One of the great virtues of Darwin's On the Origin of Species is that it presents the case for evolution in a simple, clear, and logical way. Darwin tells us in the introduction what he intends to do, and he does it. He carefully builds up the case for the existence of inherited variations, explains why there is competition between individuals, and describes how the struggle for existence can lead to a slow accumulation of small, favorable modifications that we eventually see as adaptations. As he writes in his final chapter, “this whole volume is one long argument.”

In the 150 years since The Origin was first published, Darwin's argument has been challenged, supplemented, and updated many times. The first major update was the outcome of the early 20th-century advent of Mendelian genetics, which provided a mechanism of heredity that was quite unknown to Darwin. Initially genetics muddied ideas about evolution, but in the 1920s and 1930s, mathematical reasoning was used to show how mutation and selection of genes in populations could explain the Darwinian evolution of phenotypic traits. Population genetics became the basis of the “Modern Synthesis”—the gene-centered view of evolution that solidified in the 1940s and dominated evolutionary thinking for the following half-century. Now, in the light of the discoveries of molecular biology, this version of Darwinism is being reexamined and reassessed. Knowing something about what goes on between genes and the phenotypic traits they influence means that developmental processes can take their rightful place in evolutionary thinking. But how should they be incorporated? What difference, if any, do they make to our basic ideas about evolution?

The Mermaid's Tale: Four Billion Years of Cooperation in the Making of Living Things is one of many recent books to offer a new take on evolution. Its authors, Kenneth Weiss and Anne Buchanan, are researchers in anthropology and biomedical genetics in the Department of Anthropology of Penn State University. According to its first sentence, “this book is a biography of life,” but anyone expecting an historical account of life on Earth over the past four billion years will be disappointed. In their preface the authors explain that this biography is about a set of general principles that they believe can account for the changes in organisms that have occurred over evolutionary time, as well as for what happens within cells, among the cells within an organism, and among individual organisms.

From the outset, Weiss and Buchanan tell us that they are going to stress the principles of “cooperation” (a word that they include in the subtitle) and “chance.” Darwin and Wallace's theory of evolution through natural selection depends on competition, which they say “has been applied too widely and too often in a simplified way to adequately account for the characteristics of life on its other time scales.” Cooperation, rather than competition, is the key to understanding life. In addition, we are told, “The role of chance, usually dismissed as incompatible with adaptive change, has been underestimated and misunderstood.” At this point in the book I began to worry about where Weiss and Buchanan were coming from and where they were going. After all, cooperation has been a topic of discussion among evolutionary biologists from Darwin (1859, chapter 7) onward, and it is fundamental to late 20th-century theorizing about the transitions to new levels of biological complexity (Maynard Smith and Szathmáry 1995). Similarly, Darwin (1859, p. 81) recognized that some variations would have no selective effect and would fluctuate in frequency, and the role of chance was being discussed by orthodox evolutionists even before Sewall Wright's studies of the 1930s (Beatty 1992).

The remaining six of Weiss and Buchanan's eight principles of life—inheritance with memory, modular organization, sequestration, coding, contingency, and adaptability—are presented and discussed briefly in chapter 3, which is followed by two chapters on inheritance and change over time. Here they try to show how their principles apply to genetics. Genetics is seen as a collage of codes. We are told not only about the usual DNA triplet code for polypeptide chains but also about the sequence codes in DNA-protein binding, the codes for amino acid capture (transfer RNAs), and so on. In fact, according to the appendix of The Mermaid's Tale, there are no fewer than 15 different codes in DNA and RNA. So, is this an appropriate way of looking at the relation between genes and phenotypes? It can be argued that because of all the splicing and editing that goes on in the cell, the coding concept is inappropriate and unhelpful even when talking about “the” genetic code (Sarkar 1996). I can see no value in extending it still further. Indeed I feel that talking in these terms fosters the idea that the genome is a set of deterministic instructions, although it is clear from elsewhere in the book that this is certainly not the message the authors want to convey.

The following section of the book is devoted to the time scales of development and ecology. Cells, cellular structures, and cell signaling are described in some detail, and the authors make a gallant attempt to use the framework of their eight principles. For example, they describe mitochondria and chloroplasts as “actively produced, replicably organized, modular, sequestered, interacting, and semi-independent but cooperating units inside the cell.” The chapters about life on the eco-scale cover the immune and nervous systems, where it is easier to apply their principles of coding, modularity, sequestration, contingency, and cooperation.

In the final section of The Mermaid's Tale the authors try to show why and how evolutionary theory has to change. Here they make their “ecodevoevo” synthesis. After a rather labored account of the difficulty of distinguishing between evolutionary change due to chance and that due to natural selection, they suggest that luck plays a big part in deciding which of the many viable combinations of alleles survive. Who would disagree? They look at prezygotic selection, behavioral adaptability, habitat choice, niche selection, and other trendy topics, saying that they are all consistent with their general principles. Indeed they are, but they are also consistent with old-fashioned Darwinian selection theory. Weiss and Buchanan then correctly insist that it is cooperation, rather than competition, that predominates at all levels of organization, but they ignore the evolutionary problems associated with this. Names such as William Donald Hamilton and Robert Axelrod are missing from the book's lengthy reference list.

Although there are good and interesting parts in The Mermaid's Tale, it does not form a very coherent whole: unlike The Origin, it is not “one long argument”. It also contains too many errors. For example, we are told that Caenorhabditis is a flatworm, and that the central dogma refers to the concept that a stretch of DNA equals a protein; a figure illustrating “sequestered organelles in a hypothetical cell” shows a chromosome and chromatin as distinct, separate entities. The book is aimed at a general readership, so references to original sources are generally found only in the few endnotes and in the good, descriptive “suggested reading” list. The latter is marred by citations such as “Lamarck (1984)” or “Darwin (1900),” which are presumably the dates of reprints. Editors should not allow this!

The Mermaid's Tale is not the first book-length attempt at a synthesis of ecology, development, and evolution. The benchmark for many years to come will probably be Gilbert and Eppel (2009). According to Weiss and Buchanan, the mermaids of their title are impossible creatures because they are assembled from parts coming from different sequestered branches of developmental and evolutionary trees. In contrast, Gilbert and Eppel suggest that probably all development is codevelopment-that every individual develops as a community made up of organisms from different evolutionary lineages. This does not produce mermaids, but recognizing this type of cooperation really does change how we have to think about evolution.