1. History
2. Classification
3. Variation
4. Natural selection
5. Origin of species
6. Embryology
7. Behaviour
8. Current status of neo-Darwinism
1. History
Charles Darwin was by no means the first evolutionist. Ideas of transformation of species can be found in the classics ('It's all in Lucretius' was Matthew Arnold's comment on Darwin) and in the writings of 18th-century thinkers such as Buffon, Diderot, Goethe, and Charles's grandfather Erasmus Darwin. But all those ideas are vague or covert. The French biologist (he coined that word) Jean Baptiste de Lamarck was the first to present an articulate and explicit theory of biological evolution in his Philosophie zoölogique (1809). Other notable precursors of Darwin were the Scottish publisher Robert Chambers (1802–71) with his anonymous Vestiges of the Natural History of Creation (1844), and Alfred Russel Wallace, whose independent discovery of the mechanism of natural selection prompted Darwin to abridge a vast, unfinished manuscript as On the Origin of Species by Means of Natural Selection, or the Preservation of Favoured Races in the Struggle for Life (1859). Natural selection, a materialistic explanation for adaptation and the diversity of life, was Darwin's main contribution. Coupled with a candid and persuasive argument, it was enough to convince most scientists of the truth of evolution, and to capture the imagination of most late 19th-century thinkers.The latter part of the 19th century was a period of exploration of evolutionary theory. Its biological ramifications formed the mainspring of much late Victorian science, and Darwin's theory soon became influential in many other fields, notably anthropology, sociology, politics, philosophy, and psychology. As this list implies, Darwinism was seen by many as a coherent world-view.
In 1900, Gregor Mendel's work on inheritance, originally published in 1866, was rediscovered, and the science of genetics was born. Mutation theory soon replaced natural selection as the most promising field of research into mechanisms of change, and for a while Darwinism was at a low ebb. But towards the end of the 1920s, mathematicians and geneticists — chiefly R. A. Fisher, J. B. S. Haldane, and Sewall Wright — showed that genetic theory and natural selection were fully compatible, so founding population genetics, which became the central area of evolutionary research. The integration of population genetics with more traditional fields of evolutionary interest such as anatomy, palaeontology, and systematics (classification) was pushed forward in the late 1930s and early 1940s in books by T. Dobzhansky (a geneticist), E. Mayr (a systematist), and G. G. Simpson (a palaeontologist). Julian Huxley's Evolution: The Modern Synthesis (1942) gave an alternative name — the synthetic theory — for neo-Darwinism.
In 1953 Francis Crick and James Watson transferred interest to the molecular level with their model of the structure of deoxyribonucleic acid (DNA), the material basis of heredity. Exploration of the implications of the Watson–Crick model soon resulted in the breaking of the genetic code, unravelling of the mode of translation of the genetic message, and development of other branches of molecular biology. Ideas from molecular biology, though broadly consistent with neo-Darwinism, are one of many sources of a new ferment in evolutionary thought. Darwinism and its modern descendant are by no means fossilized theories, embedded as true foundations by scientific progress. Where the current ferment will lead, or end, is impossible to guess. But the ingredients of the brew can be matched with some of the main headings of Darwin's argument as presented in The Origin: variation, natural selection, instinct (behaviour), fossils, classification, embryology — and the origin of species, the title of his book, but a topic that he hardly tackled.
2. Classification
The basic unit in biological classification is the species. Attempts to define species have been made for centuries and no definition has yet been found to cover every case. Virtually all definitions emphasize reproduction — species are those aggregations of organisms within which mating and reproduction are normal and successful. The modern abstraction covering this concept is 'gene pool': a species is a set of organisms sharing a set of genes, and the sharing is manifested in the mixing of the genes of two parents in the fertilized egg. Genes of one species are not mixed with those of another, because mating is not attempted, or is unsuccessful through the sperm failing to fertilize the egg, or through failure in development of the embryo, or through sterility of the hybrid offspring. The molecular basis for success or failure in reproduction is the exact matching of maternal and paternal chromosomes which is necessary in the cell division producing egg or sperm cells.These criteria are broadly applicable in all sexual plants and animals. In asexual organisms, reproducing by simple fission, for example, it is difficult to form a rational concept of species, since the only link between organisms is a historical one of more or less remote common parentage (see Fig. 1). We recognize this by resemblance between the descendants.
The basic tenet of a theory of evolution is that the relation between species is also historical and due to past common ancestry, manifested in resemblances between species. To evolutionists, the distinct gene pools of today's species are selected and modified fractions of the gene pools of more or less distant common ancestral species. One central task of a theory of evolution is to explain how discontinuities in gene exchange may arise — how a species may split (see section 5 below).
Above the species level, it is common experience that there are groups of species which seem to go together: birds, for example, or beetles or cats. Biological classification formalizes this fact of life by giving these groups Latin or Latinized proper names (birds are Aves, beetles are Coleoptera, cats are Felidae), and by ranking those names in a hierarchy of more or less inclusive categories (Aves is a class — category — within the phylum Vertebrata; Coleoptera is an order in the class Insecta; Felidae is a family in the order Carnivora and class Mammalia). These names and ranks are conventions, but from Aristotle to Linnaeus (18th century) to today, those who propose classifications usually believe that they express something real, an order in nature. Before Darwin, that order was commonly rationalized as 'the plan of the Creator' or the imperfect reflection of unchanging and ideal essence. According to Darwin and his followers, the relation between birds, or between beetles, or between birds and cats is not abstract but historical or genealogical, due to common descent (see cladistics).
The central concept of classification is homology. When a child learns to recognize birds, the criteria used are, at root, the same as those used by the scientist: the feathers, beak, wings, and so on are 'the same' in a sparrow and a swan, whereas the wings of a beetle or the beak of a turtle do not make those creatures birds — the 'sameness' is different or inessential; technically, it is analogy rather than homology. The task of biological classification is thus to distinguish homology (informative sameness) from analogy (misleading sameness). And the interest of classification to the evolutionist is that homologies, the characters of groups, are seen as evidence of common ancestry, so that the hierarchy reflects real historical relations. A research programme is implicit here, to reconstruct the history of life through the common ancestry inferred from homologies. Since Darwin first proposed that 'Our classifications will come to be, as far as they can be so made, genealogies', this programme has been followed with enthusiasm. The programme concerns the main outlines of life's history; it integrates classification with morphology, palaeontology, and embryology. This work is an extrapolation from evolutionary theory, and the fact that the work can be carried out does not materially affect the theory itself. The theory depends on more basic matters, an explanation of how species originate. Darwin approached that problem first through his study of variation.
Fig. 1. Population structure in sexual and asexual species. a. A sexual species, like our own, in which each individual has two parents of different sexes (dark or light blobs). b. An asexual species in which the parent reproduces by dividing into two, as in simple plants and animals like amoeba; each individual has only one parent. According to the theory of evolution, the history of life is a pattern like this, in which blobs are species, not individuals. If the population is to remain constant from generation to generation, only half the individuals in b may reproduce. In a, a parent produces more than two offspring only at the expense of others.
3. Variation
Lacking a sound theory of inheritance, Darwin studied it through the experience of those with a practical interest in it: animal and plant breeders. Of the manuscript from which The Origin was abridged, the only part that Darwin actually published was the two-volume Variation of Animals and Plants under Domestication (1868). In The Origin Darwin argued by analogy. Under domestication, breeders observed variations and were often able to perpetuate them by selective mating. Over many generations, the breeds or varieties of dogs, cabbages, pigeons (Darwin's pets), and so on are the result. Darwin's opinions on the causes and inheritance of variation are chiefly of historical interest now. He wrestled with the problem of whether variation is spontaneous, or is provoked by the 'conditions of life' (i.e. the environment), and favoured the environment. And he tried to analyse the mode of inheritance, whether or not it is 'blending', offspring being intermediate between the parents. Blending seemed empirically true for many characteristics, and it has the consequence that a new variation would be swamped or diluted over the generations. These problems, of the origin of variation and its inheritance, were not solved until the 20th century, with the development of genetics and, later, molecular genetics.There is no space here to go into genetic theory, and its outlines must be baldly stated. The material basis of heredity is DNA, a ladder-like molecule which carries a message in the form of a 'four-letter' code, the letters being four chemical bases, each of which may occupy any rung in the ladder. The message is read in triplets, or 'three-letter words', and the words are of two kinds, 'stop' or 'amino acid X'. Twenty different amino acids are encoded by triplets, some by a single triplet, others by up to six different ones (see Table 1). Thus the genetic message specifies sequences of amino acids terminated by stop signs, and, when translated and acted upon, the result is proteins, which are chains of amino acids. Genes are sections of DNA which specify a discrete amino acid chain.
DNA is carried in chromosomes, in the nucleus of every cell. When sex cells (eggs or sperm) are produced, the chromosomes pair off very precisely, each pair containing one chromosome from each of the two parents, and within each pair parts are exchanged at random (crossing-over). The newly mixed chromosomes then separate in a cell division which produces sex cells containing one member of each pair, so that the number of chromosomes is halved. The number characteristic of the species is restored when the egg and sperm nuclei, each carrying a half-set, unite in the fertilized egg. Inheritance is not blending: genes are passed on unchanged from generation to generation.
With that background, the sources of heritable variation can be specified. First, the genes of each new individual are a new assortment, because they combine two half-sets (from egg and sperm) each shuffled in the crossing-over between chromosome pairs in the parental cell divisions that produced them. This is a form of variation generated by shuffling existing material, as each deal of a shuffled pack of several thousand cards would be unique. Secondly, new variation may be introduced by mutations: accidents or mistakes in replication and repair of DNA or mishaps in cell division. Mistakes in DNA can alter one or more bases in the molecule (letters in the code), and may alter one or more amino acids in the protein specified. Mishaps in cell division may alter parts of chromosomes — lengths of DNA — which may be inverted, deleted, duplicated, or transferred to another chromosome. The effect of mutations on individual organisms ranges from negligible or undetectable to lethal. The effect of mutations on the reproductive capacity of an individual (i.e. the formation of viable eggs or sperm) also varies from undetectable to sterility. Known mutations in DNA turn up with frequencies from about 1 in 104 to 1 in 108: they are rare, but frequent enough for every human individual to carry several mutations which have arisen during his or her life. Most of these will be in somatic (body) cells, not in eggs or sperm, and so cannot be passed on to the next generation.
That statement recalls the other question that bothered Darwin — whether variations acquired during life can be inherited. The question refers to what is now called Lamarckian inheritance, for Lamarck proposed that the effects of use and disuse and other changes during life are heritable, and are a cause of evolutionary change. In fact, Darwin held much the same belief. The orthodox answer to the question is no — acquired characters are not inherited. One reason for this is that offspring inherit only an egg or sperm from each parent, and at least in animals those cells are sequestered very early in embryonic life, before the environment takes its effect. A second reason is that translation of DNA into protein — construction of the organism — is held to be a one-way process, with no feedback from the organism to the DNA. This notion is the 'central dogma' of molecular genetics. The one known exception to that dogma is reverse transcriptase, an enzyme that transcribes RNA, the messenger nucleic acid, into DNA, the message store. But to get environmentally induced information into DNA requires more, a way of getting such information into RNA (reverse translation rather than transcription). No such mechanism is yet known. But there has been support for a neo-Lamarckian model of evolution, and various ways in which the environment might direct or influence the genes have been suggested.
Ideas on variation have gained impetus from techniques for separating variant protein molecules, and so estimating the proportion of variant genes in individuals and species. The result is surprising: there is much more hidden variation than was expected. Hidden genetic variation in one individual is due to different forms of homologous genes in the half-set of chromosomes inherited from each parent. In humans, if the few that have been tested are a fair sample of the whole, the proportion of variant genes in each individual is at least 6 per cent, about normal for vertebrate species; in plants and invertebrates variation is often higher, 15 per cent or more. With the discovery that the genes of species are less uniform than expected, the problem has shifted from Darwin's — where does variation come from? — to a new one — why is there so much variation? Orthodox neo-Darwinism demands that it be due to natural selection.
Table 1. The genetic code
4. Natural selection
Darwin's main contribution was natural selection, the survival of the fittest, a materialist explanation for evolutionary change. There have been several presentations of natural selection theory as a deductive argument. Here are three:(i) all organisms produce far more offspring than are necessary to replace them, yet numbers of each species remain roughly constant. Therefore, there is differential mortality, a struggle for existence.
(ii) All organisms manifest hereditary variations. Therefore, those organisms inheriting variations useful in the struggle for existence will be more likely to survive to pass on those variations.
(iii) The environment is not constant. Therefore, those hereditary variations advantageous in a changing environment will be selected, and species will change to remain in harmony with the environment. This is adapted from an 1870 formalization by A. R. Wallace. It emphasizes the effect of changing environments, and the observed constancy in numbers of individuals
.
A briefer form, due to the philosopher A. G. N. Flew, is:
(i) geometrical rate of increase + limited resources → struggle for existence.
(ii) Struggle for existence + variation → natural selection.
(iii) Natural selection + time → biological improvement. This emphasizes the theoretical limits on environmental resources rather than constant populations and changing environments, and introduces the notion of 'improvement'. What is improved is not specified.
Another brief form: (i) all organisms must reproduce. (ii) All organisms exhibit genetic variation. (iii) Genetic variations differ in their effect on reproduction. (iv) Therefore, variant genes with favourable effects on reproduction will succeed, those with unfavourable effects will fail, and gene proportions will change. This emphasizes variant genes, and reproductive success rather than environmental factors.
If natural selection can be presented as a deductive argument in which the conclusion follows logically from the premisses, then it must be true if the premisses are true. If the premisses in one, or all three, of these formulations are empirically true, natural selection must occur.
But we have to be clear about what is selected. Natural selection concerns differential survival or, the other side of the coin, differential extinction. What are the units that survive or become extinct: are they genes, or fragments of genes (e.g. triplets or single nucleotides), or chromosomes, or genotypes (genetic constitutions of individuals), or phenotypes (individual organisms, each expressing its own genotype), or groups of organisms, or species? Wallace's formulation of selection refers to species, Flew's perhaps to individuals, and the third form to genes: are they all correct? All descriptions of natural selection invoke interactions, of individual organisms with the inorganic environment, with other individuals in reproduction, with individuals of other species in predator–prey interactions, and so on. Descriptions of natural selection also invoke replication or multiplication, of genes, or of individuals carrying favourable mutations. DNA and chromosomes mutate and replicate, but since they do not interact with the environment, selection cannot act on them directly. It is individual organisms or phenotypes — life histories — that interact with the environment, but the genotype of every successful organism, the set of chromosomes, is broken up and reshuffled by crossing-over when it produces eggs or sperm, so that individuals (genotype + phenotype) do not reproduce themselves exactly.
In the network of ancestor–descendant relations over the generations (Fig. 2), the units that survive differentially, or are selected, are bits of DNA of unspecifiable length. If there is selective change over a number of generations, some bits of DNA in the original population will be represented by many copies in the final population, and others will not be represented at all, because they became extinct when the organisms carrying them failed to reproduce. In order to summarize the change, we have to use the abstraction 'gene pool'. Over the generations, there is a change in the gene pool of the population or species. Of course, no one has ever seen a gene pool or dipped a finger in one, and the tangible effect of the change will be a change in the phenotypes in the population.
One definition of 'gene' is 'a portion of chromosome which survives enough generations to act as a unit of natural selection'. That definition may sound vacuously circular, as if we have to understand natural selection before we can understand genes, yet understanding natural selection depends on understanding genes. But it at least emphasizes the point that, because chromosomes are randomly broken up in each generation by crossing-over, there is no particular unit that survives intact over many generations. Indeed, there is a theory of 'hitch-hiking' natural selection; a neutral or disadvantageous mutation may spread because it happens to be adjacent to an advantageous one, so that the two survive as a unit without being separated by the chances of crossing-over.
Because some organisms exhibit parental care, or social organization in which more or less extensive kinship groups take part, the family or kin group may have sufficient cohesion to act as a unit of natural selection, kin selection as it is called. Beyond kin groups, the next steps up the genealogical hierarchy are populations and species. Like kin groups, they are linked by descent, by a shared gene pool, and might act as units of selection, or of differential survival. It is doubtful whether populations of the same species ever interact so that one gene pool becomes extinct: mixing of the gene pools through interbreeding is the likely outcome (though some interactions between human populations may have resulted in extinction, by genocide without interbreeding). But different species do seem to interact in this way. One thinks of the extermination of native species in the Galápagos and other closed habitats after the introduction of animals like pigs and goats. In such cases, one speaks of species selection (see section 5 below). This, on its own, can only decrease diversity, not generate it.
The observation explained by natural selection is adaptation, the apparent design of organisms for the environments in which they live. The 'argument from design' was one of the chief pre-Darwinian props for natural theology — evidence for a wise and benign Creator. Darwin's theory of natural selection cut the ground from that argument (see also Hume, David). In essence, neo-Darwinian selection theory is that mutations arise at random, with no feedback from the environment to direct or influence the type of mutation that is 'needed', and interaction between the environment and organisms bearing the mutation determines success or failure. Wallace's formalization of natural selection emphasizes changing environments as the driving force, but an alternative view is possible, of a static inorganic world which is explored by life through natural selection, so that the environment is in a sense created by organisms. For instance, it is difficult to conceive of environmental change which would lead originally aquatic plants first to colonize the land, or terrestrial animals to take to the air: the land and air were there, but not part of the habitable environment until organisms made them so. Of course, once there are plants on land, or insects in the air, we can easily conceive of the advantage of becoming a cow or a swallow.
But talk of cows and swallows introduces the problem with natural selection. Two entirely distinct aspects of the theory must be distinguished. One is concerned with things like cows, swallows, and giraffes' necks. The other is concerned with selection in populations, the bread and butter of population genetics. As for this latter sort, there is no doubt that it works. The deductive form of the argument for it proves that it ought to work, and there is experimental and observational evidence that it does. Classic instances of natural selection in populations include bird predation of light and dark forms of the peppered moth in industrial and rural Britain; the development of resistance to antibiotics in bacteria, to insecticides in insects, and to rat poison in rats; and the relation between human sickle-cell trait and malaria in Africa. These instances cover two sorts of selection. In antibiotic or insecticide resistance, the gene pool of the population is altered: genotypes to which the agent is lethal are eliminated, and those that can survive it increase. This is directional selection, where there is a shift from one state to another. Melanism in moths and sickle-cell trait are instances of stabilizing or normalizing selection, which results not in a shift, but in maintenance of the norm by elimination of variants. In both instances it happens that the norm maintained is a polymorphic population, and this is sometimes called balancing selection. Another type of selection is so obvious that it almost escapes attention; instances are the poor reproductive success of people suffering genetic defects which place them far outside the norm. Those defects are eliminated by purifying selection. The terminology seems varied and complex, but in general we can think of selection as either eliminating variation (purifying, directional) or maintaining it (balancing, stabilizing), and also as either promoting change (directional) or maintaining the status quo (purifying, stabilizing).
To neo-Darwinians, the important types of natural selection are directional selection (an explanation of inferred change) and balancing selection as an explanation of observed variation. The basis of change is the replacement of one form of a gene, or portion of chromosome, by a new (mutant) form. This will take place when the mutant form confers a relative advantage, and spreads through the population. The rate of spread depends on the degree of advantage or selection pressure (roughly the proportion of selective deaths per generation), and on whether the mutation is recessive (expressed only in those inheriting it from both parents — homozygous for it), dominant (expressed if inherited from only one parent — heterozygous), or intermediate (expressed more strongly if inherited from both parents rather than one). Change is irreversible only when the original form of the gene has disappeared; that is, the mutant becomes fixed, or universal in the population. Fixation is achieved most rapidly when the mutation is neither dominant nor recessive but intermediate, which is probably true for most mutations considered at the molecular or protein-producing level.
Balancing selection will occur when heterozygotes, inheriting different forms of a gene from each parent, are more successful than homozygotes. Here, selection will eliminate homozygotes in each generation, but will not alter the equilibrium proportions of the two (or more) forms of the gene. Balancing selection is the neo-Darwinian explanation for the high incidence (averaging 6–15 per cent of sampled proteins — see above) of heterozygosity observed in natural populations. The assumption is that heterozygotes are fitter than homozygotes because they have a wider array of resources to meet variations in the environment. An alternative explanation is the 'non-Darwinian' proposal that the observed protein variants are selectively neutral; they confer no real advantage or disadvantage and are maintained purely by chance, by random genetic drift. On this neutralist theory, harmless or even slightly deleterious mutations may spread and become fixed (or eliminated) purely by chance. As with natural selection, experiments show that genetic drift occurs, especially in small populations where biased samples are more likely, and there is a highly developed mathematical theory of how quasi-neutral mutations may behave in populations.
So far, it has proved impossible to discriminate between these two mechanisms, neo-Darwinian selection and neutralist drift, as general explanations, though neutralism has had remarkable recent successes. In essence, the neo-Darwinian expects that a given variation is correlated with some environmental variable, and the neutralist expects that it is not. When one such environmental variable is identified with one genetic variant, as malaria has been with sickle-cell trait (abnormal red blood cells) in humans, this may seem a triumph for neo-Darwinism, but a myriad protein polymorphisms remain to comfort the neutralist. The generality of balancing selection as an explanation for genetic variation took a sharp knock when it was found that bacteria, which lack sex and so cannot be heterozygous, are just as polymorphic as sexual organisms. At least the neo-Darwinist seems to have a task, or a research programme: to dissect the environment into factors correlated with variations in organisms. Neutralists have no such programme, and instead neutralism comes into its own in explaining evolution that is invisible to selection by the environment, changes in stretches of DNA which are not translated into the phenotype (see section 5 below).
Natural selection, the idea that organisms are moulded or 'designed' by the environment, is also used as an explanation in quite a different sense, when applied to cows, or swallows, or giraffes' necks. Here, explanations are under virtually no empirical control, because the environmental factors invoked are necessarily in the past. In this mode, explanations take the form of conjectures or narratives, in which organisms are analysed in terms of function, and demonstration that one design is more efficient than another is sufficient to explain its origin by natural selection, or environmental conditions under which particular features could be advantageous are postulated or imagined. Critics argue that such exercises of the imagination demonstrate the emptiness of selection theory. In explaining everything, it explains too much; as an explanation of design in nature it seems hard to distinguish from the all-seeing Creator of Paley and other natural theologians. On this view, natural selection is equated with a vacuous Panglossian optimism — 'all is for the best in this best of all possible worlds' — or with the explanation of the action of opium by its content of 'dormitive principle'. One response to these critics is that they do not touch the status of natural selection: it could still be as effective in the global sense as it is in the experiments of population geneticists. Nevertheless, no one has yet reported the origin of a new species by means of natural selection.
Fig. 2. Diagrammatic model of part of the history of a species, as envisaged by advocates of quantum speciation. The parent species is more or less stable through time, and is continually budding off potential new species in the form of small, inbreeding populations, which may be isolated by geography or by genetic accidents such as chromosome mutations. Most of these incipient species soon become extinct, but a few (one in this example) may succeed.
5. Origin of species
Although this was the title of his book, Darwin scarcely addressed the problem of how new species arise. His mechanism, natural selection, concerns change or transformation of species through time, but this alone will not produce new species; it will merely modify and preserve old ones. New species appear only if the number of species increases, if a species splits into two or more. Darwin proposed what he called the principle of divergence: by analogy with artificial selection, and by appealing to 'many and widely different places in the economy of nature', he argued that the most divergent members of a species tend to be preserved, and gradually diverge into varieties, subspecies, and eventually into distinct species. The kernel of his explanation was gradual adaptive divergence.The neo-Darwinian theory of the origin of species — of speciation — differs from Darwin's chiefly in the emphasis placed on geographical isolation as a necessary precursor of species formation, and in emphasis on genetic isolating mechanisms. Neo-Darwinian speciation theory is by no means monolithic, and there are arguments for speciation without geographical isolation, and without Darwinian gradualism.
That species can arise at one stroke there is no doubt. Doubling of chromosome number in a cell is a fairly common accident: it happens if the chromosomes divide, as before normal cell division, but the cell then fails to divide. Such cells are unable to give rise to normal eggs or sperm, because the chromosomes will associate in fours, not in pairs as is necessary in production of sex cells. But in a hybrid between two species, itself sterile because of incompatibility between the two sets of chromosomes, chromosome doubling will restore fertility (each chromosome now has a partner to pair with), and self-fertilization can then initiate a new species intermediate between the two parents of the hybrid. Many species of plants and a few of animals have evidently appeared in this way, and some have been created or recreated in the laboratory.
Speciation by multiplication of chromosome number (polyploidy) is instantaneous; the new species is descended from a single self-fertilizing individual; accident — macromutation — is the cause; and geographical isolation is unnecessary. In all these four ways, it differs from the classic neo-Darwinian model of speciation, which demands geographical isolation, gradual selective change over long periods of time, and involves populations, not individuals. Between these two extremes, there is a third model of speciation which has received much recent support and interest. This model is inelegantly called 'punctuated equilibrium' — 'quantum speciation' is an alternative. One group with evidence bearing on this model includes geneticists who have found that similar or closely related species usually differ in chromosome arrangement, implying fixation of chromosome mutations during their history. For instance, human chromosomes differ from those of chimpanzees by inversions of parts of nine chromosomes, and fusion of two. Whereas identical point mutations in DNA recur at measurable rates, each chromosome mutation is virtually unique, for each depends on the coincidence of at least two accidental breaks in a chromosome, followed by rearrangement and fusion. Organisms heterozygous for a chromosome mutation (inheriting it only from one parent) usually show reduced fertility, but those inheriting it from both parents (homozygous) are potentially fully fertile. Chromosome mutations can therefore act as genetic isolating mechanisms, favouring mating among carriers of the mutation (hence homozygous offspring) and so initiating speciation. But since chromosome mutations are each virtually unique, the only way homozygotes can be produced is by inbreeding among the offspring of the individual in which the mutation originally occurred. This model — 'chromosomal speciation' — resembles the chromosome doubling mode of speciation in several ways: no geographical isolation is necessary, accident rather than natural selection is the cause, inbreeding among the descendants of a single individual is necessary, and a new species may arise in a few generations, with the establishment of a population homozygous for the mutation.
A second set of arguments in favour of quantum speciation comes from palaeontologists, who generally fail to find evidence of gradual transformation in the fossil record. Instead, fossil species appear suddenly, persist unchanged over more or less long periods, and disappear as abruptly as they came. Data of the same sort was available to Darwin and his geological mentor Charles Lyell (1797–1875). Lyell built a theory on them, of piecemeal creation of species by 'a power above nature'. Darwin, who believed natural selection was the power in question, appealed instead to imperfections in the fossil record, and hoped that future discoveries would show the gradual transitions he expected. Today, some palaeontologists have at last given up that hope and taken the fossils at face value, as a true record of the mode of evolution. They infer that speciation occurs rapidly, in small populations, so that transitions between species are evanescent. For if change does not occur during the recorded history of species, it must occur elsewhere, in the unrecorded history of small founder populations. From these ideas comes a theory of species continually and randomly throwing off small offshoots, of which most perish but a few succeed as new species (see Fig. 2). As in the chromosomal theory of speciation, natural selection is not responsible for the appearance of these offshoots, but operates at a higher level in selecting among them. In populations, random mutations throw up material on which selection acts; in just the same way, it is argued that random speciation throws up material — species — on which selection acts. Major evolutionary changes are not due to the action of selection on mutations of genes or chromosomes, but of species selection on a mass of species, thrown up by a random process.
A third group of scientists whose ideas fit in here are molecular biologists. Techniques recently developed make it possible to work out the sequence of nucleotides in DNA and in the messenger nucleic acid, RNA. The letters of the genetic code may then be compared with the amino acid sequence in the protein synthesized on the message in the messenger RNA. The genetic code has a lot of redundancy. In eight of the sixteen boxes in Table 1 the third letter of the triplet makes no difference; the amino acid is specified by the first two letters alone. This means that mutations in the third position of many triplets will be 'silent' and will not alter the amino acid. Amino acids specified by triplets with different first (arginine, leucine) or first and second (serine) letters may also sustain silent mutations in the first and second positions. When the RNA and amino acid sequences are compared for human and rabbit beta-haemoglobin genes, for instance, two-thirds of the differences between the RNAs are silent, not reflected in the protein; the same comparison between mouse and rabbit alpha-haemoglobins gives virtually the same proportion of silent and non-silent changes. If all these differences represent mutations fixed during the history of the species compared, two conclusions can be drawn. First, the rate of 'silent evolution' must be greater than the rate of phenotypic change (we come back to this point later); second, the silent mutations must have been fixed by random drift, not by selection, since a change in DNA which is not reflected in the organism cannot be selected. If the majority of inferred change in DNA is due to mutations drifting randomly to fixation, then small populations should be frequent in the history of species, since drift is more effective in small populations, and most effective in inbreeding.
Thus three independent lines of enquiry — chromosomes, fossils, and molecular sequences — converge in a view of speciation which differs profoundly from neo-Darwinian orthodoxy. The nature of this difference or disagreement is interesting. It concerns three main things: time, or number of generations; number of individuals; and natural selection. But these three things, the first two in quantity, are the staples of population genetics. Neo-Darwinism depends on the synthesis of natural selection with experimental and mathematical population genetics, which themselves depend on avoiding random effects by considering many generations and many individuals. Population genetics can explain stability, or variety, or gradual change within species. The thrust of the above ideas on speciation is that those factors may all be irrelevant to the origin of species. In other words, the principles of population genetics are directed at the wrong level in the hierarchy: they explain the behaviour of genes in populations, but are inappropriate when extrapolated as an explanation of the history of life. It is, after all, a fairly fundamental criticism of neo-Darwinism to propose that natural selection is not relevant to the origin of species.
6. Embryology
Before Darwin, the word 'evolution' had a different meaning in biology. It referred to the unfolding of form in the development of the embryo, and in particular to the notion of preformation — that the adult organism is preformed in the fertilized egg. One complaint among Darwin's critics was that he and his followers misused and misappropriated the word 'evolution'. But in the neo-Darwinian theory, something approaching the doctrine of preformation seems to have reappeared. The theory is concerned almost exclusively with genes, but what interests most biologists is organisms and the form of organisms. Neo-Darwinism has, as yet, little to say on how form is generated; it is assumed simply to be programmed in the genes, as if the adult and its development are preformed in the DNA, or the four dimensions of a life history are mapped in the two dimensions of the linear information store.Darwinism and neo-Darwinism are concerned with historical transformation of organisms. Despite more than a century of work in evolutionary biology, it remains true that the only transformations of which we have empirical knowledge are those observed in the life histories of organisms — the acorn into the oak tree, or the egg into the caterpillar, pupa, and then butterfly. It is remarkable that these transformations remain almost as mysterious today as they were in Darwin's time. The orthodox, genetic explanation of the genesis of form invokes control genes, which are thought of as switching on and off one or more structural (protein-specifying) genes. A hierarchy of control genes is envisaged, with 'master genes' which switch on batteries of lower-level control genes. Mutations in such control genes could obviously produce relatively large changes in the adult organism. But these control genes remain theoretical, and theoretical entities can, in theory, accomplish anything. Indeed, an immediate inference from the reductionist viewpoint of randomly mutating genes is that anything is possible, and it is natural selection that sorts out the actual from the possible.
Yet the regularity and uniformity of embryonic development in animals and plants seem to demonstrate that anything is not possible. One symptom of a reaction against the reductionist programme of neo-Darwinism and molecular biology is the insistence of some biologists that the transformations of embryology demand 'something more' than proteins and interactions between proteins: the miracle that demands explanation is not necessarily how the eye (for example) evolved in the ancestors of vertebrates, but how it evolves from nothing in every vertebrate today. The embryologists and others who take this stand have produced no fully coherent alternative to neo-Darwinism, but appeal to non-genic inheritance (through the cytoplasm of the egg, inherited from the mother), to possible neo-Lamarckian modes of change, or to the self-organizing and self-regulating powers of the developing organism. This revolt against the tyranny of the genes may be summed up as an interest in epigenetics (the sequence of events that happens in the embryo after fertilization).
A neo-Darwinian response is that these complaints are nothing new, smack of vitalism, and demonstrate a confusion between proximate causes, the embryologist's proper domain, and ultimate causes, the domain of evolutionary theory. Epigenesists might reply that neo-Darwinian theory offers no better explanation of the self-organizing capacity of the organism than contingency, a lucky chapter of irretrievable historical accidents. Like so much else in neo-Darwinism, the argument degenerates here into incommensurables, the conflict between chance and apparent necessity.
7. Behaviour
One of the modern growth areas within neo-Darwinism is sociobiology. It is concerned with evolutionary interpretation of behaviour, especially in social interactions within species. Darwin initiated work on these lines with explanations, in terms of natural selection, of insect societies (ants, bees) which entail self-sacrifice or altruistic behaviour by individuals or castes. Modern sociobiology is a fusion of population genetics, ethology (study of behaviour), and game theory (see Von Neumann, John), interpreting behaviour in terms of strategies which will have selective advantage if they increase the chance of survival of individuals or their kin, who share their genes. Though relatively uncontroversial when applied to birds or butterflies, these ideas have raised a storm when, as was inevitable, they are extrapolated to human societies. The controversy, which has marked political overtones, centres over the extent to which behaviour is genetically determined, a topic on which agreement seems no more likely than in the argument over the relative contributions of heredity and environment to intelligence.In order to be subject to natural selection, traits — whether behavioural or structural — must be heritable, or under genetic control. To bring some behavioural trait within neo-Darwinism, all that is necessary is to argue that it is adaptive under certain circumstances (few traits resist ingenuity), and to postulate a gene for it. There is little harm in theoretical genes for altruism or homosexuality, but critics of sociobiology see such exercises as examples of the reductionist tyranny of the genes, or of the emptiness of a theory that can explain anything.
8. Current status of neo-Darwinism
In the preceding sections, various new directions in, and criticisms of, neo-Darwinism have been outlined. The basics of neo-Darwinism, as of Darwinism, are heritable variation and natural selection. Molecular and population genetics have supplied much insight into the source of variation and the mechanism of selection. There is no doubt that generation of genetic variation is an intrinsic property of sexually reproducing organisms as we know them. Nor can one doubt that natural selection censors variation, and maintains organisms in harmony with their environments. That might seem to be the end of the matter. Nevertheless, there have been persistent arguments over the status of neo-Darwinism, both as legitimate science and as a comprehensive explanation of life.In considering these arguments, it is essential to distinguish the two disparate aspects of neo-Darwinian theory. The first is that evolution has occurred: relations between species are material, not immaterial, and historical, due to common ancestry. This aspect is more general than Darwinism or neo-Darwinism. The particular contribution of those systems is a mechanism — natural selection — to account for evolution.
The general aspect of the theory, that evolution has occurred, has a curious philosophical corollary: acceptance of it entails a revolutionary change in the ontological status of its subject matter. Concepts such as 'ravens' and 'swans' are traditionally considered classes or universals ('all swans are white' and 'all ravens are black' are classic statements about universals in logic). Evolution changes that. Biological species, and groups of species related in a closed system of descent, become individuals, and names like Homo sapiens or Mammalia become proper names, the names of individuals, or 'chunks of the genealogical nexus', a nexus that is necessarily unique. If evolution has occurred, it follows that there can be no real laws of evolution, for laws concern classes or universals, and evolution implies that all of life is one individual system, like the solar system. In other words, one consequence of evolutionary theory is that comparative biology becomes historical science, and there can no more be evolutionary laws than there are historical laws.
The general scientific method is to propose hypotheses about nature, and to test them by deducing what each hypothesis forbids, turning to nature to see if the forbidden consequence actually occurs. Then the question is what forbidden consequences, or potential tests, may be deduced from the proposition that evolution has occurred. Various possible observations have been suggested, such as the discovery of human fossils in Cambrian rocks, which could falsify evolution and so demonstrate its testability. But none of the potential tests that have been proposed appears to be directed exclusively at evolution. For instance, Cambrian human bones would falsify not evolutionary theory but the general comparative method in biology, codified before Darwin as the theory of 'threefold parallelism' — between the succession of fossils in the stratigraphic record, the transformations of embryonic development, and the hierarchy of natural classification, based on homologies. Evolution may seem to be an inescapable deduction from that theory, but the biologists who proposed and upheld it in the 19th century would not agree. Those who argue that evolution is falsifiable seem to demonstrate the weakness of that position by the conviction, which shines through their prose, that they know the truth. The only thing forbidden by evolution is that there should be species unrelated to others by descent, and since we have no access to unambiguous records of descent, no decisive test is forthcoming. Nevertheless, evolution explains important discoveries unforeseen by Darwin, such as the universality of the genetic code, and comparisons of nucleic acid and protein sequences.
One may reject testability as a criterion, and appeal instead to the fruitfulness of evolutionary theory as a research programme. That criterion has no necessary connection with truth, for phlogiston chemistry and Ptolemaic astronomy were fruitful research programmes. Or one can point to the explanatory power of the theory, the quantity of disparate areas of knowledge it unites. Darwin used this argument: 'It can hardly be supposed that a false theory would explain in so satisfactory a manner ... several large classes of facts.' A different strategy for evolutionists is not to analyse or defend their own theory, but to attack the opposition, alternative explanations of life. Only two seem to be on offer. The first is spontaneous generation or creation, necessarily on multiple occasions and locations rather than the one required by evolution. The second is the notion that the earth has been seeded from space, and again seeding on multiple occasions is required to distinguish this from evolution. The difference between the three explanations, when considered in terms of testable predictions, explanatory power, fruitfulness, or whatever, seems to come down to the principle of parsimony, or Occam's razor: evolution is superior because it does not multiply entities unnecessarily. For creation and seeding both imply repeated intervention (intelligent or not), whereas evolution implies no more than one such intervention. When treated as a criterion of truth, this has been called the 'best-in-field fallacy'. Fallacious it may be, yet evolution has no real competitors. (See falsification.)
Turning to mechanism, the specifically Darwinian aspect of neo-Darwinism, we are on surer ground with natural selection. Darwin offered several potential tests of the mechanism: for example, 'If it could be proved that any part of the structure of any one species had been formed for the exclusive good of another species, it would annihilate my theory.' But he also wrote: 'I am convinced that natural selection has been the main but not the exclusive means of modification.' That is also the stance of modern neo-Darwinians, who allow genetic drift and other non-Darwinian means of modification such as polyploidy and less radical chromosomal saltations. In these terms, natural selection is clearly beyond criticism, for it is expected to explain only that which is explicable by it. Cases which are beyond the most ingenious selective explanation have another cause at hand, in random effects due to drift or inbreeding. Comparison of DNA sequences indicates that 'silent' substitutions are about four times as frequent as non-silent ones (with an effect on the phenotype): in other words, the majority of evolutionary change is immune to natural selection. This important conclusion is summed up in 'Kimura's rule' (Motoo Kimura is the father of the neutral theory of molecular evolution): molecular changes that are less likely to be subject to natural selection occur more rapidly in evolution. One corollary of this is the 'molecular clock', the inference that in each lineage DNA evolves at a roughly constant rate. This contrasts with evolution of form or phenotype, which seems to occur in jerks in some lineages, hardly at all in others (so-called 'living fossils'), and never at a constant rate. If, at the molecular level, the majority of evolutionary change takes place in spite of natural selection rather than because of it, natural selection seems to be false as the general explanation of evolutionary change.
The effect of Darwin's theory was to give a mechanistic explanation of apparent intelligent design: the appearance was illusory, for the creative power was in natural selection. The effect of the large neutral or 'non-Darwinian' component discovered in molecular evolution is further to downgrade the role of design: the apparent creative power of natural selection was illusory, if it is outweighed by chance effects. In other words, the role of chance increases. It is this that seems to be the nub of the matter. The reductionist line of neo-Darwinism, aiming to reduce the diversity of life to the laws of chemistry and physics, leads to randomness. Instead of the relentless mill of natural selection, 'daily and hourly scrutinising, throughout the world, the slightest variations ... silently and insensibly working ... at the improvement of each organic being' (Darwin's words), buttressed by the equations of population genetics, modern neo-Darwinians seem to be turning on the one hand to pure randomness (neutral theory and the molecular clock), and on the other to explanations which approach the miraculous (to use a provocative word), unique events perpetuated by inbreeding. The latter has echoes in the biblical account of the genesis of mankind. Natural selection has been called a mechanism for generating the improbable and, if the role of natural selection is diluted, the improbability of the results increases. The ultimate in reduction is a recently fashionable idea, that the whole of life is an accidental excrescence, a by-product of selfish DNA, whose structure is such that it survives, multiplies, and diversifies. That view seems to lead nowhere, but there are two lines of research that might lead out of the impasse. One is into the structure of the genome (the genetic equipment of the individual), where surprising things like 'jumping genes' and spontaneous amplification of bits of DNA are being discovered. Perhaps the permanence we attribute to DNA is also illusory. The other line is in epigenetics, the link between the genes and the organism. It is here that the mystery of transformation can be approached most directly.
(Published 1987)
— Colin Patterson
- Bibliography
- Bowler, P. J. (1984). Evolution: The History of an Idea.
- Dawkins, R. (1996). The Blind Watchmaker.
- Dennett, D. C. (1995). Darwin's Dangerous Idea.
- Gould, S. J. (1977). Ever Since Darwin.
- — — (1980). The Panda's Thumb: More Reflections on Natural History.
- Mayr, E., and Provine, W. B. (eds.) (1980). The Evolutionary Synthesis: Perspectives on the Unification of Biology.
- Ridley, M. (1985). The Problems of Evolution.
- — — (1995). The Red Queen: Sex and the Evolution of Human Nature.
- — — (2003). Nature vs Nurture: Genes, Experience, and What Makes Us Human.
