking caused by nutritional imbalance during gestation. There is research underway into the prospect – still a little distant – of targeted methylation as a means to the permanent silencing of particular genes in the treatment of cancer and viral infections. A whole new field of pharmacological research, with enormous potential for novel and patentable drugs, has been opened up. Alongside the science, the pseudo-science proliferates. On the web, you can read articles claiming that mental effort can cause epigenetic change to ward off or induce cancer; advertisements sell vitamin supplements said to work through epigenetics. Practising epigeneticists try to police the boundary between science and myth while at the same time defending themselves against a residual genetic orthodoxy that continues to look on epigenetics with unease. Yet science fiction can go where scientists fearful for their reputations may hesitate to tread. And sometimes, science fiction proves a better judge of future possibility than established science fact.

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. 38 No. 21 · 3 November 2016

The distinguished geneticists Brian and Deborah Charlesworth charge me with overstating the reductionism of classical genetics and underestimating its contribution to the study of development (Letters, 20 October). They also dismiss the work of Needham, Waddington and the Cambridge Theoretical Biology Club (TBC), as ‘contributing nothing’. They correctly point out that one of the founders of modern genetics, Thomas Hunt Morgan, was a developmental biologist before he turned to fruitfly genetics in the early 1900s, and that Sewall Wright, one of the trio who created the fusion of Mendelism and Darwinism which formed the 1930s Modern Synthesis, was a proponent of epigenetics. But they neglect to add that Morgan turned from development to genetics because genetics seemed amenable to experiment while development was too complex, and that Wright, whose population studies made him receptive to complexity, was dismissive of the ‘bean-bag genetics’ of the classicists, who treated genes as individual units rather than as parts of an interacting system.

The Charlesworths’ point that the experimental techniques available at the time were inadequate for turning the TBC’s theoretical insights into practice is absolutely right. But to dismiss Waddington – who coined the term ‘epigenetics’ and is their predecessor at Edinburgh – so cavalierly is ungenerous, especially now that epigenetics has become such a productive research programme.

To return to my main argument, redefining evolution as ‘a change of gene frequency in a population’ is a reductionism too far, depriving living organisms of playing any part in their own destiny. The bringing together of molecular genetics, epigenetics and development as part of the endeavour of the Extended Evolutionary Synthesis helps restore Darwin’s vision, in the closing words of The Origin, of ‘an entangled bank, clothed with plants of many kinds, with birds singing … insects flitting about … worms crawling … and to reflect that these elaborately constructed forms, so different from each other and dependent on each other in so complex a manner, have all been produced by laws acting around us.’

Steven Rose
London WC1

Vol. 38 No. 20 · 20 October 2016

According to Steven Rose, classical genetics views development as ‘little more than the carrying out of a molecular program embedded in the genes’ (LRB, 8 September). In fact, one of the creators of 20th-century genetics, Thomas Hunt Morgan, was a distinguished developmental biologist. Sewall Wright, a founding father of the Modern Synthesis, which Rose discusses in a rather dismissive fashion, devoted much of his time to studying the role of genetic factors in controlling mammalian development and coat colour, including the role of interactions among different genes, and the fact that a single gene can affect several different traits. In an article from 1934, Wright wrote:

The present states of knowledge of embryology and physiology impose certain limits on speculation. We are sure, for example, that development is an epigenetic process. The genes cannot stand in the simple one to one relation to morphological characters of a preformationist theory … Each character is affected by many genes and each gene affects many characters.

Together with R.A. Fisher, Wright developed the mathematical theory of the control of complex traits (such as body size) by the joint influences of non-genetic factors (including effects of the maternal environment) and of variants in many different genes. This theory has been repeatedly tested and supported by genetic experiments, and forms the basis of contemporary ‘genome-wide association studies’, which are identifying the genetic factors underlying many human diseases. More than seventy years before the publication of the first paper on the Human Genome Project in 2001, it was well understood by researchers in evolutionary biology and genetics that there is no such thing as a ‘gene for’ a given trait.

Rose emphasises the importance for biology of understanding development, but doesn’t mention that since the 1980s a detailed understanding of development has emerged from application of the methods of modern genetics and molecular biology to the analysis of embryonic development the very ‘ruthless reductionism’ that Rose disparages. Needham’s and Waddington’s Theoretical Biology Club, to which he devotes considerable attention, contributed nothing to this understanding; the experimental techniques available in the 1930s were simply inadequate to unravel the complex networks regulating gene activity that are involved in development. Similarly, without modern work by geneticists and molecular biologists we would know nothing about the ‘epigenetic’ modifications of DNA bases and histone molecules, which interact with DNA to modify gene expression.

Brian Charlesworth, Deborah Charlesworth
University of Edinburgh

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