Life on Earth has arrived at a threshold. After more than 3 billion years of evolution, the genome has, in the past few thousand years, wired itself in just the right, reflexive way to create creatures capable of wondering about the emergence of form from the formless. In just the past few decades, these creatures have started to acquire detailed knowledge about how this process actually works – and how it might be changed. The chances that this point should have been reached right now, while you are reading this book, seem infinitesimal, given the scale of the Universe in time and space. However, it is legitimate to ask whether the incredible length of these odds might not be more imaginary than real: it could be that 3 billion years of evolution really are required for self-awareness to emerge on any planet, given that such a property might demand genomes of more than a certain amount of complexity.
On the other hand, the increase in complexity is neither linear nor steadily upward. In terms of gene numbers, at least, genomes have increased dramatically in size only three times in the history of life, discounting events surrounding the origin of life itself (of which we know next to nothing). These increases are connected with obvious transitions in the external state and habits of organisms. The first increase happened more than a billion years ago, with the first appearance of integrated, eukaryote cells (ones with nuclei). This was by far the largest single increase in gene number, from around 1,000 to 4,000. The second event – a further increase to around 10,000–15,000 genes – was connected with the invention of organized, multicellular creatures, particularly animals, around 600 million years ago. The third event saw one or more stepwise increases in gene number, to around 30,000, and was associated with the evolution of vertebrates some time before 550 million years ago.
Within these categories, however, gene number says little about complexity, still less about the evolution of human sentience: human beings do not have significantly more genes than do other vertebrates, such as mice or puffer fish. Conversely, were complexity simply a matter of numbers, self-aware organisms could have evolved at any time since vertebrates first evolved. And we could go further: our ignorance of the relationship between internal genomic complexity and external traits prevents us from claiming, beyond doubt, that there is something special about the size of vertebrate genomes (rather than, say, bacterial genomes, or invertebrate genomes) that predisposes them to evolve self-aware creatures, such that smaller genomes would invariably result in, for want of a better term, dumb animals.
So, given the present state of knowledge, we cannot know for sure that sentience cannot arise in creatures with far fewer genes than are routinely found in the genomes of vertebrates. Self-awareness could therefore have arisen at any moment in the past 3 billion years – which makes it all the more special that it has happened to emerge, on the now.
It should now be apparent that there is no direct relationship between gene number and complexity. Sentience lies not in an increase in the number of genes, but in a qualitative change in how genes are organized into regulatory networks. And because each network has a history, and is not created anew in each generation, it cannot be said that the transition to sentience was equally probable or possible in any one of the tens of millions of different networks that existed at any given moment in the history of life, irrespective of the number of genes they contained. The change happened just once, by virtue of a small change in the connectedness of the genomic network peculiar to the immediate ancestor of modern humans. That network, because of its particular history and connection pattern, happened to be sensitive to whatever small change it was that occurred, such that larger changes immediately followed – including the transition to self-awareness.1
The change need not have been very great: a mutation that created an operator sequence where none before existed, or perhaps – as in FOXP2 – a mutation that created a new region in the protein with the potential for chemical activation by an enzyme. Once made, the change would have been followed by a flurry of other changes, so that several million years after the event it might be very hard to spot precisely what the original change was, even were we able to compare the complete genomic networks that specify humans with those of closely related, non-sentient species.
Our closest extant evolutionary relative is the chimpanzee (Pan troglodytes), whose genome sequence is being prepared as I write (late 2003). Comparison of the human and chimpanzee sequences will yield much of great interest, but the secrets of humanity are not likely to be among the immediate benefits. In fact, the chimpanzee and human genomes will look so much alike in detail that the degree of physical difference between the two species will be hard to credit.
The answer will surely lie in a concerted approach – working out those genes, and those modules, that are related to individual development. Chimp adults look very different from human adults, but we look much more similar as babies, foetuses and embryos. Small changes during early development may translate into larger ones later on. However, any changes that we find need not mark the causes of differences between the two species – the ‘smoking gun’ of humanity – but could instead represent the consequences of a whole host of subtle changes in development, sexual behaviour and life-history evolution of the kind I discussed in Chapter 12, all precipitated by an unknown number of genetic changes that will be extremely hard to spot. After all, one small change in a DNA sequence looks very much like another, and it is estimated that 18 million base-pair differences separate the human genome from the last common ancestor of chimps and humans, quite apart from any differences that might be peculiar to the ancestry of chimps.2 Not one of these changes will be specially marked out for our benefit in flashing neon lights as the one, crucial change from which all humanity must flow.
In any case, geneticists will be less interested in extracting meaning from small-scale base-sequence comparisons than in looking for those changes that could have implications for how genes regulate one another. Large-scale alterations in gene order or chromosome structure are known to affect regulation, because they split up operon-like clusters of genes, forcing once widely separated genes into closer proximity. Even though the chimp sequence is not complete, we already know that long sections of at least nine chimp chromosomes are inverted (that is, turned back to front) relative to the same sections in the equivalent chromosomes of humans. Such inversions might be related to significant regulatory changes. In addition, one of the human chromosomes is known to have been created from the fusion of two shorter ones that are found as separate chromosomes in chimpanzees and other apes. This last change is likely to have occurred in the ancestry of humans, after our lineage split from that of the chimpanzees at least 7 million years ago. Many of the other changes, however, could be peculiar to chimps and have nothing to do with the human condition. It is worth noting that chimps have been evolving away from humans for precisely the same interval that we have been evolving away from chimps, and their genomes will have had the same chance as ours to accumulate change.
In the final analysis, it could be that any further questioning along these lines is fruitless, in that no individual detail of the chimp–human genome comparison is ever likely to reveal the genomic basis for that elusive quality we know as human. On the other hand, a complete description of any genomic network, particularly for a multicellular animal, would allow unprecedented insight into how form is created from the formless. Efforts in this direction are already under way. Once we have learned to describe one species in terms of its characteristic genomic network, others will soon follow. The comparison of the properties and connectedness of the networks of different species will enable us to explore the relationship between the general properties of networks and the forms of animals and plants. This will then give us the language we need to articulate fully the question of why there are so many different species of animals on the planet, rather than fewer or still more species; and why the species we see have adopted the forms they have.
A complete description of the genomic network for humans might not expose the seat of the soul, but it might open the way to a new kind of exploration whose prizes could be of incalculable worth – and which could also create immense potential dangers. For it is our lot as humans to be curious, and to be the first creatures to have evolved, as far as we know, the ability to make a conscious decision to initiate radical changes to our own genetic constitutions. We will soon be able to change the human genomic network.
Human-induced changes to the human genome are nothing new. The invention of agriculture and a sedentary lifestyle around 10,000 years ago led to stresses on the human frame that have measurably altered the human genome. The first farmers were smaller, less healthy and more prone to disease than their immediate forebears, who were hunter-gatherers. Various parts of the human genome show signs of natural selection in response to diseases, some of which did not exist before humans domesticated animals and started to live in close proximity to them. These diseases include plague (carried by the bacterium Yersinia pestis) and tuberculosis, both of which originated as diseases of animals and which have subsequently had a profound influence on human history. Empires have fallen and history altered thanks to smallpox, syphilis, measles and typhus. More recent instances of this kind include the influenza that came from poultry and killed more people in 1918-19 than in all the previous four years of war combined. The recent epidemic of the flu-like disease Severe Acute Respiratory Syndrome (SARS) is also believed to have come from poultry, and the ongoing epidemic of HIV-1 may have been transferred to humans through the consumption of primates in West Africa. An epidemic of Type II (non-insulin-dependent) diabetes is currently sweeping populous nations in Asia, with the potential to cause radical genetic change in the next century or so.3 All these changes will have caused shifts in the frequency of different varieties of genes (alleles) in the human population – but none, as far as we know, has altered the fundamental genomic network that is common to all humanity. Furthermore, none of these changes, large and occasionally catastrophic as they have been, were achieved through conscious human action.
The first deliberate attempts to change the human genome have been made in the past few years, in the form of gene therapy. This is an experimental medical intervention in which people with syndromes resulting from known defects in particular genes are treated with synthetic versions of functional genes – administered by infecting the patient with genetically modified viruses – in an effort to alleviate their particular condition. Gene therapy has been of limited success so far, and it is possible that it will never be applicable to more than a few syndromes, and even then only in specific circumstances in which no other treatment is available. Neither does gene therapy constitute an attempt to alter the human genomic network in any fundamental way. Indeed, nothing could be further from the minds of clinicians, who are seeking only to repair defects in the existing network, not to create new ones.
The urge to create, rather than simply repair, will come with developments in computational biology, perhaps along the lines pursued by the Seattle group, only very much larger and more complex. Given that it is already possible to simulate the activities of limited networks of genes in a computer, it is therefore possible in principle to imagine a digital description of any such module, or even the complete network that specifies the development and maintenance of an organism such as the fruit fly, or a mouse, or even a human being. That is, it should be possible to write a computer program that recreates human development entirely within a computer memory.
This approach might have great benefits for medicine. It might allow scientists and clinicians to model the formation of various organs entirely by computer, realizing them in laboratory conditions, and using the results to heal people born with various defects or who have suffered amputation or organ failure later in life. The network approach could grow new limbs for the limbless, create new skin for burns victims and give eyesight to the blind. It is also possible that the network approach can be used to alleviate the symptoms of genetic diseases, whether caused by single-gene defects (cystic fibrosis, phenylketonuria); more subtle, possibly regulatory interactions (insulin-dependent diabetes, heart disease, Alzheimer’s disease and other forms of neurodegenerative disorder); or even wholesale gain or loss of entire chromosomal regions (Down’s syndrome, in which patients have an extra copy of Chromosome 21). In a sense, though, this therapy is no different from present-day single-gene therapy in its motivation: to heal a defective network, but not to change it.
I predict that the first efforts to change the network will be driven by less high-minded ideals than the relief of suffering. Cosmetic surgery provides an instructive model for what might happen. Originally developed to alleviate the pain and disfigurement suffered by burns victims, it was then applied to less life-threatening problems such as the removal of birthmarks. Although cosmetic surgery is still used to remove suffering, it is perhaps best known as an instrument for the elective fulfilment of desire – to have a more attractive body than the one offered by nature, or even a body of a different gender. It could be that the modification of genomic networks might also become an instrument of vanity, but one that offers far more potential for change than cosmetic surgery – which does not, after all, alter genes or networks.
At first, people might use network modification for rather predictable ends, for example to alter their metabolism to allow them to eat more without gaining weight, or to drink alcohol or consume drugs without ill effect (or detection by the authorities); to improve muscle tone, correct wayward eyesight, improve resistance to common infections (especially sexually transmitted ones), or change the size of breasts or genitalia. Criminals will enjoy being able to alter their faces, although it might be very much harder to morph the face of one person into the likeness of another, or to alter fingerprints. DNA fingerprints – in contrast to the prints actually on your fingertips – are unique personal markers regularly used today in forensics. They are features of highly repetitive, junk DNA, and might not be affected by any kind of network modification. Given that network analysis will not reveal the secret of humanity, it might be particularly hard to effect changes in behaviour which might, for example, improve intelligence, remove criminal tendencies, or alter personality in ways that cannot already be achieved by using drugs such as antidepressants.
Once network modification has been widely adopted for therapeutic and recreational use, it might be used in more radical ways, reflecting tendencies in fashion or even politics. Solidarity with an oppressed minority – or allegiance to a favourite sports team – might be expressed by wearing skin of a different colour, for a day or a season. Military researchers and technologically minded terrorists might find a way for people to endure harsh conditions for long periods, assume unusual athletic abilities, see in the dark or even spit poison. Changing sex, adding new limbs or even growing feathers might reflect sincere personal choices, with religious or spiritual significance.
None of these changes need be permanent, so that once they have been made, the person need not live with them throughout life, nor pass them on to the next generation. Gene fixes need be no more permanent than, say, a tattoo is today. The subject of germline modification, on the other hand, in which parents effectively choose the attributes of their children, is a controversial one – as scientist and futurologist Gregory Stock found when he organized a conference on the subject.4 In vitro fertilization (IVF) technology is gaining in popularity, especially among people who choose to wait until their thirties and forties before starting a family, when the risks posed by infertility and genetic disorder are higher than for couples who start younger. People who delay reproduction are also likely to be educated, wealthy and informed about the choices on offer, which now include the gender of their offspring, and screening embryos to ensure that a child is not born who suffers from any of a range of genetic disorders.
IVF and embryo screening – and even cloning, reports of which are still unconfirmed in humans – are examples of conscious intervention and choice in reproduction, but none of them entail or require any kind of genetic modification. It may become possible, however, to impose genetic change on unborn children. This might initially involve in vitro gene therapy to alleviate single-gene defects otherwise missed by screening, but genomic network modification might also become an issue. People might wish to use network modification to change the attributes of their children as a way of expressing their own desires or beliefs. Whatever the outcome, such interventions will raise formidable ethical questions, most notably whether parents have the right to impose fundamental and possibly indelible changes on the form and even the personalities of people who would be in no position to raise objections, nor make informed choices of their own.
Of course, there could be a loophole. Rather than imposing their designs on their direct, biological offspring, a parent might design a child entirely from scratch. If it is possible to create a computer model of the specific genome network of a human being, it might be possible to design humans with any desired trait, whose genome would not be constrained by parentage. The design might be synthesized as DNA, packaged into an egg and brought to term in an artificial womb. This implies the existence of technology far superior to anything we have at present, although the germs of it can be seen in technology used in IVF, cloning and the support of extremely premature infants.5 The effects of such a strategy on human society can hardly be imagined. Considered at the most superficial level, however, parents would not exercise a completely free choice, but would tend to select variants on a rather restricted range of designs on the basis of fashion. You could imagine a situation in which children in a particular year group would all tend to look like popular entertainers or sports stars of the day.
Somewhere along the line, these children, created entirely artificially, would acquire sentience by virtue of their construction according to the stock human genomic network. However, it is legitimate to wonder whether the computer representation of the human genomic network used to create these children might not itself acquire a semblance of sentience, as a result of the connectedness inherent in the program. Would a virtual human grow up in a computer – would it have a personality?
As we learn how to design, create and modify humans, we will do the same for many animals, plants and microorganisms, changing the world around us irrevocably, for good or ill. New lives, new organisms will be created to cater for our slightest whim, our every convenience. Solving the ethical questions posed by the potential to exercise this kind of power and control on the world around us will require a degree of detachment and maturity not evident in current debates about genetic modification or assisted reproduction. Genetic modification (GM) of crops is equivalent to gene therapy in humans, but the debate on the desirability of GM tends to entrench positions conditioned by, or in opposition to, business or political interests, rather than making progress through a detached consideration of the advantages or disadvantages of the technology to economies and markets. Reaction to any kind of assisted reproduction, conception or cloning reflects either parental selfishness or desperation, or the will of opponents to impose a narrow ethical view on others, whether or not this advice is welcome.
But GM and IVF are as nothing compared with the effects of the genomic network modification that is to come, and the standard of ethical debate must rise to meet it. A novel problem raised specifically by genomic network modification – not evident in gene therapy, GM, IVF or cloning – is that we could be able to modify human networks in such a way that we might lose the indefinable quality of humanity that makes us special: that same edifice upon which our ethical, legal and moral codes all stand, and on which our lives and loves are based. To what extent will the products of modified networks be new species, inhuman, or even ‘posthuman’?6 What will their relationships with unmodified humans be like? Will they be our servants, our masters, both – or neither?
Many of these questions may seem overly speculative or fanciful; they are certainly far from our current needs. I would argue, however, that these technological changes could happen faster than we might think, and certainly much faster than some of us might like. There is an urgent need for debate on these issues, but not the kind of debate that acts only to amplify the religious, moral or political views of the present. Such views will be of no help in the sensible management of the novel challenges that the future will throw at us, and may actually obstruct clear thinking about what is best.
The greatest immediate challenge is posed by wilful public ignorance, cozened by woolly thinking, sentimentality and, above all, lazy journalism. It should now be clear that every method to assist reproduction or improve fertility that has so far been invented does nothing more than nurture processes that occur naturally. When a couple choose the sex of their baby, they are only selecting from a number of embryos that have already been created by natural processes – the same story of sperm-meets-egg that has gone on for billions of years. These embryos are genetically and genomically no different from any other human – yet journalists cannot seem to refrain from using headline language, such as ‘designer babies’, and referring luridly to Frankenstein. It occurs to me that if people really thought about the implications of designing a baby – from scratch, on a computer, gene by gene, connection by connection – their feelings of horror might be justified. If we are to secure the future for our children, we cannot afford the luxury of being mildly titillated at the expense of rational thought.
We must try to raise our game to imagine the unimaginable, with cool heads, and we must do it soon. We have to devise rational ways of thinking about situations that might arise in the future in which our descendants, accustomed to living in a world of widespread and accepted genetic change, might have to make moral decisions that we can scarcely even imagine. What will society be like in a century or so? Will we be able to grow extra limbs or even change sex as easily as changing clothes? Will ‘designer babies’ be a reality? What would be the effects of such changes – such implied power – on the things that really matter to us now: friendship, love, the whole business of two people setting up home under one roof?
Will our descendants, capable of altering everything we traditionally regard as definitive of humans, still think of themselves as human in any sense that we can now recognize? Will they become a higher order of being – or will they become like Homo erectus, a creature that walked like a man but thought like a wild beast? To have engineered a situation in which the brief spark of humanity currently resident on Earth winks out would be a tragedy as immense as the backdrop of time against which evolution is set.
In the Book of Genesis, God gave Jacob a vision of angels ascending to heaven and told him how his descendants would inherit the Earth. But does this licence extend to becoming angels ourselves? The transformation of human beings from apes into angels may sound like pure science fiction, something we needn’t worry about any time soon. But the clouds are already gathering. Genetics was established as a discipline only a century ago, and now we have the draft sequence of the human genome. This sequence does not, in itself, tell us what it means to be human.
Now is the time to start learning what does.