The Crafoord Prize 1999

The Royal Swedish Academy of Sciences has awarded the 1999 Crafoord Prize for biology jointly to Professor John Maynard Smith, University of Sussex, England, Professor Ernst Mayr, Harvard University, Cambridge, MA, USA, and Professor George C. Williams, State University of New York, USA.

The three researchers are being awarded the Crafoord Prize for their pioneering contributions to broadening, deepening and refining our understanding of biological evolution and related phenomena such as the formation of species and their adaptation to changes in their environment.

The Academy of Sciences citation runs:
“… for their fundamental contributions to the conceptual development of evolutionary biology.”

The Crafoord Prize, consisting of a gold medal for each laureate and USD 500,000 to be shared, will be awarded in September 1999 at a ceremony at the Academy of Sciences.

Evolutionary theory in a modern perspective
How different species relate to each other and their descent from common ancestors have long been accepted, but opinions have differed concerning the processes leading to the family tree. Ernst Mayr is a leading figure in the creation of the modern version of the theory of evolution now termed “the modern synthesis”. Here knowledge gained from genetic, systematic, paleontological and ecological research has been integrated to form a coherent theory of evolution. Mayr has contributed to many research fields in this process but is perhaps best known for refining the biological concept of species, thus clarifying what is meant by a species and the conditions under which it may be formed. John Maynard Smith and George C. Williams have been interested chiefly in the evolutionary processes leading to the fact that species continually change – with continual adaptation to environmental factors – or sometimes do not change. George Williams was among the first to establish that adaptations normally come about through natural selection favouring those individuals in a population who possess such characteristics that they have more offspring than others. In this way he put paid to the idea that adaptations arise “for the good of the species”, an idea that has been current since Darwin exposed a lack of clarity in his thinking on this point. John Maynard Smith is more of a mathematician than the other two and introduced, among other things, game theory analysis to biology, coining the term “evolutionarily stable strategies”. Game theory is a mathematical method that may be used to analyse how an actor – “player”- should behave where he cannot have complete information about what another actor – “opponent”- is going to do. Maynard Smith has shown that game theory analysis can explain why, for example, it is uncommon for members of the same species to kill one another in combat. Since no combatant can have complete information on the others’ capacities, the optimal course for both parties is to refrain from escalating the combat. The result – the evolutionarily stable strategy – becomes a ritualised struggle (as among black grouse or ruffs) without bloodshed.


When genetics developed at the beginning of the twentieth century, the theory of evolution ran into problems. While it was fully accepted as a theory of the relationships of different species to one another and of their descent from common ancestors, disagreement remained great regarding the processes leading to such family trees. The discovery of mutations – random changes in the genetic make-up – caused many biologists to consider the mutations as the driving force in evolution rather than the natural selection advocated by Darwin. Lamarck’s conviction that acquired characteristics can be handed down to offspring persisted, gaining ammunition when the ideas of the Soviet geneticist Lysenko were given the status of official Soviet state biology during the Lenin era. Lysenko maintained that hardy cultured plants could be produced through inheritance of acquired characteristics. The problem remained that of defining how natural selection functioned – selection among individuals or within a population, or selection at some higher level (“for the good of the species or group”).

Concerning species and their origins
Ernst Mayr was one of the biologists who made decisive contributions to bringing order into this confusion of thought, work that eventually led to the amalgam of knowledge from genetics, systematics, ecology and palaeontology that came to be known as “the modern synthesis”. In his book Systematics and the Origin of Species, 1942, he develops the theories of concepts of species and their formation that by and large hold good today. It was here that he coined what is termed the biological concept of species. This is based on the notion that two species cannot interbreed: they are reproductively isolated. The concept of gene flow is thus central to Mayr’s concept of species; there must be mechanisms that prevent gene flow between two species to ensure that the distinctive character of each is maintained. Such mechanisms may be geographical separation or temporal separation such that the two species never meet because they live in different places or reproduce at different times. Alternatively, reproductive isolation may be conditioned by behavioural, physiological or biochemical barriers that prevent pairing or successful fertilisation across species barriers.

From this definition of the biological concept of species it becomes meaningful to enquire what happens at the very beginning of the process by which a species forms, when a population has just split into two (or more), while there is still no theoretical obstacle to gene flow between them. Mayr concluded that the commonest and most important mechanism must be that the populations become geographically separated, termed allopatric formation of species (subsequent research has shown, however, that a species may form without geographical separation, particularly among plants). A population can be split by all kinds of event from some individuals among a migratory bird species flying to the wrong place to large-scale geological upheavals such as ice ages and rock folding. Whatever the reason for a split may be, it need not be long before the formation of a species has become a fact. This is because of the probability that the sub-populations will live in partly different environments and natural selection will therefore lead to differences between them. If the populations subsequently come into contact, gene flow between them need by no means have become impossible for them to continue to diverge, completing the formation of a new species. It is enough for the offspring formed through pairing across the population boundaries to have some impaired vitality or fertility compared with the offspring of pairing within each population. There is then “reinforcement”, i.e. the behaviour and structures that function as barriers preventing gene flow across the population boundaries are strengthened and species formation is completed.

The process through which populations gradually change under the influence of natural selection is termed adaptation. Darwin himself understood that if for example one species lives in a colder climate than another it will develop adaptations that facilitate life in cold environments, e.g. thicker fur. But Darwin failed to show exactly how this gradual adaptation takes place. This point remained unclear until very recently – and still does in many quarters. George. C. Williams’ book Adaptation and Natural Selection, 1966, is definitely the most influential work in bringing order into this uncertainty. Williams shows with razor-sharp analysis that the incomparably most important causal mechanism for adaptation must lie in the biological variation between individuals that exists in all populations, at least when we consider the species that reproduce sexually. Accordingly, the different individuals in a population differ, in the above example with respect to resistance to cold, in respect to e.g. their fur thickness. If fur thickness affects an individual’s success in life – measured as the number of offspring it produces during its lifetime, and if fur thickness is at least partly genetically determined, then the proportion of individuals bearing genes that give thicker fur will increase in every generation: the frequency of these genes increases. The population in question will then gradually change towards a greater average fur thickness – it gradually becomes better adapted for life in cold climates.

Total absence of planning
This view of adaptation – which still essentially holds good – has profound consequences for our view of evolution. One is that notions of adaptations arising “for the good of the species or the group” – however attractive they may be – must be abandoned. On the contrary adaptation takes place by individuals becoming successful at the expense of other members of the species: as far as we know evolution contains no consideration that could involve safeguarding “the good of the species”. Another consequence of this principle of individual selection which Williams was one of the first to elucidate is that the adaptive process lacks planning of any kind. It is the individuals that at any point reproduce most successfully that, by definition, are the best adapted. Evolution is thus governed by small, opportunistic steps taken at any instant in the direction determined by the individuals who are better adapted than others of their species just then. It may very well be that this direction later proves to be disastrous, but the evolutionary process cannot predict this. It is therefore not surprising that the normal fate of a species through geological time is to die out. It has been estimated that over 99% of the species originating on Earth have done just this. What was once a functional adaptation later became an example of a fatally flawed design.

Game theory for adaptation
But if natural selection works at individual level, how come the world is not full of sabre-toothed beasts, swelling biceps and aggressive individuals that miss no opportunity of killing a member of the same species? John Maynard Smith has made pioneering efforts to explain this apparent paradox using game theory. This mathematical method is used to analyse interactions between actors – “players” – who cannot know what their opponents are planning to do but are capable of adapting their behaviour in the light of experience (experience may here stand for information stored in the genes just as well as in the brain). Game theory, then, can handle adaptive systems be they animal behaviour, the placing strategies of finance yuppies or the plans of military campaigners.

Consider two combatants who can elect to escalate a battle or to end it without bloodshed. The decision to escalate must be taken by each combatant without being certain whether it will have fatal consequences. This problem was analysed mathematically for the first time by Maynard Smith and Price in a classic article in the journal Nature with the intriguing title The Logic of Animal Conflict, in which the authors showed that it is often best for both parties to a conflict to refrain from escalation. Both benefit themselves by ritualising the struggle to demonstrations of power and threats rather than actually fighting. The observation that rivals seldom kill one another in nature thus does not conflict with the principle that natural selection works at individual level.

In his Evolution and the Theory of Games, 1982, Maynard Smith broadens his perspective to include all types of interaction between individuals. He shows that stable coalitions of individuals – e.g. lifelong marriages among swans and geese – may develop from an amazingly simple rule of thumb usually expressed as “tit-for-tat”; the rewarding of good with good but bad with (limited) bad or, in typical English metaphor, “you scratch my back and I’ll scratch yours”. Evil must be rewarded with just limited evil for the reason that one should avoid an escalating spiral of retaliation: a sinner must, so to say, be able to repent and return to the straight and narrow. If two adversaries in an adaptive system – in our case two individuals in a population – both apply this rule of thumb, a stable coalition based on cooperation will develop. Maintaining this coalition is an evolutionary stable strategy (ESS) for both adversaries in that it is not worth either of them breaking it by starting to use a somewhat different strategy (e.g. starting to “cheat a little”); qualitative alterations in behaviour are required for breaking the coalition to become worth while.
John Maynard Smith was born in 1920. He is professor emeritus at University of Sussex, UK, and FRS.
Professor John Maynard Smith
The White House
Kingston Ridge, Kingston
Lewes, Sussex BN7 3JX, UK

Ernst Mayr was born in 1904. He is professor emeritus at Harvard University, Cambridge, MA, USA.
Professor Ernst Mayr
207 Badger Terrace
Bedford, MA 01730, USA

George C. Williams was born 1926. He is professor emeritus at State University of New York, USA.
Professor George C. Williams
7 Yorktown Road
Setauket, NY 11790, USA