John Maynard Smith is a fellow of the Royal Society and is still a professor of biology at the University of Sussex. He has published a raft of papers and books, of which the most influential are The Theory of Evolution, Evolution and the Theory of Games, The Major Transitions in Evolution, and The Origins of Life: From the birth of life to the origin of language. He has just finished a new book about animal signalling.
If you were starting out now, what would you study?
The really exciting issues in evolutionary biology are concerned with trying to integrate what we know about developmental genetics with evolutionary theory. Understanding developmental genetics means understanding how one gene influences another. The molecular people take a na茂ve view that in some way these regulatory genes are responsible for the structures they induce. But the Hox gene, or any other regulatory gene, could switch anything on 鈥� it鈥檚 just a question of which promoters are sitting there waiting to be turned on and off. On the other hand, to pretend, as the palaeontologists do, that developmental genetics doesn鈥檛 tell you anything is equally false. We really have to get these two professions together.
Advertisement
So will genomics change our understanding of evolution?
The idea that once you鈥檝e found the gene that switches on X, you understand how it evolved is rubbish. At the moment if I gave you the sequence of a gene, you couldn鈥檛 even make the protein that it codes for: you could get the amino acid sequence, but you couldn鈥檛 fold it up let alone tell me what it would do once it was folded. So by itself genomics is deeply uninformative. That鈥檚 why it鈥檚 a question of trying to integrate what we know about the regulatory genes for development 鈥� which comes largely from knocking them out and seeing what goes wrong 鈥� with what we know about the history of the corresponding structures in the fossil record. The aim would be to construct a story of the evolution of morphology that wasn鈥檛 just descriptive or adaptive, but was also mechanistic.
What do we need to do to understand the regulatory control networks?
It鈥檚 hard to know whether we are going to find general principles as we piece it all together. Everything could just be different from everything else, but I鈥檓 sure it won鈥檛 be. The analogy with a computer program seems to be a good one. You have to understand things like procedures and the transfer of information from one procedure to another. Understanding developmental controls is going to be essentially understanding that logic.
I don鈥檛 think it鈥檚 mathematics in the sense that there鈥檚 a pure, abstract body of theory we can use, but it is a theory of information 鈥� writing instructions for making things happen. There鈥檚 so much detail now that I think it鈥檚 not too soon to start doing it. It鈥檚 tremendously exciting! If I were 40 years younger, this is what I would be doing.
You have described the evolution of life, from the simplest replicators to complex human societies with language, as seven 鈥渕ajor transitions鈥� in which there is increasing complexity in the way information is either stored or transmitted. Some people might see this as evidence of design鈥�
If you think that there was some inevitable process leading to an increase in complexity, that鈥檚 quite wrong. It is simply that if you start with something that was simple enough to arrive by chance, there鈥檚 nowhere to go but up.
Is life now as complex as it can get, or might there be an eighth transition?
It鈥檚 something that science-fiction writers are better at thinking about than scientists, who are rather limited in their imagination. But the marriage of DNA, programmed intelligence and silicon is hovering on the horizon. Technically, that would be a major transition because it would be a new way of transmitting information between generations, and storing it. But if we come back in a hundred years鈥� time, will the prostheses continue to be computers on our desks, will they be personalised bits of us, or will we find only silicon beings surviving? That remains to be seen.
So human evolution might progress by a sort of symbiosis with computing technology. But are we biological humans still evolving genetically?
Of course we are 鈥� particularly in things like disease resistance. And the widespread use of prostheses, such as glasses without which I wouldn鈥檛 have survived, must have evolutionary consequences. The most obvious one is the accumulation of deleterious genetic traits. We have the choice of coping with that with still more prostheses, or by trying to influence our own genome. Eugenics is a dirty word, but I don鈥檛 think it should be, I think we are going to have to think quite seriously about it. The words 鈥渆ugenics鈥� and 鈥渇ascism鈥� are regarded as almost synonymous and I think that鈥檚 just plain silly.
Will genetic engineering alter the human genome?
It鈥檚 very difficult because of the mixture of ethical and societal issues. The hard, immediate question is this: is it ever justified to select between embryos in order to ensure that a child does not have a nasty heritable disease? My own view is, yes it is, but that鈥檚 not a scientific opinion, it鈥檚 a moral opinion. But for the foreseeable future, it鈥檚 not a technique that could be used on a sufficiently wide scale to have population-level effects. And at the moment we can only select genes, we can鈥檛 transform them. If that becomes possible, things change.
If we were able to tinker with individual genes, should we use this on ourselves?
If it did not have crippling side effects, I think I would. Why not? But I would do it to avoid having a child with a particular hereditary disease, not to have someone who could run 100 metres in 9 seconds.
More prosaically, what are you studying at the moment?
I鈥檓 working on tuberculosis. There are a number of problems that arise from the evolution of antibiotic resistance. The two major evolutionary questions that my current research bears on are: why recombination, or in other words, why genetic exchange? And what are species? There are some very curious things that we don鈥檛 understand. The first is that bacteria go all the way from having no recombination, not swapping genes at all, to recombination being so frequent that if you look at the sequenced genes within a population, they are in linkage equilibrium. On the other hand, Mycobacterium 鈥� the thing that gives you TB 鈥� never recombines. It is very puzzling.
The exchange of DNA between microbes during recombination creates genetic variation, which makes it a sort of 鈥減roto-sex鈥�. Does this tell us something about the evolution of sex?
Well, it ought to. If it turned out to be true that, say, the parasitic microbes were all sexual and the free-living ones weren鈥檛 鈥� or even if there was a strong correlation, or some other association of recombination rate with some ecological factor 鈥� then it would indeed tell us something. But we don鈥檛 know enough cases in detail to look for that correlation. We will have to find out how much recombination there is in more bugs than we know about at the moment. That information is accumulating, the trouble is that it鈥檚 tending to accumulate only for nasty pathogens.
It is amazing that TB can acquire antibiotic resistance with absolutely no recombination. How does it do that?
It looks like independent mutation, but we don鈥檛 know. We have recently discovered that there are 鈥渕utator genes鈥� in natural populations. It is too early to say what role these are playing but they must play a role 鈥� if you鈥檙e going to become resistant to many antibiotics by a number of different genetic changes, you sure as hell have to have a high mutation rate.
And what about the species question? What are species?
It鈥檚 a very curious fact that people find it impossible to do without species names. But what do we actually mean? When it comes to clonal bacteria, for example, it鈥檚 not obvious that there are species out there. For several pathogens we know not only the level and frequency of recombination but also something about the population structure. We should be able to look at the details of individual bugs and ask 鈥� do they fall into groups that look like separate species, or is there just a continuum? And how does that depend on the amount of recombination going on? When we know the answers to these questions we should understand something about species.
Might it also say something about speciation in the macroscopic world?
My hunch is no, that the mechanisms are so different. The essential difference being that you and I can鈥檛 reproduce without sex. But sex isn鈥檛 an essential component of reproduction in bacteria. That means the answer is going to be different.
Perhaps your best-known work involves using game theory to shed light on the evolution of different behavioural strategies in animals, the 鈥淗awks and Doves鈥� idea. How has this area developed?
If you look at the animal signalling literature now, it鈥檚 entirely based on game theory. I鈥檝e just finished a book on the evolution of animal signals where we talk about religion quite a bit. You mustn鈥檛 think it is confined to human beings: religion, meaning ritual behaviour functioning to create emotional commitments 鈥� there is plenty of it. You find it in a group of hunting dogs about to go out for the day, in a group of birds about to migrate, and in some very odd circumstances in chimpanzees. Chimpanzees go in for a thing called a rain dance. Usually the adult males perform it: they jump up and down, they shout, they pull branches off trees, they go berserk. Nobody really knows what the function is.
There is one anecdote about a rain dance that really fascinates me. A group is going through the forest and they come to a waterfall, and the alpha male, only, proceeds to perform a rain dance. He splashes, he shouts, he throws rocks 鈥� it鈥檚 a big deal. What I think is going on is that he is recruiting a force of nature to strengthen his own personal position, increasing his own prestige by allying himself with something 鈥渙ut there鈥�. Isn鈥檛 that what priests do?
No compromise
JOHN MAYNARD SMITH was born in 1920, the son of a surgeon. His father died when he was just eight years old and he moved from London to the west country with his mother to live with her family. It was here that the young Maynard Smith showed his first real interest in biology, becoming an avid bird-watcher 鈥� a passion that remains strong to this day.
Later, he was sent to Eton. They taught mathematics well, he recalls, but he disliked the snobbish, arrogant and anti-intellectual atmosphere. The family had assumed that John would join his grandfather鈥檚 stockbroking business but he was determined not to. Unfortunately, the wilful 16-year-old had no alternative plan to offer the family, so he committed himself to the first profession that came into his head 鈥� he would become an engineer.
By the time he graduated from Cambridge in 1941, the second world war was in full flow, engineers were in high demand, and Maynard Smith took up a post as a 鈥渟tress man鈥� (calculating the stresses in aircraft wings) at a small firm called Myles Aircraft.
After the war, he decided to return to education. He chose biology and University College London, so he could be under the tutorship of J.B.S. Haldane. Haldane鈥檚 attractions were numerous. As well as being a renowned evolutionary thinker, he was also a Marxist, an atheist, a divorcee, and particularly despised by the schoolmasters at Eton 鈥� a circumstance that had long since convinced Maynard Smith that here was a man worth knowing.
Haldane proved a superb mentor, if a little intimidating with his rugby player鈥檚 physique and notoriously short temper. Nevertheless, Maynard Smith spent 15 happy years at UCL, first learning, then teaching and doing research. He moved to the University of Sussex in 1965 where he has remained ever since.
Colleagues comment that Maynard Smith鈥檚 great strength is his clear thinking. Intellectually, he roams widely, striding into complex terrain and identifying the crucial landmarks and underlying connections. He has used his mathematical training to tackle some of the big questions in evolution. Game theory, for example, explains why most animal conflict is just posturing, and predicts the balance between aggressive and submissive behaviour in a given population. His notion of 鈥渆volutionarily stable strategies鈥� pioneered a mathematical approach to understanding animal behaviour. Maynard Smith has also provided insights in many other areas, including the evolution of sex, parental investment and kin selection. He has never steered clear of controversy, as his long-running spat with proponents of group selection attests.
Many people are persuaded by Maynard Smith鈥檚 recent arguments that life is essentially about information and that evolution can be seen as a series of 鈥渕ajor transitions鈥� with that information being either stored, transmitted or applied in ever more complex ways.
But despite his huge influence, all the honours, citations and textbook references, Maynard Smith is still challenging the status quo. In 1999, he and his colleagues published evidence that mitochondrial DNA undergoes recombination-mitochondria from your father and mother can swap genetic material. It may sound obscure, but if they are right it casts doubt on all the research that uses mtDNA as a molecular clock to unravel evolutionary history, including the dating of our oldest common ancestor 鈥� 鈥渕itochondrial Eve鈥�.
Maynard Smith is frustrated but not surprised that the establishment chooses to ignore these findings.