Biological Emergences: Evolution by Natural Experiment (Vienna Series in Theoretical Biology) 0262182572, 9780262182577, 9781429460958

Citation preview

The Vienna Series in Theoretical Biology

Biological Emergences Evolution by Natural Experiment Robert G. B. Reid

Biological Emergences

The Vienna Series in Theoretical Biology Gerd B. Müller, Günter P. Wagner, and Werner Callebaut, editors The Evolution of Cognition, edited by Cecilia Heyes and Ludwig Huber, 2000 Origination of Organismal Form: Beyond the Gene in Developmental and Evolutionary Biology, edited by Gerd B. Müller and Stuart A. Newman, 2003 Environment, Development, and Evolution: Toward a Synthesis, edited by Brian K. Hall, Roy D. Pearson, and Gerd B. Müller, 2004 Evolution of Communication Systems: A Comparative Approach, edited by D. Kimbrough Oller and Ulrike Griebel, 2004 Modularity: Understanding the Development and Evolution of Natural Complex Systems, edited by Werner Callebaut and Diego Rasskin-Gutman, 2004 Compositional Evolution: The Impact of Sex, Symbiosis, and Modularity on the Gradualist Framework of Evolution, Richard A. Watson, 2005 Biological Emergences: Evolution by Natural Experiment, Robert G. B. Reid, 2007

Biological Emergences Evolution by Natural Experiment

Robert G. B. Reid

A Bradford Book The MIT Press Cambridge, Massachusetts London, England

© 2007 Massachusetts Institute of Technology All rights reserved. No part of this book may be reproduced in any form by any electronic or mechanical means (including photocopying, recording, or information storage and retrieval) without permission in writing from the publisher. For information on quantity discounts, email [email protected]. Set in Stone by The MIT Press. Printed and bound in the United States of America. Library of Congress Cataloging-in-Publication Data Reid, Robert G. B., 1939– Biological emergences : evolution by natural experiment / Robert G.B. Reid. p. cm. — (Vienna series in theoretical biology) Includes bibliographical references (p. ). ISBN-13: 978-0-262-18257-7 (alk. paper) ISBN-10: 0-262-18257-2 (hardcover : alk. paper) 1. Evolution (Biology) 2. Variation (Biology) I. Title. QH366.2.R45 2007 576.8—dc22 2006049132

10

9

8

7

6

5

4

3

2

1

To Clio.

This is the age of the evolution of Evolution. All thoughts that the Evolutionist works with, all theories and generalizations, have themselves evolved and are now being evolved. Even were his theory perfected, its first lesson would be that it was itself but a phase of the Evolution of other opinion, no more fixed than a species, no more final than the theory which it displaced. —Henry Drummond, 1883

Contents

Series Foreword Preface xiii

xi

Introduction: The Re-invention of Natural Selection 1 Paradigm Drift

27

2 Prologue to Emergence

67

3 Evolution by Association

95

4 The Physiological Arena

137

5 Development and Evolution 6 Epigenetic Mechanisms 7 Orthogenesis

179

223

267

8 The Re-invention of Emergence 9 From the Particular to the General 10 An Emergence Theory

363

11 A Biological Synthesis

401

Notes 437 Bibliography Index 505

465

289 329

1

Series Foreword

Biology is becoming the leading science in this century. As in all other sciences, progress in biology depends on interactions between empirical research, theory building, and modeling. But whereas the techniques and methods of descriptive and experimental biology have evolved dramatically in recent years, generating a flood of highly detailed empirical data, the integration of these results into useful theoretical frameworks has lagged behind. Driven largely by pragmatic and technical considerations, research in biology continues to be less guided by theory than seems indicated. By promoting the formulation and discussion of new theoretical concepts in the biosciences, this series intends to help fill the gaps in our understanding of some of the major open questions of biology, such as the origin and organization of organismal form, the relationship between development and evolution, and the biological bases of cognition and mind. Theoretical biology has important roots in the experimental biology movement of early-twentieth-century Vienna. Paul Weiss and Ludwig von Bertalanffy were among the first to use the term theoretical biology in a modern scientific context. In their understanding the subject was not limited to mathematical formalization, as is often the case today, but extended to the conceptual problems and foundations of biology. It is this commitment to a comprehensive, cross-disciplinary integration of theoretical concepts that the present series intends to emphasize. Today theoretical biology has genetic, developmental, and evolutionary components, the central connective themes in modern biology, but also includes relevant aspects of computational biology, semiotics, and cognition research, and extends to the naturalistic philosophy of sciences. The “Vienna Series” grew out of theory-oriented workshops organized by the Konrad Lorenz Institute for Evolution and Cognition Research (KLI), an international center for advanced study closely associated with the University of Vienna. The KLI fosters research projects, workshops, archives, book projects, and the journal Biological Theory, all devoted to aspects of theoretical biology, with an emphasis on

xii

Series Foreword

integrating the developmental, evolutionary, and cognitive sciences. The series editors welcome suggestions for book projects in these fields. Gerd B. Müller, University of Vienna and KLI Günter P. Wagner, Yale University and KLI Werner Callebaut, Hasselt University and KLI

Preface

Charles Darwin described The Origin of Species as “one long argument” for evolution by natural selection. Subsequently Ernst Mayr applied the expression to the continuing debate over Darwin’s ideas. My explanation of why the debate lingers is that although Darwin was right about the reality of evolution, his causal theory was fundamentally wrong, and its errors have been compounded by neo-Darwinism. In 1985 my book Evolutionary Theory: The Unfinished Synthesis was published. In it I discussed Darwinian problems that have never been solved, and the difficulties suffered historically by holistic approaches to evolutionary theory. The most important of these holistic treatments was “emergent evolution,” which enjoyed a brief moment of popularity about 80 years ago before being eclipsed when natural selection was mathematically formalized by theoretical population geneticists. I saw that the concept of biological emergence could provide a matrix for a reconstructed evolutionary theory that might displace selectionism. At that time, I naively thought that there was a momentum in favor of such a revision, and that there were enough open-minded, structuralistic evolutionists to displace the selectionist paradigm within a decade or so. Faint hope! Instead, the conventional “Modern Synthesis” produced extremer forms of selectionism. Although some theoreticians were dealing effectively with parts of the problem, I decided I should try again, from a more general biological perspective. This book is the result. The main thrust of the book is an exploration of evolutionary innovation, after a critique of selectionism as a mechanistic explanation of evolution. Yet it is impossible to ignore the fact that the major periods of biological history were dominated by dynamic equilibria where selection theory does apply. But emergentism and selectionism cannot be synthesized within an evolutionary theory. A “biological synthesis” is necessary to contain the history of life. I hope that selectionists who feel that I have defiled their discipline might find some comfort in knowing that their calculations and predictions are relevant for most of the 3.5 billion years that living organisms have inhabited the Earth, and that they forgive me for arguing that those calculations and predictions have little to do with evolution.

xiv

Preface

Evolution is about change, especially complexifying change, not stasis. There are ways in which novel organisms can emerge with properties that are not only selfsufficient but more than enough to ensure their status as the founders of kingdoms, phyla, or orders. And they have enough generative potential to allow them to diversify into a multiplicity of new families, genera, and species. Some of these innovations are all-or-none saltations. Some of them emerge at thresholds in lines of gradual and continuous evolutionary change. Some of them are largely autonomous, coming from within the organism; some are largely imposed by the environment. Their adaptiveness comes with their generation, and their adaptability may guarantee success regardless of circumstances. Thus, the filtering, sorting, or eliminating functions of natural selection are theoretically redundant. Therefore, evolutionary theory should focus on the natural, experimental generation of evolutionary changes, and should ask how they lead to greater complexity of living organisms. Such progressive innovations are often sudden, and have new properties arising from new internal and external relationships. They are emergent. In this book I place such evolutionary changes in causal arenas that I liken to a three-ring circus. For the sake of bringing order to many causes, I deal with the rings one at a time, while noting that the performances in each ring interact with each other in crucial ways. One ring contains symbioses and other kinds of biological association. In another, physiology and behavior perform. The third ring contains of developmental or epigenetic evolution. After exploring the generative causes of evolution, I devote several chapters to subtheories that might arise from them, and consider how they might be integrated into a thesis of emergent evolution. In the last chapter I propose a biological synthesis. In the bibliographical introduction to the reference section of this book I acknowledge authors who inspired me to return to the fray. Here I acknowledge family, friends, and colleagues who helped and encouraged me. First and foremost is my daughter Clio, who was my front-line editor during the development of this work. Once I had produced the first draft, my personal readers were, in order of recruitment, Clio (now studying parrot behavior as a graduate student at Victoria University of Wellington, New Zealand); the zoologist Louise Russert-Kraemer of the University of Arkansas; the social psychologist Bill Livant, formerly of the University of Regina; and the biochemist Rodney Roche of the University of Calgary. Particular thanks go to Rodney, who read two subsequent versions of the manuscript and provided much help with references. I am also grateful to the molecular biologist Kevin Little, now at the University of Auckland, who was very helpful with the epigenetics chapter. The microbiologist Lee Haines was an enthusiastic informant on insect endosymbionts. The historian of medical biology Judith Friedmann introduced me to the subject of anticipation diseases, as well as making helpful general comments along the way. Kathy Wynne-Edwards gave me the full story of the Siberian hamster, an iconic illustration

Preface

xv

of how physiological evolution can work. Elisabeth Vrba had a catalytic influence on the progress of this work’s publication, as well as providing me with many of her important publications. Some of my students were directly involved in my study of evolutionary theory. At the risk of alienating the many, I would mention a few of the more recent ones: Camilla Berry, Carol Hartwig, Ben Geselbracht, Kevin Peterson, John Simaika, and Will Duguid. My regular evolutionary sparring partners are Richard Ring, Bill Livant, and Tom Reimchen, and my cheering section includes Dawna Brand, Eugene Balon, Roy Pearson, Renée Hetherington, and Gizelle Rhyon-Berry. Members of the Department of Biology at the University of Victoria, including Louise Page, Tom Reimchen, Richard Ring, George Mackie, John Taylor, Gerry Allen, and Nancy Sherwood, willingly provided reference material along with librarian Kathleen Matthews. Once the manuscript was ready to submit for publication, Gerd Müller and Werner Callebaut, editors of the Vienna Series in Theoretical Biology, responded very enthusiastically. They continued to give me solid encouragement and support through the editing and revision stages. I am also grateful for the hospitality of the Konrad Lorenz Institute on two occasions during the development of this work.

Biological Emergences

Introduction The Re-invention of Natural Selection

I regard it as unfortunate that the theory of natural selection was first developed as an explanation for evolutionary change. It is much more important as an explanation for the maintenance of adaptation. —George Williams, 19661 Natural selection cannot explain the origin of new variants and adaptations, only their spread. —John Endler, 19862 We could, if we wished, simply replace the term natural selection with dynamic stabilization. . . . —Brian Goodwin, 19943 Nobody is going to re-invent natural selection. . . . —Nigel Hawkes, 19974

Ever since Charles Darwin published The Origin of Species, it has been widely believed that natural selection is the primary cause of evolution. However, while George Williams and John Endler take the trouble to distinguish between the causes of variation and what natural selection does with them; the latter is what matters to them. In contrast, Brian Goodwin does not regard natural selection as a major evolutionary force, but as a process that results in stable organisms, populations, and ecosystems. He would prefer to understand how evolutionary novelties are generated, a question that frustrated Darwin for all of his career. During the twentieth century, Darwin’s followers eventually learned how chromosomal recombination and gene mutation could provide variation as fuel for natural selection. They also re-invented Darwinian evolutionary theory as neoDarwinism by formalizing natural selection mathematically. Then they redefined it as differential survival and reproduction, which entrenched it as the universal cause of evolution. Nigel Hawkes’s remark that natural selection cannot be re-invented demonstrates its continued perception as an incorruptible principle. But is it even a minor cause of evolution?

2

Introduction

Natural selection supposedly builds order from purely random accidents of nature by preserving the fit and discarding the unfit. On the face of it, that makes more than enough sense to justify its importance. Additionally, it avoids any suggestion that a supernatural creative hand has ever been at work. But it need not be the only mechanistic option. And the current concept of natural selection, which already has a history of re-invention, is not immune to further change. Indeed, if its present interpretation as the fundamental mechanism of evolution were successfully challenged, some of the controversies now swirling around the modern paradigm might be resolved. A Paradigm in Crisis? Just what is the evolutionary paradigm that might be in crisis? It is sometimes called “the Modern Synthesis.” Fundamentally it comes down to a body of knowledge, interpretation, supposition, and extrapolation, integrated with the belief that natural selection is the all-sufficient cause of evolution—if it is assumed that variation is caused by gene mutations. The paradigm has built a strong relationship between ecology and evolution, and has stimulated a huge amount of research into population biology. It has also been the perennial survivor of crises that have ebbed and flowed in the tide of evolutionary ideas. Yet signs of discord are visible in the strong polarization of those who see the whole organism as a necessary component of evolution and those who want to reduce all of biology to the genes. Since neo-Darwinists are also hypersensitive to creationism, they treat any criticism of the current paradigm as a breach of the scientific worldview that will admit the fundamentalist hordes. Consequently, questions about how selection theory can claim to be the all-sufficient explanation of evolution go unanswered or ignored. Could most gene mutations be neutral, essentially invisible to natural selection, their distribution simply adrift? Did evolution follow a pattern of punctuated equilibrium, with sudden changes separated by long periods of stasis? Were all evolutionary innovations gene-determined? Are they all adaptive? Is complexity built by the accumulation of minor, selectively advantageous mutations? Are variations completely random, or can they be directed in some way? Is the generation of novelty not more important than its subsequent selection? Long before Darwin, hunters, farmers, and naturalists were familiar with the process that he came to call “natural selection.” And they had not always associated it with evolution. It is recognized in the Bible, a Special Creation text. Lamarck had thought that evolution resulted from a universal progressive force of nature, not from natural selection. Organisms responded to adaptational needs demanded by their environments. The concept of adaptation led Lamarck’s rival, Georges Cuvier, to argue the opposite. If existing organisms were already perfectly adapted, change would be detrimental, and evolution impossible. Nevertheless, Cuvier knew that biogeography

The Re-invention of Natural Selection

3

and the fossil record had been radically altered by natural catastrophes. These Darwin treated as minor aberrations during the long history of Earth. He wanted biological and geographical change to be gradual, so that natural selection would have time to make appropriate improvements. The process of re-inventing the events themselves to fit the putative mechanism of change was now under way. Gradualism had already been brought to the fore when geologists realized that what was first interpreted as the effects of the sudden Biblical flood was instead the result of prolonged glaciation. Therefore, Darwin readily fell in with Charles Lyell’s belief that geological change had been uniformly slow. Now, more than a century later, catastrophism has been resurrected by confirmation of the K-T (Cretaceous-Tertiary) bolide impact that ended the Cretaceous and the dinosaurs. Such disasters are also linked to such putative events as the Cambrian “Big Bang of Biology,” when all of the major animal phyla seem to have appeared almost simultaneously.5 The luck of the draw has returned to evolutionary theory. Being in the right place at the right time during a cataclysm might have been the most important condition of survival and subsequent evolution. Beyond the fringe of Darwinism, there are heretics who believe the neo-Lamarckist tenet that the environment directly shapes the organism in a way that can be passed on from one generation to the next. They argue that changes imposed by the environment, and by the behavior of the organism, are causally prior to natural selection. Nor is neo-Lamarckism the only alternative. Some evolutionary biologists, for example, think that the establishment of unique symbioses between different organisms constituted major evolutionary novelties. Developmental evolutionists are reviewing the concept that evolution was not gradual but saltatory (i.e., advancing in leaps to greater complexity). However, while they emphasize the generation of evolutionary novelty, they accommodate natural selection as the complementary and essential causal mechanism. Notes on isms Before proceeding further, I want to explain how I arbitrarily, but I hope consistently, use the names that refer to evolutionary movements and their originators. “Darwinian” and “Lamarckian” refer to any idea or interpretation that Darwin and Lamarck originated or strongly adhered to. Darwinism is the paradigm that rose from Darwinian concepts, and Lamarckism is the movement that followed Lamarck. They therefore include ideas that Darwin and Lamarck may not have thought of nor emphasized, but which were inspired by them and consistent with their thinking. Lamarck published La philosophie zoologique in 1809, and Lamarckism lasted for about 80 years until neo-Lamarckism developed. Darwinism occupied the time frame between the publication of The Origin of Species (1859) and the development of neo-

4

Introduction

Darwinism. The latter came in two waves. The first was led by August Weismann, who was out to purify evolutionary theory of Darwinian vacillation. The second wave, which arose in theoretical population genetics in the 1920s, quantified and redefined the basic tenets of Darwinism. Selectionism is the belief that natural selection is the primary cause of evolution. Its influence permeates the Modern Synthesis, which was originally intended to bring together all aspects of biology that bear upon evolution by natural selection. Niles Eldredge (1995) uses the expression “ultra-Darwinian” to signify an extremist position that makes natural selection an active causal evolutionary force. For grammatical consistency, I prefer “ultra-Darwinist,” which was used in the same sense by Pierre-Paul Grassé in 1973.6 The Need for a More Comprehensive Theory I have already hinted that the selectionist paradigm is either insufficient to explain evolution or simply dead wrong. Obviously, I want to find something better. NeoDarwinists themselves concede that while directional selection can cause adaptational change, most natural selection is not innovative. Instead, it establishes equilibrium by removing extreme forms and preserving the status quo. John Endler, the neoDarwinist quoted in one of this chapter’s epigraphs, is in good company when he says that novelty has to appear before natural selection can operate on it. But he is silent on how novelty comes into being, and how it affects the internal organization of the organism—questions much closer to the fundamental process of evolution. He is not being evasive; the issue is just irrelevant to the neo-Darwinist thesis. Darwin knew that nature had to produce variations before natural selection could act, so he eventually co-opted Lamarckian mechanisms to make his theory more comprehensive. The problem had been caught by other evolutionists almost as soon as The Origin of Species was first published. Sir Charles Lyell saw it clearly in 1860, before he even became an evolutionist: If we take the three attributes of the deity of the Hindoo Triad, the Creator, Brahmah, the preserver or sustainer, Vishnu, & the destroyer, Siva, Natural Selection will be a combination of the two last but without the first, or the creative power, we cannot conceive the others having any function.7

Consider also the titles of two books: St. George Jackson Mivart’s On the Genesis of Species (1872) and Edward Cope’s Origin of the Fittest (1887). Their play on Darwin’s title emphasized the need for a complementary theory of how new biological phenomena came into being. Soon, William Bateson’s Materials for the Study of Variation Treated with Especial Regard to Discontinuity in the Origin of Species (1894) was to distinguish between the emergent origin of novel variations and the action of natural selection.

The Re-invention of Natural Selection

5

Figure I.1 The “three-ring circus” of evolutionary causation, under the “big top” of the environment.

The present work resumes the perennial quest for explanations of evolutionary genesis and will demonstrate that the stock answer—point mutations and recombinations of the genes, acted upon by natural selection—does not suffice. There are many circumstances under which novelties emerge, and I allocate them to arenas of evolutionary causation that include association (symbiotic, cellular, sexual, and social), functional biology (physiology and behavior), and development and epigenetics. Think of them as three linked circus rings of evolutionary performance, under the “big top” of the environment. Natural selection is the conservative ringmaster who ensures that tried-and-true traditional acts come on time and again. It is the underlying syndrome that imposes dynamic stability—its hypostasis (a word that has the additional and appropriate meaning of “significant constancy”).8 Selection as Hypostasis The stasis that natural selection enforces is not unchanging inertia. Rather, it is a state of adaptational and neutral flux that involves alterations in the numerical proportions of particular alleles and types of organism, and even minor extinctions. It does not produce major progressive changes in organismal complexity. Instead, it tends to lead to adaptational specialization. Natural selection may not only thwart progress toward greater complexity, it may result in what Darwin called retrogression, whereby complex and adaptable organisms revert to simplified conditions of specialization. This is common among parasites, but not unique to them. For example, our need for ascorbic acid—vitamin C—results from the regression of a synthesis pathway that was functional in our mammalian ancestors. On the positive side, it may be argued that dynamic stability, at any level of organization, ensures that the foundations from which novelties emerge are solid enough to support them on the rare occasions when they escape its hypostasis. A world devoid of the agents of natural selection might be populated with kludges—gimcrack organisms of the kind that might have been designed by Heath Robinson, Rube

6

Introduction

Goldberg, or Tim Burton. The enigmatic “bizarre and dream-like” Hallucigenia of the Burgess Shale springs to mind.9 Even so, if physical and embryonic factors constrain some of the extremest forms before they mature and reproduce, the benefits of natural selection are redundant. Novelty that is first and foremost integrative (i.e., allows the organism to operate better as a whole) has a quality that is resistant to the slings and arrows of selective fortune. Natural selection has to do with relative differences in survival and reproduction and the numerical distribution of existent variations that have already evolved. In this form it requires no serious re-invention. But selectionism goes on to infer that natural selection creates complex novelty by saving adaptive features that can be further built upon. Such qualities need no saving by metaphorical forces. Having the fundamental property of persistence that characterizes life, they can look after themselves. As Ludwig von Bertalanffy remarked in 1967, “favored survival of ‘better’ precursors of life presupposes self-maintaining, complex, open systems which may compete; therefore natural selection cannot account for the origin of those symptoms.”10 These qualities were in the nature of the organisms that first emerged from non-living origins, and they are prior to any action of natural selection. Compared to them, ecological competitiveness is a trivial consequence. But to many neo-Darwinists the only “real” evolution is just that: adaptation—the selection of random genetic changes that better fit the present environment. Adaptation is appealingly simple, and many good little examples crop up all the time. However, adaptation only reinforces the prevailing circumstances, and represents but a fragment of the big picture of evolution. Too often, genetically fixed adaptation is confused with adaptability—the self-modification of an individual organism that allows responsiveness to internal and external change. The logical burden of selectionism is compounded by the universally popular metaphor of selection pressure, which under some conditions of existence is supposed to force appropriate organismic responses to pop out spontaneously. How can a metaphor, however heuristic, be a biological cause? As a metaphor, it is at best is an inductive guide that must be used with caution. Even although metaphors cannot be causes, their persuasive powers have given natural selection and selection pressure perennial dominance of evolutionary theory. It is hard enough to sideline them, so as to get to generative causes, far less to convince anyone that they are obstructive. Darwin went so far as to make this admission: In the literal sense of the word, no doubt, natural selection is a false term. . . . It has been said that I speak of natural selection as an active power or Deity. . . . Everyone knows what is meant and is implied by such metaphorical expressions; and they are almost necessary for brevity. . . . With a little familiarity such superficial objections will be forgotten.11

Alas, in every subsequent generation of evolutionists, familiarity has bred contempt as well as forgetfulness for such “superficial” objections.

The Re-invention of Natural Selection

7

Are All Changes Adaptive? Here is one of my not-so-superficial objections. The persuasiveness of the selection metaphor gets extra clout from its link with the vague but pervasive concept of adaptiveness, which can supposedly be both created and preserved by natural selection. For example, a book review insists that a particular piece of pedagogy be “required reading for non-Darwinist ‘evolutionists’ who are trying to make sense of the world without the relentless imperatives of natural selection and the adaptive trends it produces.”12 Adaptiveness, as a quality of life that is “useful,” or competitively advantageous, can always be applied in ways that seem to make sense. Even where adaptiveness seems absent, there is confidence that adequate research will discover it. If equated with integrativeness, adaptiveness is even a necessity of existence. The other day, one of my students said to me: “If it exists, it must have been selected.” This has a pleasing parsimony and finality, just like “If it exists it must have been created.” But it infers that anything that exists must not only be adaptive but also must owe its existence to natural selection. I responded: “It doesn’t follow that selection caused its existence, and it might be truer to say ‘to be selected it must first exist.’” A more complete answer would have addressed the meaning of existence, but I avoid ontology during my physiology course office hours. “Adaptive,” unassuming and uncontroversial as it seems, has become a “power word” that resists analysis while enforcing acceptance. Some selectionists compound their logical burden by defining adaptiveness in terms of allelic fitness. But there are sexually attractive features that expose their possessors to predation, and there are “Trojan genes” that increase reproductive success but reduce physiological adaptability.13 They may be the fittest in terms of their temporarily dominant numbers, but detrimental in terms of ultimate persistence. It is more logical to start with the qualities of evolutionary changes. They may be detrimental or neutral. They may be generally advantageous (because they confer adaptability), or they may be locally advantageous, depending on ecological circumstances. Natural selection is a consequence of advantageous or “adaptive” qualities. Therefore, examination of the origin and nature of adaptive novelty comes closer to the fundamental evolutionary problem. It is, however, legitimate to add that once the novel adaptive feature comes into being, any variant that is more advantageous than other variants survives differentially—if under competition. Most biologists are Darwinists to that extent, but evolutionary novelty is still missing from the causal equation. Thus, with the reservation that some neutral or redundant qualities often persist in Darwin’s “struggle for existence,” selection theory seems to offer a reasonable way to look at what occurs after novelty has been generated—that is, after evolution has happened. “Oh,” cry my student inquisitors, “but the novelty to which you refer would be meaningless if it were not for correlated and necessary novelties that natural selection

8

Introduction

had already preserved and maintained.” So again I reiterate first principles: Selfsustaining integrity, an ability to reproduce biologically, and hence evolvability were inherent qualities of the first living organisms, and were prior to differential survival and reproduction. They were not, even by the lights of extreme neo-Darwinists, created by natural selection. And their persistence is fundamental to their nature. To call such features adaptive, for the purpose of implying they were caused by natural selection, is sophistry as well as circumlocution. Sadly, many biologists find it persuasive. Ludwig von Bertalanffy (1952) lamented: Like a Tibetan prayer wheel, Selection Theory murmurs untiringly: ‘everything is useful,’ but as to what actually happened and which lines evolution has actually followed, selection theory says nothing, for the evolution is the product of ‘chance,’ and therein obeys no ‘law.’14

In The Variation of Animals in Nature (1936), G. C. Robson and O. W. Richards examined all the major known examples of evolution by natural selection, concluding that none were sufficient to account for any significant taxonomic characters. Despite the subsequent political success of ecological genetics, some adherents to the Modern Synthesis are still puzzled by the fact that the defining characteristics of higher taxa seem to be adaptively neutral. For example, adult echinoderms such as sea urchins are radially symmetrical, i.e., they are round-bodied like sea anemones and jellyfish, and lack a head that might point them in a particular direction. This shape would seem to be less adaptive than the bilateral symmetry of most active marine animals, which are elongated and have heads at the front that seem to know where they want to go. Another puzzler: How is the six-leg body plan of insects, which existed before the acquisition of wings, more or less adaptive than that of eight-legged spiders or tenlegged lobsters? The distinguished neo-Darwinists Dobzhansky, Ayala, Stebbins, and Valentine (1977) write: This view is a radical deviation from the theory that evolutionary changes are governed by natural selection. What is involved here is nothing less than one of the major unresolved problems of evolutionary biology.15

The problem exists only for selectionists, and so they happily settle for the first plausible selection pressure that occurs to them. But it could very well be that insect and echinoderm and jellyfish body plans were simply novel complexities that were consistent with organismal integrity—they worked. There is no logical need for an arbiter to judge them adaptive after the fact. Some innovations result from coincidental interactions between formerly independent systems. Natural selection can take no credit for their origin, their coexistence, or their interaction. And some emergent novelties often involve redundant features that persisted despite the culling hand of nature. Indeed, life depends on redundancy to make evolutionary experiments. Initially selectionism strenuously denies the existence of such events. When faced with the inevitable, it downplays their importance in favor of selective adjustments necessary to make them more

The Re-invention of Natural Selection

9

viable. Behavior is yet another function that emphasizes the importance of the whole organism, in contrast to whole populations. Consistent changes in behavior alter the impact of the environment on the organism, and affect physiology and development. In other words, the actions of plants or animals determine what are useful adaptations and what are not. This cannot even be conceived from the abstract population gene pools that neo-Darwinists emphasize. If some evolutionists find it easier to understand the fate of evolutionary novelty through the circumlocution of metaphorical forces, so be it. But when they invent such creative forces to explain the origin of evolutionary change, they do no better than Special Creationists or the proponents of Intelligent Design. Thus, the latter find selectionists an easy target. Neo-Darwinist explanations, being predictive in demographic terms, are certainly “more scientific” than those of the creationists. But if those explanations are irrelevant to the fundamentals of evolution, their scientific predictiveness is of no account. What we really need to discover is how novelties are generated, how they integrate with what already exists, and how new, more complex whole organisms can be greater than the sums of their parts. Evolutionists who might agree that these are desirable goals are only hindered by cant about the “relentless imperatives of natural selection and the adaptive trends it produces.” A Tradition of Re-invention Natural selection has not always been regarded as an ever-present, omnipotent evolutionary force. With regard to competition, Darwin himself inferred that after novelty emerged there was a lag period before there developed a “Ratio of Increase so high as to lead to a Struggle for Life, and as a consequence to Natural Selection.”16 A change in form and function that allows novel exploitation of an existing environment, or simply immigration into a pristine locality, permits “occupation of empty niche space.” Experimentation with body forms has been described as the exploration of “morphospace.”17 Implicit in these now-conventional concepts are adequate resources, the ability to use them, and the absence of such agents of natural selection as competition and predation. Differential reproduction is inevitable, but without competition it might very well be random, not correlated with Darwinian fitness. A proposal to re-invent natural selection may seem iconoclastically presumptuous, but neo-Darwinists had no qualms about establishing a precedent when they redefined it as differential survival and reproduction. While this addressed the effects of the process, it left putative causal agents such as competition, predation, and the literal choices that are made in reproductive pairing, and co-evolution, to be tacitly implied. Natural selection is not simply the effect of evolutionary change, but a syndrome of secondary causes and effects. As such, it is a real phenomenon, based in some part on the participation of genes, and not to be abandoned for its creaky logic.

10

Introduction

To make matters worse, without a murmur of dissent from the orthodox, evolution itself was re-invented as changes in the distribution of alleles in populations, for the sole purpose of making it match the new definition of natural selection. What cloud of unknowing allowed, and still allows, this to pass without protest? One answer is that Darwin offered belief in natural selection as a mechanistic replacement for belief in Special Creation. And stable belief systems characteristically tailor facts and definitions to suit their acolytes and thus ensure their survival. In Factors of Evolution (1949), Ivan Schmalhausen initiated the kind of reconceptualization of natural selection that I consider necessary. What he calls “stabilizing selection” is still a textbook cause of genetic equilibrium in populations. But most accounts ignore the stabilization of internal organization, both developmental and physiological, which was more important to him. Schmalhausen argued that such adjustments reach an equilibrium that is difficult to escape, a condition most likely to be reached in stable external environments. Yet he could see that the evolutionary complexification of some higher vertebrates had paradoxically been accelerated. For example, among the placental mammals, brain expansion had been faster than any advances in the central nervous systems of the lower vertebrates. In the hominin lineage, intelligence and mind had emerged even more rapidly. Schmalhausen could not fully fathom how stabilizing selection had been thwarted, but he speculated that stressful environments had something to do with it. The problem still confronts those evolutionists who are awake to it. Organisms under prevailing conditions of stable equilibrium are what I and most other biologists investigate. Therefore, it is not surprising that this “normal” biology should be given primary consideration in evolutionary studies, despite contradictions, omissions, and paradoxes. Common sense also tells us that there is plenty of normal biology to keep us grownups busy, so we should get on with it, stop asking sophomoric questions, and quit worrying about the adequacy of the general theory. We can depend on the “high table”18 of selectionism to send down theoretical crumbs, predigested for easy absorption. Those crumbs are devoid of essential nutrients. Selection theory actually says nothing about the origin of the qualities that are selected. It only assesses them once they have emerged, and predicts their likely distribution in future generations, provided that the usual stability prevails. Evolutionary progress to organized complexity is considered to be no more than a coincidental by-product of cumulative adaptations, which very occasionally succeeds at beating the competition. Is there not something distinctly unnatural about an evolutionary vision of nature that cannot explain how innovation arises, does not see the need, and instead looks elsewhere to observe how its consequences are played out? Re-invention is therefore a weak response to these problems. Replacement is almost a necessity. But there are large obstacles in the way.

The Re-invention of Natural Selection

11

Reductionism Reduction is a good, logical tool for solving organismal problems by going down to their molecular structure, or to physical properties. But reductionism is a philosophical stance that embraces the belief that physical or chemical explanations are somehow superior to biological ones. Molecular biologists are inclined to reduce the complexity of life to its simplest structures, and there abandon the quest. “Selfish genes” in their “gene pools” are taken to be more important than organisms. To compound the confusion, higher emergent functions such as intelligence and conscious altruism are simplistically defined in such a way as to make them apply to the lower levels. This is reminiscent of William Livant’s (1998) “cure for baldness”: You simply shrink the head to the degree necessary for the remaining hair to cover the entire pate—the brain has to be shrunk as well, of course. This “semantic reductionism” is rife in today’s ultra-Darwinism, a shrunken mindset that regards evolution as no more than the differential reproduction of genes. Although reducing wholes to their parts can make them more understandable, fascination with the parts makes it too easy to forget that they are only subunits with no functional independence, whether in or out of the organism. It is their interactions with higher levels of organization that are important. Nevertheless, populations of individuals are commonly reduced to gene pools, meaning the totality of genes of the interbreeding organisms. Originating as a mathematical convenience, the gene pool acquired a life of its own, imbued with a higher reality than the organism. Because genes mutated to form different alleles that could be subjected to natural selection, it was the gene pool of the whole population that evolved.19 This argument was protected by polemic that decried any reference to the whole organism as essentialistic. Then came the notion that genes have a selfish nature. Even later, advances in molecular biology, and propaganda for the human genome project, have allowed the mistaken belief that there must be a gene for everything, and once the genes and their protein products have been identified that’s all we need to know. Instead, the completion of the genome project has clearly informed us that knowing the genes in their entirety tells us little about evolution. Yet biology still inhabits a genocentric universe, and most of its intellectual energy and material resources are sucked in by the black hole of reductionism at its center. The Evolution of Whole Organisms Because reductionism, whether as a simple bias or as an extreme prejudice, is the dominant worldview, less attention is paid to the evolution of whole organisms. Is the elaboration of all their complex structures based on the selective accumulation of simple, randomly occurring adaptations at the gene level? In Darwin’s Black Box

12

Introduction

(1996), Michael Behe shows that molecular complexes often defy reductive analysis. Then, without exploring mechanistic alternatives, he switches to his default cause: a supernatural intelligent designer. This is the kind of argument from ignorance that infests all the variants of creationism. In its place there should be a search for explanations of how complexity can be integrated from multiple parts. Some complicated organs seem to have been generated not only without any metaphorical selection pressure, but even without any identifiable usefulness, far less any purposeful design. For example, some jellyfish have image-forming eyes, but do not have the nervous system needed to process visual images.20 One Darwinist responds: The cubomedusan nervous system may be a ‘black box’ at present, but sooner or later (God willing) it will surely be found to have all the circuitry needed to process the information coming in from those beady little eyes.21

However, a complex system might be elaborated, or simply fall into place like the three-dimensional structure of a protein molecule, before there is a pre-existing need for it, or another complex system to complement it, or a selection pressure to force it. All experimental novelties in nature may emerge into such dubious conditions. Life might be better understood if we took a step away from reductionism to deal directly with the emergence and evolution of organismal wholeness. Most of us would agree that an organism is more than a package of randomized alleles from the gene pool, like a hand of cards from a well-shuffled pack. And it does not fly in the face of convention to acknowledge that organismal uniqueness tends to be preserved from generation to generation, despite genetic assortment, recombinations of parental characters, and gene mutability. DNA analysis of the petrified remains of Cheddar Man shows that some of his genomic integrity remains recognizable in one living descendant, with no other close matches, despite 9,000 years of sexual mixing and mutation.22 The structural dynamics of the genome as a whole are conserved from one generation to the next by the behavior and architecture of chromosomes and meiotic processes. At the species level, various obstacles to hybridization, such as geographical separation, behavioral differences, and mutual sterility, prevent genetic mixing. Within breeding populations there is a high probability that a particularly competitive organismal uniqueness will persist and spread across the generation gap. But in reality the genes themselves, despite rare exceptions in gene-transfer experiments, are selflessly subordinate to the greater integrity of the whole organism.23 I call myself a “realistic holist,” meaning that I appreciate the necessity of understanding the parts in order to understand the whole, without adducing any supernatural properties to the whole. A more neutral label has been popularized by Susan Oyama in The Ontogeny of Information (1985). She uses “interactionist” to describe someone who can work with wholes and parts without bias either way, provided that their connectedness is taken into account. While this might be more

The Re-invention of Natural Selection

13

acceptable to someone who is on the road to recovery from reductionism, I still prefer to advertise my bias toward the whole. To know how its integrity is maintained requires that we understand the quality of adaptability. To understand how its integrity is improved requires that we discover how adaptability evolves. Adaptability Adaptability is an emergent property of biological wholes. At its point of emergence under normal stable conditions, this feature may not be seen to be “for” anything in the short term; i.e., it need not realize any immediate reproductive improvement. But if it had the tenacity to persist until it could break out of the bondage of competition, or evade it by entering an environment hospitable only to its kind of adaptability, it would then be free to diversify. The organism’s ability to sustain its integrity can be increased by evolutionary complexification, whether or not it leads to immediate differential survival and reproduction. Usually, increased complexity amounts to greater physiological and behavioral adaptability, which decreases vulnerability to environmental changes and the agents of natural selection. But this may be of little advantage in comparison with established specialization in a stable environment. Other things being equal, selfmaintenance mechanisms are open to improvement through cumulative adjustments of the new emergent condition that restore a dynamic stability within the organism. But if internal adaptation comes to reinforce the status quo, how do further improvements in adaptability get over that barrier? The simple answer is that they have to jump over it or wait until it comes down spontaneously. Ivan Schmalhausen argued that greater physiological and behavioral adaptability promotes organismal diversification through the exploitation of stressful environments where competition, the main agent of natural selection, is weak or absent. In contrast, although selectionism accepts the importance of unoccupied niches, it proposes that they present a new assortment of strong selection pressures that induce invading organisms to diversify. Although adaptability confers a “blanket utility” responsive to numerous vicissitudes, neo-Darwinists do not recognize it as a necessary qualification for invading new environments.24 They usually ignore it altogether or bury it as an unexamined aspect of adaptation. Even so, however useful the adaptability of emergent novelties might be under the confused conditions of invasion, it may regress into specialization as new stable equilibria shake down. The most likely response of a selectionist to these arguments would be: “Admittedly, there are occasions when adaptability may be important for survival. So that’s what is selected.” Such a statement is true but irrelevant. There are often circumstances in which adaptability is less important for reproductive success than a competitive specialization. But adaptability, as an emergent property of life, was not produced by

14

Introduction

natural selection, and it will not be entirely lost because of low scores in the selection stakes. Moreover, as William Wimsatt (1998) has observed, greater complexity can emerge autonomously from modular systems, and “one does not need special circumstances—or selection—to form self-organizing states or properties: one needs special circumstances to prevent them.”25 It is essential that evolutionary theory discover how complexity, adaptability and wholeness evolve, because they are what led ultimately to the intelligence and freedom of thought that makes such a discovery possible. Generative Hypotheses The major operation of neo-Darwinist research is measuring the fitness of variants in the “real world” of populations in their environments. The generation of novel variants is still regarded with some indifference. Stephen Jay Gould remarks: Speculation about adaptive significance is a favorite and surely entertaining ploy among evolutionary biologists. But the question, “What is it for?” often diverts attention from the more mundane but often more enlightening issue, “How is it built?”26

That leads us in to the even more enlightening question: how did it come to be built? This question is difficult rather than mundane—always a popular reason to ignore a problem. But if you accept evolution as a historical reality, and you are given to asking naive questions, you may want to know how simple forms progressed to become larger, more complex, and better organized. How did wholes as distinctive as horses and tigers and trees evolve? If you are not content with answers that involve their generation through DNA mutations that were marginally fitter than their ancestors’, if coincidental combinations in puddles of genes don’t do it for you, or if you want something more realistic than special creation or intelligent design, you may be interested in my alternative: During the course of progressive evolution, organisms have become more complex and adaptable through a series of quite sudden, novel emergences. They are produced by natural experiments, involving the interplay of environment, organisms, internal milieux, developmental systems, and genomes. Large emergences can be characterized as new wholes that are greater than the sums of their parts, with new properties that did not exist at the lower level of their generative conditions, as well as their old features. Metaphorically, the internal working of wholes may be fine-tuned by natural selection, and fitter organisms will increase in number. But evolution is not caused by natural selection— neither metaphorically nor literally. Indeed, the whole syndrome of natural selection may hinder the success of evolutionary experiments. Furthermore, even where better-adapted variants seem to be directionally selected, they may be generated by autonomous, selfamplifying, genomic mechanisms.

The Re-invention of Natural Selection

15

There are real (if rare) conditions of life in which the agents of natural selection are diminished or absent and nature’s innovative experiments in emergent evolution have an opportunity to thrive and diversify before their own numerical success reintroduces competition. This happens in the wake of catastrophic changes, or when organisms invade pristine environments, and on the rare occasions when novel emergents are so competent that the old competition from pre-existing flora and fauna is irrelevant. There are two major kinds of evolution. Once saltatory emergent progress has provided increases in complexity, self-organization, and adaptability, i.e. progressive evolution, there is greater potential for the second kind: diversifying evolution, which used to be called “adaptive radiation.” Diversifying evolution itself has several major components. One is emergent, producing epigenetic (developmental) changes, giving rise to specialized body forms through allometric shifts—with behavioral consequences. The other is adaptational in the conventional sense. It too has two manifestations: outwardly directed genetic adjustment to the environment, and to habit.27 There is also internal co-adaptational adjustment to physiological and developmental change, which may evolve regressions that are themselves saltatory. Furthermore, if isolated lineages are founded by a small number of pioneers that share an innovation, all of their descendents, including some from back-crosses with their pre-emergent relatives, will possess their emergent properties from the outset. Invoking natural selection as a means of spreading innovation through the population may be superfluous. For those who do not find selectionism sufficient, an examination of the origins of progressive emergent qualities might make a welcome change. Of all the ideas with which I challenge my students, they usually accept this: The causal theory of evolution has to include a hypothesis that suggests how innovation is generated. Darwin and his descendants have not formulated a generative synthesis. Their hypothesis only circumscribes the demographic fate of novelties. Generative hypotheses are needed not only for adaptational mechanisms that fit organisms to particular conditions of life, but also for processes of progressive (i.e., complexifying) innovation. Since these usually have the emergent property of universal advantage through more sustainable integrity or adaptability at their inception, they need no subsequent sponsorship by natural selection. Regardless of present and future circumstances, their persistence is assured, though not necessarily as the fittest types. Here the concept of selection verges on total redundancy. Indeed, if you look at the generation of any novelty, adaptable or adaptational, as the primary process, you can see that, barring accidents, its future depends on its quality. To add the subsequent action of a directing selective process is logically superfluous—if you are dealt a handful of aces, the outcome of the game is certain from the start. Emergent innovation might proffer increased complexity, and multifunctional adaptability. It might involve divergences in behavior and in developmental pathways, novel additions to the life cycle, and allometric shifts in anatomical

16

Introduction

proportions, as occurred in the evolution of a giraffe from a short-necked okapi-like ancestor. Adaptational advantage in the environment into which it is born is another possibility. Or it might fail in any of those offices. If chance is discounted, the fate of a novelty, in terms of its ultimate differential survival and reproduction, is contained in its origin, much as the physical fate of the universe was contained in the initial conditions of the Big Bang.28 Though natural selection qua differential reproduction is always present, it is an effect, not a cause. What difference would it have made to the history of biology if someone before Darwin had come up with a generative theory of evolution, in which the concept of selection was no more than a means of describing, quantifying, and cataloguing the ultimate fate of primary causal events? Obviously not much, since we find something like it in “On Generation,” an essay published by Erasmus Darwin (grandfather of Charles) in his 1794 book Zoonomia. Lamarck also had a generative theory, according to which an inherent progressive drive caused evolution, and adaptation to particular environments through the inheritance of acquired characteristics was a lesser process that sometimes got in the way. Twenty-six years before the publication of Charles Darwin’s major book, Étienne Geoffroy de St. Hilaire tried to explain “the origin of species” using those exact words to define the problem. He concentrated on how environment could generate developmental “monstrosities” that could be advantageous or detrimental. But he took no account of relative competitiveness. Erasmus Darwin did not make his point strongly enough to persuade an unprepared readership. Lamarckism failed because most biologists intuitively realized that universal progress was wrong. Like Lamarck, Geoffroy was a victim of Cuvier’s antievolutionary antagonism. As a result, the baby generative theory was thrown out with the dirty bathwater of progressionism. Lamarck’s theory would have had a better start in life if he had made evolution a series of occasional random events. The neoLamarckists moved in that direction, and they might have pulled off a paradigmatic coup had Darwin not already established a strong alternative worldview, with natural selection the sine qua non of evolution. Innovations, however they occurred, were only raw material to be shaped by his mechanism. What Are the Options? The re-invention of natural selection as an obstacle rather than a cause, or simply as a consequence of the generation of novelty, would leave the neo-Darwinist Modern Synthesis of evolution with little foundation. Would it need renovation, or a new structure? Postmodernists want to abandon the ruins as a memorial to authoritarianism, and substitute a tent village of refugee concepts for the old monolithic paradigm. I prefer something more constructive. A dialectical approach to evolutionary causation might take natural selection as the thesis. Something else, such as the epigenetic origin of novel complexity, might be the

The Re-invention of Natural Selection

17

antithesis. Several “evo-devos” (developmental evolutionists) are hoping for such a new synthesis. But it would be too narrow, ignoring physiological and behavioral adaptability and the major emergences that resulted from symbiosis. Some theoreticians relegate natural selection to a secondary though essential process that is the consequence of a primary mechanism of their choosing. Other dissidents try to save their skins by calling on natural selection as a final arbiter that eliminates any organism suffering from disintegrative change. But this is a spurious role for selection; failure is self-interring, and a funeral undertaker is redundant. To imagine that these compromises provide any freedom of thought from neoDarwinism is a delusion that recalls the Buddhist parable about Monkey. Trying to escape the hand of Buddha (Darwin being cast here in Buddha’s role), Monkey ran and ran to the ends of the Earth, where he found an enormous pillar at the brink of the bottomless abyss. Thinking that Buddha could not rule from so far away, and that he might rule there himself, he carved his name on the pillar and returned to claim success. Buddha then displayed the finger with “Monkey” inscribed on it. To effect his escape, Evolutionary Monkey must dwell at the edge of the abyss, and contemplate its traverse, without constantly running back to Darwin for approval. Progressive Evolution If natural selection is taken to be an obstacle to progressive evolution (in the sense of increased self-organizing complexity), a start from scratch, beyond the mortemain of Darwin, is demanded. When asked “If you are going to reject natural selection as the cause of evolution, what are you going to put in its place?” I am tempted to answer that selection theory has never offered a logical generative theory of progressive evolution, so there is nothing that requires replacement. But to be less of a dog in the manger, I suggest that what causes innovative variation causes evolution, and that these phenomena, though sometimes elusive, are natural events that happen within the organism as well as without. The word “progress” has two strikes against it, even in modern discussions of evolution. Strike one stems from Lamarck’s belief that evolution was progressionistic, meaning that an inherent, inexorable trend continuously made organisms more complex. This was immediately tainted by his adduction of spontaneous generation to explain how a multitude of organisms had reached lower states of complexity than others. Lamarck was forced to conclude that their ancestors of simple living organisms must have spontaneously originated later than those of complex organisms. Lamarckist implications about progress still arouse suspicions that that evolutionary form is deterministic—i.e., that there is some kind of trend that has been bound to result in self-awareness in one lineage or another. But its determinism-quotient diminishes with the realization that progress is not inexorable but random, episodic, terminable, and indeed reversible. Strike two against the word “progress” is its sense of

18

Introduction

human betterment. Evolution was long identified with human progress in the industrial and economic realms, with division of labor, and with increasing social organization. This notion was strongly held by Erasmus Darwin. His grandson Charles believed that evolutionary progress, like economic progress, required that the competitively superior had to prevail and the inferior had to die out. The evolution of something “better” clashes with modern biologists’ self-proclaimed objectivity. And they regard progressive evolution as a reification of woolly thinking about inherent trends. However, there is nothing mysterious or even disputatious about a process of progressive evolution if it is taken to be the equivalent of increasing complexity of self-organization. Progressive evolution arises from the fundamental replicative quality of life. Where there is reproduction, it is inevitable that organisms, and the building blocks that make them up, must multiply. Where there is multiplication, differentiation is inevitable in some cases, since a replication mechanism that is flexible enough to operate in a living organism must also be mutable. A biological system consisting of multiple, differentiated modules has the capacity to integrate innovations, if it has hierarchical regulation. Since regulatory processes themselves, such as modifier gene action, and hormonal and nervous mechanisms, can also multiply and differentiate, they can keep pace with the other complexities. If selforganization is an idea that bothers you, just remember that you began as a single cell. You had more than a little help from your mother, but the rest was by your own selforganization. The genes were some of the parts that came with your assembly kit, but were not the instruction manual. Emergent wholes can become more complex simply by having a greater variety of parts. In some, the universal building blocks will fall into new structural and functional configurations that exhibit greater order, which implies greater energetic economy and which results in prolonged persistence of the system. For example, a two-dimensional sheet of cells can spontaneously form a more stable threedimensional sphere under turbulent conditions. Such spontaneous complexification of the organism has the potential to improve the maintenance of integrity through adaptability. The result is wider scope for experiments in diversification. They may work, or they may fail. If they work they may wear out, but survival of biological experiments is furthered by making new copies of complete, functioning organisms. Thus, the unique and fundamental emergent property of life is persistence of its autonomous, complex patterns through adaptability and reproduction. To conclude that there has been a trend for some lineages (our own, for example) to evolve to greater complexity is a subjective historical interpretation. Complexification certainly occurred, but it is more objective to think in terms of the occasional and somewhat unpredictable realization of potentials under the appropriate generative conditions.

The Re-invention of Natural Selection

19

There is nothing new about these ideas. Hermann J. Muller, one of the first fruit-fly geneticists, proposed in 1926 that since chromosomal material could duplicate, occasionally mutate, and reduplicate in the changed form, evolution would happen automatically. But his audience thought he was joking.29 In the same vein, he later wrote: “If selection could somehow be dispensed with, so that all variants survived and multiplied, the higher forms would nevertheless have arisen.”30 Muller’s “higher forms” had the price of admission to a new environment, not as a specific preadaptation to a specific element, but as pre-existing adaptability. Also, if the newcomer was to diversify morphologically, it needed a plasticity not found in organisms whose development had been rigidly canalized. A generative theory is needed to account for the origin, nature, and evolution of adaptability and plasticity. Selection theory only quantifies their consequences. There is much more to it than the complexification of autonomous selforganization. Occasionally, major emergent biological changes in the organism are caused by the acquisition of genetic material from foreign sources. Among the earliest prokaryotes, gene-swapping across phyletic lines seems to have been the norm. Recent advances in the human genome project have revealed the relatively recent acquisition of bacterial genes. In the case of the endosymbioses that gave rise to the eukaryotic cell, whole new genomes were acquired, organisms and all. In addition, the environment can induce physical, physiological, and behavioral changes in the organism that continue to be felt by subsequent generations before becoming genetically assimilated. Disruption of the environment—local or universal, slow or cataclysmic—is also a major influence on evolution. There is nothing mysterious about these causes either, and the Modern Synthesis of conventional evolutionary theory is attempting to subsume some of them in yet another layer of ad hoc hypotheses. But I am going to rearrange these pieces of the evolutionary puzzle and take more than a few liberties with the current rules. Metaphors In biology, metaphors are commonly used as a substitute for inclusive definitions of complex processes or ideas. Since everybody likes to discuss the causes of evolution in figurative terms like natural selection, selection pressure, and survival of the fittest, it should not be too confusing if I give a little more weight to the common expression “natural experiment.”31 Darwin himself thought in terms of a self-improving industrial workshop.32 But instead of the research and development side, he emphasized the process of competition that decided which of the workshop’s new products were to survive. Therefore, over the long haul I will be more concerned with the primary, generative causes of evolution than with re-inventing Darwin’s metaphors. Natural selection, however defined, would not occur in the absence of natural experiments in variation. On the other hand, natural experiments are conducted independently of natural selection.

20

Introduction

“Breakthrough experiment” extends the metaphor. Any invention can be so distinctively and immediately successful as to be universally accepted and applied. Or it may be a mere curiosity that barely registers until a theoretically or technically useful application is found. Natural experiments, or emergences, may have to wait for the right conditions. At that point their emergent properties become belatedly obvious because their numbers increase. In a few cases, natural breakthrough experiments will simultaneously meet the tests of qualitative novelty and improved integrity, giving them the opportunity to multiply, either in competition with existent types or in a new environment free of competition. The “natural experiment” metaphor is more than mere word play. In Evolutionary Theory: The Unfinished Synthesis (1985), I remarked that the concept could be extended to genetic engineering: if it works in the lab, why not in nature?33 Just about then, a particularly appropriate exemplifying technique was developed. Agrobacterium tumefaciens is able to insert a plasmid into host plant cells that modifies their products in its favor, and this natural, bacterial, genetic engineering mechanism was co-opted by molecular biologists to introduce foreign genes—transgenes—into plants.34 Of course, in a natural experiment earlier in its evolutionary history, Agrobacterium had acquired plasmid genes from a primitive plant host. The natural experiment with plasmids was in place before its consequences in nature or in genetic engineering were played out. These parallels between nature and the genetics laboratory lead logically to contemplation of how common complicated natural experiments might be, how they are field tested, and if and how nature clears the bench to conduct a new series. Related questions include the following: How fast are natural experiments? How long do their field tests take? Is the experimental laboratory in constant or only occasional production? The Rate and the Gait of Evolution On the face of it, evolution has taken more than 3 billion years—long enough to reassure the most conservative gradualists. Indeed, their credulity is tested by such a long passage of time. For example, it is a regular procedure to average to a constant rate the number of gene mutations that have become characteristic of species over geological time. But selectionists admit that the averaged rate may have been too slow to respond to relatively quick changes in selection pressure. That evolution is episodic is not too radical an alternative. Moreover, if molecular processes are essential to evolutionary change, it might even be heuristic to think of the speed of evolution on the same scale as the speed of chemical reactions. Most of the time is spent in stasis, with experimental mutations mostly undone by repair mechanisms or shut out by normalizing selection. But suppose that real innovative evolution is packed into a series of brief spurts—mere moments of biological time. To persist, such experiments need freedom from the major agents of natural selection. Then erratic molecular

The Re-invention of Natural Selection

21

changes can be fixed as fast as they come, until dynamic stability is re-established. Unfortunately, molecular biologists who accept episodic evolution think it is caused by intensification of selection pressure. Those who accept the Big Bang theory believe that all the cosmic criteria that were to determine the universe’s future were established in the first fraction of a second. The fastest physicochemical transitions in complex organic molecules are measured in microseconds. Biochemical reactions establish the ground of evolutionary change in milliseconds; their physiological consequences occur in seconds; their epigenetic results unfold in a few months. The complete experimental reproductive cycle and feedback between environment and organism may take little longer. A geological time scale may be needed only to account for the prolonged obstruction of progressive evolution by stases that are reinforced by natural selection. Is this the principle of punctuated equilibrium reduced to absurdity? Am I looking at reality through the wrong end of the telescope? If that is what you think, remember that punctuated equilibrium has never been exposed to such a philosophical test. Its proponents still talk about punctuations as being rapid on time scales of tens of thousands of years. As to looking through the wrong end of the telescope, the compound microscope was discovered by doing just that, and figuratively speaking we need both kinds of instrument to analyze nature’s evolutionary experiments. There is more to my proposition than a downgraded natural selection and an pumped-up punctuationism. As an alternative to the Modern Synthesis, I suggest that we work toward an Emergence Theory that accounts for generative mechanisms and the emergent qualities of their evolutionary products. Although it exists at this stage of my argument only as a framework proposal, I will put together a compendium of examples of evolutionary emergences with a variety of causes, along with some working hypotheses. Although I have been drawn to emergence theory by research into comparative physiology, functional anatomy, and symbiosis, it is not cut from the whole cloth of my own speculations. Neither the basic concept nor the name is new. The problem of the generation of order from disorder was debated by Aristotle and his contemporaries. Although not evolutionists, they wondered how humans could develop from chaotic vital fluids. Erasmus Darwin, Lamarck, and Geoffroy proffered generative theories, and other biologists have long recognized the need for one. A Generative Emergence Theory In 1923, Conwy Lloyd Morgan published Emergent Evolution. Its implications were appreciated by some of his contemporaries, and for a while it affected evolutionists. But the attention of biology was diverted by the success of neo-Darwinism and subsequently by the reductive analysis of DNA. A vague sense of emergence nevertheless

22

Introduction

continues to permeate evolutionary writing. Selectionists themselves distinguish emergent, key innovations from minor adaptational changes. Complexity theorists attempt to explain and quantify how emergent order is generated in simple systems. A unifying matrix is still needed for all the causal biological components of emergence to complement each other as a new synthetic whole. Conventional wisdom acknowledges that some revolutionary events in biological evolution have potentiated far greater diversification than others. These are the great emergences: rapid, unpredictable, progressive changes that brought life into existence and have on occasion radically altered the course of its history, passing beyond the boundaries of old restraints, and constructing new levels of existence, relatedness, and operation. Such natural breakthrough experiments have not always exploded into diversification. Instead they often have had to wait for existing stases to be disequilibrated. Since they are relatively rare, neo-Darwinism ignores them, assuming them to be fortuitous consequences of the normal activity of natural selection. But they present too many awkward imponderables to conform nicely to selection theory. The word “emergence” implies a process of radical evolutionary change, and it is rich in synonyms that apply to coming out of water onto land, to hatching from an egg, to a secret truth revealed, and to human cultures’ undergoing fundamental change. Because many biologists appreciate the need for explanations of evolutionary emergence, and have an intuitive grasp of the concepts necessary to understand them, it may now be easier to find a receptive audience than in Morgan’s day. John Maynard Smith and Eors Szathmáry have addressed a variety of important emergences in The Major Transitions in Evolution (1995). They enter most of the significant causal arenas, and examine the processes in some detail, despite an explicit selectionist bias: The transitions must be explained in terms of immediate selective advantage to replicators. We are committed to the gene-centred approach outlined by Williams [Adaptation and Natural Selection, 1966] and made still more explicit by Dawkins [The Selfish Gene, 1976].35

Venturing into unexplained phenomena that most of their fellow ultra-Darwinists have ignored as irrelevant to the real stuff of evolution, they break down the major transitions to intermediate steps that have a clear advantage over the more primitive stages. But they do not explain how the steps were generated. Moreover they admit that paradoxical situations must have existed for which it is difficult to imagine what that the selective advantage might have been, and they invoke undefined “special circumstances” to explain those situations away.36 My goal is to expound, not explain away, the special circumstances of emergent evolution. There are two ways to proceed. In Emergence: From Chaos to Order (1998), John Holland takes the low road, from the bottom up. This parsimonious approach to the analysis of emergence starts out from simple mathematical models and extrapolates

The Re-invention of Natural Selection

23

toward higher-order emergences. As an organismal biologist, I take the high road, from the top down.37 I examine familiar structural and functional diversities, not to see “what they are for,” but to see “how they are built” and how they might have originated. To make the numerous examples less confusing, I sort them out into the major arenas of evolutionary performance, and try to find common features of causation that together might make up an Emergence Synthesis. The theory that evolution really happened—the “historical” or “phenomenal” theory—was supported by evidence presented by Darwin in The Origin of Species. It is partially independent of the causal theory, which deals with the mechanism of evolution. But new evidence and new ideas about the saltatory nature of complexification challenge Darwin’s interpretation of the history of life, as well as the causal theory. In the context of emergence they certainly call into question the Darwinian and neo-Darwinist views of evolution as a gradual, “insensible,” continuous, adaptational process. In “Unpredicted factors of evolution: The theory of emergence re-visited” (1985), I suggested that emergentism could provide an epistemological framework for the construction of a comprehensive alternative to the Modern Synthesis. I could have played it safe by leaving natural selection alone and simply presenting emergence as a synthesis of generative principles that are needed to fill the present gap in evolutionary thought. This is what Brian Goodwin asks for in his 1989 essay “Evolution and the generative order.” Then, in How the Leopard Changed Its Spots (1994), Goodwin simply ignores natural selection and begins to expand on generative principles as a Science of Qualities. Whether ignored or taken to be irrelevant to the generation of novelty, the framework of selection theory remains a valid mathematical system for dealing with any demographic consequences of evolutionary change. In addition, quantified fitness gives clues to the biological significance of emergent features. Nevertheless, the need for a generative synthesis becomes more obvious if the role and the limitations of natural selection are reappraised. Historically, selectionism has obstructed the development of generative theories by asserting that natural selection is the allsufficient cause of evolution, and that the Modern Synthesis amounts to an all-sufficient theory of evolution that subsumes generative causation. Therefore, dismantling this barrier is more than an attention-getting defiance of convention. I also wish to make my intentions clear from the outset, instead of sailing under false colors and hoisting the Jolly Roger in the last chapter. But the re-invention of natural selection is an opening salvo, not my overall strategy. A fledgling emergence theory or synthesis would have to try its wings as the kind of unifying compilation that Julian Huxley provided in Evolution: The Modern Synthesis (1942). Although it would fly in the faces of both modernists and postmodernists, it is needed to address the fundamental nature of evolution and its mechanisms. It

24

Introduction

would present evolutionary history as a mixture of rapid saltations (progressive and regressive) and sudden changes that appear at critical points in continuous processes. Some of its elements are broadly predictable; others are contingent on physicochemical and biotic conditions. Natural experiment is “the art of the possible,” and successful emergences, where they have the freedom to do so, result in bursts of diversification, interspersed with long periods of stasis.38 A full-fledged emergence theory would distinguish between adaptability and adaptation, profoundly different features of life that have always been lumped together in neo-Darwinism. Furthermore, speciation, which is pivotal for Darwinian evolution, would be seen primarily as a ratchet pin that prevents evolutionary cogs from slipping. Finally, emergentism could release biology from the genocentric universe in which it is confined and restore it to one in which whole organisms and their interactions are the stars of the biological and ontological fundament. Conclusion Natural selection has already had just about had all the re-invention it needs, even from those who support it as a primary cause of evolution. Its supporters understand that the causes of variation are independent of natural selection, and occasionally they hint that natural selection might even get in the way. George Williams (1966) admits: Mutation is of course, a necessary precondition to continued evolutionary change. So evolution takes place, not so much because of natural selection, but to a large degree, in spite of it.39

What the supporters of natural selection don’t understand is that they have now minimized the importance of natural selection as a causal explanation of evolution. And the next step is to see it as a barrier to the success of nature’s experiments. Once the obscuring cloud of selectionism is removed, evolutionary innovation can more easily be worked into a general thesis. There is plenty of raw material from which a generative thesis can be constructed. It has been accumulating ever since the publication of Origin, despite—and often in open defiance of—selectionism. The essential message to keep in mind during the preliminaries is that we need to pay more attention to evolutionary origins and less to their fate. However, despite everything I have said about the limitations, obstructions, and obfuscations of selection theory, the fact remains that most of the history of life has been dominated by dynamic stabilities “imposed” by the syndrome of effects and agents circumscribed by the term “natural selection.” The most general biological synthesis must take this into account. Therefore, although clearing the decks of obstructive clutter is immediately necessary for positive action, getting to the right destination is the larger goal. Once I have dealt with the causal arenas of progressive

The Re-invention of Natural Selection

25

evolution and assembled an emergence thesis, I will reintegrate that thesis with a selectionist antithesis in chapter 11. Before proceeding to the substance of emergence, I extend my critique of selectionism in chapter 1. Then, in pursuit of evolutionary origins, I advance an argument and evidence in favor of a reincarnated emergentism, beginning with chapter 2 (a primer for several chapters on the causal arenas of emergent evolution) and continuing with three chapters on more theoretical aspects. As well as acknowledging relevant hypotheses and discoveries of modern biologists, I try to rehabilitate “mute, inglorious” historical contributions. Modern biology is often compromised by their omission. Molecular biologists in particular seem to think that the “Big Bang of Biology” came in 1953, with the Watson-Crick model of DNA, and that only post-millennial research now matters.

1 Paradigm Drift

In the past 30 years or so, ultra-Darwinians have reformulated their basic conception of natural selection, changing it from a passive to a directly active agent operating in the natural world, pushing the evolutionary process relentlessly forward. . . . They seek to transform natural selection from a simple form of record keeping, a filter that biases the distribution of genes between populations, to a more dynamic active force that molds and shapes organic form as time goes by. —Niles Eldredge, 19951 If natural selection is the filter, what’s making the coffee? —Camilla Berry, 20002

Like Darwin, many biologists would admit that the creative power of natural selection is metaphorical. Like Eldredge, they might see it only as a passive filter for genomic variations. But Darwin also promoted it as “a power incessantly ready for action”—an instant substitute for the supernatural creator.3 If he had not, his version of evolutionary theory might not have caught on in the first place. But only a fractured logic could simultaneously make a metaphor into a materialistic force, hedge on it as a “false term,” and yet endow it with Biblical creativity: “What limits can be put to this power, acting during long ages and rigidly scrutinizing the whole constitution, structure and habits of each creature,—favouring the good and rejecting the bad?”4 Natural selection has always been a dualistic principle: metaphorically active and mechanistically passive. Is Eldredge right to infer that “the re-invention of Darwin” has involved a qualitative, paradigm shift, to the point where natural selection is perceived to make the coffee as well as filter it? In this chapter I will raise my objections to natural selection in both those roles. I will presume that most of my readers consider that natural selection is the cause of evolution, and that selection theory therefore “makes sense.” What makes it difficult to argue otherwise is that natural selection is not only the foundation of the modern mechanistic theory of evolution, it is also a real phenomenon. As such, it is

28

Chapter 1

an aspect of life, and what it “makes sense” of is not evolution but population dynamics. It is therefore irrelevant as a causal explanation of evolutionary change. I do not reject the real phenomena pertaining to the syndrome of natural selection. They are what dominates life most of the time. The demographic causes and effects addressed by selection theory will be merged with emergence theory to form a new “Biological Synthesis” in chapter 11. Also, be assured that I am aware of the perils of writing about a “new synthesis” or a “new theory” when such labels are commonly stuck on propositions that are neither. Paradigms “Paradigm” is a word that slipped into fashionable parlance before being explicitly understood by all who used it. Many philosophers are still dubious about its meaning and significance. Now favored by advertising copywriters, it teeters on the brink of banality. But the concept of a novel theory that generates a period of new experimentation and theorizing has enduring value for the history of ideas. There are two kinds of paradigm in science. One arises from a significant phenomenon that is either newly discovered or newly recognized as a mechanistic entity, hence constituting a phenomenal paradigm. The second, a causal paradigm, arises from the purported explanation of an accepted phenomenon. It is the one that causes most of the arguments. The discovery of a natural phenomenon, whether it be gravity, electricity, the cell, germs that cause disease, or evolution, has a paradigmatic influence in its early years until its reality is taken for granted. Thereafter it need undergo no further revolutionary reinterpretation unless the original data are found to have been false or misinterpreted. For example, before the inception of Cell Theory it had long been recognized that plant tissues are composed of cells, because their cellulose walls are visible under simple microscopes. Then, in 1839, Schwann proposed that animal tissues were also cellular. This unifying theory of the microanatomy of living organisms took hold as a paradigm that gave direction to the study of embryology, chromosome behavior, biochemistry, and gene theory. Since these could relate cellular structures and functions to whole organisms, cell theory was a successful reduction that influenced biology for the next century. The cellular condition itself became axiomatic and uncontroversial.5 A paradigm need have little foundation in fact to be universally persuasive. Take eighteenth-century Fiber Theory, for instance. As a physiological and medical paradigm, it did not directly address evolution, but it illustrates a significant general principle. Dominating medicine for about a hundred years, Fiber Theory proposed that all organisms were made up of fluid humors and elastic fibers, whose tension was an indication of health. Friction, fermentation, viscosity, blocked pores, and corrosion by acids, alkalis, and oxygen all could cause fevers. Therapy involved bloodletting to

Paradigm Drift

29

balance the humors and tone the fibers. The fiber paradigm succeeded because it combined the ancient, traditional humoral theory, which had sanctioned universal therapeutic bleeding, with contemporary scientific knowledge of physics, chemistry, and the microanatomy of circulation. It inspired Erasmus Darwin to consider that all living forms might have evolved from a primordial fiber that differentiated according to changes in its immediate environment. Lamarck thought that the pressure of subtle humors could alter anatomy. And in Frankenstein (1818), Mary Shelley imagined that the vivification of the body fibers by electricity could resurrect a cadaver. To displace the fiber paradigm, the combined assault of Cell Theory and the Germ Theory of Disease was needed. The Modern Synthesis of evolution has this much in common with Fiber Theory: it combines diehard tradition and cutting-edge technology. The two paradigms also involve phenomena that are real, but irrelevant to the major paradigmatic premise. Friction heats, and oxygen corrodes, but fever and inflammation are effects of disease, not its causes. If the same logic is applied to Selection Theory, differential reproduction is real, but it is an effect of evolution, not its cause. Fortunately, although the Modern Synthesis is as flawed as Fiber Theory, it has practiced only metaphorical bloodletting. Evolutionary Paradigms Biological paradigms require individual, pragmatic treatment, since their conception and development have not followed a simple formula. Also, evolutionary paradigms, which attempt to establish universal biological principles, fail to obey the strict definitions demanded by some philosophers of science.6 As a general rule, any paradigm, worldview, or school of thought that has irrational components—that is, all of them—cannot be scientifically cut and dried. Evolutionary theory has not lurched from one revolution to another in the way that Thomas Kuhn portrayed the paradigms of physics in The Structure of Scientific Revolutions (1962). Instead, “the evolutionary paradigm just keeps on absorbing and expanding and forgetting where it’s been.”7 The first extensive evolutionary theory was put forth in Lamarck’s La philosophie zoologique, published in 1809, the year Darwin was born. Although it made a splash when it was launched, it was anything but watertight. The inheritance of acquired characteristics was not the contentious issue it is now. But Lamarck’s exposition of the process and mechanisms of evolution, such as spontaneous generation, universal evolutionary progress, and the idea that organisms actively responded to “needs,” had no scientific basis. Nevertheless, despite his errors, and the misrepresentations of his critics, Lamarck brought evolution to the center of attention. Many accepted that evolution might be a reality, even if Lamarck was wrong about the nature of the process and its mechanisms.

30

Chapter 1

The Darwinian Paradigm The launch of Darwin’s flagship theory was successful for reasons other than its originality of design and soundness of bottom. Evolution and the mechanisms of natural selection and adaptation had already been widely discussed by numerous biologists. But by working things out for himself, Darwin brought fresh insight to them. His dogged persistence was supported by influential friends, who assisted with the publication of The Origin of Species. It was immediately boosted by T. H. Huxley, who, after years of digging in his heels on evolutionism, became “Darwin’s bulldog.” Before any development of a general theory of evolution, taxonomists had also helped to prepare the way for it with their systems of classification. In particular, the systematics of Linnaeus suggested natural evolutionary relationships. The concept of a distinct species that was closely related to others within its genus was a powerful referent for all of biology, and its applied arts. Therefore to explain the origin of species was something that everybody already felt to be important, and it was the mainstay of Darwinism. Darwin’s readers could agree that horticulturists and animal breeders had produced distinct plant and animal strains. It followed logically that in the natural world the trend toward varieties could proceed further, to new species, and from there to the evolution of higher taxa. From a common ancestor, a diversity of species adapted to different environments could arise, and this adaptive radiation dovetailed neatly with the familiar concept of homology. Darwin thus had an audience of prepared minds. By making evolution progress fitfully through random changes in only a few lineages, instead of through a universal drive to perfection, Darwin disarmed the objections of materialists like Huxley, and then impressed them with his evidence for the historical reality of evolution. The concept of natural selection, a working hypothesis rather than an experimentally established theory, was good enough for the time being. Moreover, Darwin rounded off the Age of Enlightenment nicely by replacing divine law with natural law. These factors cooperated to elevate Darwinism to paradigm status, and some modern biologists think that no further revolution is necessary. The phenomenal reality of evolution has been established, despite some difficulties with its definition. What is still open to debate is a causal paradigm built on the proposition that natural selection causes evolution. Survival of the fittest was a concept familiar to ancient hunters, farmers, and philosophers. But it had not previously led to the concept of evolution. That had instead grown out of pre-evolutionary systematics, nature philosophy, speculation about the mutability of species, and paleontological discoveries. But by giving the historical theory substance and form, Darwin positioned his mechanistic causal hypothesis strongly. Moreover, since his historical evidence included the effects of artificial selection, he could link evolution with the analogous mechanism of natural selection. Before The Origin of Species, no one had amalgamated the two ideas so

Paradigm Drift

31

persuasively. For Darwin, evolution had to be through modification by natural selection. Phenomenon and cause were entwined so intricately that the one could not exist without the other—accepting the existence of evolution required the acceptance of natural selection as its cause. When Special Creationists argue that evolution must be rejected if natural selection is shown to be an unsatisfactory cause, they are only applying Darwinian logic. For Darwinists, it took no leap of faith to see that species varied in the makeup of their individual members. Variants that were better adapted, or more fit for the prevailing conditions of life, would be most successful at reproducing their characteristics and passing them on to subsequent generations. It was as if a metaphorical breeder were selecting some for reproduction, and rejecting less desirable forms from the broodstock. Metaphorical agents, used incautiously, can be reified into tangible forces, and lose touch with reality. Nevertheless, it is evident that some features of organisms in nature are biologically superior to others, and consequently survive and reproduce more successfully. This aspect of Darwinism is so self-evident that it “makes sense” to everybody. That it makes no sense as an explanation of how the superior features were generated in the first place doesn’t seem to be self-evident, since few notice the omission. The historical elements of Darwinism have been refined and occasionally jostled by molecular biology, but the reality of evolution has never been in dispute within biology. At first the very idea caused a stir, and debate focused on its plausibility and on the adequacy of the causal explanation. Evolution’s affront to fundamentalist Christian dogma has resurrected the same debate in every subsequent generation, but within biology the theory that evolution consisted of real historical events soon ceased to have a central paradigmatic role. It was the causal hypotheses that became controversial. Neo-Lamarckism and the Rise of Neo-Darwinism By 1882, the year of Darwin’s death, neo-Lamarckism was making the running as a new evolutionary synthesis. In addition to most of the original Lamarckian laws, neoLamarckism proposed that the environment imposed direct heritable changes on organisms. It also included natural selection. However, its role was to cull unfit variations, since the environment could cause detrimental as well as advantageous changes. Lamarck’s sense of inherent universal progress in nature was missing from neo-Lamarckism. For many neo-Lamarckists, especially E. D. Cope, and William Bateson in his early neo-Lamarckist phase, developmental evolution was especially important. This gave them the intuition that evolution was discontinuous, whereas Lamarck thought it was gradual. The appeal of something for everybody, the traditional allied with the contemporary, is always a strong combination. It therefore competed successfully with the revised Darwinism proposed by the first

32

Chapter 1

neo-Darwinist, August Weismann, who had pledged to purge evolutionary theory of everything except the all-sufficient mechanism of natural selection. In 1908, the English grandfather of ultra-Darwinism, E. B. Poulton, delivered a tubthumping manifesto that also proclaimed natural selection as the one true cause. He denounced not only neo-Lamarckism, but also the revisionism of Weismann and the biometricians, who had tried to “out-Darwin Darwin.” For Poulton, the generation of the novel qualities from which natural selection chooses was quite unimportant: “So long as individual variation is present, so long as it is hereditary, it does not signify how it is produced. . . . So long as it is there it is available, and Natural Selection can make use of it.”8 No generative theory for Poulton: to insist that the origin of variation had any importance was “the revolt of the clay against the potter.” Darwinists and neo-Darwinists have substituted one imaginary creator after another for the supernatural one: an architect for Darwin, a potter for Poulton, a sculptor for Mayr, a blind watchmaker for Dawkins. All of them ignore generative hypotheses to give natural selection the creative role. Despite Poulton, Darwinism was partially eclipsed by neo-Lamarckism, by Hugo De Vries’s Mutation Theory, and by William Bateson’s neo-Mendelism in the first decade of the twentieth century. But supporters of these ideas were quickly reconverted. T. H. Morgan’s Drosophila work began to pull him back from mutationism to gradualistic Darwinism and Mendelian genetics. Then, for his book Mimicry in Butterflies (1915), the entomologist R. C. Punnett, a close associate of Bateson, commissioned the mathematician H. T. J. Norton to construct a set of tables that would theoretically predict the effect of a slight selective superiority in a variation. Punnett, a saltationist at heart, was surprised when Norton showed that the smallest advantage would spread rapidly until it dominated the population. For the time being, the lid was nailed down on the coffin of “mutation pressure.” “Selection pressure” came to the fore. Why had mutation pressure seemed so important? Previously, the Darwinian idea that heritable characters of the parents are blended in the offspring was widely believed. This would have made it impossible for any small novelty of variation to sustain its distinct nature for long, because of homogenization with the more common characters. The only hypothetical solution for this problem was to assume that any novelties had to be saltations or large-scale mutations. The Mendelian demonstration that hereditary traits did not disappear through blending inheritance, combined with Norton’s calculation that characters with small selective advantages would rapidly spread in populations, made saltations theoretically unnecessary— much to the relief of Darwinists who shared their master’s concern that they made selection redundant. This did not alter the fact that, as Darwin pointed out, “monstrosities” often appeared spontaneously, especially in domesticated organisms. The question of how variation arose, regardless of its evolutionary potential, began to be answered by chromosome theory. It also clarified the role of sexual reproduction,

Paradigm Drift

33

which involves random exchanges of homologous chromosomal material, leading to “recombination” (mixing of parental features). Even in organisms with only a few chromosomes, there is an astronomical number of possible combinations, most of which are viable and “available as raw material for natural selection.” To this source of variation the modern concept of mutation can be added. When the structure of DNA was discovered, it was realized that point mutations (alterations in the sequence of bases in the DNA code) could bring about changes in the proteins into which the code was translated. Here, so it seemed, was the missing explanation of the origin of variations needed for natural selection to do its stuff. A new wave of neo-Darwinism rose in the 1930s from theoretical population genetics, which quantified fitness and selection, and predicted the genetic, and demographic make up of future generations. Since there will be no subsequent need to bring up Weismann’s first wave, I will apply “neo-Darwinism” to the later phase. Its supporters knew that old-fashioned Darwinism had been taking a beating. They were determined to mend it, and to turn their new version into the dominating force of evolutionary biology. Neo-Darwinist Semantics One of the reasons for the momentum of neo-Darwinism was its revision of definitions of natural selection and evolution. Neo-Darwinist Natural Selection Neo-Darwinists knew that Darwin’s synonym for natural selection, “survival of the fittest,” came down to no more than “survival of those that survive.” Therefore, they redefined natural selection as differential survival and reproduction. “Differential” referred to the relative success of variants at evading or improving predation and beating the competition. “Fitness” denoted the degree of adaptation to environment in Darwinian theory, and under neo-Darwinism it simply expressed the numerical distribution of particular variants as percentages, the most numerous being classed as the fittest. As such it is clear and uncontroversial. Refinement of the concept has vitiated the circularity of “the survival of those that survive.” But circularity remains; it has simply become “differential reproduction of those that differentially reproduce.” In the first glow of righteousness, neo-Darwinists pretended not to notice that differential survival and reproduction are effects and not causes. Natural selection is really the result of an aggregate of interactive causes, including competition (intraspecific and interspecific), predation on animals, and browsing on plants. In addition, there are sexual selection and the types of co-evolution in which the literal choice of one organism by another influences reproductive success. Relative resistance to disease and to climatic change are also qualities that result in differential reproduction. It may seem odd that a cause

34

Chapter 1

should be defined in terms of its effect, without reference to its agents, but selectionists claim that the tacit agents add up to an “integrating explanation.”9 Regardless of its awkward logic, natural selection, qua differential survival and reproduction, “makes sense” because it is a real and ubiquitous phenomenon. But these features do not automatically make it causal—a brook always has a babble, but the babble does not cause the brook. Nor does any definition of natural selection contain an explanation of what Cope called “the origin of the fittest,” referring to the source of novelty, the raw material that natural selection refines.10 However, selection theory provides a rough inductive guide to emergent evolution. By counting alleles, a notion of the quality of evolutionary novelty can be indirectly reached, although distorted by the conditions of life that prevail at the moment. Those conditions are usually a state of dynamic equilibrium, and so neo-Darwinism has always concentrated on the stability of populations. As population biologists readily admit, natural selection is largely a normalizing or stabilizing process that operates for a large number of generations without shifting the distribution curve, far less ever producing a new species. Under most circumstances, minor change, or simply no change, usually is the best way to stay in the game. The “evolutionary stable strategy,” a concept introduced to population biology from game theory by John Maynard Smith (1978), should realistically be called “non-evolutionary stable strategy,” since it describes how change is resisted in the presence of strong competition. Competition is nevertheless believed to be a significant factor that can carry a lineage out of stasis through directional selection. It supposedly leads to the differential survival of exaggerated features that give a competitive edge. In evolutionary “arms races,” predation pressure is believed to be more important than competition. The prey’s protective armor, size, or evasive skills become enhanced, and the predator’s armor-piercing claws and teeth, strength, size, and speed are improved. Since competition and predation cannot literally direct appropriate genetic change, their random mutation and selection may not be the mechanisms involved in biological arms races. Directional evolution could result from narrow epigenetic constraints, or a more fundamental process of genetic self-amplification. More interesting are occasions when competition and predation are absent. Contrary to the notion that evolution cannot occur under such circumstances, natural experimentation occurs anytime and anywhere, without immediate reference to natural selection. And some emergent types might need freedom from competition and predation to find their feet (or roots). Conventional selectionism makes much of how immigrants to isolated islands evolve rapidly, and here absence of competition and predation is important. The same conditions apply when environments have been cleared by some natural disaster. On rare occasions, under non-catastrophic circumstances, existing competition may become irrelevant if a newly emergent

Paradigm Drift

35

phenomenon is immediately superior. Business entrepreneurs, often social Darwinists, use the saltatory expression “getting the jump on the competition.” But conventional Darwinism has evolutionary change occur in “insensible” steps, to allow natural selection to always be in control. And in new environments, “new, strong selection pressures” hurry things along—gradually, but a bit less gradually than usual. Selection Pressure At the top of this chapter I asked if Niles Eldredge is right in saying that there has been a qualitative shift to an ultra-Darwinism that makes natural selection a force that actively generates adaptation. For me, quantitative shifts are more significant. One symptom of neo-Darwinist extremism is a greatly increased emphasis on the creativity of selection pressure. It infests the writing of even the most reflective neo-Darwinists. The metaphor of a pressure that eliminated inferior varieties was originally used by Alfred Russel Wallace in his Ternate essay, published as part of the Darwin and Wallace papers by the Linnean Society in 1858: Now, let some alteration of physical conditions occur in the district—a long period of drought, a destruction of vegetation by locusts, the irruption of some new carnivorous animal seeking ‘pastures new’—any change in fact tending to render existence more difficult to the species in question, and tasking its utmost powers to avoid complete extermination; it is evident that, of all the individuals composing the species, those forming the least numerous and most feebly organized variety would suffer first, and, were the pressure severe, must soon become extinct.11 [emphasis added]

An interesting contrast is found in the way Ivan Schmalhausen used “pressure” in Factors of Evolution: The Theory of Stabilizing Selection (1949). He wrote about “the vital energy of the organism (as a representative of a definite species), or its pressure upon the surrounding environment” as leading to natural selection.12 This pressure tended to be constant under stable conditions, but could increase for particular types when conditions changed. Although Schmalhausen used neo-Darwinist language, and also thought in terms of aggressive repression of organisms by extreme environmental conditions, it is clear that he saw that generative evolutionary pressures were to be found in organisms, and the effects were to be found in natural selection. Unfortunately, neo-Darwinists were to take from Schmalhausen only what they barely needed to understand stable populations, and “find the selection pressure” is all they needed to do. Selection pressure is now given a metaphorically creative sense by modern biologists. Although there are many superficial instances of such adaptations, there are many other instances of appropriate adaptations that could result from only the slightest genetic change, and yet such adaptations have not appeared. To understand the problems with selection pressure and response to need, consider this example: Female grassfinches are attracted to males with artificial crests, so there

36

Chapter 1

must be an almost tangible, strong selection pressure for natural crests. Unfortunately, “conceiving how female predilection for sexy head gear can suck a crest of feathers out of a male grassfinch’s head is rather difficult.” Nothing of the kind has in fact appeared naturally in any of the 120 species of grassfinch, although it is a common condition in many of its passerine relatives. There are many cases where the slightest nudge to an existing genotype could bring out advantageous traits. But the nudge has not been nudged, or the traits have refused to appear.13 In Chance and Necessity (1971), Jacques Monod equated necessity with “selective pressure.” Necessity drives the selection of chance events in the direction of adaptations appropriate to the particular habits and habitat of the organism. To his credit, Monod recognized that logically this is no different from Lamarck’s scorned Second Law, which proposed that organisms evolved in response to their needs. A creative selection pressure is not only superfluous, it is logically the same as Lamarck’s proposal that organisms respond to needs. Lamarck made the organism the agent of evolution—not an imaginary external force. In every generation, neo-Darwinists flog the dead horse of Lamarckism; but instead of punishing Lamarck for his Second Law, they ought to be flagellating themselves for selection pressure. In any case, the physicochemical and biotic environment can have a direct causal influence: it changes the internal nature of organisms, shapes their development, influences their behavior, and occasionally offers freedom from the usual agents of natural selection. Under such conditions, the expression “selection vacuum” or “negative selection pressure” would be more appropriate. Neo-Darwinist “Evolution” To complement their new definition of natural selection, neo-Darwinists redefined evolution as “changes in the distribution of alleles in populations.” This premise does not even require stable mutations of alleles. Existing alleles can change in their distribution from generation to generation through differential survival and reproduction without the slightest progressive evolutionary or directional effect. The new definition of evolution also opened the door to random, non-selective events like genetic drift, which resulted in increased numbers of alleles that did not have the highest fitness. Then along came the neutral theory of molecular evolution, which argued that the numerical distribution of alleles was not a reflection of natural selection. Lucky breaks such as bottleneck effect and founder effect involve random subsampling of the parental population. Although founder effect is regarded as a key to diversification in island populations, such non-selective events are regarded as insignificant special cases of evolution. Epistemologists have always upheld the Hippocratic aphorism that an adequate hypothesis should “save the appearances,” or account for all of the known manifesta-

Paradigm Drift

37

tions of the phenomenon in question. And yet the progressive or complexifying component of evolution is missing from selectionist formulae. To avoid argument, some selectionists say that increase in complexity is epiphenomenal—a possible but rare cumulative by-product of the action of natural selection. However, George Williams takes a harder line: I would maintain . . . that there is nothing in the basic structure of the theory of natural selection that would suggest the idea of any kind of cumulative process. . . . Gene substitution would slightly improve the precision of one or more adaptations, but as perfection is approached the opportunity for further improvement would correspondingly diminish. This is not a process for which the term progress would be at all appropriate.14

Perfection yes; progress no? For Williams, organisms that seem to have progressed to higher forms and functions are simply “better adapted” under some circumstances. He insists, moreover, that “organization” should only be used in that context: When a biologist says that a system is organized, he should mean organized for genetic survival or for some subordinate goal that contributes to successful reproduction.15

The slighting of progress has been rationalized by others on the grounds that it is too burdened with political economics to be worth retaining. Others say that evolutionary history needs to be kept apart from its causation for the sake of clarity. But the real reason for ignoring the issue is that neo-Darwinism can’t explain it. And there is this underlying dread: What if complexification is autoevolutionary and irrelevant to selective advantage? Neo-Darwinist Adaptationism In the introduction, I discussed the elusive semantics of the innocent word “adaptive” and said that it becomes even more slippery when innovation that is adaptable or adaptational is said to be elicited by selection pressure. “Whatever evolution wants, and can get, it does get.”16 Stephen Jay Gould and Richard Lewontin (1979) say that such adaptationism is “Panglossian,” after Voltaire’s character who said that everything happens for the best. Under such critical fire, the theoretical elite come out from behind their Looking Glass, where evolution gets what it wants, and where natural selection makes what it wants, to regroup under the flag of formal definition. But soon they retreat back through the mirror, proliferating more preposterous circularities such as “wings are adaptations to flight” or “eyes are adaptations to sight”—as if flight and sight were metaphysical absolutes that existed before the emergence of wings and eyes. Fuzzy logic is a necessary part of tentative thought experiments, but it should not continue to be used so vaguely into the maturity of a theory. Indeed the language of neo-Darwinism is so careless that the words “divine plan” can be substituted for “selection pressure” in any popular work in the biological literature without the

38

Chapter 1

slightest disruption in the logical flow of argument. Little wonder creationists find it such an easy mark. But they and selectionists have this in common: faith in the efficacy of a creative mechanism that has no material reality. “Theological Darwinians” have accepted Darwin’s view of natural selection as a secular creator.17 And throughout their writings there is now no straightforward evolution; it has acquired a fixed epithet as evolution-through-natural-selection. These semantic reinventions of Darwinism and neo-Darwinism contributed to the polemical success of the Modern Synthesis. The Modern Synthesis During the second and third decades of the twentieth century, hypotheses and opinions were shifting from saltationism back to gradualism. But the Darwinian paradigm would have lost steerage way if it had not been for popularizing works that brought theoretical population genetics into a general evolutionary context. The Causes of Evolution (1932) by J. B. S. Haldane was first to serve this purpose. Setting out to rescue Darwinism, Haldane discussed the likelihood of evolution happening faster in isolated populations, and suggested the importance of “developmental genetics.” Theodosius Dobzhansky’s Genetics and the Origin of Species (1937) broadened the treatment of evolution through genetic change, and catalyzed a more comprehensive amalgamation of evolutionary disciplines that was to take its name from Julian Huxley’s Evolution: The Modern Synthesis (1942). Ernst Mayr’s Systematics and the Origin of Species (1942) is another component of the founding corpus of the Modern Synthesis, along with G. G. Simpson’s Tempo and Mode in Evolution (1944), which merged neo-Darwinism with paleontology. Bernhardt Rensch’s Evolution above the Species Level argued that the same process that caused speciation caused evolution at the highest taxonomic levels. First published in German in 1947, it was embraced by the Modern Synthesis when it came out in English translation in 1959, because it made microevolution all-sufficient. American evolutionists and textbooks often ignore Julian Huxley’s role in the development of the Modern Synthesis. Ernst Mayr (1980) credits Huxley with the introductory application of the word “synthesis” and notes that he had often discussed it with him. But he remarks that Huxley was “luckier” than the other pretenders because he published first. Frederick Churchill (1980) considers Evolution: The Modern Synthesis as little more than an expansion of a conference address delivered by Huxley in 1936, and politely refers to the last chapter on progress as “curious,” although it was central to Huxley’s synthesis. William Provine (1988) regards Huxley as a compiler, rather than a synthesizer. He notes, incidentally, that not only Huxley felt left out by the Americanized Modern Synthesis; almost every ego connected with evolutionary biology in the 1940s, whether in Britain or America, felt injured. Niles

Paradigm Drift

39

Eldredge’s The Unfinished Synthesis (1985) argues that the Modern Synthesis made even paleontology walk the plank, only to be reluctantly invited back to dine at the captain’s table in the last decade or so. But he ignores Huxley. By a curious coincidence of metaphor, my Evolutionary Theory: The Unfinished Synthesis was also published in 1985. In it, I took a dim view of the various forms of the Modern Synthesis, and was already looking for a comprehensive theory that would reject the neo-Darwinist excesses that were already developing. Since then, I have acquired a better appreciation of Huxley’s attempts to deal with progressive evolution. Jeffrey Schwartz (1999) describes Glenn Jepsen’s account of how W. H. Bucher first mentioned a synthesis of genetics, morphology and paleontology in 1941. Then Columbia University incubated the idea that further discussions at formal meetings would let the protosynthetists “learn each other’s languages.” It hatched at a major conference in 1947, whose proceedings were published in 1949 as Genetics, Paleontology and Evolution, with the credit going to Jepsen, Mayr, and Simpson.18 Most of this happened after Huxley published the book that popularized the usage of the term “synthesis” and gave it a more comprehensive treatment than any of its subsequent rivals. Why is Huxley so ignored? Michael Ruse (1996) says that although “prodding from Huxley” helped formalize the synthesis in America, he was sidelined for being too much of a progressionist.19 Also, other evolutionists were not too happy about allometry and embryological evolution since they smacked of orthogenesis and saltation. And there was one additional “niggling worry” about Huxley: Huxley’s intentions were synthetic, and synthesizing is precisely where he succeeded, as never before. Yet . . . there was something of his grandfather about his approach to evolution. There was a part of Huxley that was never that deeply committed to adaptationism.20

Many biological studies had been developed independently of evolutionary theory. Huxley meant to integrate all of them, using progress, or what some of his contemporaries were calling “macroevolution,” as the unifying theme. I am persuaded by Ruse that Huxley was disreputable among the synthetists because he was not enough of a selectionist, and by Mayr’s (1980) inference that Huxley was too soft on orthogenesis, allometry, and progress. But I hesitate to exclude pure American chauvinism as a reason for dropping Huxley. In any event, the priority that American synthetists gave to quantitative neo-Darwinism was to subsume all the rest. Their version of the synthesis may have reduced conflicts between different specialists, and may have given them the sense of a common cause, but it also shed some of the important evolutionary causes. The most comprehensive and detailed general account of the development of the Huxleyan and American versions of the Modern Synthesis is proffered by Vassiliki Betty Smocovitis in Unifying Biology: The Evolutionary Synthesis and Evolutionary Biology (1996). Since it is not in her mandate to evaluate the biological foundation of the

40

Chapter 1

Modern Synthesis, Smocovitis sees selection theory as a positive part of the slow unification of biology. To her, the American version was more cohesive through the frequent interaction of its contributors, and Huxley’s monograph was relatively disorganized, though appealing to a wider audience outside biology. She does, however, acknowledge William Provine’s judgment that the American Modern Synthesis is easier to define in terms of what it rejected than what it synthesized.21 Let us take a break to consider how the word “synthesis” has been used by evolutionists. It may mean little more than an amalgamation of independent studies that provides encyclopedic convenience. A “unifying synthesis” may harmonize a disjointed discipline with the application of an integrating explanation, a role awarded to natural selection in the Modern Synthesis. Furthermore, if independent investigators come together to actively examine fields with which they were once unfamiliar, there is a good possibility that novel hypotheses will emerge from a fusion of previously isolated ideas. The “dialectical synthesis” of thesis and antithesis is another alternative. Applied effectively it might produce a powerful new theory. As well as resolving contradictions, it involves the heuristic association of independent concepts. Julian Huxley, though familiar with the genre, gave no hint that he had such a purpose. However, Ernst Mayr (1980) argues that the new synthesis was the fusion of two independent and sometimes conflicting research traditions: experimental genetics and population studies. Both sides benefited from their new insights, so the whole was now greater than the sum of its parts. “Such an event,” Mayr adds, “occurs only occasionally in the history of science.”22 As to the situation prior to the synthesis, he writes: I am appalled at the misunderstandings, the hostility, and the intolerance of the opponents. Both sides display a feeling of superiority over their opponents “who simply do not understand what the facts and issues are.” How could they have ever come together? Just as in the case of warring nations, intermediaries were needed, evolutionists who were able to remove misunderstandings and to build bridges between hierarchical levels. These bridge builders were the real architects of the synthesis.”23

Thus Mayr elevates the Modern Synthesis beyond the attempted unification of Huxley (who gets an honorable mention as a bridge builder—we are left to guess who the primary engineer of this rare, momentous synthesis might be). Mayr’s interdisciplinary bisociation end runs the kind of Kuhnian revolutionary interpretation that he finds distasteful (Mayr 1972). But it sails close to a dialectical synthesis. Scientific syntheses of this kind have no automatic claim to righteousness. The bogus Fiber Theory was also an invigorating fusion of two research traditions. For any kind of synthesis, all the arguments need to be aired, and awkward exceptions confronted at the outset instead of being black-boxed, ignored, or liquidated. The Modern Synthesis has failed this test, and has lost many of the comprehensive features

Paradigm Drift

41

that Huxley gave it. No amount of historical revisionism can make up for that. Yet it persists, and it can still beat the competition. Paradigms Lost?24 Three quasi-paradigms were lost when neo-Darwinism came to the fore. One was neoLamarckism. Another was Hugo De Vries’s Mutation Theory, supported by William Bateson’s neo-Mendelism. A decade or so later, a flash-in-the-pan version of emergentism suffered the same fate. What these had in common was saltatory evolution, but the early brand of emergentism also suffered from associations with mysticism. The gradualistic and materialistic Modern Synthesis has held sway ever since, aground though it might be. What might it take to finally blow it off its reef? In The Structure of Scientific Revolutions (second edition, 1970), Thomas Kuhn suggests a new paradigm may be imminent when the existing one is in a state of crisis, as marked by several signatures. First, there is the failure of the old paradigm to explain new data: The profession can no longer evade anomalies that subvert the existing tradition of scientific practise. . . . Then begin the extraordinary investigations that lead the profession at last to a new set of commitments, a new basis for the practise of science.25

Darwinism and its descendants have postponed crisis by going with the flow of changing ideas, and then ignoring the anomalies, or subsuming them into its thesis. But what of other indicators? Kuhn (1970) writes: Because it demands large scale paradigm destruction and major shifts in the problems and techniques of normal science, the emergence of new theories is generally preceded by a period of pronounced professional insecurity. As one might expect, that insecurity is generated by the persistent failure of the puzzles of normal science to come out as they should.26

If this is so, selection theory, which has been confronted by crisis for a quarter of a century, shows remarkable paradigmatic fitness. Biologists can retreat to their many professional niches, and conduct much of their research without reference to the central theory. Maybe they see no profit in participating in a vituperative ongoing debate—name calling and righteous indignation should be added to Kuhn’s signatures of crisis. Those who feel insecure want to accept the reassurances of reliable authorities, who, in turn, oblige by deploying an array of sophistries to meet this end. In the mess of the Modern Synthesis, the evolutionary elite at the captain’s table keep beating the selection gong.27 The dictum that ideas grow old and die with their originators does not apply: new evolutionary authorities with more extreme versions of the same old ideas have grabbed the vacant privileged seats. At the captain’s table, accommodation of any previously antithetical idea, referred to as “change of emphasis,” is accompanied by rounds of self-congratulation for the vitality of the

42

Chapter 1

evolutionary debate. However, the reverberating extremism of ultra-Darwinism might yet disequilibrate the paradigm enough to give emergent ideas a chance to be established. Kuhn further comments: In periods of acknowledged crisis scientists have turned to philosophical analysis as a device for unlocking the riddles of their field. The search for assumptions can be an effective way to weaken the grip of a tradition on the mind and to suggest the basis for a new one. Scientists have not generally wanted or needed to be philosophers. Indeed normal science usually holds creative philosophy at arm’s length.28

Simply quoting Kuhn is philosophical commentary, and The Structure of Scientific Revolutions is still one of the most frequently cited books in the scientific literature. It is also the Bible of the postmodernist flotilla that tries to blockade biological modernism. They declare that dominant paradigms tend to treat relative truths as absolute. Therefore they should be cut down, and new ones should be nipped in the bud, lest they too repeat the errors of the past. At best we should be allowed to choose from a fleet of alternative models.29 Biological postmodernism came in with structuralism and the New Biology of the 1970s and 1980s, but its attack on evolutionary modernism seems to have left selectionism intact. Have these assaults weakened the Modern Synthesis at all? If so, how close might it be to breaking up? In The Darwinian Revolution (1959), Gertrude Himmelfarb pointed out that Darwinists and their successors avoided confrontations by subsuming antitheses under the conventional thesis. Awkward contradictions were simply filed as special cases and then forgotten. Some were flatly judged wrong and ignored. But if antitheses are constantly subsumed, a thesis is likely to list and go off its original course; indeed the drift of the original paradigm would have revived Darwin’s chronic seasickness. Given a glimpse of twentieth-century perspective, Darwin would certainly have been relieved at the retention of natural selection as the cause of evolution. But he might have wondered why the organism had been marooned while population thinking, selfish genes, and molecular reductionism had been taken on board. Neo-Darwinism started out on the purposeful course of quantifying population theory, which seemed to give it the same kind of predictive powers as physical laws. However, it had provided nothing qualitatively new by the time it signed on with the Modern Synthesis. Under this name the superstructure was rebuilt and extra cargo embarked, but, already aground, it could go nowhere, and a lot of special cases, ad hoc hypotheses, and other flotsam and jetsam have drifted up against it ever since. The Modern Synthesis retains the same kind of appeal that neo-Lamarckism had a century ago: something to offer everybody, provided they accept natural selection as the ultimate cause. Sometimes it is credited with the active role of discovery, investigation, and analysis of antithetical phenomena. The following quotation is from a popular introductory biology textbook:

Paradigm Drift

43

The modern synthesis recognizes, and in fact first described how other mechanisms such as genetic drift and chromosomal mutations can cause rapid, nonadaptive evolution. But the major emphasis of the synthesis is on gradualism and natural selection.30

This sinks below historical inaccuracy. Not only does it reify the synthesis as an agent of investigation, it falsely infers that apparently antithetical phenomena were discovered through its inductive guidance. The anti-Darwinist Richard Goldschmidt, who promoted the saltatory causal role of chromosomal mutation and position effects, would turn in his grave—if he has ever stopped spinning since his interment. Furthermore, it shows how little improvement there has been in the 40 years since Himmelfarb’s admonition. An antithetical discovery about rapid evolution is subsumed and toweled over with “gradualism and natural selection.” The captain’s table responds that the language may be loose and figurative, but everybody knows fundamentally what is going on. The evidence is to the contrary. Most non-biologists who accept neo-Darwinism, as well as many orthodox biologists, seem to think entirely in figurative terms: Natural selection and selection pressures are real driving forces of evolution. Every generation of undergraduates is brainwashed into this way of thinking—“If it exists, it must have been selected.” Biology is constantly being misrepresented to students and the public at large, and only an arrogant elite with serious vested interests would connive at that. Some adherents to the Modern Synthesis are not completely stifled by it. In one of the most forthright treatments of the subject, Natural Selection in the Wild (1986), John Endler issues standing orders that ought to be nailed to the mainmast of the Modern Synthesis: Natural selection is a process which results from biological differences among individuals, and which may give rise to genetic change or evolution; it is not a “force” that “acts,” and does not have an “intensity.” By analogy with a chemical reaction process, natural selection has a rate and rate coefficients, which are estimated by fitnesses. The same analogy shows why “force” is an improper analogy: it is not a “force” that causes the change in reactant frequencies, but the chemical properties of the reactants.31

Endler’s analogy tells us to find the properties of the generative reactants that give rise to emergences. Identifying them with individual genes is not enough, because gene expression is instructed by higher levels in the organismal and environmental hierarchies. One difficulty in dealing with selection theory, as Endler sees it, arises from the kinds of logical confusion that I have just described. He does a fair job of clearing it up. It helps that he even considers natural selection as a result of differences, which for me translates as “natural selection is the effect—as opposed to the cause—of evolutionary change.” More explicitly, he states that origins of evolutionary novelties are external to the selectionist program of causation. Endler also cautions the faithful that

44

Chapter 1

“natural selection is not easy to detect.” Lack of detection of it, when present, is compounded with apparent detection when it is absent, and with misleading demonstrations of its existence.32 If these points too were highlighted in orders of the day for the Modern Synthesis, the paradigm might not be stuck so fast on a reef. It is one thing to regret the minor deficiencies of the Modern Synthesis as its supporters see them, or to reminisce over the loss of the neo-Lamarckist, mutationist/Mendelist, and emergentist pretenders. It is entirely another to consider what the Modern Synthesis has rejected. Here a keening chorus will not do. Those aspects of the evolution of life must be reinstated. The Loss of Life, Organism, Mind, and Evolution If the most enlightened analyst replaced Pangloss at the helm, the flagship of the Modern Synthesis would still be on the reef where neo-Darwinism put it, with no ability to maneuver, only to sit there and accumulate more ad hocsam and jetsam, ignoring inadequacies and subsuming serious contradictions as special cases of the neo-Darwinist thesis. The Modern Synthesis had no dialectical agenda for the future accommodation of antithetical ideas so as to advance to a new synthesis. It made complete castaways of neo-Lamarckism, and direct effects of environment, and it neglected developmental evolution. Simpson was browbeaten into renouncing his earlier flirtation with orthogenesis, and megaevolution—which, after all, came down to the abhorrent saltatory progress (what Mayr calls “soft” evolution). Unification of the synthesis was gained by losing evolution and replacing it with “changes in the distribution of alleles in populations.” When population thinking came along, the Modern Synthesis lost the organism as well. At the time it was necessary to correct a common misapprehension about evolution. Novelties spring from the germ lines of parents without appearing in the parental phenotypes. In the next generation these novelties are manifested phenotypically, but the individual organisms do not evolve. Or do they? Their own actions affect them and future generations. For example the nutritional health of the mother may influence the condition of her ova, and their subsequent development. Exposure to extremes of stress may directly increase the mutability of DNA and change methylation patterns. So what is the way round this paradox? What actually evolves? According to population thinkers, if there is random mating and gene flow in a population, then the entire population pool of genes is available to produce the subsamples that constitute organisms. Moreover, since individuals do not possess all of the genetic characteristics of the whole population, a comprehensive account of evolution must involve the entire gene pool, not the organismal bits. The population supposedly evolves by changing the numerical proportions of existing alleles and by incorporating novel alleles with high selective value. Another conventional reason to

Paradigm Drift

45

emphasize populations is that they persist for long enough to have evolutionary clout; individual plants and animals are too ephemeral. In Phenotypes (1994), David Rollo gives the issue more focus by writing that it is phenotypic lineages within populations that are selected and evolve. Moreover, for him, lineage evolution involves a lot more than allele redistribution and structural gene mutations. Progressive factors like improvement in developmental, physiological and behavioral organization are important. And he gives whole individual phenotypes active evolutionary roles that affect the fate of the lineage. All I need say differently for the moment is that lineages do evolve, but selection qua differential survival and reproduction is a consequence of the qualities of those lineages, not a causal agent. And no matter how much “better” one lineage is than another, it will finally establish its own change-resistant equilibrium. Individual organisms began to be ousted from any evolutionary consideration because pools of genes are easier to work with statistically, and behave less aberrantly. As a consequence, evolutionists who thought that there was more to a population or an organismal whole than the sum of its genes were dismissed as Platonic essentialists. A similar criticism was applied to anyone who thought an archetypal individual could found a new lineage, since the whole gene pool had to be seen as the real foundation. Plato had remarked that what we perceive of nature is a corruption of its true essence, as if we were observers in a cave, looking at the shadows of beings that passed outside. But it seems to me that the gene pool is the shadow play. Population thinkers have transferred the essence of the organism to the population—not a population of actual creatures interacting with each other in a real environment, but a mathematical model devoid of life. Individual genes have also had essences transferred to them. Genes are all that matter to molecular biologists because genes are what they study. It should not be surprising that they inflate reductionism out of egotism. But to help shift biology to the genocentric universe a rogue singularity was needed. It arrived in the form of the selfish gene. As originally conceived, it only cooperates with others to reach the goal of replication, and uses organisms merely as transportation mechanisms.33 Such romantic reductionism, combined with the genes-for-everything hyperbole of genomics, has greatly appealed to simplistic materialists, and to the current crew of sociobiological anthropologists and “evolutionary psychologists” on the ultraDarwinism watch. If behavior and even mind are gene-determined they are susceptible to natural selection, and the modernist program is served. This logic should not be surprising either—in earlier rejecting higher and lower we had already lost our minds.34 The phenomenon of life itself is also secondary to natural selection. Vitality has too many bothersome, essentialistic, ineffable qualities, and the conventional stance infers that it must merely be the arithmetical sum of the functional parts of a living

46

Chapter 1

organism. Survival and reproduction have been inherent features of life since it first emerged. But in the reversed Looking Glass perspective of selectionists they are the products of natural selection. When challenged, revolving-door logic whirls them out from behind the mirror to the reality of differential survival and reproduction. “If it exists it must have been selected” has an ontological contradiction that Darwin intuitively distinguished when he equated physiological adaptability with the “innate, wide flexibility of constitution” that allowed an organism to “persist in its own being.”35 Spinoza, who called the essential quality of persistence “conatus,” wrote: “The endeavour with which each thing strives to persist in its own being is nothing else than the real essence of that thing.”36 Darwin’s exclusion of natural selection as the agent of persistence in being is significant. Since he did not pursue the idea further, he probably realized that it undermined his own major premise, and like much of Darwiniana it has been ignored by the Modern Synthesis. Persistence in being, flexibility of constitution, adaptability, self-organization, and homeostasis are all aspects of the integrity of the organism that can be dealt with independently of special physiological and ecological exigencies. Adaptabilities that preserve integrity are to be found at any level from the molecular to the behavioral in an organism, and they extend out into its ecosystem. Lamarck implied that they evolved progressively and were distinct from adaptational reactions to “needs.” Neither he nor Darwin explained how the maintenance of integrity in the face of modifying influences might have originated or evolved. As they are fundamental emergent properties of life, along with reproduction, any selectionist claim that they are the product of natural selection can be reduced to the absurdity “Natural selection is the cause of life.” Attrition of the Modern Synthesis So far the casualty list of the selectionist barrage includes progressive evolution, orthogenesis, the organism, mind, and life itself. Although these victims had already fallen in earlier engagements, the ultras have not stopped trampling upon them. However, there are other elements of Huxley’s original Modern Synthesis that have been ignored or lost. Epigenetic (or developmental) evolution was once a casualty, but has now recovered, though suffering from a secondary infection of selectionism. In the 1940s, after the first flurry of reconstruction, the most important additions to the Modern Synthesis were the structure, replication, and coding of DNA. These offered elements of the generative process that had been sadly lacking. But once molecular biology provided some explanation of how new variants could be produced as fuel for natural selection, that was all that the Modern Synthesis wanted. It was unperturbed by the fact that the triumph of reductionism—the discovery of DNA structure—offered no explanation of differential gene expression. How organisms with a very similar quota of structural genes came to be so widely diverse was ignored. It

Paradigm Drift

47

had long been known that the chromosome count was of little significance; couch potatoes have fewer chromosomes than the chips they consume. So it is strange that the mapping of the human genome in February 2001 trumpeted that one of the mystifying discoveries is that a human has only 10,000 more genes than a roundworm—as if this presented a brand-new evolutionary puzzle for solution. Such problems are now being actively pursued by students of developmental evolution, the self-styled “evo-devos” (possibly an abbreviation of “evo-devo-taries”) who tarried so long before asserting the importance of developmental studies to evolution and climbing aboard the Modern Synthesis. Their reception was lukewarm, sometimes hostile, for the others at the captain’s table were mostly population geneticists with little feeling for the evolutionary implications of molecular biology. And traditionally they saw epigenetics as a threat to the ship, by its association with saltatory evolution. For Julian Huxley, embryological phenomena were central to evolution, were heritable, were acted on by natural selection, and were a necessary element of his synthesis. However, Huxley never integrated his studies of allometry fully into the Modern Synthesis, and his treatment of “differential development” was disappointingly brief, despite the weight he thought it should have. His introduction to the 1962 second edition of Evolution: The Modern Synthesis enthusiastically endorsed Waddington’s (1957) epigenetic ideas. But he allowed himself to be persuaded by Rensch (1959) that orthogenesis, part of the original synthesis, should be cast away. It is also illuminating to read the introduction to the third and last edition (1974), written by the cream of Oxford evolutionists, hand skimmed by Huxley himself. Evolutionary embryology, which Huxley had been initially determined to bring into play, is absent. Physiology? Forget it. In losing such major components, Huxley’s Modern Synthesis regressed to Mayr’s version, no longer amalgamating nor unifying, but providing a bottomless hold-all for any new idea that came along, or old ideas that became new again. To wield influence, a paradigm does not have to be comprehensive, nor does it have to be based on correct information or sound ideas. It can also allow empirical applications of false hypotheses. Only with hindsight might they be seen to have been totally misdirected. What matters most when paradigms are first established is their psychological or epistemological impact. Also in 1974, Søren Løvtrup published Epigenetics, a fresh attempt to raise evolutionary consciousness—for those who were awake. Others muttered in their bunks “But it’s all just genetics!”—another subset that could be ignored. Optimistic bulletins about the unsinkability of the Modern Synthesis kept being posted. In the 1978 Evolution issue of Scientific American it was presented as a perpetual paradigm requiring only minor adjustments. In The Evolutionary Synthesis: Perspectives on the Unification of Biology (1980), edited by Ernst Mayr and William Provine, Dudley Shapere sounded various alarms concerning the loose cannon of reductionism, the rarity of examples of

48

Chapter 1

adaptive mutations from laboratory genetics, and neglect of the discontinuities in the fossil record.37 But he left embryology and epigenetics out of his muster. Viktor Hamburger, who was well capable of making amends, merely rationalized the omissions: I do not imply a criticism of the originators of the modern synthesis for their neglect of developmental genetics. On the contrary, I would assert that it has always been a legitimate and sound research strategy to relegate to a “black box,” at least temporarily, wide areas that although pertinent would distract from the main thrust. No great discoveries or conceptual advances are possible without this expediency.38

The first difficulty with this assertion is that something put in a black box ought to be dealt with as a significant phenomenon whose mechanism is not known, not ignored as a nuisance. More to the point: what great discoveries or conceptual advances support such a judgment, and where could a main thrust take a sunken synthesis? Almost two decades later, even some population biologists and ecologists are looking beyond the ecological domain of selectionism and realizing the importance of developmental biology for evolutionary theory. I have already mentioned David Rollo. Another such ecologist is Wallace Arthur (1997): We have come to accept a theory of evolution that explains the origin and diversification of exquisitely-engineered organisms on the basis of the selective destruction of genetic/developmental variants whose initial production has been treated, for the most part, as a ‘black box’. . . . Why has this pronounced lopsidedness of evolutionary theory, with its emphasis on destructive forces, been allowed to develop?39

Arthur suspects that selectionists, having fought a long war against special creationism and Lamarckism, are still afraid that if they rest on their laurels these old enemies might come surging back. The Modern Synthesis has indeed emphasized selection as a destroyer and an eliminator. But if that redundant role were abandoned, it would still be as hard to discard natural selection as a materialistic creator, as it was for protoevolutionists to discard special creation. The lopsidedness of the Modern Synthesis was due to the abandonment, or repulsion of paleontology, the organism, epigenetic saltations, orthogenesis, genetic assimilation, and the direct effects of the environment. And what do these rejecta have in common? They participate directly in evolutionary change, and they are minimally subject to natural selection. The Rule of Gradualism One of the perceived flaws of paleontology was that is was soft on saltation. Saltationism never came close to paradigmatic influence, except when The Mutation Theory was published by Hugo De Vries at the turn of the twentieth century, and William Bateson suggested at the same time that Mendelian genetics argued for dis-

Paradigm Drift

49

continuous change. Richard Goldschmidt argued for the saltatory effects of chromosome mutations and was ridiculed. Later saltationists, such as paleontologists K. Beurlen and O. H. Schindewolf, were irritations rather than obstacles to the passage of neo-Darwinism. Despite minor buffeting, gradualism remained a bulwark of neoDarwinism and the Modern Synthesis. If a saltatory macromutation is sufficient to generate a successful new species, natural selection is allowed no part in the transition except in the spurious role of final arbiter. But if tiny variations are gradually acted upon by natural selection, to culminate in speciation, then Darwinian honor seems to be satisfied. One of the reductions to absurdity of gradualism is hylozoism, also known as panpsychism.40 According to this principle, if evolution occurs by gradual accumulation and gradual emphasis of pre-existent characteristics, then any feature possessed at any evolutionary level must have existed in some lesser state at lower levels. Mind, therefore, must somehow be a property of even inanimate matter. Alfred Russel Wallace accepted the logic of hylozoism, but since he balked at molecules with mind he was forced beyond Darwinism: If a material element, or a combination of a thousand material elements in a molecule are alike unconscious, it is impossible for us to believe that the mere addition of one, two or a thousand other material elements to form a more complex molecule could in any way tend to produce a self-conscious existence. The things are radically distinct. To say that a mind is a product or function of protoplasm, or its molecular changes, is to use words to which we can attach no clear conception. You cannot have in the whole what does not exist in any of the parts; and those who argue thus should put forth a definite conception of matter with clearly enunciated properties, and show that the necessary result of a certain complex arrangement of the elements or atoms of that matter will be the production of self-consciousness.41

This demonstrates a distinct contrast between Darwinism and emergentism, because emergent wholes do have radically distinct properties that do not exist in the parts, while adaptational aggregations do not. Wallace’s own mind, clouded with spiritualism as well as selectionism, could only interpret the emergence of the big brain as a divine gift, not as an epigenetic saltation. Although hylozoism can easily be reduced to absurdity, it has attracted strange bedfellows, such as Henri Bergson and Bernhardt Rensch, whose only other common trait was a deficiency of a sense of the absurd. Hylozoism also infects some current “antichaos” studies, with paradoxical consequences. For example, if, despite the contrary stance of emergentism, there were indeed physical laws that generate complexity and extend to the highest levels of organization, then evolution would advance without reference to natural selection, and Lamarckian progressionism would be rationalized. Gradualism has more recently been the target of a number of hull-breaching collisions, beginning in 1972 with the punctuated equilibrium of Eldredge and Gould. They argued that the unevenness of the fossil record was due to periods of relatively

50

Chapter 1

sudden speciation followed by periods of unchanging stasis. Since ontogenic and population homeostasis resisted change, some kind of genetic revolution was necessary to overcome it. And they agreed with Mayr (1963) that strong selection pressures at the boundaries of species distribution were part of the disequilibration. The result was rapid speciation followed by re-establishment of an even more robust and hence prolonged genetic homeostasis. Punctuated equilibrium was not a new idea. Although discontinuities of the fossil record had always been dismissed as accidents by gradualists, several biologists, including Darwin, had thought that evolution might have a naturally irregular tempo. In 1871, St. George Jackson Mivart had compared “intermitting periods of equilibrium” in evolution to what would occur in a stable rock pile disturbed by sufficient force, such as an earthquake. It would rearrange itself into a new configuration that would remain stable for another long phase of inertia.42 The emergence of a superior organization, though unlikely in a tumbling rock pile, was inherent in Mivart’s view of evolution. Herbert Spencer constructed a corollary hypothesis, “the law of the instability of the homogeneous” or the “principle of equilibrium,” which held that homogeneous systems were fundamentally unstable because external forces acted upon them unevenly. Functional anatomical variations within simple organisms created heterogeneous internal forces, and as the biosphere became more complex, so did organisms. Spencer defined evolution as “a change from an indefinite, incoherent homogeneity, to a definite, coherent heterogeneity; through continuous differentiations and integrations.”43 He too thought that progressive stages were followed by static periods of equilibration, which jibes with the punctuationist and emergentist points of view. Thus, not only did some Victorian biologists understand the irregular gait of evolution; they thought there were qualitative differences between rapid and slow changes, the former saltatory emergences, and the latter adaptations to the external environment. These were more radical versions of punctuated equilibrium than that initially advanced by Eldredge and Gould in 1972. However, the latter later allow the possible involvement of developmental macromutations, as well as catastrophic global events. The punctuationists’ use of terms like “quirky” and “herky-jerky” echoes Francis Galton’s “changes in jerks.”44 But although they now embrace epigenetic ideas, they can barely bring themselves to say “saltations,” the word for evolutionary leaps traditionally used by non-Darwinists. One of the original defenders of gradualism, R. A. Fisher, author of The Genetical Theory of Natural Selection (1930), argued that evolution by natural selection demanded gradual change because epigenetic saltations must be maladaptive—too many complex systems would be thrown out of order. Embryonic aberrations were not functional advances that added anything to survival and reproduction; they were hopeless monsters, doomed at or before birth. However, Fisher’s real objection to

Paradigm Drift

51

epigenetic saltation as a mechanism of evolution was the same as Darwin’s: It would make natural selection redundant. When the genetic code was discovered, it could be seen how a simple change in a single nucleic acid base could alter the primary structure of a protein, and possibly confer selective advantage. At the time, that seemed to be all neo-Darwinism needed. The importance of chromosomal position effects, gene interaction, and hierarchical arrays of regulators was ignored. Richard Goldschmidt’s hopeful monsters were laughed at as the wild speculations of a man who was not too sure about the reality of the gene in the first place. Also, Barbara McClintock’s research on “jumping genes” and her explanations of the evolutionary role of these “controlling elements” were resisted by the conservative molecular biologists of the Modern Synthesis as spurious, and even mystical, for more than 20 years after they first heard of them.45 Neo-Darwinists nevertheless agreed that there were a number of unusually important novel events in the history of life. Ronald Fisher proposed that recessives might randomly accumulate in a population until suddenly expressed as dominant innovations. Ernst Mayr (1960) attempted to accommodate them in the Modern Synthesis in his essay “The emergence of evolutionary novelties.” G. Ledyard Stebbins (1974) identified about 70 of these important emergences in the history of life, and then commented that their extreme rarity justified focusing attention on the much more common action of natural selection on simple adaptations. If incongruities cannot be subsumed, let them be discounted as rare and special cases. Darwin did just that with Naudin’s mosaic theory of heredity, which had anticipated Mendelian genetics. Luck Another phenomenon that was rejected early on by the early modernists is the effect of chance, as opposed to natural selection, on the distribution of alleles in populations. Fisher had envisioned species as being made up of large populations with unrestricted gene flow, so that low-probability events would be meaningless. Sewall Wright’s concept of genetic drift, whereby random change could, in small populations, lead to distributions of alleles affected more by luck than adaptive qualities, was opposed in the 1930s. But it is now subsumed by the Modern Synthesis, along with Mayr’s founder effect, a principle of some consequence to emergentism.46 Founder effect occurs if a small population is set up in isolation at the fringes of the parental population, with a restricted gene pool that has low genetic variability. The founders might only be a few individuals, or even one gravid female animal or one selfpollinating plant. Their unusual gene combinations produce unusual organisms, to the extent of starting a “genetic revolution.” Another view of happenstance that came in the wake of DNA sequencing research was Motoo Kimura’s contentious suggestion that many neutral mutations, lacking

52

Chapter 1

detrimental impact, but with no identifiable selective advantage, have accumulated in all genomes.47 Taking the concept of natural experimentation a step further, Susumo Ohno thought that repetitive DNA could mutate out of sight of natural selection, provided that there was an original gene performing its original function.48 This can also happen at the organ level, as we will see in chapters 2 and 4. What about the role of chance in large-scale exogenous changes? Cuvier believed the discontinuities of the fossil record were due to natural catastrophes. Under stable conditions, all organisms were so perfectly integrated internally, so adapted to their conditions of life, that any change would be detrimental, and evolution was consequently impossible. Catastrophes caused extinction and change, but not evolution—organisms that already existed elsewhere could move into the devastated environments if they were right for the part. Biologists now believe that Cuvier was wrong about evolution and right about the natural catastrophes. Yet he came close to the truth about evolution too. Adaptation reinforces dynamic stability and resists evolutionary change, and extraordinary circumstances are required for it to be overcome. Being in the right place at the right time, especially in the face of catastrophe, is another piece of flotsam that has drifted up to the Modern Synthesis. David Raup (1991) and Stephen Jay Gould (1994) agree that natural catastrophes have been a major influence on evolution by making new environments available to the lucky survivors. But as Samuel Butler might have put it, cunning is required as well as luck.49 By “cunning” he implied physiological and behavioral adaptability, as well as intelligence. Among the lucky survivors of catastrophes, the most cunning organisms would have the greatest potential to undergo substantial diversification. If this were properly understood, evolutionists would pay more attention to the emergence and operation of adaptability, instead of sidelining it as an epiphenomenon of natural selection. Attempts to define physical rules of complexification barely hint at the evolution of adaptability in organisms, because the more complicated life gets, the more the physical rules are superseded. This will become a major theme in the development of an emergence theory (chapters 2, 8, 9, 10). Exceptional conditions, such as environmental catastrophe or rare transitions to new environments, are part of the conventional story. But according to neoDarwinism natural destruction and elimination are filters, if not creative forces of selection. Nothing is said about what’s making the coffee. That is, how did qualities emerge that let the organisms pass through the filters as if they didn’t exist, and survive the consequences? The idea of release from an ecological straitjacket, combined with the concept that there is more than luck involved in subsequent biological evolution, could be a lifeboat for some of the troubled passengers on board the foundering Modern Synthesis. Indeed, the term “ecological release” has crept into neo-Darwinism, but it has done so married to “new selection pressures.”50 One might also expect neo-

Paradigm Drift

53

catastrophists to question the role of natural selection. If catastrophe provides new environments, or old ones that have been swept clear of their former occupants, they will have a clear field. The filters have no significance for the emergent newcomers, except to prevent competition and predation by those that cannot penetrate them. Selection theory ignores the causally prior adaptabilities that permit penetration of new environments, arguing that once the survivors have collected their wits and gone about their business again they are simply subjected to different and stronger selection pressures that create adaptive innovations, speed up their spread in populations, and fill the empty ecological niches. If you find this persuasive, I know of a certain bridge in Brooklyn that happens to be up for sale. Darwin himself wrote: “When the individuals of any species are scanty, all the individuals, whatever their quality may be, will generally be allowed to breed, and this will effectually prevent selection.”51 If he had explored all the implications of that dangerous idea, he might have realized that the very ground of the concept of natural selection, namely artificial selection by plant and animal breeders, was shaky. “Artificial” Selection Darwin gave artificial selection pride of place in The Origin of Species as the best analogy for natural selection. But Alfred Russel Wallace said it was the worst analogy: . . . no inferences as to varieties in a state of nature can be deduced from the observation of those occurring among domestic animals. The two are so much opposed to each other in every circumstance of their existence, that what applies to the one is almost sure not to apply to the other. Domestic animals are abnormal, irregular, artificial; they are subject to varieties which never occur and never can occur in a state of nature. . . .52

Populations of domestic animals released from human care reverted to the wild type. In contrast, natural species would diverge indefinitely, because there would always be variants that had a better chance to survive under natural conditions. But Wallace went too far when he said “never occur and never can occur in a state of nature.” Of course they can, but when they do they might not fit in with the existing dynamic stasis. Neither Wallace nor Darwin delved deeply into the more fundamental question: How could those new variants have originated before the tests of breeder or nature? However, domestic breeding has a great deal to tell us about what can happen among unusual congregations in the absence of natural barriers, and how that might be of evolutionary significance—a logical extension that Wallace failed to make. Wallace’s conception of selection pressure reinforced his opinion that artificial selection was irrelevant to evolution. Domestic varieties had to be protected from such natural interference. It also illuminates my argument that they are relevant, in that they demonstrate what is possible in the absence of such “pressure.” I wonder if

54

Chapter 1

Wallace was comparing himself, the naturalist toughing it out in the wild, to dilettante breeders speculating from the comfort of their Victorian armchairs. Wallace’s rejection of the agricultural model did not persuade Darwin. Although the latter knew that under domestication plants and animals showed greater variability, including the production of “monstrosities,” his whole causal hypothesis was founded on the analogy with artificial selection. What was missing from Darwin and Wallace’s thinking was that neither artificial nor natural selection was creative. What was selected was generated by unknown mechanisms of biological change. And in both cases, the selection resulted from the quality of that biological change. William Bateson, Hugo De Vries, and I. N. Vavilov, who were particularly involved in the genetics of domestic organisms, were focused on the discontinuity of evolution. Indeed every biologist is aware of the great range of phenotypic possibilities, such as homeotic gene expressions in Drosophila that can survive only in protected laboratory populations. Selectionists get excited about the “enormous evolutionary change” that can be effected by the artificial selection of such traits as bristle numbers. To me it seems more like non-evolutionary change, but in either case it was achieved by excluding natural selection and genetic dilution. Many experiments with laboratory populations no longer resort to conscious selection for particular traits. Instead they imitate nature by modifying the living conditions for the experimental organisms, i.e., by neo-Darwinist lights, they change the selection pressures, to see what features are correlated with differential survival or reproduction. No matter how successful such experiments are, they do not explain the origin of the correlated features. Moreover, the many predictions based on hypothetical changes of selection pressures are hit-or-miss, and for the concept to have any theoretical rigor there would have to be more hits. The conventional response to such arguments is that the bottom line is always differential reproduction, so my distinctions don’t matter. But this merely demonstrates the reverse Looking Glass logic of both Darwinism and the current formal definition of natural selection. The effect is differential reproduction, but we should be looking for and then discriminating among causes. Under unnatural circumstances where a supernatural agent—in this case the human experimenter or breeder—is literally selecting the organisms to be mated, and consciously removing natural influences, the result is bound to be differential reproduction. So there is no contradiction to worry about? Now about that bargain bridge. In domestic animal breeding programs, “sports” appear from time to time. These are emergent novelties, or hopeful monsters that are distinctly unlike their parents. Their hopes are realized by breeders who deliberately protect them from harmful influences. They also ensure that these rare emergents reproduce with their own kind, protecting them from genetic dilution. Inbreeding in minute populations, whether in the wild or by breeders, brings out quite unusual phenotypes. “Under domestication,” Darwin

Paradigm Drift

55

remarked, “it may be truly said that the whole organisation becomes in some degree plastic.”53 And he knew that some of the products were “monstrosities” that were bred from to introduce new strains. In nature, mere monstrosities did not count. Dmitry Belyaev (1979) found that if Arctic foxes were selected for breeding entirely on the basis of tameness, hormonal bias brought out dormant genes and produced a wide range of sports. Close inbreeding was not a factor in these experiments. A necessary generative condition is the high serotonin levels found in non-aggressive types, and high estradiol and progesterone that improve fertility and attract males to females. The pineal gland is particularly plastic under the conditions of domestication, and its modification has extended the breeding season.54 Consequent behavioral changes were similar to those found in tame dogs, such as tail erection and wagging, ear drooping, and licking their human handlers. Piebald sports occurred, some remarkably like border collies and hence less than ideal for the manufacture of fur coats. Belyaev’s program excluded foxes with the most competitive or “naturally adaptive” traits, such as aggression and fear of humans. And his observations demonstrate how artificial selection is a much better model for the transformation of radical evolutionary experiments—hopeful monsters—into successful emergences in the absence of natural selection and the diluting effects of large populations. He called the process he discovered “destabilizing selection,” to contrast it with his mentor Ivan Schmalhausen’s “stabilizing selection.” The latter was part of the process of “autonomization,” whereby environmental influences on the organism were brought under organismal control. Populations were also stabilized, except at the stressful fringes. (What biologists imply by the word “stress” is that physiological and behavioral responses to environmental extremes are forced to operate at or beyond their limits, resulting in nervous, hormonal, energetic, and reproductive exhaustion.) The syndrome of stabilizing selection results in a population which throughout most of its range has had its stress minimized through co-adaptation of the parts that make up physiological systems. L. L. Whyte (1965) used a similar concept, which he called “internal selection,” to mean an increase in internal organismal efficiency that decreased the effects of physiological stress. Despite the fact that this aspect of physiological evolution fits quite nicely into the selectionist paradigm, population biologists have lost sight of Schmalhausen’s physiological and embryological inferences. They take stabilizing selection to be no more than regular culling of the more extreme allelic variations, and subsume it under normalizing selection. “Destabilizing selection” is a paradoxical term since in the case of Arctic foxes the selection is a deliberate human action, and the purpose is not differential reproduction, but optimal (economical) production of the chosen breeding population. And indeed natural influences that would usually repress emergents that differed hormonally and behaviorally are absent. Nor is the stress of an extremely harsh

56

Chapter 1

climate an issue. What Belyaev was observing was an assembly of animals, which already had certain behavioral and physiological characteristics, into an unnatural population that would not be able to survive wild conditions. It wasn’t so much that the population was being destabilized as it was that its constituent organisms were being destabilized, physiologically and behaviorally, because of hormonal disruption of epigenetic regulation. Variations produced under these circumstances were not entirely random, nor were they always beneficial to the animals or their breeders. Despite the unfortunate choice of destabilizing selection as its name, Belyaev had identified a phenomenon of general evolutionary significance. Although his research required conscious human interference, Belyaev supposed that the same phenomenon could occur naturally, especially under stressful conditions.55 What he did not say is that, although destabilization could occur under natural conditions, normal wild types and the agents of natural selection would have to be reduced for there to be any interesting consequences. Belyaev’s prediction that destabilization would occur in the wild under stressful conditions has been borne out by studies of Siberian hamsters, emperor penguins, and naked mole rats, whose case histories will be examined in chapter 4. Socialization produces the same effects in animals that are not under severe stress in the wild, though environmental influences remain important. For example, the proportions of different castes of eusocial termites are hormonally manipulated through differential feeding of the juveniles. Predation stress results in the production of more soldier termites. And their day-to-day behavior, including nest building, is pheromonally affected. Despite the power of analogy between natural and artificial selection, neoDarwinists now tend to emphasize the differences between the processes, since they think, like Wallace, that conditions of artificial selection would rarely occur in nature. (I regret that having to call a real act of intentional selection “artificial,” and a metaphor “natural” makes the discourse hard to follow.) Like Wallace, they have not thought it through. If they had looked closer at the significance of artificial selection, and like Belyaev had sought to find similar natural conditions, they could have avoided a vexing epistemological problem raised by Karl Popper. In 1959 Popper argued that selection theory is unsatisfactory because it does not suggest nullifying experiments, due to the difficulty of controlling differential survival and reproduction. Quite to the contrary, such experiments have been conducted, for perhaps as long as 12,000 years, with striking results that do indeed refute selection theory. Ever since humans started breeding plants and animals from marginally useful wild stock, removal of the agents of natural selection has released a cornucopia of large seeds, grains, succulent fruits, and numerous domestic animals whose “hopeful monstrosity” now appears normal to us. These may have been generated in part by the artificial selection of broodstock that interacted genetically, epigenetically, physiologically, and

Paradigm Drift

57

behaviorally in novel ways. This was not premeditated, since it resulted from intentional prevention of genetic dilution, for the purpose of preserving desirable novelties. None of this was generated by selection; instead it resulted from a series of natural experiments protected from natural selection by the breeders. The goal of plant and animal breeders was never to cause progressive evolution in their fields and flocks, or even to originate new species—much more sensible to improve a sheep by starting with a sheep and stopping with a better one. However, it is now possible to genetically engineer new species, and since the mechanisms of epigenetics are becoming better known, it might be possible to scramble them enough to produce emergent novelties. Better, perhaps, to confine such experiments to computer models. But, as I will argue later, the origin of species is little more relevant to evolution than natural selection. The major focus of emergentism is progressive improvements of adaptability in the wild that precede speciation. There have been many phases in the course of evolutionary history where adaptability allowed organisms the freedom to persist in stressful fringe environments, to survive natural disasters, and to push on into new territory in the company of similar adventurers. Whatever causes the progressive improvement of adaptability has the effect of wider distribution and greater diversification. That would seem to be a major evolutionary phenomenon, but it is barely considered by the Modern Synthetists. So what do they make of the causes and effects of evolution? Causes and Effects The cause of artificial selection has to do with the intention and action of the breeder, for example, putting a thoroughbred mare to stud with a race-proven stallion. Selectionists, however, turn a real intention into a metaphorical one. Causal analysis of domestic breeding demonstrates that breeders are not responsible for the emergence of desirable changes in domestic populations, such as flower color, coat patterns, and novel morphologies. True enough, the action of the breeder to protect the selected broodstock from the stresses they might find in nature, and to prevent genetic dilution, hurries the program along. However, what caused those novelties to arise in the first place should be far more significant for the evolutionist. This brings us back to a question I asked in my introduction: What cloud of unknowing allowed the Modern Synthesis to confuse cause and effect in evolution? The miasma began to develop in the late eighteenth century, with Erasmus Darwin’s materialism. Trying to escape the earlier chaos of prime or final causes, such as “The Intention of the Creator,” Darwin followed David Hume’s example and put cause and effect in a chronological relationship: CAUSE AND EFFECT may be considered as the progression, or successive motions, of the parts of the great system of Nature. The state of things at this moment is the effect of the state of things,

58

Chapter 1

which exist in the preceding moment; and the cause of the state of things, which shall exist in the next moment.56

Neo-Darwinists would admit that differential reproduction is the effect of a variety of earlier causes including mutation, competition, and sexual selection. Then they would argue that differential reproduction changes the proportions of alleles in the next generation of the population, and thus becomes causal. There is certainly a change in the probabilities of allelic mixing as a consequence, which matters to an organismal biologist as well as a population biologist. But—and it’s a big BUT—these new causes are really the knock-on (or domino) effects of the original causes of generative change. The deepest logical pitfall here is the confusion of differential reproduction with simple organismal reproduction. Reproduction, especially if it has a sexual mechanism, but without the “differential” qualifier, is itself an important generator of evolutionary change. It is the point in evolution where the gametes that contain the organism’s evolutionary past are combined in a way that can rearrange the evolutionary future. The mechanisms and processes that worked before are innovatively born again into the organisms of the next generation. Our neo-Darwinists would probably seize on the previous sentence to say: “Whatever works, or what is adaptive, is what is selected. Therefore there’s more of it in the next generation; energetic inefficiency and inability to beat the competition are diminished until they reach the point of no return—so, what’s your problem?” My response is that “whatever works” was generated in the first living organisms. The history of interactions between those organisms and their environment, along with the experimental nature of reproduction, are causally sufficient to explain evolution. Improved usefulness or adaptiveness arise as qualities of some of those experimental evolutionary novelties. Thus, evolutionists should concentrate on the generation of change and regard differential survival and differential reproduction as epiphenomenal to evolution. When George Romanes described natural selection as the ultimate cause of evolution (1892–1897), he was trying to get around the problem of tautological definitions of natural selection like “survival of those that survive.” In 1961, Ernst Mayr reemphasized the separate categories of “proximate causes” and “ultimate causes” in his analysis of biological causality. Thereafter natural selection was classed as the ultimate cause of evolution, and the causes of the changes that were selected became proximate.57 It is an unfortunate semantic accident that “proximate” hints at triviality, and “ultimate” gives a sense of importance, almost as strong as “final cause”— a Creator, or Aristotle’s “Prime Mover.” Mayr has taken a hard line on this, coming around finally to equating ultimate cause with evolution. In One Long Argument (1991) he admits to being puzzled by an embryologist who says that the evolutionary significance of development is not intelligible within neo-Darwinism. And he responds that studying the mechanisms of proximate causation “has never been the job of the

Paradigm Drift

59

evolutionary biologist.”58 Thus, at the end of the twentieth century Mayr put himself on the same platform where E. B. Poulton pontificated at its beginning: “So long as individual variation is present, so long as it is hereditary, it does not signify how it is produced. . . . So long as it is there it is available, and Natural Selection can make use of it.” Mayr further remarks that in some publications on evolution and development in the 1990s proximate and ultimate causes are “hopelessly mixed-up.”59 But they cannot be quite so mixed up as those publications that confuse cause and effect. In this book I am going to argue that the proximate causes exhaust the search for evolutionary causation, and that the ultimate cause of natural selection belongs to non-evolutionary biology. Ultra-Darwinism Niles Eldredge thinks that there has been a paradigm shift in the causal theory of evolution, and that natural selection is once again taken to be the all-sufficient cause. It should be worth re-examining what he calls “the Bible of ultra-Darwinism,” Adaptation and Natural Selection by George Williams (1966). Its author wanted to remove worrisome accretions such as Waddington’s genetic assimilation, and vague, or informal concepts of organization and progress. His statement of intent is “to purge biology of what I regard as unnecessary distractions that impede the progress of evolutionary theory and the development of a disciplined science for analyzing adaptation.”60 Ergo, evolutionary theory can progress (by improvements in the application of selection theory to adaptation) but evolution cannot. He immediately infers that before the emergence of life there existed molecules in the primordial soup that could catalyze their own replication: This is a common chemical property. Even a water molecule can catalyze its own synthesis. Only rarely would a molecule be formed that would produce chance variations among its “offspring,” and have such variations passed on to the next “generation,” but once such a system arose, natural selection could operate, adaptations could appear, and the Earth would have a biota. The acceptance of this account of the origin of life implies an acceptance of the key position of the concept of adaptation and at least an abstract criterion whereby life can be defined and recognized. We are dealing with life when we are forced to invoke natural selection to achieve a complete explanation of an observed system. In this sense the principles of physics and chemistry are not enough. At least the one additional postulate of natural selection and its consequence, adaptation, are needed.61

The second paragraph of this quotation would seem more than enough to justify Eldredge’s judgment about ultra-Darwinism turning natural selection into a creative force. But Williams takes it further by echoing Poulton’s earlier sentiment that the generative processes of evolutionary change don’t matter:

60

Chapter 1

We must take the theory of natural selection and use it in its simplest and most austere form, the differential survival of alternative alleles, and use it in an uncompromising fashion when a problem of adaptation arises. . . . The principle of natural selection is not, as a rule, used by biologists in an adequately disciplined fashion. It is usually applied to problems like that of long-term morphological changes, as seen by paleontologists, or to problems of ecotypic specialization (usually climatic), and cladogenesis. These phenomena make easy demands on a theory of adaptation. Most of the conclusions on patterns of speciation would be much the same whether based on Lamarckian, nineteenth-century Darwinian, or modern genetic concepts. . . . Darwin’s or even Lamarck’s concepts form a perfectly adequate basis for explaining most of the phenomena of systematics.62

The Williams Bible was followed by the New Testaments of Dawkins’s The Selfish Gene (1975), Wilson’s Sociobiology (1975), and Maynard Smith and Szathmáry’s The Major Transitions in Evolution (1997). Their ultra-Darwinism burns with the same flame of uncompromising austerity that attracts the moths of “evolutionary psychology.” If not a new paradigm, it is certainly a potent ideology with a fine hubris. Its elitist preemption of criticism is exemplified by Williams’s remarks about progress: I believe that my point of view on the subject of progress and of changes in the mechanism of adaptation is really the prevailing one in the laboratory and the field and in the technical literature of biology. It is mainly when biologists become self-consciously philosophical, as they often do when they address nontechnical audiences, that they begin to stress such concepts as evolutionary progress. This is unfortunate, because it implies that biology is not being accurately represented to the public.63

Score one to Williams for the prophetic accuracy of the opening sentence. As to misrepresenting evolutionary biology, the risks of self-consciously philosophical exposition would seem to apply to ultra-Darwinism as easily as to progress. But it is progress that has been lost. Richard Dawkins, inspired by Williams, finds progressive evolution acceptable to the adaptationist “where progress implies improvement in whatever adaptation a specific lineage happens to exemplify. In this sense, evolution is deeply progressive, and must be so if Darwinian natural selection is to perform the explanatory role that we require of it, and that it alone can perform.” This is quoted from an undated flyer for Dawkins’s video lecture “Is Evolution Progressive?” It may be flogging a dead horse to say it again, but since the video was “produced with the cooperation of the Royal Society, London, and the British Institute of Biology” it deserves another whack. Dawkins is essentially agreeing with Williams that progress is nothing but adaptation, and that it must be so defined to justify the creative role of natural selection, and its dominance of the evolutionary paradigm. What of Williams’s accusation that biology is being misrepresented? The ultras proclaim behind the Looking Glass that metaphorical forces are creative; when challenged, they come out to admit that they are not. And all the time they whisper that the poor dears can’t grasp the complexities of the argument, and any dissent

Paradigm Drift

61

should be stifled because it will give support to those Special Creationists. Reductionist metaphysics misrepresents evolution as the product of the action of metaphorical forces on molecular abstractions. Misrepresentation is a chronic condition of ultraDarwinism. Religious Fervor Paradigms whose popularity is underpinned by polemic, consensus, and belief have a lot in common with religion. They regularly need new prophets to reinterpret them in modern language, and they need puritanical zealots to keep them from terminal moribundity. Re-inventing Darwinism as ultra-Darwinism only demanded an exorcism of demons, not a qualitative shift. Its roots go all the way back through molecular reductionism, selfish-gene-ism, the Williams purge, population thinking, the Poulton manifesto, and the Weismann purification, to the Biblical language of The Origin of Species that I cited in the opening paragraph of this chapter. Where a principle is polymorphic in its formal definition, and common usage, it is easy to lose sight of reality and tempting to reify figures of speech like natural selection and selection pressure. Religious metaphors such as the cleansing of the temple are common. Williams uses “the light and the way” for natural selection in Adaptation and Natural Selection (1966). This Biblical misquotation is repeated by E. O. Wilson in Consilience: The Unity of Knowledge (1998). Richard Dawkins concurs that Williams’s 1966 book is the Bible of ultra-Darwinism.64 Dawkins, in turn, has disciples who regard him as their prophet: In some ways, Richard Dawkins has been the Martin Luther of biology. He’s the guy who cut through all the theological mysticism that grew around the true evolutionary church and asked, “What’s the big question?” The big questions are the questions you can answer. Any question you can’t is by definition tiny and uninteresting.65

As you may recall, Luther preached salvation by faith alone; reason was “the Devil’s Whore”—so much for reasonable questions, big or small.66 There is no denying the evangelical success of the ultras. Like the Scottish Covenanters, they stand in the open with sword in the one hand and the Bible (of Darwinism in this case) in the other. It is by no means a last stand: their conventicles are drawing curious congregations. Who then stands in opposition to the ultras? Eldredge occupies the middle ground along with the “naturalists,” who are less prone to make assumptions and less “theoryladen.” They are, he says, closer to the original ideal of the Modern Synthesis, ever ready to absorb new ideas about molecular biology, population biology, speciation, phylogenesis, and ecosystems. “We see these entities as simple outcomes of the dual fact of organismic life: economics and reproduction.”67 Without denying his success at exposing evolutionary extremism, I would de-emphasize Eldredge’s particular entities and try to teach him more about “the facts of organismic life” by addressing

62

Chapter 1

symbiosis, physiology, behavior, environmental causation, epigenetics, and evolution itself. The “catastrophe, constraints, and contingency” crowd are not that far from moderate neo-Darwinism either. Therefore the crisis will not be resolved by any of these modernist factions. Now Where? If modernists can’t pull it off, can postmodernists? They argue that the authoritarian Modern Synthesis should not be replaced with another dictatorial doctrine but should simply be abandoned in favor of a smorgasbord of relativistic ideas such as structuralism, post-Lamarckism, and complexity theory. However, not only does the scalpel of forensic epistemology antedate postmodernism; it can also be turned on postmodernism. Francis Bacon, who warned against natural theology, saw both the ultra-Darwinists and postmodernists coming: They who have presumed to dogmatize on nature, as on some well investigated subject, either from self-conceit or arrogance, and in the professorial style, have inflicted the greatest injury on philosophy and learning. For they have tended to stifle and interrupt inquiry exactly in proportion as they have prevailed in bringing others to their opinion: and their own activity has not counterbalanced the mischief they have occasioned by corrupting and destroying that of others. They again who have entered upon a contrary course, and asserted that nothing whatever can be known, whether they have fallen into this opinion from their hatred of the ancient sophists, or from the hesitation of their minds, or from an exuberance of learning, have certainly adduced reasons for it which are by no means contemptible. They have not, however, derived their opinion from true sources, and, hurried on by their zeal and some affectation, have certainly exceeded due moderation.68

I once read this Bacon quotation to a literary postmodernist, and he responded: “Spoken like a true modernist!” I’m not sure if he meant me or Bacon. However, to remain whimsical for just another moment, I see myself as an oxymoron: a postmodernist with a mission. It might be tempting to evacuate the grounded flagship of neo-Darwinism with a zealous fleet of philosophical shrimp boats serving tasty treats, but I believe they would eventually raft up with the Modern Synthesis and be subsumed. Postmodernism gives fair warning that any new theory might settle into another dynamic stasis resistant to change. Accordingly we need to ask what the minimal requirements for a different evolutionary synthesis might be, as opposed to a genteel shift of emphasis, or a flotilla of fresh hors d’oeuvres. I will offer you a different cuisine, on a table of emergentism designed to assemble relevant but disparate ideas closely enough to form some novel associations. It will present evolution as a distinctly different kind of process than that proposed by Darwin and the neo-Darwinists. A

Paradigm Drift

63

general acceptance of evolution as a saltatory as well as a gradual phenomenon would be a radical departure, especially if saltations were understood to occur in months rather than millennia. However, evolution cannot be viewed as an exclusively saltatory process, since many novelties emerge at critical threshold points along continuous evolutionary lines. The causal theory of the Modern Synthesis must be replaced, since its crucial agency, natural selection, reinforces stasis and obstructs evolution. This is not to be judgmental about the selection syndrome itself. It is bound to happen when any resource is limiting, whether or not evolutionary change has occurred. And it is bound to be re-established soon after evolutionary change gets the opportunity to be expressed. In this sense, evolution is the author of its own fate, but selection is never the cause of evolution. Even in lineages where natural selection could be said to produce directional adaptive change, there may be prior directive, or self-amplifying molecular mechanisms involved, so that directional selection may be epiphenomenal, rather than the primary cause of change. As a rule of thumb, the simplest way to deal with the reverse Looking Glass logic of selectionism is to insert a negative into every positive claim it makes: “Natural selection is [not] the cause of evolution” etc. Selection theory would still apply effectively to quantitative demographics, and to change in community structure, and to our understanding of how things settle into energetic stases and then stay that way. These should be recognized as important aspects of life as a whole, but their quantifications would cease to be regarded as causal explanations of evolution, and none too soon. Søren Løvtrup (1975) observed that we should see evolution as an emergent process, and that history would judge the phase of evolutionism dominated by neo-Darwinism in the same way as we now see the quaestiones disputatae of the Medieval Scholastics, such as how many angels could dance on the head of a pin.69 Even if we all shared his sense of history, the conventional scientific approach to improving evolutionary theory would still be to establish one’s expertise in a particularly relevant field and to amass hard data that could not be refuted. But research specialization can come down to counting and sorting the angels on the pin-head. Their dance can be so mesmerizingly intricate that it is easy to forget that there are larger arenas of evolutionary causation, and that the significance of a particular one may be quite trivial overall. With due deference to innovative, specialized research—the only professional avenue widely open to academic biologists—a synthesizing overview cannot emerge from minor local adjustments of facts and figures. Although Darwin had a good foundation of observation and argument, his historical placement in the development of biology was a great advantage to getting across a general theory to an eager audience. Now, we might ask “What has become of the scientist more concerned with the impact that the work of others has on him, than with the impact that his work

64

Chapter 1

has on others?”70 My own initial approach is to propose an integrating framework that has room to accommodate the available evidence and to validate some useful historical ideas that have been cast away by the Modern Synthesis. This I will commence in the next chapter. Meanwhile, since there is not enough room within the Modern Synthesis for radical change, my first order of the day has been a clean sweep. That many emergences have identifiable adaptive advantage at their point of origin complicates the argument that emergentism is a distinctly different approach to evolution. But it does not all come down to “natural selection did it.” To perceive a phenomenon as adaptive is not to explain it. A blinkered search for an adaptive advantage will usually find one that meets the ecological demands of the moment. But it may be only one aspect of a greater adaptability that enhances the fundamental integrity of the organism, allowing it to endure, regardless of ecological conditions, but without conferring identifiably superior fitness. In Materials for the Study of Variation, With Especial Regard to Discontinuity in the Origin of Species (1894), William Bateson warned: No one who has ever tried to realise the complexity of the relations between an organism and its surroundings, the infinite variety of the consequences which every detail of structure and every shade of instinct may entail upon the organism, the precision of the correlation between function and the need for it, and above all the marvelous accuracy with which the presence or absence of a power of a structure is often compensated among living beings—no one can reflect upon these things and be hopeful that our quantitative estimates of utility are likely to be correct. But in the absence of such correct and final estimates of utility, we must never use the utility of a structure as a point of departure in considering the manner of its origin. . . . It thus happens that we can only get an indefinite knowledge of Adaptation, which for the purposes of our problem is not an advance beyond the original knowledge that organisms are all more or less adapted to their circumstances. No amount of evidence of the same kind will carry us beyond this point. Hence, though the Study of Adaptation will always remain a fascinating branch of Natural History, it is not and cannot be a means of directly solving the problem of the origin of Species.71

It would not be too much of an oversimplification to say that the subsequent history of the drifting paradigm was largely devoted to attempts to prove such dissent wrong. It also ignored his suggestion that what was needed was evidence of a new kind, and more knowledge of the principles of evolution: “It is submitted that the Study of Variation gives us a chance, and perhaps the only one, of arriving at this knowledge.”72 Bateson used the word “variation” in the same sense that I use “emergence”: as that process of evolutionary innovation characterized by discontinuity. Although Darwin believed that variation was virtually continuous, he knew that a proper understanding of variation would open new fields of evolutionary explanation. However, “variation,” as a biological term, has been deprived of that significance by the Modern Synthesis, coming down to minor changes which are merely fuel for the ecological engine of

Paradigm Drift

65

“evolution.” And variation has also been corrupted to imply molecular changes so small as to result in phenotypic continuity. It is thus is no longer available as a synonym for emergence. But I hardly need to change Bateson’s exhortation to put it thus: to get to the essence of evolution we have to explore the generative conditions and causes of emergence. In them we might find the explanatory power that ultraDarwinism claims for natural selection. From the outset I have taken the calculated risk of outraging conventional wisdom about the causes of evolution, before making the case for generative emergentism. Some dissenting evolutionists prefer to avoid controversy by paying lip service to the notion of the ultimate causal role of selection while giving priority to the proximate generation of evolutionary novelty. Contra-selectionist and pro-emergentist issues could be dealt with separately, but it is more satisfactory and less evasive to contrast them before finally re-integrating them. Nevertheless, if my attempted revision of natural selection were totally discarded, the need for generative hypotheses to explain how novelty arises would still be past due.

2 Prologue to Emergence

There ought to be some kind of formal structure that captures the essence of emergence. —Jack Cohen and Ian Stewart, 19941 We will not understand life and living organisms until we understand emergence. —John Holland, 19982

Whenever I reveal my lack of faith in selectionism, I am asked “What do you put in its place?” The straightforward answer is “emergentism.” Yet for anyone who realizes that selection theory is little more than a theory of demographics, or “book-keeping,” the answer is “There’s nothing to replace, because there is no causal theory of evolution.”3 Selection theory, if overdependent on metaphorical circumlocution, is adequate to population biology. However, it fails to account for the generation of evolutionary change. Although this was clearly stated by St. George Jackson Mivart in 1871 in his book The Genesis of Species, the void has never since been filled. Now, more than a century later, theorists such as Goodwin, Cohen, Stewart, and Holland are awake to the need for generative principles of emergent evolutionary novelty. Through the Looking Glass, abstractions and metaphors are creative forces. In contrast, the emergentist focuses on the organism, looking in at its genetic, developmental, and physiological structure and looking out at its environment and its responses to it. We don’t have to abandon Darwinism altogether; it established the historical theory in the first place. Moreover, emergences often have significant demographic consequences—or dominate the ledgers—resulting in new ecological equilibria. These effects must eventually be reintegrated into a larger biological synthesis, as complementary to the causes of evolutionary change, and a full account of their history. While the Modern Synthesis was under construction, evolution, the organism, mind, and life itself were abandoned. Any attempts to retrieve them are worth a try, but how shall we go about it? Emergence involves the discontinuous production of novelty, including greater degrees of complexity. Some introductory remarks about that are in order.

68

Chapter 2

Complexity Modern complexity theory is sometimes called “anti-chaos theory,” since it tries to make sense of how local complex stabilities can persist in chaotic non-biological systems. Complexity theory sensu lato, as it pertains to biological systems, has quite a long tradition within biology, in the writings of E. S. Russell, D’Arcy Thompson, Ludwig von Bertalanffy, Joseph Needham, Karl Weiss, Ivan Schmalhausen, C. H. Waddington, and the early emergentists. But whatever direction complexity theory has come from, it has not been from Darwinism or selection theory. Complexity theorists take an interest in biological complexity and its evolution, but they usually work with computer models rather than real organisms. And although they often allow a secondary role for natural selection, their quest is not for selection pressures that produce complexity, but for intrinsic principles that might govern its autonomous emergence and development. They might, for example, ask: “Does the behavior of gases and water, in forming temporary, semi-stable vortices and turbulences, inform us of the nature of evolutionary complexification? Does the behavior of a purely physicochemical system in dynamic equilibrium add to what we already know about dynamic equilibria in living systems?” The short answer is “Probably not.” Any complex system can be compared to any other complex system, and equations can be derived, without offering any idea of how a new emergent level arises from the lower generative level. Biological innovations are not bound by the laws of physics in the same way that purely physical systems are. In fact, interpretations of complex physicochemical systems have probably been informed more by the application of organismal ideas than the other way round. All the same, explorations of non-equilibrium thermodynamics and chaotic attractors are fascinating and stimulating.4 If their models demystify the complexification of organisms to some extent, good and well. Most biological complexity theoreticians seek to avoid controversy by accommodating natural selection before moving on to the main business. For example, Michael Conrad (1990) writes that evolutionary organization results from “self-complication” and asks “Why does evolution work?” He writes: “The reason is not to be found solely in the magic optimizing power of variation and selection. It is as much due to the organizational structure that undergoes the variation. Evolution works because this organization is amenable to evolution, and because this amenability itself increases in the course of evolution.”5 Conrad equates organizational improvements with adaptability, which has the additional qualities of “reliability” and “stability.” So be it. They are self-sufficient and improvable properties of the kind that have characterized life since it first emerged, without any necessity for the redundant judgment of natural selection. For Conrad, however, progressive organization can only persist by “hitchhiking along with the advantageous traits whose appearance it facilitates.”6

Prologue to Emergence

69

Although he is on the right road, a more holistic interpretation would see organismal progressive organization not as the hitchhiker but as the whole vehicle and its contents. It not only carries traits that are immediately advantageous, but has multifunctional features that will be advantageous under different circumstances. They are not picked up one at a time along the way. Furthermore, organizational progress might have no adaptational features in the short term. Its quality is its integrity, which is sustained through greater adaptability. The organism can now go on doing the same thing when conditions change, and do different things when conditions stay the same. But such self-sufficient improvements may not amount to a competitive phenotypic fitness in a stable ecosystem. Furthermore, specialization may lead to regression of former adaptability. For example, most fish that live permanently in fresh water have lost their ancestors’ adaptability to deal with salt water as well. In a similar vein, Stuart Kauffman’s book Origins of Order: Self-Organization and Selection in Evolution (1993) begins as follows: We must understand how such self-ordered properties permit, enable and limit the efficacy of natural selection. We must see organisms in a new light, as the balance found, the collaboration achieved, when natural selection acts to further mold order which pre-exists. In short we must integrate the fact that selection is not the sole source of order in organisms.7

My own conviction is that natural selection is neither the sole source, nor any source of order at all except demographic equilibrium. Any increase in order improves organismal integrity, and so will persist if its energetic requirements are met. Thus, it will be perceived to have selective advantage. But its energetic requirements may not be met if there is strong competition from the pre-existing organisms in a state of ecological dynamic stability. Persistence may be better realized in fringe environments where competition is low. Once established, the new emergents should hypothetically shake down into an even more resistant equilibrium. However, some lineages have been able to periodically overcome that resistance. The Reduction of Complexity In the context of complexity, it is necessary to talk about “reductionism,” a word that is often used very loosely because “everybody knows what is meant.” If that were really so, there ought to be a clear distinction between reduction and reductionism, but biologists tend to lump them together. For example, “the triumph of reductionism,” as the discovery of DNA structure is often called, should be “the triumph of reduction.” “Reduction” is an epistemological tool for understanding complex structures by analyzing their components. Its successful application is demonstrated by Watson and Crick’s 1953 discovery of DNA’s structure, which suggested the basis of reproduction

70

Chapter 2

and the genetic code. The problems of reproduction and heredity, originally asked at the organismal level, were reduced to the molecular level, and came to fruition when returned to the organism. Reduction is a neutral epistemological method, whose value is to be measured in how usefully it can be referred back to the original problem. However, reduction with “-ism” tacked on represents a biased worldview. In Consilience (1998), E. O. Wilson remarks: Beyond the mere smashing of aggregates into smaller pieces lies a deeper agenda that also takes the name of reductionism: to fold the laws and principles of each level of organization into those at more general, hence more fundamental levels. Its strong form is total consilience, which holds that nature is organized by simple universal laws of physics to which all other laws and principles can eventually be reduced. This transcendental world view is the light and way for many scientific materialists (I admit to being among them), but it could be wrong. At the least, it is surely an oversimplification. At each level of organization, especially at the living cell and above, phenomena exist that require new laws and principles, which still cannot be predicted from those at more general levels. Perhaps some of them will remain forever beyond our grasp. Perhaps prediction of the most complex systems from more general levels is impossible.8 [emphasis added]

Wilson values reductive methodology and understands the obstacles to predicting complex systems from their simpler constituents. But despite his reservations, the sentence that I have italicized infers that to understand complexity, final reduction to physical laws is necessary. Apotheosis of reductionism to “the light and the way” is transcendental, but not original.9 In the past it was holism that was committed to the flames for its transcendental qualities. Wilson rescues holism as “consilience by synthesis,” but subordinates it to the reductionism of “total consilience.” It is an easy step from Wilson’s relatively balanced view to a doctrinaire reductionism that infers that the only way and light is through “upward causation” driven by physical laws. Its natural disciple is the molecular biologist. Some students of complexity are critical of reductionism, having a greater appreciation of emergent properties of organismal levels that cannot be predicted from a knowledge of molecules. J. H. Woodger (1929) warned against “intolerant abstraction” that abandoned the whole once it had been reduced.10 Along the same lines, William Wimsatt (1997) remarks that the difference between a biophysicist and a theoretical biologist is that the first is only interested in the invariate aggregative properties of organisms once they have been put through a blender and fractionated by centrifugation. He believes (regrettably, not always a valid proposition) that the biological theoretician is interested in how those properties interacted in the organism to make it a whole greater than the sum of its parts before it went through the blender. Wimsatt calls the “nothing-but” atomism that he understandably dislikes “vulgar reductionism.”

Prologue to Emergence

71

Systems Reduction Evolutionary complexification has parallels with the developmental complexification in the individual organism as it grows from zygote to adult. In How the Leopard Changed Its Spots (1994), Brian Goodwin writes: The sciences of complexity lead to the construction of a dynamic theory of organisms as the primary source of the emergent properties of life that have been revealed in evolution. These properties are generated during the process known as morphogenesis, the development of the complex form of the adult organism from simple beginnings such as an egg or a bud. During morphogenesis, emergent order is generated by distinctive types of dynamic process in which genes play a significant but limited role. Morphogenesis is the source of emergent evolutionary properties, and it is the absence of a theory of organisms that includes this basic generative process that has resulted in both the disappearance of organisms from Darwinism and the failure to account for the origin of the emergent characteristics that identify species.11

I will assess this salient in chapter 10. But Goodwin and several others who follow similar themes often use expressions such as “emergent order,” “emergent properties,” and “emergent evolution” without distinguishing, defining, or even indexing them, as if they are epiphenomena of universal rules or algorithms of complexity. Instead of reducing complexities to the fundamentals of physics and chemistry, they apply systems reduction. This means that they examine simpler organized systems that appear to have the same qualities as the complex ones. These are easier to analyze, while sufficiently ordered to help make sense of the really complicated ones. This is much more useful than simplistic molecular reduction.12 But even if we knew of an algorithm that dictated how a new level of complexity is obtained at a given emergent level, it would not necessarily apply to, and so predict or explain the generation and nature of the next emergent level up. The power of systems reduction also attracts Jack Cohen and Ian Stewart, authors of The Collapse of Chaos (1994). They keep vulgar reductionism strictly at the end of a barge pole while they quant among emergent phenomena. Nevertheless, for them, natural selection is the evolutionary “metarule” that causes complexification. Flaunting the redundancy that lies in “survival of the fittest,” they declare: “The complexity of a living organism is the result of an evolutionary game played over huge periods of time according to a rule so simple that it is really just a logical tautology. Winners win.”13 Then they remark, quite unselfconsciously, “If a theory is so flexible that it can ‘explain’ anything whatsoever, it’s probably nonsense”!14 If this is heavy irony, I have to admit to being caught out. One of Cohen and Stewart’s models of an emergence anticipates what is to come in this book, and at the same time illustrates their paradoxical view of natural selection as the cause of evolution. An imaginary creature experiments with the accumulation of hydrogen and takes to the air as a balloonist. “To begin with—no competition up there—they would thrive, just as the winged creatures did along our evolutionary

72

Chapter 2

track.”15 Subsequently they imagine an environment where the emergent forms fill up the available space. Now, “the scene will be set for genuine evolution, by competition, selection, and reproduction of those . . . that survive.”16 Plus ça change—by this Looking Glass logic the generative process, emergence, and the thriving of its creature in a space free of natural selection is not “genuine evolution.” (I hope this is more heavy irony.) In the real world, by way of contrast, evolution tries through natural experiment, and succeeds in the absence of natural selection that enforces stasis. Surveying and explaining the natural experimental methods of evolution, and discovering how the opportunities for success present themselves, are the tasks of emergentism. Origins of Emergentism The roots of emergentism lie in the classical Greeks’ concept of order coming out of chaos, and their debate over whether it happened by chance or necessity. Was it the occasional effect of unpredictable combinations of simple forms, or the result of design in nature? Aristotle made much of the nature of wholes, especially those that were greater than the sums of their parts. Emergent properties are what make them so. The great second-century anatomist and physiologist Galen also distinguished between the resultant and emergent qualities of wholes.17 The question of how chance or necessity might be involved in the production of more complex wholes remains open, even among materialistic evolutionists. Emergent change could be purely a rare and random accident that increases complexity of structure and function. Or it could be an expression of the operation of natural physical and biological tendencies toward complexity. The eventual answer may be found to combine both. Hegel had an emergentistic, metaphysical vision of the revolutionary progression of life from non-living to living to conscious and then to spiritual. Also, Kant perceived that in the development of an organism simple parts interact to produce a progressively complex series of emergences of functional forms, in contrast to a machine that is an assemblage of pre-manufactured parts. Writing in the Enlightenment era, Kant was influenced by the thoughts of Buffon and his compatriots on natural adaptation and the effect of the environment, but he did not apply his analysis of emergent development to biological evolution. And he came to believe that living phenomena were irreducible because they were purposeful, rather than that they naturally evolved through discontinuous stages of emergent complexity. Kant’s distinction between the assembly of mechanical components and certain kinds of complex systems carried over into the nineteenth century, when John Stuart Mill (1843) pointed out that chemical compounds have novel features that cannot be predicted from a knowledge of their elements. George Henry Lewes (1874–1875) noted that these “emergent” qualities are distinguishable from additive “resultants”. And

Prologue to Emergence

73

emergent qualities are especially characteristic of organisms. When biological elements are compounded, novel features are generated. Only the natural combining qualities of the components need be considered. Although this should have made any additional vital force redundant, the vitalists loved it, and transcendental emergentism became the bedfellow of Henri Bergson’s doctrine of the élan vital. This vital spark sprang from a universal primordial consciousness, to energize evolution. Even so, the transcendentalists tried for a degree of objectivity. Henry Drummond, though an evangelical member of the Free Church of Scotland, eschewed creationistic vital forces in The Descent of Man (1894): When we pass from the inorganic to the organic we come upon a new set of laws—but the reason why the lower set do not seem to operate in the higher sphere is not that they are annihilated, but that they are over-ruled.18

Drummond also realized that greater complexity brought greater adaptability. The self-styled realist Samuel Alexander regarded godliness as a primordial property of matter from which deity would finally emerge, life and mind being the intervening stages. Yet in Space, Time and Deity (1920) he came down to earth sufficiently to establish that emergences had properties that overruled the demands of the lower levels of organization: The higher quality emerges from the lower level of existence and has its roots therein, but it emerges therefrom and it does not belong to that lower level, but constitutes its possessor a new order of existent with its special laws of behaviour.19

Compare Drummond and Alexander with John Holland (1998): If we turn reductionism on its head we add levels. More carefully, we add new laws that satisfy the constraints imposed by laws already in place. Moreover these new laws apply to complex phenomena that are consequences of the original laws; they are at a new level.20

C. L. Morgan and Emergent Evolution The chief progenitor of evolutionary emergentism, Conwy Lloyd Morgan, was a zoologist, comparative psychologist, and student of T. H. Huxley. His first major contribution to theoretical biology, in 1896, was his independent formulation of “genetic assimilation,” as a version of it was later called by C. H. Waddington. J. M. Baldwin and H. F. Osborn coincidentally wrote about the phenomenon in the same year. Morgan began to lose faith in Darwinian evolution by natural selection, shifted toward emergentism, and he had a powerful influence on Samuel Alexander. While Morgan’s Emergent Evolution (1923) established the central idea, it was deficient in examples and explanations. The Gifford Lectures, on which the book was based, demanded a natural theological context, and he struggled with general metaphysics and the role of the Divine. While his training, research and historical

74

Chapter 2

placement uniquely qualified him to provide a materialistic biological interpretation, he never got round to it. And by declining to commit to sudden, large-scale, qualitative evolutionary change, he was actually less adventurous than Alexander. Morgan concluded that an emergence might have the appearance of saltation but was best regarded as “a qualitative change of direction or critical turning point.”21 Thus, continuity was preserved, even where it looked as if a novelty had leapt into existence. This is a legitimate but not exclusive example of emergent phenomena to which I will refer, in deference to Morgan, as critical-point emergence. This may strike a familiar chord with those aware of Per Bak’s principle of self-organizing criticality, which posits that gradual cumulative changes may reach stages where novel characters are revealed.22 These critical-point emergences are equivalent to the threshold effects that are important in developmental evolution. Moreover, some may be predictable if they follow a linear progression. Therefore, unpredictability is not a sine qua non of emergence, as the early transcendental emergentists believed. However, despite Morgan, some emergences are both unpredictable, and saltatory. These are the ones that grab the attention of emergentists, and so some philosophers of science call them “strong emergences.” They call critical-point, threshold events “weak emergences.” They are not terms that I will use, since they invite invidious comparison. As to causation, Morgan had a more complete interpretation. The causes of emergence could be either “immanent” (= intrinsic, arising autonomously from the dynamic structure of the organism) or “transeunt” (= extrinsic, having an external cause). The latter would obviously depend on environmental conditions, but Morgan did not delve further into such contingencies. I am now attempting what I think Morgan could and should have done 80 years ago. The original conceptions of emergent evolution, while never totally denounced as heresy, had three strikes against them. First, there was the proposal that emergences were caused by vital inherent tendencies. Second, it made much of spiritual emergent levels at a time when transcendental theories were being eclipsed by materialism. Third, there was intense competition from the second wave of neo-Darwinism, based on gradualistic theoretical population genetics. Christopher Caudwell wrote a clear, simple, and non-vitalistic view of emergent evolution in 1935. But it was not published until 1986. This parallels the case of Friedrich Engels’s The Dialectics of Nature, written in the late nineteenth century but not published until 1926. Both of these Marxist authors had found an emergentism that they could reconcile with materialism. Caudwell, who died during the Spanish Civil War in 1938, saw evolution as a series of steps: . . . at each step the environment has become different—there are different laws, different problems, different obstacles at each step even though any series of steps despite its differences

Prologue to Emergence

75

has certain general problems, laws and obstacles in common. Each new step in evolution is itself a new quality, and this involves a newness which affects both terms—organism and environment.23

Recent Approaches to Emergence In Evolutionary Theory: The Unfinished Synthesis (1985), I discussed the early history of emergentism and holistic biology, and suggested that the concept of emergence could be the foundation of a novel evolutionary synthesis. Shortly before that, in 1983, the neurophysiologist Roger Sperry had written about the evolution of mind as an emergence, a point of view that had persisted among some psychologists and anthropologists, during the half-century eclipse of biological emergentism. The emergence of cognition, or mind, still enjoys a large independent literature. Sperry described his own recognition of emergence almost as an epiphany: My long-trusted materialist logic was first shaken in the spring of 1964 in preparing a nontechnical lecture on brain evolution in which I was extending the concept of emergent control of higher over lower forces in nested hierarchies to include the mind-brain relation. I found myself concluding with the then awkward notion that emergent mental powers must logically exert downward causal control over electrophysiological events in brain activity. Mental forces were inferred to be equally or more potent in brain dynamics than are the forces operating at the cellular, molecular and atomic levels.24

This illustrates what Sperry called the “interactionist” relationship between emergence, hierarchical structure, and causation. It is not difficult to understand his point. When we think about something we can conjure up sensory memories, and with a little practice learn to stimulate adrenaline release or even alter our heart rates. Later commentators on emergent evolution have suggested a variety of ways to assess its causes, properties and epistemological value, which I deal with in chapter 8. For the time being, I will simply add that several biological authors besides Jack Cohen, Ian Stewart, and John Holland (quoted at the beginning of this chapter) are aware of the importance of emergence and the need for its formal treatment. Susan Oyama (2000) surveys the challenge of an unknown terrain thus: In what ways does the emerging organism evoke, seek, produce, or reliably have supplied for it the very stimuli and conditions it requires for further development? How does it interact with them? What are the ways this entrainment can be derailed? What are the possibilities for compensation when such derailment occurs? Surely these possibilities are themselves dependent on conditions that may vary. How do some early variations become amplified, while others are damped? Propensities, potentials, stability, and lability are characteristics of systems in progress, and it is up to us, using the more productive lines of inquiry from behavior genetics, physiology, ethology, and related disciplines to analyze developmental systems of interest to discover how various inherited interactants are related.

76

Chapter 2

When an environmental variable is found to be confounded with genotype, we have located an aspect of developmental order, an association that may well order future development. These are the associations we must investigate in order to understand how each structured developmental state arises, for it is this state that will be an important determinant of subsequent interactions. Such confoundings, then, are not dross to be eliminated by better experimental or statistical control. They may be the very gold we are seeking, and the point of control is to illuminate the ways mutual selectivity works.25

To give her due priority, I note that Oyama was already discussing these matters in her 1985 first edition of The Ontogeny of Information: Developmental Systems and Evolution. Brian Goodwin has also staked a claim in this venture, and Stuart Kauffman writes that a theory of emergence is the proper goal of complexity studies.26 John Holland, a computer scientist and a psychologist, approaches a mathematical theory in Emergence: From Chaos to Order (1998) and concludes that the next steps to be taken in the development of a theory of emergence will have to proceed along the following route: Usually the persistent patterns that arise in these generated [emergent] systems are not easily anticipated on direct inspection of the generators and constraints. The most lucid examples of emergence arise when these persistent patterns obey macrolaws that do not make direct reference to the underlying generators and constraints. [Examination of a] minimum of three levels will be necessary to arrive at a representative body of theorems relevant to emergence: mechanism > agent > aggregate. Formally we might only consider two levels, treating all higher levels as recursions of these basic relations. With no very sound argument to back me, I think the threelevel study will be more revealing and more productive as a first step.27

It is important that Holland says “body of theorems” rather than a single formula. Even at a third level there is neither a universal rule nor equation of emergent evolution that applies to the whole. For two that are highly distinct, compare the emergent complexification of eukaryotes by symbiosis and the autonomous emergence of mind as a novel quality of the hominid brain. “Aggregate” is an unfortunate term, since it is commonly used for resultant wholes in contrast with emergent wholes that are greater than the sum of their parts, and I am sure Holland does not intend otherwise. There is a profusion of mechanisms and novel macrolaws (= emergent rules of organization) in all organismal hierarchies, whether they be molecules, unicells, symbiotic complexes, multicellular organisms, populations, societies, demes, communities, or ecosystems. So they need to be sorted out and clarified. In the following five chapters, I combine facts and concepts pertaining to emergent biological processes before proceeding to theoretical generalizations in chapters 8–11. Instead of dealing with each of the hierarchical levels that I have just mentioned, starting with the simplest and ending with the most complex, I deal with causal arenas that match traditional divisions of biology. In each of these, hierarchical structure can be discerned.

Prologue to Emergence

77

All biologists agree that organisms are hierarchically constructed, and that such a perception is epistemologically useful. But I must caution that some reductionists reject the implication that hierarchical structures contain irreducible emergent properties. As Jaegwon Kim (1999) presents their argument in “Making sense of emergence,” emergent properties are only meaningful if they are mechanistically causal, and, if they are, their hierarchical levels must be reducible to physics. Hence, they conclude that any talk of an emergence doctrine is empty. The hollowness of such inferences is exposed by Peter Corning in his 1998 essay “The synergism hypothesis”: The term “emergence” is especially popular among die-hard reductionists, because it implies that wholes are merely epiphenomenal effects of laws and causal processes that can be fully illuminated at lower levels.28

The problem of reductionist-emergentists is that, in attempting to model complex physical systems, they assume that all emergent phenomena, whether empty or fruitful, can be explained from the bottom up. However, they will not succeed in this goal, because the operation of new constraints and interactions at new emergent levels is refractory. In fact they have given themselves an excuse for not trying! My thesis is that emergent properties do exist, and that they have a mechanistic causation that operates up, down, and across hierarchical levels. They might be reduced to physical principles by anyone with the competence to deal with all of the hierarchical levels involved, and the urge to do so. But they cannot be predicted from the bare laws of physics without smuggling the properties of higher emergent levels down to the foundational laws. For example, the operation of a nerve might be “reduced” to the action of electrons or ions. But to go from electricity to the nerve requires passage through several levels of organismal complexity, none of whose emergent properties can be predicted or even projected from physics. My own pragmatic preference is to go straight to the generation of biological change, taking hierarchical structure into account as I proceed. Indeed, my next chapter on symbiosis and association does go from simple subcellular structures to complex societies. The chapter on physiology shows how functional anatomy has gone from a simple state of conformity with environment to an independent, organized state of homeostasis, through evolution within existing hierarchical levels as well as by the addition of new ones. Developmental biology illustrates how hierarchical structure and its emergent levels are rebuilt from simple origins in every generation. This can then be related to evolutionary history. For each causal arena I intend to outline what there is, how it got that way, and how it can be explained by evolutionists. Without a few more conceptual guidelines my exposition might seem aimless or dislocated. Therefore, imagine we are going on a field trip, and our first stop is the emergence circus. As we set out, I offer the new explorer the following program of

78

Chapter 2

emergent evolutionary performances, and a field trip checklist will come at the end of the chapter. Outline of an Emergence Theory When I discussed the flaws of the Modern Synthesis, I argued that any alternative had to include a better generative hypothesis than the action of natural selection on genetic variation by mutation. If the necessary set of principles can be identified, we will have advanced toward a theory of emergent evolution. Emergence is commonly associated with the discontinuous, unpredictable generation of brand-new biological features, with increased complexity and self-organization. But there are varying degrees of discontinuity and novelty. Although non-biological phenomena can emerge in time with novel qualities, my version of emergentism largely deals with all processes of evolutionary change, together with their generative conditions, mechanisms, and emergent properties. It combines the emergentism of the 1920s with those of later commentators, as well as my own opinions. The numerical success of new emergents in populations depends on unpredictable but affective contingencies, or accidents if you will. These also have to be taken into account, and this aspect of emergentism finally interfaces with the demographic theory of selection. If most of the history of life and the ordinary activities of organisms are non-evolutionary, or merely trivial in their evolutionary content, it would be constructive finally to clarify what how evolution and equilibrium have complemented each other in the history of life, and that will be considered in chapter 11. For the moment I will begin to correct the reverse Looking Glass logic of the Modern Synthesis by negating its axioms: Evolution is [not] caused by natural selection. Cumulative adaptation is [not] progressive evolution. Evolution is [not] exclusively gradual. Evolution is [not] slow. The origin of species (i.e. speciation) is [not] the pivotal process of evolution. Evolution is [not] change in the distribution of alleles or genotypes in populations. Now we need something to put in their place—an equivalent series of positive aphorisms to anticipate my thesis:

Prologue to Emergence

79

Evolution happens through natural experimentation. While it involves reproduction and differentiation, it is distinct from differential reproduction. Progressive evolution is a cumulative, hierarchically ordered series of emergences with qualitative novelties of morphogenesis, physiology, behavior, and association. Emergences produce wholes that are greater than the sum of their parts. Emergences are sudden on the biological time scale, happening at the speed of biochemical, physiological, developmental, and behavioral reactions. They may lead to adaptational specialization, which limits evolutionary options, or to greater organismal adaptability, which proffers greater freedom of choice for individuals. Emergence may be directly caused by environmental change, though ultimately genetically based. The actions of individuals are crucial in shaping the evolution of their descendants. Arising from the previous two points, many emergent conditions that are now regarded as genetically heritable are preceded by non-heritable events (behavioral actions, environmental effects and developmental, physiological and behavioral responses.) The uniqueness of the phenotype, and hence the dynamics of the underlying genome as a whole, is more relevant to evolution than individual genes and proportions of alleles in populations. Emergences may be all-or-none saltations, or critical-point (threshold) innovations, both complexifying and regressive, that punctuate allometric trends or orthogenetic drives. (“Regressive” may seem contradictory; but sometimes the loss of an inhibiting character opens up new possibilities for experiment.) Spurts of emergent evolution are followed by diversifying evolution (adaptive radiation), which peters out into long periods of stasis reinforced by the agencies of natural selection. These equilibrium phases temporally dominate the history of life. Reproductive isolation preserves emergent novelty. Speciation may simply be a byproduct. The aspect of reproduction also distinguishes biological emergence from physicochemical emergence.

80

Chapter 2

To supplement the above bare-bones synopsis, there follow some capsule commentaries on principles or phenomena relevant to the modes of emergent evolution. Modes of Emergent Evolution We have seen that evolution is defined by selectionists as changes in the distribution of alleles in populations. More enlightened modern synthetists recognize the importance of phenotypes, which equate to whole organisms. And some define evolution as transgenerational changes in organisms, i.e., they are only evolutionary if they are heritable. But what does evolution mean to an emergence theorist? I have already adumbrated two kinds of evolution: progressive and adaptational. Progressive evolution is equivalent to increase in complexity, which is discussed above. The most important point is that progressive evolution brings increased adaptability, which allows organisms to persist in an ever wider variety of environments. This kind of evolution is most likely to be saltatory and to result in emergent properties that will govern future evolutionary events. The non-biologist is very much interested in examples of progressive evolution, such as the emergence of complex organisms from bacterial ancestors, and the emergence of humans from the lower primates. Adaptational evolution occurs in relation to specialized habits and distinct habitats. Its most radical forms are divergences that result from emergent improvements in adaptability, combined with some kind of release from the constraints of hypostatic natural selection. It produces specialized body forms and habits such as are found in most classes of organism, and is therefore the most visible to the observer, and most identifiable to non-biologists as evolutionary adaptation. Nearly all aspects of progressive and adaptational evolution are heritable, and require a DNA component as part of the system. Emergence theory will engage the question of when DNA enters into the evolutionary process. It will also subsume hereditary factors which are not DNA-based. Moreover, environmental factors that consistently and persistently influence the operation of hereditary mechanisms must also be accounted for. The modes of emergent evolution involve natural experiments that result in a combination of discontinuous, saltatory evolutionary emergences, and self-amplifying drives that reach critical-point emergences—in contrast to slow, gradual evolution through the accumulative action of natural selection on random gene mutations. Emergentism primarily attends to the progressive evolution of coordinated complexity. Diversifying evolution is not to be seen as a simple process of adaptive radiation, but as the result of a combination of causes and effects including epigenetic and physiological emergence, and orthoevolutionary mechanisms other than directional selection, as well as adaptation in the traditional sense. Natural selection is taken to be a real syndrome of intermediate causes and effects arising from emergent

Prologue to Emergence

81

qualities. It leads to adaptation and ends in rigid specialization. And, in doing so, it obstructs progressive evolution by establishing dynamic stases affecting development, physiology, associations, and ecosystems. At the same time it reinforces the foundations, especially physiological coordinative conditions, from which new emergences can spring. When new emergent properties appear the consequences that are most apparent in the fossil record are diversifying evolution, but this may not occur until the restraints of natural selection are removed. Those emergent innovations that are able to escape those restraints are not bound by old rules—an emergent homeotherm need not slow down when the ambient temperature drops. The primary causes of emergent evolution are epigenetic (embryological), symbiotic-associative, physiologicalbehavioral. And the environment affects all of these primary causes, through direct physicochemical changes, genetic assimilation, and the disequilibration of ecological stasis. Catastrophic change not only weakens or removes the obstacle of natural selection; it might also physically initiate new epigenetic evolutionary experiments. For example, heat shock from sudden climatic change or catastrophic volcanic and bolide impacts might increase mutability, through the epigenetic influence of stress proteins. The major evolutionary emergent properties of animals amount to increased adaptability in individuals. Through emergence to higher levels of physiological regulation, a more sophisticated homeostasis confers greater freedom of choice and action. The freedom of greater adaptability has historically led to an increased potential for diversification (adaptive radiation) of the novel, emergent types. Eventually the emergence of mind allows analysis of options before commitment to action. Charbel El-Hani and Sami Pihlström note in their 2002 essay “Emergence theories and pragmatic realism” that “the early pragmatist tradition is characterized by the frequent use (by philosophers like James, Dewey, Mead, Charles Hartshorne, Sidney Hook, and Ernest Nagel, among others) of notions such as creativity, freedom, evolution and novelty, which quite naturally find a place in emergentism.”29 I enthusiastically combine such optimistic notions throughout this book, although I find the biological processes and consequences of emergences more interesting than their ontological analysis. But how can we assess the potential of any organism to undergo emergent evolution? Modularity Progressive evolution of complexity is possible because of biological modularity, which involves a multiplicity of varied organic holons, or modules, such as exons, genes, multigene modules, protein domains, biochemical pathways, cells, and organs. These units can be shuffled, replicated, transposed, and mutated. Therefore, redundant copies are necessary components of evolutionary experiments. More important for the proper development and functioning of the organism are their

82

Chapter 2

integration, interaction, and self-organization, which involve a multiplicity of modifier genes, gene-regulating proteins, hormones, receptor molecules, sensory organs, neurons and their dendritic connections, and a circulatory system. Also implicit is a hierarchical ordering that allows them to be regulated efficiently en masse while leaving them some freedom to operate and to change independently. This is important both for the evolutionary potential of the organism and its adaptability as an individual. It can be generalized that, if a self-reproducing entity has a modular structure, it can become more complex, it can remain the same, or it can regress. In evolutionary history there are striking examples of all those options. Adaptability Progress, or complexification, can be largely equated with increase in the adaptability of organisms, a feature that maintains organismal integrity, one of the emergent qualities of life itself. In 1896, James Baldwin called it a “blanket utility,” because right from the beginning it can survive a range of challenging circumstances. As Darwin surmised, improvements in the ability to “persist in being” are self-sufficient, in contrast with qualities that are naturally selected according to environmental conditions. Since the full potential of adaptability is out of sight of natural selection, its fitness rating is only as high as that of its operating features under limited circumstances. Specific elements of adaptability may become specialized if internal and external conditions remain constant, with the result that adaptability often regresses. Adaptations on the other hand are genetically fixed, inflexible, and appropriate only to certain pre-existing internal or external environmental conditions. Wholes That Are Greater Than the Sums of Their Parts Emergences are innovations that in their most radical form constitute new levels of organized complexity. Their novelty comes from novel relationships among preexisting systems, combined perhaps with contingent catalytic additions—in lay terms they produce wholes that are greater than the sums of their parts. For example, at the molecular level the whole of a hemoglobin tetramer loads and unloads oxygen much more efficiently than four separate hemoglobin monomers. This is because the subunits, if separated, cannot cooperatively change shape to accelerate the process. At the organismal level, such emergent wholes are well exemplified by endosymbioses that generated eukaryotic cells with a constellation of new relationships. When multicellularity was generated it possessed only minor emergent properties. Yet the differentiation of the many cells realized the Aristotelian concept of “separation of offices and the concurrence of efforts,” which he compared to the harmonious efficiency of a well-organized city state. Some emergences are not all-or-nothing saltations of this kind, but may arise at critical points in continuities. For instance, if, over many generations, continued allometric growth causes a continuous increase in

Prologue to Emergence

83

wing size, in an animal that can already flap and glide, there will be a critical point where lift exceeds drag, and true flight emerges. This kind of emergence is quite predictable from aerodynamic principles. But the biological pattern of growth that reaches the threshold of flight is harder to reduce. New organismal wholes obey the fundamental rules of physics, chemistry, and biology, but their own novel emergent properties can override those of the lower hierarchical levels. For example, the brain’s neural emergent property thought can override reflex, such as a rush of hormones produced by an older emergent level. Emergences create material wholes that are greater than the sum of their parts. For example, in symbioses, a compound toxic to a host organism may be used by its symbiont to make useful molecules. Emergences are a mix of the predictable and the unpredictable, as I will establish in chapter 10. Since a higher level of organized complexity has novel emergent qualities, the search for causal evolutionary continuity from the physicochemical up to the conscious is counterproductive. Because of emergent discontinuities, a unified theory of the evolution of biological complexity will not be found at the physical level. While a search for commonalities at all emergent levels is desirable, the novel properties of each level require individual focus. Key Innovations Another word of caution: Biologists who have been trying to translate what I have been writing about emergences into their own more familiar language are probably asking “Is he referring to key innovations?” It is typical of Western epistemology to particularize from a confusion of possible causes. So it is not surprising that to clarify emergent evolution we might seek crucial, catalytic, novel phenomena that are the keys to success in entering new environments, or to getting the jump on the competition. It is also necessary to know all the generative conditions of emergence. They amount to all of the molecular, organismal, and environmental qualities that are necessary for an emergence to occur. These are roughly equivalent to the “initial conditions” of cosmologists and complexity theorists. “Key innovation” is useful for a generative condition that comes along later than the others either as a crucial catalytic variation or a new contingency. New contingencies are to be found in molecular, physiological, and environmental relationships, as will become clear from examples in the following chapters. Key innovation can also refer to a particular quality of evolutionary emergence. However, total focus on a single crucial novelty is to see a tree and ignore the forest. For the qualities that result from an emergence I am going to borrow the expression “constellation of emergent properties,” a metaphor that several authors have already applied. Therefore, as well as cautiously identifying key catalysts and key effects, I am going to attempt to delineate the minimal requirements for generation, however many there might be. And I will do what I can to sketch all of the consequent properties. There might be

84

Chapter 2

only one or two, if any, that are of immediate advantage, but they could not exist without the others, and it is often the case that the others eventually participate in the diversifying evolution that follows major biological emergences. Evolvability Modern selectionists, who have little constructive to say about adaptability and progress, are now daringly asking themselves “Does evolvability itself evolve?”30 There is a marked contrast between the answers of the Modern Synthesis and those of emergentism; even the question itself has a different meaning in the two contexts. For the selectionist it is axiomatic that natural selection causes evolution, so the question means “Does evolvability have selective advantage?” Moreover, all that evolvability means to a selectionist is the ability to generate potentially useful variations that can be selected. If that is factored in, the question becomes “Does the ability to generate potentially useful variations have selective value?” That simply returns us to the conventional view of neo-Darwinism that genetic variety in a population is useful in fluctuating conditions. And genetic variety need only involve a mosaic or polymorphic population with an existent wide range of alleles. An underlying ability to generate genetic novelty might be seen as a wasteful property that would tend to be culled regardless of conditions. The question of evolving evolvability demands more than an impoverished selectionist answer. Something along the lines of macroevolvability or progressive evolvability is wanted. To an emergentist, the answer starts with the emergent properties of the first life forms. In addition to maintaining their integrity by being adaptable organisms, they also had the evolutionary potential to become more complex through multiplication. Prokaryotes could only try simple natural experiments, and unicellular eukaryotes could only get a little more complicated. The pace did not really pick up until sexuality and multicellularity originated, and it accelerated yet again when terrestrial organisms with high metabolic rates and more sophisticated feedback mechanisms emerged. In other words, evolvability was itself an emergent property of the first simple organisms, and in some lineages it has indeed evolved progressively at an accelerated rate despite the obstacles placed in its way by the agents of natural selection. In this light there seems to be a certain inevitability of evolution, making the word resonate with its old meaning of “unfolding.” However, complexity was potentiated rather than contained within the founding organisms, and neither a final predetermined goal nor a predictable route is implied. Emergentism and the Historical Theory of Evolution Much of the evidence for the reality of the phenomenon called “evolution” remains the same as it was in Darwin’s time, but a shift of interpretation comes with

Prologue to Emergence

85

emergentism. Evolvability is something that Darwin thought important, though the previous section shows just how differently evolvability can be interpreted in the context of progressive evolution. And emergentism treats evolutionary tempo and speciation differently. For a start, they are treated as causally separate categories. Beginning with a Bang Evolutionary emergences may require long and gradual acquisitions of the appropriate generative conditions, and yet occur almost instantaneously, as in the case of endosymbiotic associations of prokaryotes. Some epigenetic emergences, although subject to environmental influences, are fundamentally the result of molecular changes that occur in fractions of a second, or cell movements that take only moments, in marked contrast with gradual evolution on a geological time scale. Indeed, the original cosmic Big Bang established all of the familiar physical rules for the subsequent development of the Universe in a fraction of a second. Gregory Benford writes: The Big Bang would also be better termed the Enormous Emergence—space-time snapping into existence intact and whole, of a piece. Then it grew, the fabric of space lengthening as time increased. The crucial new element here is the vibrant role of space-time itself. This, ancient Newton missed.31

Environmental change has often had an all-or-nothing, sudden impact as well, when, for example, aquatic organisms emerged onto land, when climates changed rapidly, or when bolides impacted. The “Big Bang of Biology,” as early Cambrian evolution has been called, may not be measurable in microseconds, but the time scale of the emergence of the relevant novelties may be measurable in tens of years, rather than the tens of thousands of years proposed by the most radical of neo-Darwinist punctuated equilibrists. The interpolation of periods of non-evolutionary stasis is what brings the total lapsed time into the geological time frame. Speciation Darwinists believed that speciation was the product of adaptational divergences of populations of an existent species, and that the establishment of new species was the crux of evolution. There is now a spectrum of explanation that ranges from the Darwinian position to the idea that speciation is a coincidental process that may not involve adaptation until after new species are in place, and perhaps not even then.32 Although the species concept has undeniable practical value for categorizing organisms, the “species problem,” as a taxonomic issue, could be a quaestio disputata— a “How many angels can dance on the head of a pin?” sort of question. It cannot, however, be rejected, since it is interwoven with sexual reproduction and population diversification.

86

Chapter 2

Bolder neo-Darwinists distinguish distinct kinds of “speciation event”—for example, the experiments in body plans conducted at the time of the Cambrian explosion. These speciation events might be better called “emergences,” and they cannot all be filed as Darwinian speciation and forgotten. The appearance of what a taxonomist would call a new species is not a definitive step in evolution; it may only be epiphenomenal to emergence. Reproductive barriers can be erected by geographical isolation, behavioral differences, capricious likes and dislikes within populations, or by mutual sterility between populations. By protecting new types from genetic dilution and incompatible hybridization, they are significant from any point of view. But emergentism particularly recognizes the importance of preserving not merely novel gene variants but unique organisms, along with their genetic and epigenetic combinations. The emergent organism may found a new phylum or only a new genus, depending on its emergent properties, and environmental contingencies that affect subsequent diversification. Or, appearing in the wrong place at the wrong time, it may be doomed to scant success or failure. But regardless of its evolutionary potential it always belongs to a “species.” Eventually diversification will peter out into a number of species locked into ecostasis. Another consequence of reproductive isolation combines some of the features of genetic drift and founder effect. If emergent types appear in a small population through an unusual genetic combination that then increases through drift, such creatures might realize their hopes by finding themselves in like company, and their emergent properties will characterize their future lineage. Natural selection need have no hand in determining the universality of those emergent qualities within the new population. Emergentism and the Causal Theory of Evolution Although a real phenomenon, the syndrome of causes and effects known as natural selection is not the cause but a consequence of natural experiment. Its agents hasten specialization, and fine-tune epigenetic, physiological, and ecological homeostasis. In so doing it has the positive role of consolidating the position of the new emergent, and thus it parallels the role of normal science when new paradigms emerge. There are several distinct categories of natural experiment that constitute the primary causation of evolution. They are experiments in symbiosis and association; experiments in physiology and behavior, and experiments in epigenetics and development. They affect all hierarchical levels in the organism from the molecular to the whole organism in its relationship with its environment. They all operate within internal milieux and external environments, and interact through feedback and feedforward loops. The operational details of these aspects of emergent evolution will be surveyed in the following five chapters.

Prologue to Emergence

87

Natural Experiment as a Metaphor This is subject to the same criticism as natural selection, i.e., that it seems to imbue nature with purpose. Intention is not inferred in either case, but since such figures of speech are easily abused, natural experiment needs clarification. An expression that I began to use as a glib response to the selection metaphor, it requires more substance if it is to stay the course. Nature has no intentions of playing scientist, selector, or anything else, but changes occur, and they are not always random. Selectionists would have no problem with the statement that the experiments involve spontaneous changes in DNA—they would retort that natural selection takes care of the rest, and that’s all that needs to be said. Therefore the inadequacy of this oversimplification must be exposed. Natural laboratories exist in different places, at different hierarchical levels. The biosphere is a heterogeneity of ecosystems, organisms, and their internal environments, cells, and molecules. In some of those laboratories, trials are less erroneous than in others. At interfaces between environments, the accumulation of molecules, or unicells, or multicellular organisms, make closer associations or symbioses more probable. At the gene level, accidental changes in modifiers may result in nothing significant in prokaryotes, but if they alter epigenetic programs in a developing plant or animal they may have major consequences. If they are disintegrative they do not count, and if they affect organisms by increasing their self-organization, they persist. If they have adaptational qualities they survive and reproduce differentially. When multicellular plants and animals began to differentiate, their interaction with each other and with the environment at large generated specific instead of random responses. Animals able to tolerate or accommodate to environmental change experimented behaviorally. With more resilient homeostasis, they extended their horizons. With more complex nervous systems they began to make choices. In the previous chapter I agreed with Alfred Russel Wallace that artificial selection is a bad analogy for natural selection. But to what extent does it provide a valid analogy for natural experimentation, or for the laboratory where such tests are carried out? Human control of plant and animal breeding emerged 10,000 years ago, and has advanced considerably since then. It began with whatever improvements could be obtained. Dogs that were tolerable, or even attractive for their tameness and playfulness, were able to hunt and herd; they could also be eaten in a season of woe. Domestic sheep and were first brought together on the basis of their natural herding behavior, and goats may have been camp followers of early nomads. The first grains were probably chosen consciously for their size and threshability. Natural qualities beneficial to humans were both accidentally gathered and deliberately picked. Novel broodstock organisms were brought together in the same place at the same time, and competitors, predators, wild browsers, and random back-crossing with the wild populations were excluded. Thus, protection from the agents of natural selection,

88

Chapter 2

enhancement of nutrition, and prevention of genetic dilution were guaranteed by the action of the breeders. These conditions are all available in the natural laboratories of new environments made available through organismal adaptability; and sometimes nature, like Hercules, catastrophically cleanses the Augean Stables. But these are only the circumstances under which experiments succeed. What generates the experimental changes themselves? With some justification, it has been suggested that human breeders establish the genetic propensity for change in their breeding populations through artificial founder effect. Belyaev’s Arctic fox broodstocks were not inbred, yet behavioral interactions among them and their handlers could alter their reproductive patterns through hormonal effects. This does indeed occur in nature. When it comes to adaptability, protective artificial conditions can bring out exotic cognitive features in such animals as dolphins, parrots, crows, and the great apes. Wallace was correct in thinking that such dilettantine qualities would be quashed by natural selection. But we should go further, to ponder the conditions where natural adaptabilities of physiology and behavior could be generated, and not be quashed, but immediately flourish, or hang on until propitious environmental changes occurred. To fully realize the implications of this argument, we have to understand that the whole of biospheric heterogeneity, and the interactions of its component organisms, provide for natural experiment. Wherever we enter this whole, to analyze evolution with our eyes open, we can see that local experiments have consequences that reverberate throughout. Molecular breeders may have refined their selective artifices to the actions of single genes, but to identify spontaneous mutations of DNA as the generative causes of natural experiments is to deny the meaning of the rest. Thus, the question put by John Holland (1998)—What circumstances make the unlikely more likely?—does not have a simple answer. As well as a multiplicity of conditions that generate the experiments, the nature of the laboratory determines their results. Emergence and Environmental Contingencies Emergences that result from natural experiments can be fully appreciated only in an environmental context, whether in the internal milieu or the world at large. Some environmental contingencies are biological, such as the availability of potential symbiotic partners that can interact to form wholes greater than the sum of their parts. The presence and population density of other conspecifics, or different species, can stimulate developmental changes in some organisms. Other contingencies are random physical events, such as change in climate, and catastrophes. However, some conditions that await the progressing organism are fixed and predictable, such as the availability of freshwater and terrestrial environments with distinct physicochemical features: salinity, acidity, oxygen content, gravitational pull, and light intensity, for example. This physicochemical heterogeneity of the environment may physiogeni-

Prologue to Emergence

89

cally change organisms as they move, accidentally or by choice, from one phase to another. Thus, some natural biological experiments are contingent on organismenvironment interactions. Conversely, some environmental conditions such as the oxygenation of the biosphere are contingent on biological events. These “geophysiological” adjustments support James Lovelock’s Gaia Hypothesis.33 Other significant environmental contingencies are random geophysical events, such as volcanic eruptions, earthquakes, and the movement of tectonic plates. Like the cataclysmic impacts of bolides, these can alter the course of natural experimentation and evolution by clearing the bench for new trials. Last First Words What’s in a Name? Emergence is the name of a process that may be non-biological or biological. It is manifested as the sudden appearance of a new quality, “sudden” having a time span ranging from microseconds to tens of years. The new quality results from a unique combination of generative conditions. Some of these may be rare or unlikely; some may already be in place—essential to the persistence of the complex pattern. Hence, the run-up to the apparent leap may be protracted into geological time. The generative conditions might build in a particular direction through self-amplifying mechanisms interacting with habit and habitat. This too takes time, until a threshold is reached where the new property emerges. If a strong dynamic stability prevails where a novel emergent occurs, its presence may go unnoticed until that stability is undone. If there is no resistant hypostasis the novelty will immediately increase in numbers. There are four phenomena that distinguish biological from non-biological emergences. The first is biological reproduction, which gives emergent patterns of living complexity some guarantee of persistence, despite their thermodynamic vulnerability. That is linked to the second general quality of living emergents: dynamic integrity. They are self-maintaining. Reproduction combines with integrity to add a third emergent quality: the ability to conduct natural experiments in emergent evolution. The fourth phenomenon is also experimental: organisms can match their integrity against a variety of environments, and so induce both physicochemical and biological changes in themselves. Increase in the complexity of organisms, or the acquisition of new emergent qualities, equates with increasing freedom to direct specific activities. Goal-directed sequences, from biochemical pathways to distinctive behavior, finally contribute to purposeful action or inaction.34

90

Chapter 2

Emergence by Natural Selection My opening broadside for an Emergence Synthesis asserted that natural selection, though a real phenomenon, obstructs evolution. The selectionist counterattack is that emergence, though a real phenomenon that they might reluctantly call macroevolution, is generated by natural selection. They claim that in populations, adaptive alleles, and groups of alleles that interact to produce microevolutionary, adaptational, phenotypic traits, are accumulated and concentrated through natural selection. This increases the probability of combinations that constitute qualitative emergent change. The dissonance of selectionist emergentism is illustrated in The Biology of Ultimate Concern (1967) by Theodosius Dobzhansky, an evolutionist who frequently peered beyond the hand of Darwin into the abyss. Instead of “emergence,” Dobzhansky used “transcendence,” being careful to say that he meant a real phenomenon as distinct from any kind of vitalistic essence: The biological evolution has transcended itself in the human “revolution.” A new level or dimension has been reached. The light of the human spirit has begun to shine. The humanum is born. It remains to consider briefly some of the misgivings which arise in connection with the above account of the evolutionary transcendence giving rise to man. Those who see an unbridgeable gap between the humanum and the prehuman state question the presence on the animal level even of rudiments from which the humanum could arise. Now, the point which the believers in unbridgeable gaps miss is that the qualitative novelty of the human estate is the novelty of a pattern, not of its components. The transcendence does not mean that a new force or energy has arrived from nowhere; it does mean that a new form of unity has come into existence. At all events, no component of the humanum can any longer be denied to animals, although the human constellation of these components certainly can.35

So far, so good, especially since there is still fierce debate over what humanness actually is. Of course the generative conditions must exist in the pre-human state. Dobzhansky prefaced those conclusions with the qualification that natural selection is “automatic, blind, . . . lacking foresight, [and] opportunistic.” Accordingly, “in radical evolutionary reconstructions, the emerging product is an appalling mixture of excellence and weakness. That this is the case with man is almost a platitude.”36 Now, I have already admitted that the syndrome of natural selection affects the processes of internal and ecological adjustments that follow emergence, and lead to dynamic stabilization. Yet immediately after his inspiring remarks on transcendent materialism, Dobzhansky retreated back to selectionist romanticism: . . . natural selection is in a very real sense creative. It brings into existence real novelties— genotypes which have never existed before. Moreover, these genotypes, or at least some of them, are harmonious, internally balanced, and fit to live in some environments. Writers, poets, naturalists, have often declaimed about the wonderful, prodigal, breathtaking inventiveness of nature. They have seldom realized that they were praising natural selection.37

Prologue to Emergence

91

Natural selection does not bring novel genotypes into existence. Organisms bring novel genotypes into existence. The existence of those organisms is not due to natural selection. Their existence is due to the simple emergent qualities of life: selfmaintenance, self-organization, and reproduction. Competition and natural selection are epiphenomenal to those, so the “writers, poets, and naturalists” in question had not missed the point. The “gradual accumulation of good genes” is a more recent variation of Dobzhansky’s error. It cannot be omitted as part of the process of emergence and its consequences; yet it is a genocentric oversimplification. It overlooks the fundamental similarity of all organisms at the DNA level. If selective accumulation of adaptational mutant genes were responsible for major phenotypic differences, there would be major DNA differences between ourselves and the bonobo, our nearest chimpanzee relative. But we differ by only 1.3 percent. The selectionist argument refuses to accept the saltatory nature of emergences, and the fact that the components of such interactions have to be generated before they are selected. Nevertheless, in chapter 8 I will offer the emergentistic selectionists another round of debate. My argument that the success of evolutionary experiments is obstructed by competition and predation has been heavy-handed, in order to make my point. And some critics have remarked that there are many other routes that selection can take to ensure the success of particular novelties, instead of their repression. My response is that the origin of such adaptiveness contains its fate, making natural selection a redundant concept. Therefore, understanding the origins and the qualities of novelties is the way to understand evolution. As to the proof of the pudding, sample the courses of associative, physiological, behavioral and developmental evolution that I am about to serve, before deciding that the short commons of selectionism offer the more satisfactory meal. An Emergence Synthesis In constructing the following list of contrasts between the emergentists’ and the selectionists’ approach to evolution, I have taken a contrarian stance. But the fact remains that differential survival and reproduction are measurable effects, and the agents of natural selection include real biological phenomena. They cannot be omitted from a complete synthesis. The chapters that follow immediately deal largely with the selectionist’s proximate causes of evolution, and might be considered a contribution to what Goodwin (1994) calls “a science of qualities.” Ultimately, however, we will be confronted by the need to synthesize emergentism and selectionism, which will be realized in chapter 11.

92

Chapter 2

An Emergence Formula Until now, I have avoided the question of whether there can be a general formula that applies at all emergent levels from cosmology through life to mind. The epigraph from Cohen and Stewart—“There ought to be some kind of formal structure that captures the essence of emergence”—expresses the complexity theorists’ wistful hope for the grail of a universal formula. Holland (1998) establishes a way station for such a quest, but his hopes are likely to be curtailed. Solé and Goodwin (2000) show that particular emergent levels are open to mathematical formulation, but they note that because of the non-linear nature of emergence itself, particular principles can neither predict nor be sufficient to higher emergent levels. The great allure of emergentism to some reductionists that Corning (1998) mentions comes from wishful thinking that from the mathematics of game theory, chaos theory, and non-equilibrium thermodynamics will come a predictive formula that will finally conquer the messiness of life. It is a false hope. The progressive evolution of a particular lineage might very well be made to fit a theoretical curve based on a simple equation in which increased complexity is a function of reproduction and differentiation in time. But we should not make the same error as Lamarck, about evolutionary progress being a simple time function. Although all life increased in complexity to some degree, in many evolutionary lines it ceased or regressed, and where it did not, its progress was irregular. Even if lineages that ceased to progress, or became extinct, are ignored for the sake of clarity, a simple curve of progressive evolution still could not project the irregularity of saltatory and critical-point emergences. And although causal contingencies might be expected, only hindsight allows them to be identified and placed on the curve. Therefore, even if such a general unifying formula could be approximated, emergence is such that a perfect knowledge of biological origins could not contain or predict the particular nature and timing of future evolutionary novelties. Life emerged in a universe whose cosmology is subject to physical laws. But life, without disobeying such principles, is not predicted or contained by them. One apposite developmental problem is pinpointed by Walter Gehring’s description of his quest for determinants of anterior and posterior structures in Drosophila larvae. A mathematical model of segmentation based on the Turing reaction-diffusion equations required only one determinant. But Gehring discovered two; he had not been “chasing a ghost”: . . . in this case nature did not find the most parsimonious solution. As we shall see, evolution requires a lot of tinkering, redundancy, and double assurance, that only later is streamlined through natural selection.38

While the quest for a generative theory of emergent evolution may be arduous, it must begin with the first step. That it is directly relevant to evolution rather than to its demographic consequences is a start. Putting life, mind, the organism, progressive

Prologue to Emergence

93

Table 2.1 Emergentism

Neo-Darwinism

Addresses the causes of evolutionary innovations.

Ignores the causes of evolutionary innovations.

Addresses progress, which is equated with adaptability and increased complexity.

Ignores progress. Subsumes adaptability under adaptation. Doesn’t like complexity.

Combines saltatory and gradual processes.

Denies saltatory processes.

Natural selection is an obstacle to evolution.

Natural selection is the cause of evolution.

Includes autonomous organismal cause.

Denies autonomous causes.

Stresses importance of organisms, and populations as made up of individuals with unique phenotypes.

Stresses importance of populations as gene pools, and reduction of phenotypes to selfish genes.

Environment directly affects physiology, behavior, and development.

Environment only sets conditions for adaptation.

Biological changes at the organismal level precede genetic fixation.

Genetically heritable changes are the fount of evolution by natural selection.

evolution and freedom back into theoretical circulation should have some restorative effect along the way. Table 2.1 summarizes the dissimilarities between the neo-Darwinist Modern Synthesis and an Emergence Thesis. “Natural selection is an obstacle to evolution” is more dramatic than punctilious. Evolutionary change can occur anywhere at any time. It is on the whole a random process, although I will be narrowing down the generative conditions for emergent evolution to those that are the most likely. What natural selection, or the stasis that it reinforces, actually obstructs is not qualitative evolutionary change, but the successful establishment of emergent novelty in sufficient quantity to make the kind of impact that might feature in the fossil record. Such establishment is due either to superior competitiveness, in which case it is wrongly subsumed as an effect of selection, or to freedom from competition. Commonalities of Emergentism and Neo-Darwinism Speciation is not a pivotal event of evolution in my brand of emergentism. Nevertheless, I share with neo-Darwinism the concept of speciation as a ratchet pin that prevents regression of novelty through genetic dilution by interbreeding with the older forms of organisms. It is also an inevitable concomitant of emergent qualities that permit population expansion and diversification. Evidence for the historical reality of evolution is similar in both theses, but its tempo, patterns and modes are subject to different interpretations, as demonstrated in

94

Chapter 2

the above discussion of how emergentism affects the historical theory. Emergentism has the potential to resolve and rise above the present crises of evolutionary thought. Antonio Lima-de-Faria (1986) presents an extensive table of differences between his “autoevolution” and neo-Darwinist evolution with which I largely agree. However, my immediate goal is to keep the contrasts simple, and to broaden them later in chapter 10. A more leisurely exploration of generative emergent causation follows, and by the time we get to chapter 8 we should be in a better position to assess its formal requirements as well as its limitations. We are going on an intellectual field trip, and the first part of the quest is a visit to the evolutionary circus. There are bound to be a few thrills and spills, but don’t panic! You are in good company. The faint-hearted went back home before the end of the last chapter, and never even got to the ticket office. As far as the circus is concerned, you might want to watch out for the following: mechanisms common to all causal arenas the generative conditions from which emergences spring and any features common to them all factors that catalyze emergences, provided that appropriate setting of generative conditions is in place particular contingencies, predictable or unpredictable that have affected generative conditions and the success of their emergences the adaptable constellation of multiple properties characteristic of new emergences the course of emergent evolution that has progressively led to greater self-organization, independence, and freedom of choice. I make no apology for my zoological and anthropocentric bias in the last of these points. The field trip has a shortcut. The traveler who doesn’t have the experience to appreciate the performances, or the endurance to tolerate the sweat and the greasepaint at the three-ring circus, can pick up the rest of the expedition at the exit of chapter 7, and continue with the quest in chapter 8. chapter 9 reviews the major circus performances, and leads into the goal which I hope you will find in chapters 10 and 11. Bon voyage!

3 Evolution by Association

Symbiosis, like sexuality, is a powerful force that has brought together preadapted genetic combinations, at least initially, in the absence of mutation. Symbiotic complexes are important sources of raw material for genuine evolutionary innovation. —Lynn Margulis, 19811 From the long view of geological time, symbioses are like flashes of evolutionary lightning. To me symbiosis as a source of evolutionary novelty helps explain the observation of “punctuated equilibrium” of discontinuities in the fossil record. —Lynn Margulis, 19982 If it is indeed in social organization that we find emergent evolution most manifestly at work . . . then we shall be cautious in accepting even the advice of the king of the termites on our own social problems. —H. S. Jennings, 19273

Association and Emergence Associations within and between organisms are the alpha and omega of evolution: involved in the emergence of life and in the latest adjustments to human culture. In the early stages of life on Earth, symbioses altered the biosphere and established major ecosystems. Such associations emerge quickly and generate new wholes that are greater than the sums of their parts. Therefore, it is not surprising that Lynn Margulis invokes discontinuity of innovation. Yet the first epigraph misses one fundamental point: Symbioses are not simply “important sources of raw material for genuine evolutionary innovation” (as if they had to wait for natural selection to approve before they actually “evolved”)—they are in and of themselves genuine evolutionary innovation. One of the natural experimental processes that Margulis discusses is a mix-match principle that applies to associations at many levels of organization, from molecules

96

Chapter 3

and cells to organisms and societies.4 That can be complemented by the concept of repetitive differentiation, whereby existing units are multiplied, varied, and reassociated in novel groupings. Life emerged when organic molecules mixed and matched in competent combinations. Primitive cellularity came from protobiont coalescences. Exons, the functional subunits of genes, can also be mixed and matched to produce a variety of proteins. Also, protein domains, once they have been synthesized according to the exon codes, can be mixed and matched to make novel protein molecules. Intimate endosymbioses of prokaryotes gave rise to the eukaryotic cells of protoctistans, fungi, plants, and animals. When eukaryotic cells formed multicellular associations, they could then differentiate and undergo epigenetic evolution. Logically and chronologically, the causal arena of evolutionary association claims priority for consideration as a major component both of a generative theory of evolution, and of the new biological synthesis that is needed to replace the Modern Synthesis. As Margulis infers, protosymbionts, as independent organisms, already had their own genetic systems, functionally attuned to their independent modes of life. Then, the sudden intercourse of two different genomes and cell types generated a broad constellation of emergent properties. This “flash of evolutionary lightning” contrasts starkly with eons-long accumulations of adaptational mutations. The component organisms of symbiosis were “preadapted” only in the sense that they already formed functional wholes that could persist in their own being. But in endosymbionts they had novel properties that arose from complementarity. Margulis puts it plainly in the language of emergentism: “The tendency for independent life is to bind together and reemerge in a new wholeness at a higher, larger level of organization.”5 Endosymbioses that gave rise to eukaryotic cells clearly demonstrate progressive and saltatory change. Nevertheless, emergent symbioses may be more fragile than their formerly independent constituents, and subsequent adjustments are needed to reach the most dynamically stable state, where homeostatic flexibility is achieved with minimal energy output. Once multicellular eukaryotic foundations were laid, epigenetic evolution could build upon them through intercellular and sexual associations. Thereafter, symbiotic interactions between individual organisms continued to make intermittent but major contributions to the complexity of the biosphere by founding new ecosystems. Social associations within and between individual species of insects, birds and mammals made further evolutionary emergences and divergences possible. Despite a growing trend among biologists and psychologists to reduce the properties of associations, including human society, to genes, emergentists have always understood that every evolutionary emergence contains novelties of organization that were not in the simpler levels, and were therefore not predictable from them. Herbert Spencer Jennings, quoted in the third epigraph to this chapter, cautioned particularly

Evolution by Association

97

against extrapolation across phyletic gulfs, from organisms such as social insects whose behavior is extensively gene-based, to humans whose behavior is largely cognitive. By including all of the different kinds of association in one chapter, I hope to construct a minor synthesis of associative evolution that fits the framework of emergentism. However, intercellular association in multicellular organisms is largely in the developmental arena, and will not be elaborated in this chapter. Decoding the Terminology of Associations In Evolution by Association (1994), Jan Sapp notes that A. B. Frank (1877) coined “symbiosis” for close interspecific relationships, excluding parasitism. The following year, de Bary broadened the definition to include parasites and societies, and he also implied that the evolution of symbioses did not require an active role for natural selection. The title of Sapp’s book avoids semantic ambiguity, and he keeps it simple by omitting intercellular associations, sex, and social interactions.6 However, despite serious attempts at formal definition by others, usages still vary. Some modern biologists prefer “mutualism” for symbioses in which both host and guest benefit. Yet mutual advantage is implicit in endosymbiosis, the most intimate kind of association. “Symbiogenesis” is used for the emergence of symbiosis. The term “symbiont” is used for the organism that takes up residence in a host. Yet there is no single word for the combination of host and symbiont. Margulis uses “symbiotic complex” and occasionally refers to a “chimera.” “Composite,” “consortium,” and “association” are useful general terms, but they miss the symbiotic nature of the complex. How about reviving the word “symbiote” for the symbiotic whole? Sapp cites the discussion of the Committee of Terminology set up by the Society of Parasitologists in 1933.7 They concluded that the choice of terms was “a matter of taste and usage rather than of correctness.” They discarded “symbiote” over the objections of their philological consultant, and chose “symbiont” as a word that could apply to either member of the symbiosis. “Symbiote” and “symbiont” continued to be used interchangeably, for a while, but scientific usage now favors “symbiont” for the smaller member of the complex; “symbiote” appears more frequently in science fiction. I do not like to neologize, but I am going to use “symbioplex” as the obvious combining form of Margulis’s “symbiotic complex.” I’ll try not to overdo it. Some symbioses put a curious twist on the inheritance of acquired characteristics. In my 1985 essay “Unpredicted factors of evolution: The theory of emergence revisited,” I used “the acquisition of heritable characteristics” for emergences involving gene transfer or genome pooling between previously unrelated organisms. My purpose was to draw a distinction between the phenomenon and the conventional interpretation of the Lamarckian inheritance of acquired characters. Eugene and Elisabeth Wollman

98

Chapter 3

first appreciated this distinction in 1925, and they called it “paraheredity.”8 In Evolution: The Modern Synthesis (1942), Julian Huxley had already used “allopolyploidy” for the combination of karyotypes (chromosome complements) by hybridization in plants. Lynn Margulis and Dorion Sagan (2002) now refer to these kinds of combinations as “the acquisition of genomes” in their eponymous book. The familiar modern expression “gene acquisition” implies an extrinsic source for “new” genes. It applies to the natural transformation and conjugation experiments routinely conducted by prokaryotes, where genes are the major cellular components acquired. The word “symbiosis” means an association involving two or more different species, and is the kind of relationship that has priority in this chapter. However, intraspecific associations are involved in sex, multicellularity, and societies. These will therefore be included in the following outline of the categories of association. Interspecific Associations Endosymbiosis This signifies an intracellular association, such as the presence of mitochondria in our own cells, and the additional acquisition of chloroplasts by plants. Viruses, bacteria, algae, and fungi may be also be found in the cells of multicellular organisms, and on occasion the symbionts have their own lesser endosymbionts within them. In endosymbioses there is not only an exchange of energy and molecules between the symbiotic cells, genes have been transferred, resulting in a near monopoly of proteinsynthesizing information by the nucleus. The mitochondria and plastids have also had host translocation proteins inserted in their membranes, such that metabolite transfer is facilitated. Many endosymbionts are “holobiotic”—passed “vertically” from one generation to the next in the ova.9 They are not re-introduced from a pool of free-living, independent microorganisms in every generation. Some symbioses involve “horizontal” transmission, which means that the symbionts are not carried in the eggs, but can exist as independent organisms and re-invade the next generation of hosts. For example, when gardeners plant peas and beans they help establish a symbiosis by applying a commercial soil inoculant of Rhizobium bacteria that enter the seedlings, and then fix nitrogen for the synthesis of amino acids and proteins. The dinoflagellate symbionts of corals are endosymbiotic in the digestive epithelia; but they can be released into the sea to be horizontally transferred to juvenile corals. In a less intimate association the symbiont may live extracellularly within the host organism. Such symbionts may be passed vertically on the surface of the ovum, or horizontally, entering the larva or juvenile from a nearby symbiotic source. Like corals, giant clams have dinoflagellate symbionts belonging to the genus Symbiodinium.

Evolution by Association

99

Indeed the clams probably acquired the symbionts from corals in the first place. As in the case of the corals, the clam larvae have no symbionts at first, but after a few days of juvenile growth they obtain them from the seawater that they have filtered. The source of these free-living planktonic dinoflagellates is incompletely digested fecal matter from adult clams in the same vicinity. When they are eaten by the juvenile clams the symbionts are not digested, but take up residence in translucent extensions of the digestive gland that extend into the blood sinuses. Thus, they are extracellular, and not intracellular, as in the corals. There are a variety of strains of Symbiodinium, some of which are preferred by juvenile corals and some by juvenile clams, and if several are acquired, only one succeeds.10 This reinforces two points: loose symbioses can be precarious, and under competition only the most compatible strain of symbiont is established. In lichens, algae are associated with the fungi that make up the main structure, but the different cells do not interpenetrate. The reproduction of both algae and fungi is largely vegetative. “Soredia”—small bodies that may be released from the parent plant—contain both of the symbiotic partners, and establish independent lichens. Sexual reproduction can occur in the fungus, resulting in characteristic fungal fruiting bodies. This kind of symbiosis has evolved independently many times between different algae and fungi. The ability of fungi to acquire mineral salts, combined with algal photosynthesis, is greatly to the advantage of the symbiotic whole, and the algae may be protected from browsers by fungal toxins. Where blue-green prokaryotes are present they contribute nitrogen fixation.11 Lichens are characterized by a physiological resilience that allows them to live in barren terrestrial environments where water and mineral nutrients are scarce. Many are cold-hardy. Since the fungal pigments protect the algae from damage by ultraviolet light, lichens may have been the first plant-like organisms to successfully invade dry land, and once there they prepared it for other plants by making soil. Thus far, these have been examples of physically intimate, mutualistic associations in which both partners exchange nutritive or other services. Other associations can be very loose; their benefits may relate to nutrition, defense, or reproduction. For example, some sea anemones that attach to the shells of hermit crabs eat fragments of the crabs’ food, and protect them from predators with their stinging cells. They may also extend the original shell apertures, saving the growing crabs from the trouble of having to find a larger home. There are numerous “cleaning associations” where the cleaners keep the hosts tidy by browsing on their epifauna, and they may also share in their meals. In the sea, prawns and fish are common cleansers of larger fish, and on land birds often provide the service to larger mammals and reptiles. These relationships merge with “commensalism,” meaning “eating at the same table,” referring to animals that lead separate lives but share their food, and perhaps a degree of mutual protection.

100

Chapter 3

Some species “domesticate” other species. Not only humans keep house plants, pets, farm animals, and crops. Ants, for example, often supervise the behavior of aphids, “milking” their “honeydew” (sugary fecal secretions). Invariably, organisms interact intra- and extra-specifically in their particular “community,” or subsection of their ecosystem. Another generalization about societies that is useful in many contexts is Aristotle’s aphorism that all collective life involves the separation of offices and the concurrence of efforts.12 He not only discussed how this can be effectively achieved in human society, but also applied the model to the coordinated functioning of human functional anatomy.13 Human social and cultural evolution does not arise from the accumulation of behavioral adaptations, but from the adaptability of our species. Adaptability is a constellation of qualities that are physiological and behavioral, with links to development. Adaptability allows the organism to respond effectively to changing circumstances. It makes the human organism a generalist rather than a specialist. And although adaptability is grounded in the genes, it operates at higher levels of organization, and can vary from individual to individual and group to group in its expression. The ultimate biological manifestations of that adaptability appear in the central nervous system, the brain, and the conscious and subconscious mind. These are what free us from the automatic responses of our primitive ancestors. All kinds of association, whether intimate endosymbioses, or happy families, are sources of emergent adaptability, and to seek their explanation at the level of the structural gene is futile. Symbiocosms Terrestrial and aquatic biotopes are often composed of mosaics and mixes of different symbioses that demonstrate a degree of interaction far beyond simple cooperative, competitive and predator-prey relationships. These group relationships have been called “symbiocosms.”14 As an erstwhile student of digestive physiology I have always been interested in the mixture of cellulose-digesting micro-organisms that occur in cattle, and in insects such as termites. However, a remark by microbiologist Lee Haines refocused my attention to non-cellulolytic symbioses in insects, and extended it to the literature concerning symbiocosms. She called the tsetse fly, Glossina, which she studies, “a soup of symbionts.”15 Tsetse is best known as a vector of trypanosome parasites that affect humans, their cattle, and game. Thus, its medical and economic importance has made the full exploration of its symbionts desirable. The bacterial symbionts include Wigglesworthia glossinidia, found in the cells of the bacteriome, an organ composed of expanded epithelia of the foregut. Another endosymbiont, Sodalis glossinidius, is found in the cells of the mid-gut. A third endosymbiont, a species of Wolbachia, occurs in the cells of the ovaries, milk gland (part of the reproductive system), fat body, and hemolymph. The first two endosymbionts have a mutualistic

Evolution by Association

101

relationship with the host, providing essential nutrients not contained in the blood diet of the insect. They, in turn, benefit from the energy resources of the host. If they are absent, the insect is unable to reproduce successfully.16 Wolbachia is more of a parasite than an essential endosymbiont, and can have deleterious effects on reproduction. All of these microorganisms are transmitted vertically via the eggs. Thus, the term “symbiocosm” could be extended to include organisms such as the tsetse fly. The presence of endosymbionts is universal in insects that are specialized feeders on plant juices and blood. The emergence of Aphidoidea, the aphids or greenflies, at the end of the Permian, c. 250 MYA (million years ago), seems to have been triggered by the acquisition of Buchnera, a proteobacterial symbiont that could compensate for the deficiency of essential amino acids in plant saps.17 Minor bursts of diversification within the aphid order can be traced to the acquisition of additional endosymbionts. There is a strong correlation between the evolution of symbionts and their hosts, which can be mapped quickly by identifying DNA base sequences, or protein primary amino acid sequences, and comparing these to gene and protein databanks that have accumulated since the human genome project was initiated. The focus of my interest in the symbiocosm was our greenhouse, when our aubergine plants became infected by aphids. (This is in the target-of-opportunity tradition of T. H. Huxley, who ventured only once into entomology: when greenflies infested his Pelargonium.) The aphids were placed on the aubergines by ants of the species Lasius alienus that had found pastures new for their “cows.” They then proceeded to feed upon the honey-dew that the greenflies exuded. Ants too have endosymbionts, some being Buchnera strains, with which they were most likely infected via aphid honeydew. Formica, and the wood ant Camponotus have endosymbiont strains of the genus Blochmannia, which is related to Wigglesworthia. Wolbachia is widespread among the entire family of the ants.18 Our greenhouse symbiocosm also has a population of ant lions, the larvae of Myrmeliontidae, which are related to lace-wings. Their adults could be mistaken for damselflies by non-entomologists. But the squat bodies, relatively enormous jaws, and menacing behavior of their larvae are better known, since they have become models for science-fiction monsters. They make ant traps in fine, dry soil or sand, and lurk under the surface of the bottom of the pits. (The ant lions seem to have figured out Per Bak’s collapsing-sand-pile problem a long time before he did.) When ants fall into the pits they cannot get enough purchase on the loose substrate to escape, and by flicking sand at them, the ant lion makes it harder. Then its bite injects a poisonous, paralyzing saliva. The toxin secreted by the ant-lion endosymbiont Enterobacter aerogenes is a homologue of GroEL, a variant of chaperonin, one of the heat-shock proteins.19 Curiously enough, homologues are also found in aphid endosymbionts, but they clearly have no anti-ant action.20 Toxicity relates to neural blockage, probably by binding of the molecule to nerve cell membranes, and not to the original chaperonin function.

102

Chapter 3

Another greenhouse pest from which our succulents and cacti suffered is the mealybug, a species of Pseudococcus, an aphid relative. One of the South American species of mealybugs has attendant silvanid beetles that solicit their honey dew.21 For some years it has been known that the polyploid cells of the intestinal bacteriomes of mealybugs have endosymbionts that provide essential nutrients. But now it is discovered that those endosymbionts contain other endosymbionts. These could be parasitic, but the researchers suggest mutualism.22 The Pacific coast termite Zootermopsis angusticollis was resident in the wood flooring under some of the greenhouse benches. Termites have long been known to possess cellulolytic bacteria and protoctists as well as nitrogen-fixing bacteria. In some of these insects, symbiotic nitrogen fixation might equal that of free-living bacteria in the same environment.23 Symbiotic filamentous fungi are present in the guts of some ants and termites.24 The leafcutter ants, genus Acromyrmex, which make leaf-litter gardens in which they grow fungi, also carry the spores of the fungi in their guts. These are in addition to a wide spectrum of bacterial endosymbionts.25 The cuticles of some fungalgardening ants also bear Streptomyces, which are actinomycete prokaryotes that protect their fungal crops from the parasitic fungus Escovopsis. As a bonus, Streptomyces also produces growth stimulants for the ant-garden fungi.26 Coincidentally, various species of this genus produce antibiotics that are widely used to treat human diseases. A less common, but still unwelcome visitor to our greenhouse tomato plants is the whitefly Bemisia tabaci. It not only feeds on the plant sap, it also vectors the yellow leaf curl virus. Its symbiont Sodalis produces a chaperonin that is believed to protect the virus from attack by proteolytic enzymes during its passage through the whitefly’s gut.27 Another of our plants, Abutilon pictum, a mallow that is sometimes called “flowering maple,” is permanently infected by viruses. In this case the viruses do no harm, but cause a leaf color variegation that has been deliberately selected by horticulturists.28 We feel that the micro-ecosystem of ant society and insect-bacterial-viral symbioses truly elevate our greenhouse to the status of symbiocosm. Co-evolutionary Associations Some present-day biologists argue that from the evolutionary point of view the most appropriate catch-all term for biological associations is “co-evolution.” Taken literally, this is perhaps legitimate, but the classical instances of co-evolution involve different organisms that live independently of each other much of the time, while influencing each other’s adaptations through a literal selection process. Therefore, I will persist in using “association” for the most general category, and I will use “co-evolution” in its traditional sense (referring, for example, to the co-evolution of flowering plants and pollinators). While this appears to be a literal instance of natural selection, plants with flowers had first to emerge before they could be directionally selected for exaggerated colors and scents by potential pollinators. Once specialized, flowers only permitted

Evolution by Association

103

pollinators of anatomical proportions suitable for carrying pollen to other members of the same plant species to reach the nectar. There are, moreover, exploiters who simply bypass mutual selection by biting into nectaries from the outside. The constraints on these co-evolutionary arrangements re-emphasize the limitations of selection, both metaphorical and literal. No novelty is produced; existent types are modified and specialized until they have gone as far as they can go, and to get there they may have been driven by molecular mechanisms that were indifferent to natural selection. Symbioses may be precarious. In adopting symbionts or co-evolving with others, organisms make themselves hostages to fortune—harm to one is harm to both. The unicellular flagellate Euglena can digest its green symbionts in times of need, but in so doing it loses photosynthesis and may not get it back again. When the Cretaceous was terminated by an impacting bolide, near-extinction of the flowering plants was probably affected by the near-extinction of their insect pollinators. However, symbionts safely tucked away inside host cells are better able to meet such challenges. Parasitism Parasitism is occasionally lumped with symbiosis, but it deserves a separate category since only one participant gains advantage, and the host is adversely affected. Parasitism is, however, an interesting evolutionary association in its own right, perhaps involving environmentally effected metamorphoses. It also illustrates just how radically morphogenesis can be altered. For example, the crustacean parasite Sacculina starts out as a normal nauplius larva, but once established in its crab host it becomes a pervasive amorphous mass of tissue like fungal mycelium. Furthermore, the complexification of life cycles and the proliferation of intermediate forms found in parasites shows how high are the energetic costs of such precarious associations. Viruses are parasites of a kind that can have highly disruptive effects on whole species and ecosystems. They have co-evolved with their hosts, and sometimes migrated to others where their deleterious effects were magnified. But they may also contribute to saltatory emergences, since retroviruses can introduce foreign genetic material to their hosts. We know about this mainly from genetic alterations of bacteria, but some instances involve viral transgenesis between distantly related vertebrates. Because of technical difficulties, it had not been possible to assess just how widespread the phenomenon is. However, the recent completion of the human genome project indicates which genes have been introduced by retroviruses, and many bacterial genes have been found as part of the human makeup (as opposed to the totality of genes that originated with the ancient prokaryotes in the early lineages of eukaryotes). Some of those retroviruses may have entered into a more intimate and non-pathological association of virus and host. Endogenous retroviruses (ERVs) of mammalian cells are now part of the mammalian genome, and the most familiar ones function in

104

Chapter 3

placental development, either in implantation or in preventing the maternal immune system from rejecting the placenta.29 These kinds of retroviruses mutate much faster than the host genes, and so have a greater potential to increase genetic variability. Intraspecific Associations Sex Much ink has flowed on the subject of the evolution of sex and its selective advantages; less will flow here. Sex brings organisms together; it would not be stretching the category too far to say that the diploid sexually reproducing organism is a symbiosis of haploids derived from two different parents. For the act of fertilization, sexual associations are sometimes loose, and temporary associations, sometimes in permanent proximity. Some female invertebrates and fish go so far as to carry quasiparasitic dwarf males around with them all the time. Sexual associations have many physiological and behavioral implications, depending on the evolution of intercellular messengers, and the organs that produce them; attractants that range from molecules to epidermal color patterns, enticing secondary sexual organs, and behavioral modifications. Therefore, the evolution of the nervous system has also been involved. Sexual reproduction is also a major source of variability, and is therefore one of the significant experimental evolutionary processes of nature. Its effects on the relationships between organisms can be detrimental—mating dogs are oblivious to motor vehicles. However, family associations correlated with sex have had an important evolutionary role, especially in the vertebrates. Sex among the higher land plants is sometimes called a “better adaptation to the dry terrestrial environment” in the sense that their gametes are less prone to desiccation. Yet the fluid transportation mechanism of animal gametes has required the very close associations that have had so many evolutionary consequences. Sex and Chromosomes Before the different kinds of founding eukaryotic cells went off to find their fortunes and establish their kingdoms, they had had already acquired the talismans of mitosis and meiosis. Chromosomes are necessary both for the equal distribution of DNA between daughter cells during mitosis, and also for the production of gametes by meiosis, as the preliminary stage of sexual reproduction. The precursor of the chromosome was the prokaryote genophore—a large loop of double-stranded DNA, which is normally a tangled mass. Unlike the eukaryotic chromosome, it has is very little protein associated with it. The genophore carries the genetic information for making essential structural proteins and enzymes. Prior to the fission of the cell the

Evolution by Association

105

genophore is duplicated, and usually two identical cells are produced. Prokaryotes reproduce asexually, replicating themselves along with any genetic changes that have occurred in them. But the primitive eukaryotes had several genophores and plasmids from various symbionts—the complexity of the eukaryotic genome had increased to the point where packaging was crucial. Did one of the original symbionts become the eukaryotic nucleus, and sequester genetic material from the other members of the association? The following segment on chromosomes is largely based on Origin of Eukaryotic Cells (1970) and Margulis’s subsequent publications. Eukaryotes have multiple chromosomes that usually consist of histone proteins, bound to DNA, arranged linearly along them. There is also satellite DNA at the ends of the chromosomes—telomeres—and in the centromeres, which control mitosis and meiosis. Telomere attrition occurs in proportion to the number of times the cell line has undergone mitosis, and this diminution may be correlated with senescence.30 Chromosomes have multiple functions, such as helping to regulate the expression of their genes. In a multicellular organism the functions of differentiated cells require the repression of most of their chromosomal DNA. Chromosomes also ensure to some extent the preservation of whole organismal properties based upon a hierarchical polygenic organization—alleles are not randomly mixed when gametes are formed. Also, by condensing into compact rods during mitosis and meiosis, their spatial distribution is easier to control. Imagine trying to shuffle and deal a deck of cards whose individual lengths are proportionate to their numerical value. The long, tangled threads of the interphase chromosomes have to be reduced to manageable units for consistently even dispersal. When endosymbiotic experiments were being initiated by primitive prokaryotes, redundant DNA must have constitutively churned out proteins and wasted energy and molecular resources. Regression and differential expression required differential histone binding and methylation mechanisms, which were part of the shakedown into dynamic stability. However, they were flexible enough to allow subsequent change, and major steps in the direction of chromosome evolution and epigenesis. In protoctists there is a variety of ways of dispersing DNA evenly, if not absolutely equally, between daughter cells. The cell can make multiple copies of DNA and take the chance that each daughter cell gets at least one complete set. The ciliates have meganuclei that do the protein-synthesizing work of the cell, and multiple micronuclei that carry sets of DNA that exist for reproductive purposes only. There are also some variations on the theme of a mechanical spindle that pulls the chromosomes apart during mitosis. Although animals have spindle-organizing centrioles, and plants do not, the mechanics of mitosis are similar. There is no good evidence to support the idea that the framework of moving spindle fibers originated with ancient chromatin, i.e., the protein and nucleic acids of the chromosomes. Instead the spindle apparatus came from an independent source, perhaps the

106

Chapter 3

symbiotic spirochete-like prokaryote. The basal bodies, or kinetosomes of locomotory structures, such as flagella and cilia, collectively known as undulipods, contain their own organizing chromatin which can generate both undulipods and mitotic spindles. Ciliated cells in the bronchi of the lungs for example, do not mitose, because the kinetosome is completely committed to undulipod function. When these cells wear out they have to be replaced through mitosis of unciliated cells whose daughter cells then differentiate cilia. The division and separation of the basal bodies of the flagella/cilia could have originated the spindle that pulls the chromosomes apart. Did spindle and centromeres co-evolve, so that the attachment of the chromosomes and the separation of the chromatids could be consistently effected? It is a more parsimonious idea than the older hypothesis that proteins called “tubulins,” which form into microtubules, were universally distributed from a common ancestor. There was then independent evolution of undulipods, basal bodies, centrioles, and spindles in various evolutionary lines, giving rise to different kinds of undulipods. Microtubules constituted functional structures such as animal muscle fibers, as well as many other cellular kinetic functions in all of the kingdoms of organisms. Whichever hypothesis is right, and Occam’s razor shaves the options down to the Margulis-Sagan version, both are complemented by the idea that the chromosomal centromeres, and spindle organizing elements could have arisen from the replication origin and terminus of the prokaryote genophore, where it attaches to the cell wall.31 That seems a reasonable argument for the derivation of the centromere, but what about the centrioles—the bodies that participate in spindle formation? The endosymbiotic version accounts for those as well, as part of the constellation of emergent properties of the thermoplasma-spirochete symbioplex. The microtubular apparatus emerged only once, originally as a locomotory system, but diverged in function to become a spindle apparatus for host cells that had built up a nucleus with duplicated genophores, derived from several symbionts. Equal division might have been possible as long as the genophores remained few and small. But chromosomes could be enlarged if they acquired proteins that could condense and reduce them to manageable proportions for reproduction. These emergent protein structures had the potential to regulate gene expression in future multicellular lineages as well as condensing the chromosome structure. When mitosis occurs in a cell some internal structures, such as the microtubular cytoskeleton and internal membranes, including the nuclear membrane, break down and leave the cell as an unobstructed stage for the dance of the chromosomes. Then they are all reconstructed when mitosis ends. Even speculation about the evolutionary details of such molecular events boggles the imagination. That a hopeful monster of a cell leaped into existence with everything simultaneously in place is out of the question. As the gradualists argue, we have to look at everything step by step. But the steps are

Evolution by Association

107

themselves saltatory emergences, and identifying their advantage does not explain their genesis. Once the proto-eukaryote had assembled a useful complement of chromosomes that could be equally divided among its daughter cells, it could then clone itself in a stream of copies. Populations of the new eukaryotes would all have consisted of very similar organisms. But finally the associative emergence of sexual reproduction provided for greater experimental variability. With chromosomes and mitosis in place, sexual reproduction was a comparatively simple emergence. Diploidy could first have appeared as a failure to complete mitosis at telophase, the stage where the doubled complement of chromosomes are usually divided into two daughter cells. This would probably have spontaneously increased the size of the eukaryotic unicell as well—there is some correlation between cell and genome sizes; large genomes need more space if they are to synthesize more proteins.32 Diploidy could also have been introduced through conjugation of two unicells with homologous sets of chromosomes. In attempting normal mitosis the colinearity of the DNA and histones in the homologous pairs could have been partially responsible for synaptic pairing of the kind found in meiosis I prophase, although in most of the familiar expressions of meiosis there is now a special protein apparatus that pulls them together. The next essential step in meiosis was to separate the pair of chromosomes in the tetrad, without breaking apart at the centromeres as they do in mitosis. This modification of the mitotic division restored the chromosomal complement of the daughter cells to the original haploid condition. It must however be remembered that each chromosome has already divided into sister chromatids held together at the centromere. In primitive multicellular plants and animals the haploid condition dominates the life cycle. For them, having diploid cells is no great necessity, except for the brief period of chromosomal exchange involved in crossing over. That crucial moment makes every offspring of the act of sexual association a potentially novel emergent type. The conventional explanation for the dominance of diploidy in the life cycles of higher terrestrial plants and animals is that it provides back up copies of genes in case of damage from ultraviolet radiation in organisms directly exposed to the sun. It also creates redundant genetic material for random natural experimentation and the possibility of differentiating the sex chromosomes into dissimilar structures like the X and Y in insects and mammals. With sexual reproduction, the genetic variability that the eukaryotes lost when they departed from the simple prokaryotic state was restored in a new form. Sexual association of different individuals increased the potential for evolvability through recombination of parental traits, gene duplication, and chromosomal mutation. Chromosomal mutation provides for the position effects that Richard Goldschmidt thought responsible for large phenotypic changes, i.e., hopeful monsters. That the position of a gene, or smaller sequence of a DNA molecule is important in its function

108

Chapter 3

is no longer in doubt, especially through the work of Barbara McClintock on transposable elements. Current research on the microtopography of DNA strands will illuminate this further. In addition to chromosomal mutations, the duplication of individual chromosomes, and of entire karyotypes contributes to evolution by providing redundant DNA and proteins that can be differentiated to serve different offices. There is a last but not least detail regarding the evolutionary significance of chromosomes, based on the work of Neil Todd (1970, 2000, 2001) and Robin Kolnicki (1999, 2000). This has given rise to the karyotypic fission theory, brought to the fore by Margulis and Sagan (2002). Todd’s original hypothesis was that another kind of chromosome breakage was possible. All the chromosomes could simultaneously break in two at the centromere-kinetochore, thereby doubling their total number. (The centromere portion is DNA that is continuous with the rest of the chromosomal DNA; the kinetochore portion is made of proteins that attach to the mitotic spindle.) Mammals, the group that Todd surveyed, have disparate numbers of chromosomes. Within one deer genus, Munjiacus, the diploid number of chromosomes varies from three pairs to 23. Conventional wisdom gives the primitive mammal high numbers of chromosomes that are then gradually reduced in number by chromosomal fusion— sorted, of course, by natural selection. How is it that within one genus the chromosomal numbers are so widely different in the different species? Todd’s theory of karyotype fission now proposes that the original mammals had seven pairs of chromosomes. The molecular mechanism of fission involves an acceleration of the reproduction of the centromere-kinetochore relative to the reproduction of the rest of the chromosome. That acceleration can be the result of physicochemical changes that affect the whole cell, so it happens in all of the chromosomes, making them more likely to break when they attach to the mitotic spindle. Todd additionally posits that fissioning events, which are unashamedly saltatory, were responsible for diversification of the emergent types. This might have come from a greater degree of variability that Margulis would subsume under her mix-match principle. Multicellular Associations Simple multicellularity emerged independently in several lines of evolution. PreCambrian algae were probably the first. Theoretically, the multicellular condition has the advantage of being energetically more efficient for its cellular members, there being a division of labor among them. Before there can be a division of labor, there has to be a mitotic division that does not end with total uncoupling of the daughter cells. A colonial protist such as Volvox survives without intimate links between its cells. Cells with rigid walls can be held together by sticky exudates. But the cells of most multicellular plants and animals have intimate links. “Cell adhesion molecules” allow cells of the same type to recognize each other and stick together. The extracellular

Evolution by Association

109

portions of the class of molecules call “integrins” can attach, detach, and anchor cells to each other, as well as to physical extracellular substrates. Since cell membranes are fragile, the underlying fibrous cytoskeletons must be involved. Integrins provide part of the cytoplasmic scaffolding that can order cellular biochemical pathways, as well as inducing the formation of ultramicroscopic junctions that attach the cells to each other, sometimes so tightly that water cannot penetrate between them. Other kinds of junctions allow intercellular exchange of molecules. “Membrane receptor molecules” initiate uptake of specific dissolved molecules or particulate matter, or trigger cascades of hormonal responses in the cells. This allows not only communication between adjacent cells but also between cells that are widely separated in the organism. Some such cell surface molecules are found in primitive unicells. Once they became involved in the formation of intimate, multicellular associations, the repetitive differentiation of the genes coding for their synthesis became an important feature of developmental evolution, and eventually generated the emergence of the immune system in vertebrates.33 While sexual reproduction compensated for the loss of the ancestral genetic flexibility of the prokaryotes, multicellularity conferred greater evolvability. It led to the diversification of body plans, and to the progressive evolution of self-regulating physiology. Thus, it contrasts with population adaptability based on the random process of gene acquisition, possessed by primitive unicells. The progressive evolution of the multicellular condition is a topic that I will come back to in chapter 5. Meanwhile we will move on to the next major evolutionary performance involving intraspecific associations. Societies If endosymbiosis is the alpha of associations, the omega is society. The entire spectrum of associative evolution, from primitive symbioses through parasitism to human societies, was regarded as grist to the mill of emergentism by the philosopher-entomologist William Mortimer Wheeler. However, like C. L. Morgan, Wheeler thought that even such large emergences as life and mind were the cumulative and imperfect end results of series of relatively small emergences.34 Herbert Spencer Jennings, who made emergentism “the Declaration of Independence for biological science” (that is, independent of natural selectionism and Laplacean reductionism), also cautioned against forced comparisons of emergent systems operating at different phyletic levels: If it is indeed in social organization that we find emergent evolution most manifestly at work, if it is here that that which is new in principle most frequently and conspicuously appears, then we shall be cautious in accepting the advice of even the King of the termites on our own social problems; we shall use discretion and take his advice at most as suggestion toward experimentation. For any organism of society separated from others by steps in emergent evolution, the only possible method for progress is by trial and error.35

110

Chapter 3

After 80 years this advice has not yet been heard by sociobiologists and “evolutionary psychologists” who argue that gene-based heritable behavior patterns analogous to those of insects and birds must also be present in humans. This is not to say that useful comparisons between systems at different emergent levels cannot be made. E. O. Wilson, successor to Wheeler in the study of social insects, says he bases his synthesis of sociobiology on a quantitative holism that involves the “recognition and study of emergent properties.”36 He infers, for example, that human self-knowledge is shaped and constrained by emergent emotional properties of the hypothalamus and limbic centers of the brain such as guilt, fear, love and hate—but he believes that it was natural selection that made the hypothalamic and limbic systems. In so doing, Wilson surrenders both sociobiology and emergentism to ultra-Darwinism. He has also been accused of deterministic bias in the analysis of human societies, and his natural heirs in “evolutionary psychology” go even further in the same direction. Notwithstanding, sociobiology has a place as a distinct subdiscipline of the general topic of evolution by association. Social interactions produce emergent phenomena, which are by definition not driven (or “upwardly caused”) by genes. Wilson has sought to identify such emergent phenomena, and has compared them in insect societies with the operational system of differentiated cells in a multicellular organism. While remembering Jennings’s epigraphic warning about taking the advice of the Termite King, such system comparisons might also offer more focus on features of emergence common to other evolutionary arenas. Social insects have long been of interest for their apparent altruism or and willingness to sacrifice their energy and their lives for the sake of the society. In many cases, individuals are sterile, leaving reproduction to queens and drones. This has given rise to the concept of the “super-organism,” with queens and males representing the sex cells and worker insects representing the somatic cells, selflessly sacrificing themselves to the higher goal of reproduction. The idea was first mooted by William Wheeler in 1911, elaborated by A. E. Emerson (1939), and R. W. Gerard (1942). But it is now primarily associated with E. O. Wilson (1985). Interactions between social insects certainly produce emergent behavior with properties not found at the individual level. Among the vertebrates, striking social behavior with emergent overtones is found in the movements of schooling fish and flocking birds. But the most important social associations for our own species have been family and tribal groups that culturally evolved into modern societies. I will have more to say about societies later, but several generalizations are appropriate at this point. One is that societies consist of groups belonging to a single species. In some cases the members of the society are all members of the same family, i.e., they are eusocial.

Evolution by Association

111

Now, having considered the various degrees of intimacy and mutualism that occur in associations between organisms, what might we make of their ecological effects during life’s history. The Emergence of Ecosystems through Symbiogenesis When Lynn Margulis suggests that the evolutionary importance of symbiosis has been ignored by the Modern Synthesis, I agree, and I can think of several reasons. One is that the Modern Synthesis subsumes anything it wants to salvage without bothering to resolve contradictions in a truly synthetic manner, or to give credit where due. In the case of symbioses, the Modern Synthesis simply argues that natural selection builds them, approves them or eliminates them. A tacit reason for the Modern Synthesis to underplay symbiosis is that evolutionary symbiogenesis really belongs with saltatory emergentism. Margulis has a global view of the importance of symbiosis in the life of Earth, and her treatment of the most significant one—eukaryotic endosymbiosis—has been exhaustive. Yet her thesis has underemphasized the historical impact of symbiogenesis in the foundation of all of the major ecosystems. Thiobios This is an ancient trophic system established by the archaean prokaryotes that used volcanic sulfur as a primary source of energy. Once animals diversified through the differentiation of multicellular forms, the thiobios could be expanded through bacterial-animal associations that exist in some bivalves, gastropods, annelids, pogonophores, and echinoderms. Hosts created ideal physiological environments for such symbionts, and in return they received simple organic products synthesized by their symbionts, using the energy of sulfur. The bacteria may be found inside host cells, or in blood spaces, loosely associated as epiflora on the outer surface of the host or in aerated chambers formed in organic sediments by infaunal metazoans. Several different species of sulfur-oxidizing bacteria have been discovered in such relationships, which are now well known in several families of bivalve mollusks. Although Tom Fenchel and Rupert Riedl had recognized the significance of the thiobios in 1970, the reaction of benthic biologists was indifferent until the flurry of research began on marine thermal rift communities. The thiobios is at its most conspicuous in deep water volcanic thermal vents that supply a mineral source of sulfide. “Conspicuous” is not meant literally, since the first examples were found by submersibles at a lightless depth of three kilometers in the Galapagos Rift only two decades ago. There, rich communities of the huge symbiotic pogonophore tube worm Riftia have grown up along with vent clams and mussels, supporting a tertiary trophic level of gastropods, crustaceans, and fish. These are oases of energy in an otherwise barren, benthic wasteland. The thiobios also exists in a more diffuse form on

112

Chapter 3

continental shelves where there are organic sinks with high sulfide levels produced by bacterial decomposition. Marine Photosynthetic Ecosystems Although the thiobios is an ancient and interesting ecosystem, photosynthetic systems dependent on the energy of light have a much more universal distribution. The largest biomasses of the marine food chain depend on the symbiosis between chloroplasts and plant cells, especially phytoplankton. Moreover, the impact of photosynthesis on the oxygenation of the oceans and the atmosphere was one of the largest ecological and evolutionary impacts in the history of Earth. Most benthic ecosystems depend upon light as their primary energy source. However, in the tropics, while light is not limiting, soluble plant nutrients such as nitrogen compounds and phosphates are. Consequently the tropical benthos is relatively barren, with the exception of coral reefs. The modern coral reef, now ubiquitous throughout warm seas, has a complex topography, with a richly diverse community of plants and animals. All this is founded upon a symbiosis between dinoflagellate zooxanthellae and the reef-building coral cnidarians. Again the host provides cellular living space and metabolic nutrients such as ammonia, carbon dioxide, and phosphates that the algae can photosynthetically convert to carbohydrates, amino acids, fats and nucleic acids. Some of these are returned to the host. In the mineral nutrient-deficient coral reef environment, the photosynthetic zooxanthellae acquire nitrogen and phosphorus from the animal food of the corals. The food consists of small invertebrates and fish that are captured by the “nematocysts” (stinging cells). Nematocysts may be symbionts that first evolved as independent unicellular microsporidians.37 Coral zooxanthellae have been passed on as symbionts for other animals, such as sea slugs that feed on coral.38 Tridacnidae—the family of giant clams that live in the reefs—also feed on zooxanthellae and particulate debris released from heat-stressed corals as phytoplankton, though they are resistant to digestion. Symbiogenesis between dinoflagellates and the clams has resulted in a drastic allometric shift in the clam’s shell and soft anatomy, providing a greater area of tissue to house the symbionts where light is most available. Photosynthetic symbioses occur sporadically in other kinds of marine animals such as flatworms and polychetes, but only in the coral reef do they provide the foundation of an ecosystem. Terrestrial Plant-Fungus Ecosystems It wasn’t until plants spread to land that their functional-morphological complexity began to emerge. The littoral was already occupied by biofilms and mats of blue-green prokaryotes, simple algae, bacteria, protozoa, and perhaps fungi. The first truly terrestrial emergence was made by the kind of association of algae and fungi found in

Evolution by Association

113

lichens. These constitute a polyphyletic group, meaning that their members emerged independently numerous times. A similar, but even looser, association persists in the great majority of the higher land plants.39 Familiar woodland mushrooms are the fruiting bodies of symbiotic basidiomycetes that invest tree roots as mycorrhizae, enabling nutrient exchange between partners. Once their foothold on land was stabilized by symbiotic fungi, the plants were at liberty to experiment with morphogenesis and reproduction. In Symbiotic Interactions (1994), Angela Douglas argues that mycorrhizal symbioses must have been established at the points of origin of the major groups.40 In other words, they were pivotal emergences of land plant evolution. In some instances, mycorrhizal symbioses involve a third party, such as squirrels that dig for the fruiting bodies of truffles, and help spread their spores by breaking them open in the safety of a tree. From their germination new mycorrhizal strands invest new trees that the squirrels will eventually climb to enjoy the fruits of their labors and help them propagate once again.41 Cellulolytic Symbioses Do humans who drink milk and eat meat know that the major resource for these habits depend on a symbiosis between cellulolytic micro-organisms and their cattle hosts? Perhaps more than 400 million years ago, in the terrestrial environment, a new food pyramid emerged from a symbiosis that utilized cellulose. A generative condition of much of plant evolution was the possession by some algae of cellulose walls. These helped them survive first of all in fresh water, by resisting excessive water intake. Then, when plants invaded land, their rigid cellulose cell walls and vessels, supplemented later with woody fibers, provided the framework for the vertical architecture of bushes and trees. The first animals that arrived on land, probably flatworms, oligochete worms, gastropods, crustaceans, and proto-insects, found a primary resource that was partly inaccessible. Little cellulose food had been available to their marine ancestors, and although some could digest simple cellulose, most of them had lost the appropriate enzymes. Fortunately for the incipient terrestrial food pyramid, this is not so for soil fungi, bacteria and protozoa that were to become symbionts. Herbivorous invaders of the land foraged in decaying plant material; primarily because plant cover and vegetable mold provided some protection from the sun and from desiccation. In addition, the breakdown products of decomposing plants are more accessible to simple digestive systems than native cellulose, and the fungal mycelia that run through the decaying plants are also digestible. The lifestyle and simple morphology of the land isopod crustaceans that are variously called “wood-lice,” “slaters,” or “pill-bugs” are probably very similar to those of the primitive crustaceans that first made landfall.

114

Chapter 3

Some of the bacteria, with their group adaptability in responding to novel and harsh environments, were or became resistant to digestion, and established themselves as gut symbionts in the invertebrate pioneers. Some protists did the same, including one of the most symbiotically complex creatures, Myxotricha paradoxa, which is found in a primitive Australian termite. The protist host cell has the usual eukaryotic mitochondria and undulipods. Two kinds of spirochete bacteria attach to the outer coat at specific sites, one type amounting to about half a million cells per host cell, organized to provide a coordinated locomotion for the whole organism.42 With such symbionts, crustaceans, insects and slugs and snails could get much more out of the plants they ate, and these new herbivores spread rapidly. From this secondary terrestrial trophic base, a third level of carnivores was extended. Biochemically this required no great novelty since they already had the enzymes needed for digesting animal food. It was only a matter of cranking up the proportions of peptidases to make the digestion of animal protein more effective. Specialized predatory behavior and raptorial mouthparts were secondary. Nitrogen-Fixing Symbiosis Nitrogen-fixing symbiosis is largely confined to plants, many of which live in environments with few available mineral nitrogen compounds. Here, some plants, such as Venus flytraps, sundew, pitcher plants, and butterwort, are insectivorous, actually digesting animal protein to acquire usable nitrogen. Many, including peas, beans, vetches, alders, and other woody flowering plants, have symbiotic bacteria that fix atmospheric nitrogen to make ammonia. When the germinating seeds pick up the appropriate soil bacteria. (Rhizobium and its relatives in the case of the legumes) the young host plant grows nodules that house them. Although the symbiosis is energetically expensive for the host, it has a large ecological impact by contributing to the invasion of barren wastelands and providing for a subsequent succession of nonsymbiotic plants. The symbiosis is probably key to the emergence and diversification of the common and pea and bean family Viciidae. Now that some of the major impacts that symbiogenesis has had on evolutionary history have been outlined, some philosophical aspects of association can now be explored, beginning with another aphorism of Aristotle. The Whole Is Greater Than the Sum of Its Parts The many routes to greater complexity, and to new wholes that are greater than the sums of their parts, will be illustrated in this chapter and in the next two chapters. The symbiogeneses that occurred during cellular evolution were among the earliest and most striking examples of the principle. That an emergent association has the potential to be a whole that is greater than the sum of its parts is virtually a truism.

Evolution by Association

115

But it is unfortunate that such a concept should pass into triteness without analysis, since it is not only relevant to symbiosis, but also central to the development of emergentism. Reductionists have traditionally argued that ignorance of the whole is mere ignorance of some of the parts, and that the emergent properties of a whole can be ignored if one knows all the features of the parts. John Stuart Mill pointed out more than a century ago that the holistic properties of water cannot be predicted from a mere knowledge of the characteristics of oxygen and hydrogen. But, argue the reductionists, it is so simple: water and its properties are merely the qualities of hydrogen and oxygen in combination with each other! The emergent quality of water is not that easily explained away by simplistic holism, any more than reductionism. The emergent properties of water or any other whole are difficult to predict from a knowledge of the nature of the components. It makes it easier to describe them by working back from the known combining features. Although unpredictable combinatory uniqueness is a mark of some emergent properties, evolutionists of all stripes have to operate by hindsight, because the events under investigation have already happened. And generating concepts of future complexities is not in the usual repertoire. Taking care not to smuggle our existing knowledge of the outcome into the exercise, it might be possible to post-predict, i.e., to estimate if known complexities would have been predictable from a knowledge of the historical conditions that generated them. But it would be hard to imagine the possibility of mutualistic, symbiotic wholes in a universe where they had never occurred. Even when the evidence from looser associations such as lichens and coral was staring us in the face, biologists, including myself, initially resisted the radical notion that mitochondria were symbiotic. But in our enlightened state we finally see that this endosymbiosis beautifully illustrates the principles of complexification by association. Complexification by Association It is easy to see the mutualistic potential when a protomitochondrial, independent organism takes residence in an anaerobic prokaryote. Simply tabulate the biochemical characteristics of the mitochondrial symbiont and the rest of the eukaryotic cell, ignoring, for the time being, the fact that genes and functions were exchanged and some may have regressed over the course of time. Among other things the cellular cytoplasm uses simple carbohydrates, in the absence of oxygen, to make a small amount of the high-energy compound adenosine triphosphate (ATP). The mitochondrion removes from the cell ammonia and ammonium ions, toxic compounds that results from amino acid metabolism, and turns them into other amino acids, or initiates synthesis of urea and uric acid (which are less toxic than ammonia). This perhaps was part of a primitive ability of the promitochondrion to use inorganic sources of nitrogen for the construction of amino acids, proteins, nucleotides, DNA,

116

Chapter 3

and RNA. The mitochondrion also removes oxygen from the cell to produce copious ATP. Remembering that oxygen was a toxic compound in the ancient anaerobic biosphere we can see that the symbiosis turns detrimental, or negative qualities into positive ones. Even by simple arithmetic, the whole, in concert, is greater than the sum of the parts. Symbiogenesis thus provides a way of emerging to new wholes that are greater than the sum of their constituent wholes. The condition of new wholeness that emerges from symbiogenesis is largely due to complementarity of the biochemical, physiological, and behavioral functions of the pre-symbiotic individuals. Margulis (1981) and Douglas (1994) provide long lists of emergent metabolic properties of symbioplexes. Having looked at the spectrum of associations that range from the most intimate and mutualistic to the loose and lopsided, and having discussed generalities about emergent wholes, we return to the chronology of their evolution. Where to begin? The first major biological emergence was that of life itself, and the nature of the first cells that appeared. This is a matter of molecular association, which will be considered in the next chapter. At this point it is enough to note that the first protobionts were simple systems that did not have the full range of genetic faculties that we associate with modern bacteria. They had to reproduce fast enough to keep up with wear and tear, aided by a catalytic capacity to energize essential molecules for polymer synthesis. Their reproduction also had to be faithful enough to ensure survival of functional phenotypes. In the absence of any defensive digestive or immune systems, protobionts with simple phospholipoprotein membranes could freely coalesce and become more complex systems with emergent properties arising from the complementarity of previously independent features. Mere compatibility could simply have been enough to ensure their persistence. However, the greater quality is adaptability. Eukaryotic Emergence The emergence of eukaryotes through endosymbiosis between primitive prokaryotes was the most important complexifying saltation in the evolutionary history of living organisms. How it occurred is still a matter of speculation, since the most appealing mechanism—phagocytosis—is not employed by modern bacteria. Margulis originally suggested that predatory heterotrophic bacteria were prime candidates as endosymbionts.43 These normally lyse and enter the prey cells, seal the breach, and then digest their surroundings at their leisure. If the assault was limited to lysis, entry and resealing, with subsequent destruction of the prey inhibited, sharing of resources could have ensued. But this explanation creates the problem of how the double membranes of mitochondria arose. Therefore phagocytosis, which temporarily combines the membranes of both predator and prey remained appealing. One popular current idea suggests that enlarged Thermoplasma-like Archaea were the original kind of host cell.44 This is partly based on the fact that Thermoplasma is more

Evolution by Association

117

accessible since it lacks a rigid cell wall, and the discovery of Archaea genes in the eukaryote nucleus.45 Archaea have both the interior “endomembranes” that could have given rise to nuclei and endoplasmic reticula, and a cytoplasmic skeleton that could have provided the contractile threads necessary for phagocytosis.46 The archaean Archeoglobus has many protobacterial genes, relicts of transitory transgenetic brushes with other bacteria over the course of time.47 Margulis and Sagan (2002) have put an interesting twist to the idea of an archaean Thermoplasma-like host by combining it with a spirochete. It not only gave the symbiosis a metabolic complementarity between sulfide production and oxidation, it also added undulipod motility. They postulate that the original host was a heatresistant archebacterium whose metabolism used the energy of sulfur, with hydrogen sulfide as an end product. A spirochete that used the sulfide as an energy source became endosymbiotic with the archaebacterium. Not only did they pool their metabolic resources, the sulfide “scrubbed” free oxygen, which was beginning to appear in the environment at the time. Thus, the spirochete which was completely vulnerable to the toxic oxygen was protected.48 The symbioplex could now seek the most mutually beneficial gradients of sulfur and its compounds, acidity, and oxygen. A corollary of the model also suggests how the nucleus arose. To do so it draws upon the nature of the “karyomastigont,” i.e., the complex consisting of nucleus, the cilium (or undulipod) and the fibrous protein connection between the two. Margulis and Sagan call it an “emergent structure” of the original symbiosis.49 Imagine that the part of spirochete symbiont that intruded into the host retained its surrounding cell membrane, containing its genophore. The undulipod was left at the surface of the host where it continued to be a locomotory organ. How the genophores of the archean and spirochete coalesced within the protonucleus is unknown, but exchanges of the smaller gene-containing plasmids between bacteria is common. And there are variations on the theme, arising from differential growth rates of host and symbiont, giving rise to some cells with multiple karyomastigonts, some with multiple undulipods, some with multiple nuclei and some without nuclei. All these variations still exist in nature. And once the genophores became chromosomes there was a karyomastigont fibrous protein system and a set of centrosomes ready to participate in mitosis and meiosis. The hypothesis is both parsimonious and predictive, and illustrates how a single emergence can have so many immediately beneficial properties and great evolutionary experimental potential. The later acquisition of protomitochondria would have compounded the oxygenscrubbing function already present in the original host. Another interpretation of the mutualistic benefits of the endosymbiosis involving mitochondria, is that it could have been based on hydrogen production instead of respiratory exchanges and detoxification.50 Some α-proteobacteria, of the kind generally agreed upon as the best

118

Chapter 3

protomitochondrial candidate, excrete hydrogen and carbon dioxide as metabolic byproducts of the oxidation of organic molecules. (Mitochondria now retain the hydrogen ions for the production of ATP.) Also, some Archaea depend upon hydrogen and carbon dioxide as the only source of energy and carbon for their metabolism. The two could thus combine in a state of instant mutualism. At that point the mitochondria need not have detoxified oxygen. But they were poised to use oxygen as a hydrogen acceptor, once phagocytosis had provided a more satisfactory alternative source of energy for both symbiont and host. This degree of functional complexity was only surpassed when another endosymbiosis gave chloroplasts to some of the eukaryotes. Prokaryotes that became chloroplasts were already symbioplexes with complementary photosystems derived from two previously independent prokaryotes.51 The endosymbiotic hierarchy may have been further compounded: traces of nuclear material in some algal chloroplasts suggest that after photosynthesizing unicellular eukaryotes had been established, some of them, in turn, became symbionts in other algae. Karl Niklas notes that “something old, something new, something borrowed, something blue” takes on special significance for algal marriages.52 Photosynthesis originated before the emergence of eukaryotic algae, but it is not clear how quickly nor to what degree it oxygenated the old reducing environment. While there would have been temporary oxygenated pockets, there was a large, inorganic sink that required filling before free oxygen could spill over. That surplus would initially be used up by the oxidation of organic debris from dead bacteria. But if organic remains were sequestered as fossil material the balance would shift to the side of free oxygen. Algae were the first organisms to tilt the proportions enough to finally ensure the general accumulation of free oxygen in the water and in the atmosphere. Thus, they had double emergent properties: their own physiological qualities and their ability to convert the anaerobic environment to an oxygenated one. This environmental contingency would have been catastrophic for many obligate anaerobic organisms, clearing the bench for new experimental forms. Their deaths would have initially wobbled the supply of free oxygen. But eventually the oxygenation of the biosphere was to make causal waves that would surge through time, leading to the evolution of air-breathing animals. Their high-oxygen-dependent metabolic rates would eventually support a homeothermic physiology, and the emergence of sophisticated nervous systems and mind. The notion that chloroplasts were once independent, free-living micro-organisms had been suggested by W. Pfeffer in 1881, and elaborated by A. F. W. Schimper in 1885. In the same decade, K. Altmann had speculated that mitochondria had similar origins.53 But considerable resistance to these historical ideas persisted until Lynn Margulis established the theory of serial endosymbiosis in 1970. Even then it took a decade of persuasion, and the unearthing of new molecular evidence, to supplant the

Evolution by Association

119

“direct filiation theory.” That alternative theory had postulated that organelles were secondarily derived from the outer cell membranes by a slow and complex process of invagination and involution—thus explaining why mitochondria and chloroplasts had double sets of membranes. Although the hypothesis that the organelles acquired the extra set of membranes when they took up residence in the host cell is more parsimonious that the direct filiation theory, it was the latter that lent itself better to Darwinistic gradualism. Some complained that endosymbiosis was a cheap explanation: “The endosymbiosis hypothesis is retrogressive in the sense that it avoids the difficult thought necessary to understand how mitochondria and chloroplasts have evolved as a series of small evolutionary steps.”54 The critics then compounded their polemic by accusing the endosymbiosis hypothesis of playing into the hands of the bogeyman of special creation. But the selectionists’ real worry was that endosymbiosis is a supreme example of saltation: a prokaryote can only enter the association as a whole organism, and the new whole is immediately a doubly complex structure with an array of emergent qualities. As Darwin feared, if the monster is successful at birth, with intrinsic features responsible for its survival and reproduction, natural selection has no part in its construction, except in the adjustment of a stable internal equilibrium. Apart from the debatable origin of organelle membranes, and the parsimonious appeal of endosymbiosis, it was already known that organelles have their own DNA, and independent reproduction within the cell. Then, DNA sequencing evidence began to show that chloroplasts and mitochondria are much closer to extant, independent prokaryotes in their DNA structure than they are to the chromosomal DNA of the eukaryotes that house them.55 Their functional morphology is also more bacterial than eukaryotic. Organelle DNA is organized like bacterial genophores, not in proteinassociated chromosomes. They undergo binary fission like bacteria. The translation mechanism and ribosomes are typically bacterial, and their outer membranes have biochemical characteristics similar to bacteria. The new eukaryotes with mitochondrial and plasmid symbionts largely abandoned prokaryotic transformation, the ability to absorb naked genes, and plasmid transfer by conjugation. Nevertheless, transposable elements have persisted all the way to humans. Viruses have been variously described as pathogenic, neutral, or a source of variation. Viral transduction remains a dark horse, detrimental in some cases, yet a potential universal source of molecular saltation. In addition to retaining these features of saltatory evolution, the eukaryotes were the first to acquire true sexual reproduction, and along with it the potential for chromosomal and genic mixing, to gain a wider spectrum of variation within individuals and populations. The new cells were more adaptable as individuals than their prokaryotic ancestors. Primitive eukaryotes sorted themselves out into the forerunners of the major lines of evolution: the Protoctista (or Protista, as they used to be called), single-celled and

120

Chapter 3

multicellular, some photosynthetic and others not; the photosynthetic and multicellular Plantae; the unicellular and multicellular, non-photosynthetic Fungi; and the multicellular Animalia. Along with the prokaryotic Monera they make up Whittaker’s five kingdoms of living organisms.56 Some of these major groups are polyphyletic, and the Protoctista were variously ancestral to the plants, fungi and animals. However, for the sake of simplicity, we will accept the five kingdoms as they stand, albeit arising from the two founding domains: Archaea and Bacteria. The primary concern for the moment is how they all entered into associations that are among the most significant emergences of evolutionary history. Symbiogenesis Can general principles be inferred about the generative conditions for symbioses? There are two major requirements: first the participants for potential symbioses must be present together, and in sufficient quantities for natural experiments to be possible. A common affinity for the same environmental locus is obviously significant. Common nutritional needs, predation, infection, or parasitism can get two different types of organism into intimate, non-mutualistic proximity. But for them to become symbiotic, defensive mechanisms of the potential hosts, and detrimental parasitic or predatory effects by the potential symbionts must be absent, repressed, or removed. Since these are likely to be rare occurrences, the necessity for large numbers of potential participants is all the greater. Before the symbiogenesis of the eukaryotic condition, congregation of the future partners at an environmental interface must have been a generative prerequisite. If, for example, endosymbiosis involved a host cell with anaerobic or weakly aerobic metabolism, and a proto-mitochondrion with aerobic metabolism, they may have met at a redox layer with oxygenated water on one side and anoxic water on the other. In the oxygenated water there would also have been the proto-chloroplasts responsible for the oxygen output in the first place, and already inured to oxygen toxicity. As much as being in the right place at the right time, the crucial rallying cry for potential associates in this symbiosis is to be in the same place at the same time. That several different unicellular types did so allowed their consequent mixing and matching. That the right place did not come into existence for a billion years or so after the origin of life helps to explain the long delay in eukaryote emergence. It also emphasizes that however plastic the genetics and biochemistry of prokaryotes, their evolutionary potential is feeble in comparison with sexually reproducing, multicellular organisms. The mix-and-matchability of their cells was involved in the evolution of new organs, but diminished as those became specialized. All was not lost, however, since migratory organizing cells could create new combinations.

Evolution by Association

121

One feature of symbiostasis, control of symbiont numbers by the host, may provide for the establishment of new symbioses. Excess numbers of Symbiodinium, the symbiont of corals, which may become toxic at high temperatures, are expelled by the host. While swimming on tides flooding over a heat-shocked coral reef I have several times experienced a zone of warm green water containing free coral dinoflagellates. Global warming now appears to have intensified the effect to the point of destroying some reefs.57 Normally these regions in high summer always contain numerous symbionts available for setting up shop in new hosts, as well as fodder for filter feeders that would otherwise find the tropical sea rather barren of food.58 This is how the ancestors of giant clams obtained their symbionts. In thermal vent regions of the ocean where volcanic sulfide escapes into oxygenated water there are so many free-living sulfur-oxidizing bacteria that they coat every surface, including the exposed gills of filter feeders that collect them as food. This multitudinous proximity, and high probability of ingestion by larvae, must have been important in establishing thiobiotic endosymbioses. All of the bivalves in the vent regions are symbiotic. In contrast, few bivalve types, and even fewer of the other invertebrates elsewhere in the ocean, have the symbiosis. Despite all of the other potential advantages and opportunities, most bivalves close their shells tight in the presence of sulfide, or die if they cannot somehow avoid it.59 Two ancillary emergent features for this association are the abandonment of the quintessential bivalve character of clamming up, and maybe having the ability to tolerate or detoxify sulfide. Like all other emergences, the sulfide-oxidizing symbiosis had to meet several generative conditions that were long in accumulating. However, the symbiogenesis itself is both saltatory and mutualistic, since the first bacterial cells could immediately begin to exchange nutrients and combine resources with the host. In the sea, freeliving sulfur-oxidizing bacteria are close to the redox interface between oxygenated and deoxygenated environments, and this fluctuates according to the production of sulfide and the turbulence of the water. The clam host offers a saltatory environmental improvement for its bacteria by regulating that redox interface and ensuring continuity of metabolic needs. It should now be clear that one of the most universal processes of symbiosis among primitive microorganisms, and then between them and animal hosts, is ingestion. For the process to succeed, the microorganisms must reproduce faster than they are digested, or become indigestible. This route for the establishment of symbionts has been established time and again, and it still happens in the case of horizontal transmission, where the symbionts are not passed to the next generation via the egg, but are recruited from the external environment. Fungal-algal symbioses could have been established simply by congregating at marine or freshwater shorelines. They must also have mingled in multi-organismal strands in shallow ponds subject to desiccation. As for cellulolytic symbioses, the

122

Chapter 3

symbionts were in the beginning universally distributed among decomposing plant material, and regularly ingested. A minimal resistance to digestion was needed to allow it to set up accommodations in the host’s gut—especially if the main chamber where it could operate was in advance of the normal stomach, as in ruminants. The novel proto-symbionts were then recycled through the feces of the host to be spread through the population of animals—as much a Lamarckian as a Darwinian development, since all of the animals ingesting the bacterial spores with their vegetable diet were simultaneously affected. Later there came a number of developmental experiments in guts that could accommodate the symbionts efficiently. At the most complex end of the spectrum of association are familial and social groups, whose success also requires the participants to be in the same place at the right time, to exchange not only nutritive and emotional energy, and not only information, but wisdom. Epigenetic Effects of Symbioses Major complexifications during the early history of life were due to the establishment of symbioses in themselves. The subsequent major phase of complexification involved cell multiplication and differentiation in multicellular organisms. Symbionts also contribute to this later phase through epigenetic effects on their hosts that influence gene expression or cell division. (“Epigenetics” refers to the normal and evolutionary processes of development.) The intestinal bacteriome that is found in many fluid feeding insects consists of large cells whose polyploid condition echoes that of the root nodule cells of leguminous plants. Flavinoids from the young host plants stimulate their Rhizobium symbionts to produce “nod factors,” glucosamines that in turn stimulate the production of “nodulin” in the plant. This stimulates growth of the nodules that house the symbionts, as well as membrane formation in the intracellular symbiosome, which is distinct from normal cell membrane.60 Thallus formation by the fungal component of lichens is stimulated by the symbiotic algae. Another example of an interspecific though not quite symbiotic epigenetic effect is the curious morphogenesis of sea lettuce genera, Ulva and Monostroma, which have presumably had characteristic lettuce-leaf thalli for most of their evolutionary history. Yet, this development requires the presence of epigenetically stimulating, free-living bacteria. In axenic (bacteria-free) culture the sea lettuces develop as masses of threads instead of their usual single broad thalli.61 A similar but more intimate epigenetic influence comes from symbiotic bacteria in Ardisia, a tropical flowering plant with fluted leaves. The leaf form and the successful growth and reproduction of the plant depend on the presence of the intercellular bacteria that are passed vertically in seeds. Diversification of the genus and its relatives may largely be the result of symbiosis with this and other similar bacteria.62

Evolution by Association

123

Not only plants are profoundly affected by the epigenetic effects of symbionts. Intestinal bacteria influence the functional development of the mouse gut.63 Multiple mechanisms, including digestive enzyme activation, immunity, mucosal defense, nutrient absorption, and normal postnatal gut development, all have some dependence on the presence of the commensal bacterium Bacteroides thetaiotamicron. How far back in the evolutionary history of symbiosis between bacteria and eumetazoans did such epigenetic influences originate? They are certainly not genedetermined by the host, nor even innate in the conventional interpretation of the term. Influencing the nature of their accommodations through epigenesis may be a common attribute of symbionts. Could the profound allometric expansion of the siphons of giant clams, where photosynthetic symbionts are housed, have been triggered by the dinoflagellates? Could the proliferation of bacteriocytes or even paedomorphosis in sulfur-oxidizing bivalves have been stimulated by their symbiotic bacteria, or by high tissue levels of sulfur compounds? Symbiostasis As Lynn Margulis says, symbiogenesis is like “a flash of evolutionary lightning.” Dazzling as that is, a lot of the foot-slogging needed to advance our knowledge of such emergences involves the establishment of the dynamic stability between the associated participants. In some cases, such as endosymbiosis, stasis is a desirable consequence of emergence. While symbiogenic experiments are saltatory, they are crude at the time of origin, and mutually beneficial adjustments are made by both parties as time goes by. Evolution progressed through epigenetic emergences built upon multicellularity that had emerged from endosymbiotic unicells, and foundational symbiostases were established at each level. Members of associations like endosymbiosis and differentiated multicellularity became so mutually attuned that they have hardly ever been undone. However, when associations are initiated, even if they are immediately mutualistic, the emergent wholes may lack the special competitive qualities of their formerly independent constituents, and may lack a cohesiveness that guarantees their continued co-existence. One disadvantage to symbiosis is that the partners become hostage to each other’s fortune. An important aspect of symbiostasis is limitation by the host of the population of unicellular symbionts. Commonly, the reproductive rate of the symbiont is lower in the host than that of similar free-living micro-organisms, and the control mechanism is usually assumed to be the limitation of symbiont nutrients. Giant clams farm their symbionts to some extent, their amebocytes culling and digesting the senescent cells, and delivering them to the giant kidneys of the clam for final processing.64 Surplus cells are also extruded through the digestive system, whence they can enter juvenile hosts. In sulfur-oxidizing symbioses, the hosts probably lyse aging bacteriocytes and

124

Chapter 3

the symbionts that they contain. Aphids somehow control the rate of division of their bacterial symbionts, although they increase in size. And the bacteria of mature aphids are lysed along with the mycetocyte cells that contain them.65 Other adjustments, in the first eukaryotes for example, included the surrender by the mitochondria of most of their protein synthesizing capacity to the host nucleus. Nuclear genes became organized into exon subunits that could be combined in a variety of ways, and after chromosomes emerged they were mixed and matched to make up linkage groups that sustained interactive gene coordination, and hence organismic wholeness. In some instances, symbioses have resulted in marked changes in the physiology and morphology of the hosts. After they got symbiotic dinoflagellates, giant clams stood on their heads and expanded their posteriors into massive siphonal pastures for their symbionts. A correlated growth of giant kidneys dealt with the new metabolism. Bivalves with sulfur-oxidizing symbionts sometimes underwent paedomorphosis or found other ways of increasing the gill volume to house their bacteria. Radical changes in gut anatomy followed the establishment of cellulose-digesting bacteria in terrestrial herbivores. Nitrogen-fixing bacteria established epigenetic feedback pathways that resulted in root-nodule formation. I have hinted that the symbionts themselves are epigenetic catalysts of physiological and anatomical change in animals. And all manner of anatomical arrangements have been made by terrestrial plants that optimize the photosynthesis of their chloroplasts. So after the establishment of the symbiosis, in addition to the minor changes and adjustments that contributed to symbiostasis, there have been emergent, epigenetic processes of allometry, and life cycle alteration that will require further examination in chapter 5. At the beginning of this chapter I proposed to survey symbioses, sex relationships, multicellular associations, and societies. Symbiosis, despite its many manifestations, forms a coherent topic, which demonstrates how very dissimilar types can interact for the benefit of the whole. But, in reserving the major discussion of cellular interactions for chapter 5, I have left a weak link between symbioses and societies. In the following section I reinforce that link with a discussion of how some authors have compared cellular associations with animal societies. I also assess the heuristic qualities of such comparisons. Societies, Cells, and System Reductions First, I wish to introduce the useful vocabulary of E. O. Wilson’s Sociobiology. I do not intend this to be a general preamble to, or endorsement of, sociobiology as he conceived of it, although the topic has a legitimate place in general zoology. Second, I evaluate two examples of system reductions that compare societies to cellular associations. System reduction, which is derived from general systems theory, is a method for understanding a complex system by analyzing a simple system that is sufficiently similar to allow legitimate inferences to be made.

Evolution by Association

125

Wilson uses the traditional definition of “society”: “a cooperating group of conspecific organisms.”66 An “aggregation” is a group with minimal communication and no organization, brought together by extrinsic conditions to enjoy some mutual benefit such as keeping warm in winter. A colony is a highly integrated group with specialized members, and Wilson applies the term to associations such as hydroid cnidarians with differentiated zooids, as well as social insects. These associations are “eusocial,” meaning that the individual members are genetically identical or highly similar because they are siblings. The original sense implied by a human colony, a transplantation that has a similar social structure to a parental society, is excluded. “Group” is a flexible term indicating a subset of a population whose members interact more strongly than they do with other members of the population. Wilson limits symbiosis to interactions between different species. “Social symbiosis” is used for social animals that associate with different species, such as ants that have domesticated aphids and scale insects that supply them with honeydew. These relationships can be divided up into social mutualisms, commensalisms, and parasitisms, which are self-explanatory. Wilson defines the following qualities of societies. Since he subsequently compared the insect society “superorganism” to an organism composed of cells, I add my own comments on cellular and social parallels in brackets. Cohesiveness—the physical proximity of members of the group that enhances communication, distribution of food and defense. [The comparison with multicellular systems is self-evident.] Connectedness—in the absence of unpatterned connectedness a communication signal is passed to all of the nearby members of the group, an alarm call in a flock of birds, for example, without discrimination even as to species. If the group has a hierarchical structure, signals can be confined to particular subgroups—the bellow of a rutting stag to other mature males, for example. [Within an organism a widely broadcast hormonal signal is only picked up by cells that have the appropriate specific recognition sites. Neurons have hard-wired connectedness, plus a phenotypic ability to experiment with dendritic connections.] Permeability—the degree of how open groups are to intermingling. Some social groups never accept newcomers, others accept them freely. [Prokaryotes are permeable to the genes of other individuals, not necessarily of the same type, but, by and large, permeability at the organism level through exchange of cells is confined to sexual reproduction and symbiogenesis. It could be argued that changes in the movement of organizer cells in development that contribute to evolutionary change might be included in this category. Permeability increases gene flow, but in social groups it tends to destabilize their interpersonal structure.]

126

Chapter 3

Compartmentalization—the organized complexity of both societies and organisms depends upon the degree of independence of the subgroups. Under attack, wildebeest leave their calves to be defended by the mother alone. Wilson contrasts them with zebra, which form defensive family units with the stallion taking a prominent protective role. He also cites ants that build semi-independent nests that exchange individuals with each other and have the potential to become mother nests if the original maternal site is destroyed. [Cellular compartmentalization of this kind is weak in lower animals and plants that can regenerate whole organisms from damaged parts. In higher animals artificial cloning is possible, but the original organism cannot lose essential compartments or organs without losing its integrity or being adversely diminished. Indeed, compartmentalization is part of a general trend of progressive evolution, contributing to the emergence of a sophisticated physiological homeostasis, and ultimately mind.] Differentiation—of roles in the case of societies—of cells in organisms. Ontogenic differentiation in societies of ants and marine animal colonies is a function of population density. [This is true to a limited degree in cell specialization in multicellular organisms as well, although it is also linked to evolutionary status. Human societies do not differentiate like ants, but there is division of labor and specialization of some functions that has equivalent social advantage.] Integration—a complex social system implies mutually beneficial integration of behavior. [Caste specialization in social insects parallels cellular differentiation in multicellular organisms, depending upon the elaboration of communicative signals in the form of chemical secretions, and on their selective recognition, followed by stereotyped responses. These reflexive insect responses to chemical information are not a fundamental part of the social integration of our species. There is evidence that pheromones have some socio-physiological impact, such as the coordination of menstrual cycles. In human society, however, an industry has grown up around blocking the effect of body odors. Moreover, where behavior is strongly colored by emotional responses to hormones, social disintegration is just as likely as coordination. More positive effects are found at the level of interpersonal relationships and the mutualism of families, which will be discussed shortly. Integration can also be assessed by the amount of time spent by individuals in contributing to social welfare—high in social insects; low in mammalian and bird societies; variable in humans, depending on social structure. Cells are, of course, committed to organismal welfare except under pathological conditions. Along with compartmentalization and differentiation, integration is part of the phenomenon that Aristotle identified as the separation of offices and concurrence of efforts. This he applied both to the harmonious interaction of the parts of an organism and to the organization of human society.]

Evolution by Association

127

A more recent comparison of social animals and cells appears in Bernard Crespi’s 2001 essay “The evolution of social behavior in microorganisms.” It is well known that many microorganisms form biofilms on a variety of water-covered surfaces such as underwater rocks, ships’ bottoms, and human teeth. Biofilms are rich sources of food for some aquatic organisms. Crespi, however, is interested in integrations of cells in biofilms that can be compared to properties found in the societies of higher animals. They may be cooperative, with divisions of labor that provide benefit to all members of the biofilm. These differentiations involve changes in gene expression, as they do in multicellular organisms. Biofilms have means of defending themselves from invaders, including the secretion of acellular polymers that form a matrix which also protects from desiccation and the infiltration of external toxins. They may be able to exclude cells of the same species that do not belong to the clone. These processes may involve intercellular chemical messengers, as may their diurnal rhythms, and the release of seeding cells at appropriate times.67 Population control is possible in some microorganism films or colonies through responding to the concentration of chemical signals that they have released into the near environment, a mechanism known as “quorum sensing.”68 In some cases biofilm structure is three-dimensionally complex, allowing for physiological exchanges, and for the accommodation of complementary species. Thus, as Crespi argues, biofilms provide shelter, along with nutritional, reproductive and defensive support, as do the constructions of ants, termites, gophers, mole rats, and humans. Even without close communication within biofilms, microorganisms may engage in group foraging, as when myxobacteria attack bacterial prey en masse.69 Pfiesteria dinoflagellates synchronously release fish-killing toxins, and feed on the cadavers— they are thus a particular menace to aquacultured fish.70 Some microorganisms cooperate reproductively. For example, while Rhizobium is fixing nitrogen in plant root nodules, it is incapable of reproduction.71 But if the nodule decays the bacteria return to full viability. Meanwhile, free-living members of the genus, which may benefit from ammonia leakage from the nodules, are always in reproductive readiness. Under limiting nitrate conditions, some cells in cyanobacterial filaments of Anabaena become nitrogen fixers, losing both photosynthetic and reproductive capacity.72 This kind of self-sacrifice for the good of the whole is taken further by bacteria such as Escherichia coli. When they are attacked by bacteriophages they stop synthesizing the antidote for one of their own poisons and thus commit suicide before the bacteriophage babies come to term.73 However, some cunning phages block the response and prevent the death of the bacteria. Moreover there are always cheaters that get into the act for a free ride. I have deliberately slipped into unrestrained metaphorical language, involving self-sacrifice, suicide, and cheating—things that humans do. This lack of semantic selfcontrol is a common failing among sociobiologists and “evolutionary psychologists.” Yet Crespi clearly defines what he means by some of these terms:

128

Chapter 3

Altruism: behavior that involves a fitness cost to one individual or cell (the altruist) and a fitness benefit to another individual or cell (the recipient of the altruistic act). Cheating: engaging in behavior that exploits the cooperative behavior of conspecifics by imposing fitness costs on them, while providing fitness benefits to the cheaters. . . . Cooperation: multiple individuals or cells engaging in a common task for mutual benefit.74

Without quibbling about questionable and undefined terms within the definitions, it is clear that they are applicable to bacteria and human beings. However all of the properties that emerge at higher levels are missing from the definitions. The human complexities of neural conditioning, hormone levels, morals, education, and purpose are absent. Now, the answer to this objection is that the definitions, as they stand, inadequate or not, do apply at all levels, and permit commonalities to be demonstrated at every hierarchical level. If that happens to be an intellectually productive exercise good and well. However, the next logical step for vulgar reductionists is to seek genes for cheating, genes for cooperation, genes for altruism, and so on in bacteria. The hidden agenda of this semantic reductionism is to “prove” that behavior can be reduced to those genes from every higher level. And the cheapest shot is that we too are the “victims” of those genes. Aristotle compared the division of anatomical labor in organisms to the responsibilities of governing offices in the city state. In like manner, the cell theorist Rudolf Virchow compared cellular activity to human society, regarding disease as internecine warfare.75 Such exercises in metaphor-making spring from a knowledge of human society, i.e., they are applied from the top down. It may be heuristic to discover how much putative higher-level complexity exists at lower levels, yet I do not find any novel conclusions about the nature of cellular interactions in the examples discussed above. The nature of cellular interactions does not generate the nature of societies. Moreover, it is counterproductive to semantically reduce the complexity of human society to something that cells or genes do, in the name of reductionist ideology. While it is interesting to seek general principles that apply to life at all its levels of complexity, their discovery should not divert us from comprehensive analysis and synthesis. At many levels during progressive evolution from the primitive protobiont level there have been episodic emergences of new associations involving self-organized complexity, with novel rules of operation. Figurative commonalities like permeability, integration, differentiation and integration occur in cells, insect societies, and human associations. However, the similarities that have been listed above are more of an interesting academic exercise than an answer to any outstanding biological questions. To understand emergent evolution it is more productive to identify and understand the differences—this is where sociobiology and emergentism part company. Ricard Solé and Brian Goodwin have recently reviewed the emergent behavioral properties of social insects in Signs of Life (2000). Here they extend the comparison of societies and multicellular organisms to the organization of complex brains. Many of

Evolution by Association

129

the features are similar in both systems. However, connectedness is a strongly contrasting property: neuronal interconnections are permanent—insect contacts are fleeting. Yet ants and termites engaged in nest building will keep coming back to the same place through positive pheromonal feedback, and local concentration may result in the construction of pillars and arches, in contrast to the random or chaotic deposition of building pellets. This may be compared to a long-term memory in a neuronal network. Both are long lasting emergent phenomena resulting from the mutual feedback between insect/neuron and the larger environment/brain.76 Such comparisons demonstrate that although the features of individual social insects or neurons may be known, behavior at the emergent level of interaction is not reducible to those features. Another interesting element in this comparison is the chaotic aspect. Although the behavior of individual social insects is chaotic, ordered cooperation may emerge, for example in the form of pillar building. Where it will happen is unpredictable. Solé and Goodwin suggest that the network of neural connections in the brain also has chaotic features. By the time of birth, most of the characteristic neural connections between sensory input and integrating nuclei have been permanently established. Yet there are billions of additional dendritic contacts, to be added at the unconscious level, which provide the raw material for thinking and learning and memorizing. To the extent that they are initially a tabula rasa they are chaotic. Yet a conscious signal, such as a visual pattern, can create an attractor in that chaos—an idea; a thought. With sufficient reinforcement, the idea will be stabilized subconsciously, to be retrieved when consciously stimulated.77 The program of identifying the genetic determination of human behavior that E. O. Wilson proposed in the last chapter of the first edition of Sociobiology (1975) has been pursued unsuccessfully. It is easy enough to find genes whose mutants somehow interfere with some kinds of behavior; but it is outrageously simplistic to declare that those are the genes for the activity. Nevertheless, tentative suggestions that human social activities have a genetic basis have become bald assertions of genetic determinism in Wilson’s Consilience (1998). We are told that what we assume to be freedom of human thought and action is nothing but the commingling of multiple gene-determined algorithms. By the way, William Whewell (1840) originally defined consilience as “‘jumping together’ of knowledge, an intellectual emergence produced by linking facts and fact-based theory across disciplines to create a common groundwork of explanation.”78 But not much leaping around is possible on a Procrustean bed built out of a reductionist premise that definitions, facts and theories have to be cut to fit. Unfortunately, Wilson has meanwhile helped spawn an “evolutionary psychology” intent on identifying the genetic determination of all human behavior. Regardless of any new ideas that might come from system reductions, we should concentrate on the qualitative differences between emergent levels, and address the

130

Chapter 3

question of how the differences might have emerged. Animals have followed three different lines of social evolution. Two of these, found in insects and birds, are similar in that they involve predetermined behavior patterns. The difference between the two is that some insects have additionally evolved societies that are largely constituted of non-breeding sibling workers, for example among social bees, ants and termites. Extreme cases of polymorphism are found in some social insects—again arising epigenetically from environmental (nutritional) influences. And the caste structure of their societies can be manipulated by differentially feeding the larvae. Bird societies are more flexible, since all of their members are generalists who retain their reproductive potential. The third type of society is found in mammals, and reaches its highest level of mutual benefit in humans, who combine a long period of infantile helplessness and family care with societal differentiation. The latter is not stereotyped and heritable, and it is not all beneficial, especially where the human faculties of objective observation and logical analysis are inhibited by the stasis of tradition. There is a tradeoff between these stultifying influences, and their occasional advantages, along with the general benefits of education. In this regard, Richard Dawkins’s (1982) development of the “meme” concept of enduring influential ideas is relevant. He feels that new memes, being subject to competition, are bound to improve through conscious selection. To me, the lively spark of a Good Idea is equivalent to an emergence. Its settlement into a persistent ideology with checks and balances that inhibit further new Good Ideas is equivalent to the establishment of a stable equilibrium by the syndrome of natural selection. Even in the world of commerce, this gets in the way of profitable innovation. As to social traditions, politics, and organized religion, it’s no wonder we have never emerged from the worst of those. The passage of memes from one generation to the next has been described as a kind of inheritance of acquired characteristics modulated by a kind of natural selection. Yet that kind of inheritance, as Lamarck complained, can get in the way of real evolutionary progress. However, this is not to be a treatise on educational philosophy nor political science, and I will henceforth confine myself to simple familial and social mutualism. There can be no question that such mutualism has been an essential part of our evolution. A pregnant mother has to have sufficient energy and dietary essentials to bring a healthy fetus to term. Such resources were necessary for the developmental evolution of humans to reach its present level. Then the post-partum life history continues with a helpless infant that cannot be independent of parents or social group for some years. Familial care, or more precisely, maternal care, is not a unique emergent property of humans, being highly developed in other mammals, and birds, and to a lesser extent in the other animal groups. Human parental care required a prolongation of an existent condition, with the added features of intelligent flexibility and patience. Hormonal reinforcements made parents more benign, particularly

Evolution by Association

131

fathers, who in most other species are conspicuous by their absence from the family circle, if not an actual threat to the welfare of the young. Exceptions in other mammals require hormonal changes. Djungarian hamster females and their young cannot survive without the care of the males, and this is achieved by mutually attractive hormones. Some such hormonal adjustments are now known for marmots that mate for life. Oxytocin would appear to make the female fall in love at first sight and vasopressin does the same for the male. Similar mixological experiments involving cocktails of oxytocin, dopamine, GABA, serotonin, and norepinephrine make humans behave like marmots on occasion, and in a less genteel manner. This tempts some observers to say that human behavior is “nothing but” the consequence of gene-determined molecular combinations. But feedback from higher emergent levels of cognition can influence and override the lower levels. Human associations are dynamically connected by communication through language. However, the status of language as a novelty of human evolution is debated by two camps of linguists and social scientists, who take either emergentistic or Darwinistic stances. Argument has centered upon the teaching of sign language to chimpanzees and gorillas, and there seems to be a Darwinist imperative among the researchers to demonstrate that there is a continuity of linguistic ability from the chimpanzees—our close biological relatives—to humans. If it can be shown that, despite their anatomical handicap for speech, the great apes have the ability to assimilate a large vocabulary, and to use word combinations creatively, then the whole progression of linguistic emergence may be interpreted as a gradual improvement of that ability, with any small improvement in vocalizing anatomy being naturally selected and accumulated in accordance with the conventional neo-Darwinistic interpretation. The debate is biased by anthropomorphic public opinion supporting “ape-ability,” which has nothing to do with the merits of the arguments, and a lot to do with wishful thinking about our cute relatives. In contrast, the emergence camp is accused of being anthropocentrically obsessed with human uniqueness. My own opinion is that human linguistic ability is a qualitative critical-point emergence correlated with cerebral expansion and reorganization, one of the major anatomical emergences of fetalization. It is not a product of quantitative accumulation of little bits and pieces that have all conveyed sufficient advantage to have gradually produced human speech. Along with language came logic. Chimpanzees, gorillas, monkeys, cetaceans, dogs, cats, parrots, and crows can communicate intraspecifically and with humans by making symbolic sounds. They also seem able to consciously identify an immediate problem of survival. And some of them can apparently stop and think, and then act. Instinctive responses are the fastest responses to such problems, but they lack the adaptability that can change the reaction in midstream. Yet birds, and even insects, which depend heavily on instinct, seem to be able to break out of it occasionally with

132

Chapter 3

novel solutions. Even so, humans are the only species that can perceive a problem, think it out to a logical, and novel conclusion before acting, devise an appropriate behavioral response, and then execute it, or hold back, according to circumstances and further analysis of the likely consequences of the action. An intellectual might approach a problem by rapidly assessing the possible hypotheses, designing thought experiments to test them, mentally working through the likely results, and finally, because of all the anticipated failures, do nothing. From the anthropomorphic point of view, a bright-eyed crow sitting on a branch and cocking its head might be doing exactly that, but, as the emergentist C. L. Morgan warned, we know from personal experience that humans do it; it would have to be proved that that’s what the crow was really doing. Logical abilities contribute immediately to the overall adaptability of humans and their social groups. Individuals benefit from observing the example of others. The vocal communication of appropriate action is part of the constellation of emergent qualities of Homo sapiens. Visual and auditory signals that transfer information have appeared in many social animals, including insects and birds. But other more sophisticated forms of communication such as sign language have not evolved to any extent in social animals in the natural environment. In humans, signing comes after the fact of language, which requires a vocal apparatus, and a range of auditory nuances, integrated with the innate logico-linguistic functions of the neocortex. Only then could the invention of signing could be passed down to the great apes under laboratory conditions. Alfred Russel Wallace wondered what would be the selective advantage of a brain that could do calculus long before it was ever needed? However, a prenumerate human has the unconscious logical ability to analyze rate changes, and voluntarily integrate this ability with vision and action, when, for example, a running hunter throws a spear at a prey also on the run. Although it was formally translated by Newton into calculus, it already existed as an emergent, usable, albeit unformulated quality of the human brain. It might never be used, and if used would have been subject to modification by practice. Another early advantageous use for the emergent big brain was a quantitative leap in the number of meaningful symbols or words that could be remembered and vocalized. There was also linguistic framing of logic. Although logic and an innate grammar, the latter concept itself anathema to Darwinist behaviorists, occupy different parts of the brain, neuronal connections assure their synergy. Another aspect of mind also has a social context. Self-consciousness is not necessarily confined to humans. But we also recognize that others are similarly self-conscious, and can both empathize with others, and take advantage of that ability. To begin to communicate linguistically, it helps to be aware that others have the same kinds of minds, and can perceive the same questions and answers.

Evolution by Association

133

Another unique emergent function of the neocortex of Homo sapiens was the combination of memory with the cultural feature of education. Learning from your own mistakes is a matter of basic logic and memory, though it comes at a price. Learning from the mistakes of others is another adaptability with high utility and low energy cost, even where there is no immediate advantage, and the system operates out of sight of natural selection. All of these adaptable features were inherent qualities of the emergent brain of our species, and their potential was further realized by education and tradition. Instinctive behavior patterns in insects and birds probably started with a degree of behavioral flexibility that became genetically assimilated. That is, if any advantageous activities could be reinforced by gene-based features, the link between the two would become increasingly fixed. Plasticity was traded off for speed of response. However, there did not have to be a genetic fixation for the human traditions to degenerate into the inflexible stasis of rote learning and ritual. The history of medicine provides numerous examples. Quick-witted folk healers would have been alert to new knowledge of herbal medicines and toxins, but oral traditions tend to lose the meaning they had for their discoverers, and to become cluttered with distortions. Written traditions fared little better. The intellectual flowering of Hippocratic medicine 24 centuries ago degenerated into a rigid conformity that was briefly shaken up by the medical founders of the Alexandrian Schools a century later. But medicine did not enjoy a full renaissance until Galen, after a four-century hiatus. Fortunately, we humans have the emergent saving grace of humor to help us through strangulating conventions. This adaptable manifestation of our innate logical ability puts unrelated ideas together to provide the novel answers and surprised laughter that finally break us free. The fixation of human traditions is connected with an issue that is significant for the formal treatment of evolution by association as a subset of an emergence synthesis. Since I know of no word that covers all the biological and sociological aspects of this issue I have to invent “symbiostasis.” At one end of the association spectrum, sociobiology recognizes the development of change-resistant dynamic social stabilities—social homeostasis. At the other end of the spectrum, where endosymbiosis resides, dynamic biochemical and physiological stabilities are involved. Symbiosis and Major Evolutionary Transitions Some of the associations outlined in this chapter are among “major transitions in evolution” discussed by John Maynard Smith and Eors Szathmáry (1995). For neoDarwinists to venture thither is a new salient, and they are to be congratulated for tackling the “difficult thinking” involved in explaining natural selection as the cause

134

Chapter 3

of emergent novelty—even if it sometimes leads them into a logical quagmire. I have no argument with their estimation of the lapsed time periods, and no fault to find with their overall historical account. However, I would expect that other ultraDarwinists find the following hard to swallow: We offer the following escape from this paradox [that life originated in a much shorter time than it took for the eukaryotes to emerge]. During the origin of life, innovations were compared with relatively inefficient competitors. Established prokaryotes are in contrast vigorous competitors. Innovations (e.g. loss of the cell wall and origin of the nucleus) are likely to cause disruption and a transient loss in fitness. Thus, special circumstances—in fact a series of them—must have been needed for these innovations to become fixed, despite the existence of potentially winning competitors.79

Maynard Smith and Szathmáry explicitly exclude natural selection from the buildup of generative conditions for the emergence of life, since biological reproduction had not yet been acquired. Here they argue that protolife thrived when competition was absent, and when competition strengthened, it obstructed subsequent major innovations. If that is so, their mandate of explaining transitions in terms of selective advantage is redundant. Accepting that the eukaryotic condition found in living organisms did not emerge fully formed, and failing catastrophic destruction of the competition, the protoendosymbioses must have survived through greater adaptability, in environments that were marginal for existent prokaryotes or where they were not in competition for the same resources. A reductionistic focus on a step by step accumulation of innovations never perceives the range of holistic emergent properties that can be simultaneously generated by integrative complexification. To most Ultras, there is a prolonged evolutionary ferment or struggle for existence and progress is a mere epiphenomenon. In contrast the emergentist view is that molecular innovations are instantaneous. They immediately work in the context of the modified whole, or make no difference, or cause its disintegration. But the primitive prokaryotic experimental tools for such innovations were limited. Furthermore, the emergence of eukaryotes may have crucially depended on various prokaryotes being in the same place at the same time, the place being an interface between aerobic and anaerobic environments. And the time taken to get to that point was prolonged by the slow buildup of sufficient aerobic conditions. Even where competition is brisk, a new wholeness may place emergents beyond it, or in a position to exploit new environments where the traditional competition fears to tread. There is no need to adduce an imaginary force that brings pressure to bear, and picks and chooses from among the old and the new. However, as Darwin well knew, the very success of novel organisms re-introduces competition among them, reimposes dynamic stability, or regression, and constructs barriers to further progress.

Evolution by Association

135

Program Notes In the preceding chapter I issued a checklist of things to look out for at the circus. After our session at the Symbiosis Arena, we can fill in some of the blanks. 1. Mechanisms that perform in all three rings of the evolutionary circus. Although we have only attended to one of the rings, mixing and matching of independent organisms was a crucial part of the original evolutionary process. The same idea applies in other arenas at the level of gene expression, cellular differentiation, and physiological and social organization. Therefore the phenomenon of repetition and differentiation of molecules and multicellular associations are also important, and will be on show in the physiological and developmental arenas. 2. The generative conditions from which emergences spring, and common features of generative conditions. So far we have observed the importance of the aggregation of potential symbionts at appropriate interfaces. Being in the same place at the same time is an important and highly likely contingency for the establishment of known symbioses, sexual associations, and interspecific associations. Particular symbioses have unique emergent properties. But environmental interfaces will be shown to be important and predictable for physiological emergences as well. Here I wish to re-emphasize the idea that it may only have been the emergence of an oxidizing environment that provided the interface for the concentration of the unicells that were to participate in endosymbiosis, which would go a long way to explaining what took the eukaryotes so long to emerge. 3. Novelties that catalyze emergences, provided that appropriate generative conditions are present. As might be expected, these are mostly unique emergences affecting established organismal systems, and appropriate to particular conditions of life. As well as internal adjustments that followed on the heels of the emergence of endosymbioses are many that were in themselves saltatory novelties. The establishment of nuclei and the construction of chromosomes are examples. 4. Particular contingencies, predictable or unpredictable, which have affected generative conditions. In the symbiosis arena these have already been addressed under article 2. One addition might be the involvement of very large numbers of potential symbionts, as in the case of sulfur-oxidizing symbiosis or cellulolytic symbiosis.

136

Chapter 3

5. The constellation of multiple functions characteristic of new emergences. It cannot be emphasized too strongly that endosymbioses established the metabolic complexity that potentiated plant and animal physiological and epigenetic evolution, sometimes through regression of the original complexity. The role of symbioses in founding whole ecosystems, and the role of associations in founding complex societies, including our own, emphasizes how they have generated multiple functions indispensable for progressive evolution. These conclusions are also appropriate to the final article. 6. The course of emergent evolution that has progressively led to greater self-organization, independence, and freedom of choice. After the alpha of the symbiogenesis of life came eukaryotic unicells and then multicellular associations. From their differentiation emerged the morphological and functional complexities that carried evolutionary progress forward to the omega of the meeting of minds. Next we will be looking at physiological emergences, which depended on primordial associations to build the biochemical base for homeostasis. It will become yet more obvious that the division of the causes of evolution into the associative, the physiological, and the epigenetic is arbitrary, if convenient. Moreover, these arenas of causation all exist in an environmental context, and will be found to depend to some extent, for both their origin and survival, on environmental contingencies.

4 The Physiological Arena

. . . the highest organic development is the most complete division of labour and the most perfect physiological centralization. . . . Organic progress consists in increasing differentiation and increasing integration. —J. J. Murphy, 18691 The problem of the organism as a whole is the problem of the origin, development, and maintenance of the mechanisms of integration in their relation to origin, development and maintenance of the individual. This is first of all a problem of physiology, not of heredity, because . . . heredity does not account for the individual, but merely for the potentialities, some of which are realised in the individual. —C. M. Child, 19242 We have thus arrived at a definition of evolutionary progress as consisting in a raising of the upper level of biological efficiency, this being defined as increased control over and independence of the environment. As an alternative we might define it as a raising of the upper level of all-round functional efficiency and of harmony of internal adjustment. —Julian Huxley, 19423

Physiology is the study of function. Therefore physiological evolutionists must address the question of how function evolved. We might even ask if it evolved. The philosopher-biologist J. H. Woodger believed it did not, since all living things shared the same general metabolic activities in common that allow their persistence in being.4 All placental mammals, however diverse, have the same physiological organization and adaptability, which does not diversify in parallel with anatomy. As Darwin himself said, they share “an innate wide flexibility of constitution” onto which adaptations can be “grafted.”5 Placentals share a good many of their molecular metabolic activities with their primitive ancestors all the way down to prokaryotes. Therefore, up to a point, Woodger was right. But Julian Huxley was also right to recognize the evolutionary significance of a “raising of functional efficiency.” Physiological progress enabled the diversification of behavior and anatomy in the higher chordates and other animal phyla.

138

Chapter 4

Woodger thought that physiologists were not interested in evolution because metabolic activities were much the same in all organisms. The reverse of the coin is the indifference of evolutionists to physiology, despite the correlation between homeostatic sophistication and anatomical adaptive radiation. Darwin remarked, “Practically, when naturalists are at work, they do not trouble themselves about the physiological value of the characters which they use in defining a group or in allocating any particular species.”6 If the most important physiological features were held in common across phyletic gaps, it was the grafted adaptations of anatomy that were of greater evolutionary interest, because their presence could be attributed to natural selection. J. J. Murphy, who is quoted at the chapter’s head, understood that the evolution of organismal complexity was linked with physiology and development, and St. George Jackson Mivart and E. D. Cope were paying attention. Perhaps one reason why evolutionary physiology was neglected for most of the twentieth century was its early affiliation with neo-Lamarckism. A recent attempt to round the circle of evolutionary causation, beyond ecological adaptation and population change, incorporates development, yet leaves out physiology and behavior as its minor consequences.7 But I challenge anyone to extrapolate from the most perfect knowledge of mammalian embryology to explain mammalian physiology and behavior, far less their evolution. The physiology of the embryo and fetus is that of the mature mammal, since, in utero, they enjoy the homeostasis of their mother. The impossibility of predicting and explaining the nature of the adult from a knowledge of the embryo parallels the problem of predicting and explaining the nature of more complex forms from simple, primitive organisms. This is because evolutionary emergences have novel properties unknown at the lower hierarchical levels of form and function. Since evo-devos assume that physiology is a consequence of development, physiology has been left out of their current agenda. But there is an older historical reason why physiological evolution has been disregarded: the fascination with comparative morphology. Gross anatomy is what catches the eye, and its evolution can be traced in the geological record. Physiological qualities do not fossilize, yet they support the behaviors that constrain and elicit anatomical change. Another reason for the neglect of physiological evolution is the historical domination of animal physiology by medicine, which emphasized its practical relevance to humans. PostDarwinian physiologists, such as Claude Bernard, Walter Cannon, and C. S. Sherrington, took a theoretical interest in homeostasis and the integration of regulatory mechanisms, but paid little attention to the evolution of such systems.8 A search for relevant and thoughtful epigraphs took me to where physiology borders on psychology. In Physiological Foundations of Behaviour (1924), C. M. Child showed a physiologist’s appreciation for the organism as a whole, and the same understanding of persistence in being as Darwin. My third epigraph is from Julian Huxley’s Evolution:

The Physiological Arena

139

The Modern Synthesis (1942). Its major theme, progressive evolution, was identified as improvement in physiological coordination and efficiency. Unfortunately, although he regarded regulation of body fluid salinity, the amniote egg, the placenta, and homeothermy, as major physiological emergences, he did not explain their origins. Others who contemplated physiological adaptability are to be found at the interface between physiology and embryology. For example, I. I. Schmalhausen, in Factors of Evolution: The Theory of Stabilizing Selection (1949) agreed that animals had always had universal metabolic functions, but he also believed that the general features evolved: “Physiologic adaptability is, to a certain extent, a characteristic property of all organisms. It has arisen simultaneously with life itself and has developed progressively as the vital processes became increasingly complex.”9 Interaction with the environment and genetic assimilation were important parts of the process. He assumed that self-organization was perfectly capable of accommodating epigenetic change as well as physiological novelty, since it had frequently been done over the course of time. And he was also well aware of the mutual impact of behavioral and physiological evolution. Moreover, when C. H. Waddington (1957) wrote about “the strategy of the genes,” he had general physiology in mind as part of that strategy. In Internal Factors in Evolution (1965), L. L. Whyte dealt broadly with the “coordinative conditions” that integrated developmental and physiological features. Mutations or other changes were co-adapted within the existing dynamic stability by “internal selection.”10 David Rollo, an ecologist with an unusual appreciation of physiological evolution, has commented: The wisdom of the early naturalists that organisms are intricately balanced so as to carry out a diverse spectrum of different and specific functions effectively has become formalized in the ecological concept of the adaptive suite. The essence of this appellation is that each feature of the phenotype may be adaptively honed by evolution for maximal competence, but the integration of various aspects may be of even greater importance (i.e. the whole is greater than the sum of its parts).11

Although comparative physiologists and biochemists will protest with some justification that they routinely study physiological adaptations and their energetic efficiency, they have largely ignored what interested Schmalhausen and Whyte: the evolution of homeostasis from its rudimentary condition in the first living cells. And that is the heart of the matter. By equating adaptability with “adaptive suites,” Rollo comes too close to inferring that the elements of complex systems are adaptations that were individually selected, and that the whole is their sum. Many such adaptations exist—physiological specializations for particular habits work well in certain environments. Many others fine-tune the internal dynamic stability of homeostasis, but they do not generate physiological progress; nor do they offer adaptable multifunctionality. How homeostasis emerges to higher levels of organization and adaptability is the greater

140

Chapter 4

issue. Michael Conrad’s Adaptability: The Significance of Variability from Molecule to Ecosystem (1983) transcends the adaptive suite concept, but dwells little upon the physiological stage between molecule and ecosystem. However, his later essay “The geometry of evolution” (1990) anticipates my present theme: Why does evolution work? The reason is not to be found solely in the magic optimizing power of variation and selection. It is as much due to the organizational structure that undergoes the variation. Evolution works because this organization is amenable to evolution, and because this amenability itself increases in the course of evolution.12

This reassures us that Conrad is aware of the fundamentally progressive nature of evolutionary physiology. But while he emphasizes the connection of adaptability and self-organization, he insists that both must “earn the right to persist” by having some advantageous trait that can transport all the rest as “hitchhikers.” A different way of seeing this is not to separate the advantageous trait from the freeloaders, but to appreciate it as an integral part of a constellation of features without which it would have neither advantage nor existence. As I argued in chapter 1, organisms have always had the physiological right to persist, or to maintain their integrity, except when they have conducted self-destructive experiments or suffered a natural catastrophe. Any improvement in self-organization is self-sufficient, bringing a blanket utility that is already prepared for a range of environmental vicissitudes. It does not follow that adaptable generalists will always compete better than specialists; but their emergence in the first place is independent of their future competitiveness. It might puzzle the reader to note that none of the theoreticians in the above discussion are or were practicing physiologists. It is certainly bothersome to a physiologist who has been attempting to explain physiological evolution to students for 30 years, yet has never found a textbook that devotes more than a page to the general principles of physiological evolution. Before going any further I want to make a minor grammatical point. Usually the abstract word “adaptability” has neither a plural form, nor the company of the indirect article. It is nevertheless convenient to treat adaptability in the same way as the similar noun ability. Specific adaptabilities of respiration, digestion, excretion and so on, in combination with general systems of regulation, add up to all-round adaptability. It seems awkward at first, but one gets used to it, and it avoids circumlocutions. The Distinction between Adaptation and Adaptability The performance of both adaptational modifications and adaptable emergents can be observed in the physiological arena. Adaptation is a genetically fixed and inflexible quality that is appropriate to certain conditions of the external or internal environments. If these remain the same, adaptation continues to enhance survival and

The Physiological Arena

141

reproduction. Extreme adaptations, or particular specializations to the prevailing circumstances of mature ecosystems, may confer high fitness, and may exclude natural experiments in adaptability that are unable to compete. Although internal adaptations have been mostly ignored by selectionists, it is quite usual for physiologists to think, for example, of a gut being adapted to diet in its anatomy and enzymological profile. Moreover, some modern comparative physiologists or physiological ecologists have taken to totting up optimal energy budgets for physiological qualities—only to find that the cheapest is not always the most competitive. Internal adaptations, such as enzyme variants that function better in specific pH or salinity conditions than others of the same protein family, are real—all of them can be altered by mutation and thus re-adapt if conditions change. Internal selection of alleles for these altered proteins tends to stabilize the coordinative conditions of physiology, or fine-tune homeostasis. While adaptations are inflexible, adaptability is the quality of an individual organism to adjust itself effectively in response to internal and external environmental change. During its own lifetime it thereby maintains its internal milieu, i.e., modifies its own behavior, physiology and even anatomy, in the face of environmental change. Behavioral change also widens environmental range. The environment need not fluctuate, nor be unpredictable, for adaptability to be brought into play. A gannet’s homeostasis allows it to withstand the random blizzards of winter and heat waves of summer. It may launch itself from the sun-baked ledge of a cliff into flight, to find its fishy prey, which it then catches by plunging totally into the cold sea. These environments are quite predictable, and do not change per se—the gannet’s physiological adaptability allows it to stay the same when the environment changes, and do something different if the environment stays the same. To equate adaptability with “robustness,” as is quite common in the current biological literature, is to generalize too broadly, since highly adaptable organisms may not be robust in the usual sense of the word, and organisms of low adaptability that conform to environmental change may be highly robust. Compare, for example, the fate of deep-frozen prokaryotes and deep-frozen people once they are thawed. Adaptability is genetically based, in the sense that it needs the information of DNA that can be translated into enzymes and structural proteins. But adaptability is not fixed in its phenotypic expression, since the multiplicity of genes on which its functions partly depend can be switched on or off by organismal and environmental conditions, and those that are active can have an inhibitory or activatory role within a hierarchical organization. For example, the voluntary action of writing this passage includes several layers of operation. Its mental construction and editing involve creative neocortical intelligence, wakefulness determined by neurotransmitter molecules, and perhaps emotional coloring by the endocrine system. As a sentence is

142

Chapter 4

transmitted to a written form, sight, excitatory and inhibitory neural control of finger muscles, and tactile and visual feedback come into play. All of these are governed from the top down by a conscious decision, and none of them are gene-determined, although the replenishment of enzymes, neurotransmitters, hormones, and muscles depends on gene-coded protein syntheses. Nor are most of the developmental processes that give rise to the organ systems in question directly gene-determined; they too consist of hierarchically internested molecular, cellular and organ functions. The greater the degree of adaptability, the more the organism can cope with changing conditions. Why, in that case, have we paid so little attention to the question of how it so evolved? One reason is that “adaptation” is often used as a catch-all term for adaptation and adaptability as I have defined them. The word “adaptive” is also problematical since it is given the sense of “useful,” and refers to features that are “adaptable” (= phenotypically modifiable or adjustable) and “adaptational” (= pertaining to genetically fixed characteristics that are useful in particular environmental contexts). Because it is essential to make a strong distinction between the adaptable and the adaptational, I try to minimize my use of “adaptive.” Unfortunately, physiologists often refer to modifications of an individual animal as adaptations. When they say we adapt to the cold by shivering, nobody bothers that it confuses two fundamentally different kinds of response. When we get cold, “the gene for shivering” is not switched on, since it doesn’t exist. Yet popular interpretations of molecular biology, which professionals sometimes encourage, would have us believe that there is a gene for everything. And if adaptability is arbitrarily treated as a subset of adaptation, it becomes easy, not only to confuse two distinct categories, but also to lose the most important one. Michael Conrad (1983) says that it is necessary to develop a theory of fitness that “does not have to answer every question about the functional value of this or that adaptation.” Conrad continues: In fact, this is quite impossible because particular adaptations are so particular and specific. However, this is not true for adaptability. This is because adaptability is adaptation to the uncertainty of the environment, therefore adaptation is something which can in fact be characterized in a very general way, independent of the specific features of different environments. Moreover, adaptability and also reliability (or adaptations for coping with noise) are just the adaptations which are crucial from the standpoint of determining the directions of evolution, for not only are they what determines whether some particular biological organization has the “right to persist,” but they also encompass the genetic search strategies which determine the probability with which these different organizations come into existence.13

Conrad, a stickler for formal usages, defines adaptability as “the ability to cope with the unexpected disturbances of the environment,” and “reliability” has the special sense of “the ability to cope with internal noise,” or shutting out irrelevant perturba-

The Physiological Arena

143

tions within systems, whether they be cellular, organismic, or pertaining to populations, communities, or ecosystems. I have no fundamental objection to either of these definitions, and applaud his clear recognition of the importance of adaptability for determining evolutionary directions. Nevertheless, Conrad also sees adaptability as “adaptation to the uncertainty of the environment” [emphasis added], which rationalizes adaptation as the higher category and adaptability as the subset, when he should instead be explaining the mystery of how the challenge of uncertainty can be met. Since he obviously finds a real distinction between particular adaptations and adaptability, his semantic reduction of adaptability to adaptation is not careless, but an evasion of the difference, perhaps to make his comprehensive fitness theory tidier. Yet, sweeping adaptability under the rug of adaptation is not real theoretical tidiness—out of sight is not out of mind if the contradictions are fighting like cat and dog under there. Subsuming adaptability under adaptation “solves” Darwin’s problem by making “persistence in being” an adaptation and therefore amenable to natural selection. And that of course will reassure neoDarwinists that what they previously neglected as an exception can be harmoniously subsumed, without carrying the analysis any further. A parallel problem arises when physiological adaptability is equated with the variability of populations—as Conrad does in the very title of his book: Adaptability: The Significance of Variability from Molecule to Ecosystem (1983). This ensures that adaptability can be filed and forgotten in the same box as genes that may have multiple alleles, or a population that has broad genetic variations or polymorphisms. That ploy allows population biologists to ignore physiological adaptability and get on with bean counting: a case of “let’s not do it but say we did.” The adaptability of an organism— the ability to modify itself or its behavior in direct response to environmental or internal change—is a property that is qualitatively distinct from variability in populations. And if popular usage confuses categories it should be corrected. Conrad can keep variability as a property of populations, but I am going to treat adaptability as an exclusive property of organisms, and deal with it in my own terms with particular reference to the physiological arena. Since I have been largely concerned in the adaptability at the level of physiological organs, and with the general adaptability of the organism to its environment, I have ignored a significant aspect of Conrad’s Adaptability: the adaptability of biological molecules to one another. Marc Kirschner and John Gerhart compensate for this deficiency in The Plausibility of Life: Resolving Darwin’s Dilemma (2005), in which they demonstrate how the adaptability of regulatory mechanisms can not only ensure that the organism survives biomolecular change but also accommodate environmental effects as part of epigenetic and hence morphological change. The significance of the latter will be raised in chapter 9.

144

Chapter 4

Modes of Adaptability Toleration and Accommodation Some organisms can tolerate internal environmental change imposed by external environmental change, up to a limit. But tolerance is usually accompanied by behavioral adaptability such as shutting out the environment, or escaping from the hostile zone. Tolerance also may be supplemented by the kind of adaptability known as “accommodation,” where, despite internal change imposed by external environmental change, physiological responses allow the organism to function adequately. The most sophisticated kind of adaptability is homeostasis, a stable dynamic equilibrium that keeps the internal milieu relatively constant, despite external environmental changes. Many marine animals that live on the seashore tolerate the diluting effects of freshwater runoff, periodic drying, heating and cooling, and when conditions become intolerable they shut themselves off from the environment, or walk, run or swim away. But they often have accommodatory mechanisms better than mere tolerance. For example, a species of bivalve lives on the western beaches of my neighborhood fjord, Saanich Inlet. Nuttallia obscurata, called the “varnish clam” because of its shiny, yellow-brown, outer protein coat, was introduced from Japan or Korea about 20 years ago and has already spread far to the north.14 Its larvae are attracted to fresh water flushes and settle in large numbers in the high tidal zone, some of them even burrowing in the bed of an outflowing rivulet. I regularly find adult specimens alive, but tightly closed, in the lower reach of the stream. It seems as if the constant presence of fresh water finally makes them decamp and resettle further down the beach where the water is saltier. Their adaptability consists of a combination of clamming up, locomotory mobility, passive drifting, and accommodating to fresh water. They also have unusually versatile feeding habits, including ciliary feeding with the foot, a habit usually confined to juvenile bivalves. Their choice of the upper shore has separated them from most of the other clams, so they are free of interspecific competition for space, although they run the risk of discovery by terrestrial predators. As an example of accommodatory mechanisms in animals subject to internal change brought about by external change, some salmonids have a suite of kinases, energy conversion enzymes that have maximum efficiencies at different temperatures, and so can continue swimming at the same speed when they get warmer or colder. Seasonal temperatures change much too fast for random gene duplication and mutation to keep up, so the suite of enzymes had to be already in place to be effective when the fish subjected themselves to environments with a fluctuating temperature regime. Homeostatic stability of the internal milieu through the operation of regulatory feedback mechanisms makes existence in different environments possible. It also

The Physiological Arena

145

potentiated brain development, the evolution of intelligent behavior, and the emergence of mind. These, in turn, supported homeostasis by inventing clothing, fire, houses, scuba gear, and vacuum-proof spacecraft. Although the scope of behavior in animals other than ourselves usually has to be limited to what homeostasis can deal with effectively, there are exceptions of toleration, where some homeostatic elements are temporarily or permanently abandoned. For example, humans, like many other warm-blooded vertebrates, accommodate to strenuous exercise by accumulating an oxygen debt. That means that the muscles switch to anaerobic respiration and temporarily build up lactic acid. During this accommodation period, oxygen levels, metabolism, and lactic acid concentrations are all different from the resting equilibrium. When exercise stops, and sufficient oxygen is present, the homeostatic balance is restored. There are more striking examples of temporary regression of homeostatic regulation. A dehydrated camel allows its core temperature to fluctuate several degrees from its normal set temperature. Relaxation of homeothermy into toleration of change is an option that saves several liters of body water per day, which would otherwise be sweated off to keep the body temperature down to normal. It is also contingent on a climate with cool nights, allowing the excess body heat to be radiated off, leaving the camel to start the day cold, and characteristically grumpy. Even more exceptional is the eusocial naked mole rat Heterocephalus glaber, which lives in burrows that have a high, stable temperature. It has largely abandoned endothermic heat generation, with the advantage of significant savings in metabolic energy. R. D. Alexander (1991) sees this in the larger context of an “odd constellation of attributes including nakedness, slow growth, ectothermy, burrowing life-style, eusociality and specialization on underground tubers . . . unlikely to be understood unless they are considered together.”15 This is the kind of holistic, or interactionist viewpoint necessary to understand the emergence of physiological novelty, which comes out of an array of interrelated generative conditions, and confers a multifunctional constellation of new ones. Homeostasis The highly adaptable condition of mammalian homeostasis was first described by the physiologist Claude Bernard as a constancy of the internal milieu that conferred la vie libre, or independence from the external environment.16 He did not mean that the internal environment was rigidly fixed, knowing that continuous physiological adjustments were needed to maintain the apparent constancy. Bernard was not just making a vague generalization, for he had analyzed specific components such as the regulation of blood sugar. In The Wisdom of the Body (1932), the physiologist Walter Cannon coined the word “homeostasis” for this physiological condition. Like Bernard, Cannon was indifferent to evolutionary explanations.

146

Chapter 4

In teaching comparative physiology, and trying to explain the evolution of the physiological steady state, I raise the “homeostasis paradox.” Given that the internal milieu has changed historically in its salinity, temperature, oxygen capacity, and organic content, how can homeostasis evolve, if during the process it progressively becomes more change-resistant? Alexei Severtsov was fascinated with such problems of the evolution of adaptability, as was his junior colleague Ivan Schmalhausen.17 A few decades later, the theorist C. H. Waddington realized that a parallel paradox existed in developmental evolution. The consistency of developmental functions, which he called “homeorhesis,” resists embryonic change in the same way that homeostasis resists physiological change. Waddington came to the same conclusion as Schmalhausen: changes in stable internal systems were most likely imposed from without.18 Physicochemical parameters of the internal milieu have historically been disequilibrated and changed in response to environment, especially in freshwater and terrestrial animals. This straightforward environmental impact was named “physiogenesis” by the neo-Lamarckist E. D. Cope. Among those physicochemical features, body fluid salinity, oxygen content, temperature, and mechanical stresses have been changed by the environment over the course of time. Ultimately standards were set and regulated as part of organismal homeostasis. These features were initially changed by the environment, and did not evolve in a biological sense—what evolved were the homeostatic mechanisms that regulate them, especially the hormonal and nervous systems that increase overall adaptability. Those went on evolving after the physicochemical factors had reached their final set points. The unanswered question is “How did the organism genetically assimilate what began as an ontogenic effect?” The conventional hypothesis is that any genetic or epigenetic change that complements the ontogenic effect would be marginally advantageous since it would give the organism greater freedom from its environment. But this is paradoxical. Why go to the lengths of genetic assimilation which is not per se advantageous if the organism is consistently and persistently exposed to the required environmental stimulus at no energetic cost? Physiology, as the examination of function, combines the study of internal organismal biochemical activities and functional anatomy. But it should be clear from the previous discussion that physiological evolution also needs to be understood in a behavioral context. An animal might stay where it is while the environment changes climatically or catastrophically. Alternatively, it may go to a new environment that will affect the evolution of its lineage. This requires complex behavior, which in turn depends upon the adaptability of underlying physiological systems. Here is where genetic accommodation shows its worth, even when the organism has no need to depend on it until behavior and changed environment so dictate. There would be no future in exploring an environment where survival was impossible. Such adventurism

The Physiological Arena

147

is thwarted by stereotyped behavior that is predictably fit for the prevailing conditions of life, genetically based and “rigorously selected”; but it leaves little scope for accident and experiment. The evolutionary alternative is to increase physiological adaptability and freedom of behavioral choice. Along that route, intelligence may emerge, and the dangerous streak of curiosity kept under conscious control. .The grounds of the idea that behavioral change results in ontogenic change and thus catalyzes evolutionary change is discussed by Gilbert Gottlieb (1987) as an essential component of a genuine evolutionary psychology. Gottlieb, like the historical predecessors he mentions, does not emphasize physiological adaptability as the foundation of such behavioral experiments. The feedback relationship between internal physiological functions and externally directed behaviors have changed the epigenetic and mature internal milieu, made physiology more complex, and led to further behavioral evolution. Since behavioral studies are commandeered by ethology or comparative psychology, and early developmental functions are studied under embryology and epigenetics, a more interdisciplinary approach is needed to integrate development, physiology, and behavior. Roy Pearson (1986; 2004) has contributed to such a synthesis by merging homeostasis with developmental physiology as “homeodynamics.” However, I will keep it simple for the time being, and try to make the most of physiology in its own right. It is an oversimplication to see a physiological organ system as an adaptation to ecological conditions. It must also be appreciated that increased complexity of regulatory systems can provide general-purpose improvements, regardless of current exigencies. For example, a neo-Darwinist explanation of nerve myelination might be that first one axon of particular adaptive value was myelinated, and subsequent myelinations were selected according to their contribution to Darwinian fitness. In contrast, myelination of nerves could be seen as an emergent phenomenon involving much of the nervous system, activated by an epigenetic change that made all of the accessory glial cells act in the same way. It was not only a specific escape response to a particular predator that was speeded up (and then selected). Multiple sensorimotor functions were affected. Multifunctionality is an important aspect of the evolution of more complex homeostatic systems. Multifunctionality Most emergent systems have multiple functions, some of which may confer competitive advantage. Others may later increase the organism’s ability to survive changing environments. They are a mixture of qualities that existed prior to emergence, or came into being with the emergence—not post hoc adaptations to environment. They were there from the start, unnoticed by natural selection (i.e. they do not have differential effects on survival and reproduction), and hence they are

148

Chapter 4

ignored by neo-Darwinists who are on the lookout only for the key adaptation that constitutes “the explanation.” Conventional usage for such pre-existent qualities that have novel adaptive powers is the awkward word “preadaptation,” which suggests that adaptation can anticipate future conditions. A better word is “exaptation,” which implies that a primitive structure that served a particular function can convert to a novel function.19 But simple exaptation needs deeper analysis to bring out inherent multifunctionality. For example, the statement that a primitive fish lung is “a preadaptation for buoyancy,” can be replaced with “as its respiratory function was unnecessary it became exapted for buoyancy.” But although it is a semantic improvement it is not a comprehensive argument. By simple physical principles, a lung at its inception is already a buoyancy organ, before the accoutrements of a swim bladder have been added. Not only were lungs bifunctional, they were differentiated duplicates of gill pouches that originally had multiple functions of respiration, feeding, and metallic ion uptake—all derived from an alimentary tract that had many other selectively absorptive powers. The gut is a major conduit of exchange between the internal milieu and the external world, being formed of absorptive endoderm in contrast to the relatively impermeable outer ectoderm. Not only is food digested and absorbed, but salts, water, and respiratory gases can be exchanged. This constellation of functions existed immediately when guts first emerged. In the deuterostome line from which the chordates arose, a filter-feeding apparatus that doubled as a respiratory organ was added at the anterior end of the gut. It persisted in the primitive jawless fish, and is still seen in the ammocete larvae of lampreys. Its iodine-absorbing endostyle groove became the vertebrate thyroid gland. From the tops of the gill pouches was derived the thymus gland, which in fish retains a segmental appearance. The ciliated slits, which had the dual role of food filtration and ventilation, became specialized as gills. As well as exchanging oxygen and carbon dioxide, gills retained salt-absorbing or saltexcreting functions, and complemented the excretion of nitrogenous waste. Then, from posterior embryonic gill pouches, the first simple air sacs emerged as lungs and swim bladders. Our own lungs still contain epithelia derived from the endoderm of the gut. As another example of multifunctionality, mammalian hair glands secrete fluids that could immediately act as coolants, nutrients, and hair conditioners. The emergence of those secretions was correlated with dental differentiation, and the anatomical flexibility needed for grooming. If an organ is multifunctional to begin with, its morphology may diverge in several different specialized directions—hair glands might become mammaries—or it could remain unspecialized, adaptable, and multifunctional. Multifunctional adaptability is particularly important in stressful fringe environments, or at interfaces between different environments.

The Physiological Arena

149

Life at the Edge Conventionally, it could be argued that in unstable fringe environments natural selection would favor adaptability and remove inappropriate specialization. Edge effect (Cook, 1961) is an orthodox proposition that adaptable types originate at peripheral environments as a result of reduction in interpopulation migration and protection from the diluting effects of the larger gene pool. Their survival is reinforced by coadaptational differences presented by unusual biotic and physical factors. But edge effect is based partly on the concept that in evolution nothing can happen unless there is a selection pressure to force it. The less conventional idea of Schmalhausen and Belyaev, namely that in fringe environments stress destabilizes the physiology of organisms, and allows more room for physiological maneuver, would seem to argue that the fringe environment is where evolution happens. This would miss the point that adaptable organisms already exist in the parent population. Adaptability can become more sophisticated autonomously in the population at large, and have novel manifestations that might have incipient physiological advantage in stressful environments, if the more adaptable types can get to them. As far as plants are concerned, random distribution of seeds or spores makes their destiny a matter of luck. And those that are sufficiently adaptable may be lucky enough to fall on stony ground, where there is less competition for space. In contrast, animal distribution through exploratory behavior is self-directed for those that are more generally adaptable. Some of them may also have incipient superiority on stony ground, or other stressful conditions. Schmalhausen had already displayed a holistic view when he wrote the following: Physiologic adaptability to variations in the external environment enables organisms to migrate more freely and to conquer new places in nature to a much greater extent than does morphologic adaptability. This occurs because the former is established much more rapidly, and is characterized by quickly and easily reversible reactions. Hence, physiologic adaptability denotes a considerably more extensive eurybionty than does the capacity for adaptive [adaptational] modifications. Organisms can live and multiply under extremely diverse conditions. In view of the rapidity of physiologic adaptation [adaptable response], especially of behavior, an organism is transformed relatively easily even when biotic conditions vary (within definite limits). Other conditions being equal, this means a diminution in the intensity of elimination, and an increase in the number of individuals of the given species but within the limits of the physiologic adaptability of the species.20

Schmalhausen went on to say that physiological adaptability also accommodated epigenetic change. My own thinking about these matters was catalyzed by J. W. Beament’s 1961 review essay. “The role of physiology in adaptation and competition between animals.” He concluded that general adaptability, especially in insects, would make it possible for them to be far more widely dispersed than is usually found in nature. The egg of the cabbage white butterfly Pieris brassica can hatch and go through several molts as a

150

Chapter 4

caterpillar—under water—given a supply of cabbage. Beament was particularly interested by the structure of the insect cuticle, and concluded in a 1964 report that the orientation of an internal monolayer of wax molecules was sufficient to provide waterproofing. At transition points (usually above 30°C), thermal disruption of van der Waals forces that hold the wax molecules together lead to sudden water loss. The pupae of the cabbage white have additional protection in a cuticular surface wax monolayer that keeps them waterproof to 60°C. Beament also established that in aquatic insects the monolayer was inverted, so that water was kept out, and postulated that the hydrofuge waxiness of the tracheal system and its spiracles could have allowed aerial respiration in the aquatic ancestors of insects. Aquatic insects often carry a bubble of air with them which acts as a “physical gill.” As oxygen in their tracheae diminishes, it is replenished by dissolved oxygen from the environmental water. Because the nitrogen in the bubble is not used, and diffuses into the water very slowly, it keeps the bubble-gill stable, so that the insects do not have to constantly surface to replenish it. Despite the general perception of insects as little automata, many can adjust their behavior. Darwin knew that without any respiratory or locomotory specialization some proctotrupid wasps can submerge themselves underwater for up to four hours, swimming with their wings.21 Thus, the emergence of lepidopterans with aquatic larvae is not unimaginable—the early stages of the insect life cycle are where most of their evolutionary plasticity resides. But what they would need is freedom from competition and predation from insect and other specialists already in the water, and perhaps some kind of aquatic cabbage to get them started. If I were to retitle Beament’s original essay, I would call it “The limitations imposed by adaptation and competition on physiological adaptability.” Schmalhausen combined two fundamental types of emergence. The first was intrinsic increase in complexity that improves general physiological adaptability. The second was the autonomization of extrinsic factors, or what might now be described as the genetic assimilation/accommodation of useful, environmentally induced, phenotypic changes. Subsequent mediation of both of these types of emergence was by the fine-tuning of internal physiological organization, through co-adaptation of the constituent mechanisms. This aspect of stabilizing selection operated on the physiology of organisms, and by omitting that aspect, population biologists render it theoretically jejune. Physiological stabilization was an endogenous process that would occur under any circumstances. But Schmalhausen did not exclude more strictly adaptational innovations, physiological or anatomical. What came out of Schmalhausen’s synthesis of intrinsic and extrinsic emergences was that the resistance of dynamic stability to change could be overcome. And in extreme environments adaptable organisms stood a better chance of evolving further. Adventurous organisms have to be adaptable in advance of exploring new stressful territories. Filters do not exist for them, but competing conspecifics or congenerics are

The Physiological Arena

151

excluded, so that a novel genetic and physiological and behavioral mix could then destabilize the norm. Therefore, while this discourse incorporates some Darwinist and neo-Lamarckist thinking, it leaves out selection pressure, and the inheritance of acquired characteristics. When Lamarck said the organism responded to a need, he could have been anticipating the idea implicit in selection pressure. But the organism does respond to its environment, not just through physiogenesis but through its anatomical, physiological and behavioral flexibility, and that finally has an evolutionary effect. These points are illustrated by the following case histories of animals at the edge. The “furry-footed” Djungarian hamster Phodopus campbelli inhabits desert regions of Asia that have long, cold winters and short, warm summers. Kathy Wynne-Edwards and her colleagues have thoroughly studied reproductive hormonal physiology, reproductive behavior, parental care of the young, and foraging behavior.22 And they draw on existing data for excretory physiology, predation, and general features of the local environment. Phodopus campbelli has an unusual constellation of endocrinological and behavioral qualities that allow it to survive in conditions of low temperature and lack of water that would be stressful to most mammals, including its nearest relatives. While subject to wild predators, Phodopus campbelli avoids competition with the congeneric Siberian hamster Phodopus sungorus. Homeothermic animals need water, and lactating mammals, living in burrow microclimates during a short summer, get hot and desiccated. With an excretory efficiency like that of the more familiar desert rat Dipodomys, Phodopus campbelli can go without free-standing water, but has to forage for long periods of the night to find enough seeds and juicy insect larvae to provide energy and water. Paradoxically, the “cold-adaptations,” including fur and fat insulation, are necessities of survival in winter, but liabilities during the short, warm breeding season, because the lactating female, in close contact with her pups, is likely to overheat and dehydrate. Hormonal modifications in females of the species, especially in post-partum progesterone production, attract the male to be constantly present and available to keep the young warm, while she departs to forage and cool off. The male has a hormonal role to play too, since pregnancies are usually aborted if he is not close by most of the time. He has the additional role of delivering the pups and then cleaning them up. This case is especially interesting as a natural example of the destabilizing selection that D. K. Belyaev identified in domestic Arctic foxes, and which I introduced in chapter 1. Belyaev was well aware of Schmalhausen’s quest for conditions that would disequilibrate the balancing selection that blocked evolution. Tameness had been the main factor in the choice of foxes for breeding stock and along with it came unusual reproductive cycles sensitive to hormonal features distinct from those of wild foxes. Thus, the “artificial” conditions of the farm, which necessarily included freedom from

152

Chapter 4

competition, predation, and climatic extremes, destabilized the kind of dynamic population equilibrium found in the wild. Here, the agents and differential consequences of natural selection were absent. As Belyaev inferred, we should be looking to situations where wild populations, societies, or simply pairs, living in stressful conditions, have a behavioral and hormonal interplay that causes a constellation of changes that reduce stress, and in some instances increase cooperation. Stress is not a selection pressure that has caused the changes. Those were incipient in the ancestral population; the behavioral and physiological interactions of the more exploratory individuals brought them out, and the absence of competition was a more significant factor, allowing the changes to persist. Wynne-Edwards judged the Djungarian hamster to be an appropriate model for Belyaev’s destabilizing selection, when I broached the subject with her. However, she puts Phodopus campbelli into the boxes of proximate and ultimate causation, suggesting how it became differentiated from its closest relative P. sungorus by a variety of antagonistic selection pressures, extreme environment being the ultimate selective force. It is more parsimonious to think about furry-footed hamsters in their own right. They can live at an environmental edge because they have all the emergent qualities of placentals. The ancestors of Phodopus campbelli already possessed the adaptabilities needed to push the envelope: water regulation, insulation, adipose tissues, a modifiable hormonal output, and a range of adaptable behaviors. They could avoid persistent predators by nocturnal foraging, and, as they pressed on, competition for food from their own species and other small mammals was reduced. Some might have already had a sufficient rudiment of male and female togetherness that could be phenotypically reinforced by positive feedback: the more cuddling the more hormones; the more hormones the more cuddling. Not much genetic assimilation of appropriate hormonal variations would be needed to reinforce what already existed. Others of the ancestral type would have failed to reproduce under the harsh conditions. It is more to the point to examine the qualities that ensured the success of P. campbelli than what natural selection did to others that lacked them. The Siberian hamster, P. sungorus, stayed behind in more benign conditions, but some individuals show incipient traces of the hormonal requirements for success in colder, drier conditions. Furry-foot demonstrates the generalization made earlier about naked mole rats, which have also found a way to succeed under environmental stress. We must appreciate the full spectrum of generative conditions and the full constellation of emergent features. Though it is tempting to find “a key innovation that was selected,” there is in these hamsters’ lives a web of causality, involving behavior, environmental feedback, homeothermy, hormonal changes, and hamster houghmagandy. Another commonality between hamsters and other animals living at extremes is family togetherness through hormonal mediation. The emperor penguin, Aptenodytes forsteri, raises its young close to the South Pole. Females lay their eggs and then head

The Physiological Arena

153

north for the open sea. Their mates incubate the eggs during the dark freezing winter, and cooperative huddling keeps them up to 50 percent warmer than they would be on their own.23 The penguin females, fat from feeding in the faraway ocean, finally take the long trudge necessary to relieve their babysitters in the nick of time. In an environment at the opposite climatic extreme, we find eusocial naked mole rats. (That is, they live in extended familial association of parents and siblings.) Their interactions are modulated by hormones and behavior, although, unlike hamsters, “bonding” and mutual comfort seem hardly appropriate descriptions for mole rat cooperativity. The queen actively suppresses sexual reproduction in other mature female colony members. Detecting those that are in estrus, she aggressively jostles them until they are sufficiently stressed to go into anestrus. However, when she is too heavily pregnant to be a bully, some of the non-breeding females come back into estrus, so that they are ready to mate in case of accident to the queen. Mole rats in general are confined to burrow systems that have a stable temperature, and their homeothermic abilities have regressed. Being under heavy predation, they are shortlived, and they depend on their social arrangements for survival.24 It is hard to see what more animals such as Djungarian hamsters and naked mole rats could do, apart from minor adjustments. They’ve gone about as far as they can go. Schmalhausen had a word for that: “telemorphosis.” It is now appropriate to distinguish between two kinds of stressful, edge environments. There are boundaries that provide some respite from competition for physiological opportunists, but which physically resist further encroachment. These include polar surface conditions, and the diminished oxygen levels of the upper atmosphere. Bacteria, of course, prove this rule, since they can survive in outer space for a while, as well as living at temperature and pressure extremes. But their robustness is different from the vulnerable adaptabilities of insects, birds and mammals. On the other hand some interfaces were temporary barriers: between anoxic and oxygenated; between salt and freshwater; between water and land. Their penetration required a prior quality of adaptability, and diversifying evolution was subsequently the result. The gannet is an example of an adaptable animal with sophisticated homeostasis that allows it to stay the same when the environment changes, and also to change its environment spontaneously, as when it dives after fish. A more striking example of a bird that exposes itself to stressfully extreme environments is the Arctic cormorant. One would think that temperature homeostasis would be utmost importance to it. Yet subdermal fat insulation and the secretion of preening oils have regressed in that cormorant species. Not only do they get colder faster, but also their plumage is more wettable, and they have to hang out their wings to dry in freezing temperatures after a dive. On the plus side, reduced insulation means reduced buoyancy, and cormorants can dive after prey to greater depths with less effort. Owing to the anomalous properties of water, the deep temperature of Arctic seas is not much colder than that

154

Chapter 4

in more temperate zones. The potential trouble arises from having to expose a wet body surface to dry off in subzero air, where the sacrifice of body heat is unavoidable. Yet diving for about nine minutes a day satisfies the cormorants’ usual nutritional and thermoregulatory needs. In the absence of competition from other birds, they are able to tap a major resource of high-energy, fatty food. Their adaptability sustains them even in the absence of apparently crucial “adaptations” for living in the Arctic.25 Complexity theorists use metaphors like “life at the edge of chaos” to refer to biological molecules that persist through being not too stable and not too unstable. Metaphorically, organisms too can exist at the edge of chaos, between stable ecosystems and adjacent disruptive environments. Those that can meet the unpredictable challenges of physicochemical extremes elude the competition that operates in the dynamically stable conditions of the ancestral environment, and so can escape from the hypostasis of natural selection. Fish in estuaries, for example, could historically acquire new food resources, occupy new living space, and escape from predators, and competition, if they were willing and able to make the final leap. Fringe environments that are only marginally different from the parental ones are still more likely to be habitable by the most adaptable types. Here, they can congregate, and test the unexploited adjacent environment. Once they have passed into a new environment, they can diversify without hindrance from the agents of natural selection. As the pristine environment fills up, the radiating lineages become more and more specialized, and the sap of originality dries up. At the present time, interfacial environments are occupied by specialists that should not be taken as models for adaptable transitional animals on the verge of invading the adjacent, pristine land. But the specialists do give us some clues to the adaptabilities of those that went further with their explorations. Once the original adventurers passed through the interface and into a new environment, different physicochemical conditions were imposed upon them, and these potentiated further advances in adaptability. Physiogenesis One of the significant generative features of adaptability is “physiogenesis,” physicochemical change imposed on an organism by the environment. Once animals acquired body cavities and blood vascular systems, their body fluid buffered the weak cellular homeostasis from the insults of the outside environment. But initially the internal milieu conformed to the physicochemical nature of the marine environment. As long as the sea remained constant, so did the body fluid. In contrast, animals near the seashore, or near influxes of freshwater, were affected by environmental fluctuations and this is one of the preconditions for the further evolution of adaptability. This is the definition of physiogenesis first given by neo-Lamarckist evolutionist Edward Cope. Ivan Schmalhausen complicated the issue by adding the biological ability to

The Physiological Arena

155

deal with such change, but the original is more straightforward. Physiogenesis is not a metaphor like selection pressure, but a real physicochemical cause that acts immediately and continuously, eliciting tolerance, accommodation or the response of homeostatic mechanisms. The next step is for an internal dynamic stability to be established that minimizes the energy drain, while preserving responsiveness. Here internal selection or co-adaptation is involved, as Ivan Schmalhausen and L. L. Whyte proposed. But Schmalhausen realized the internal stability, both in physiology and development could be a barrier to further evolutionary change. His senior colleague Alexei Severtsov coined a catch-all term—“aromorphosis”—for the whole spectrum of changes that started with the destabilizing influence of physiogenesis and ended with improved adaptability. He believed that aromorphosis came in emergent waves, alternating with slack periods of adaptational evolution, a kind of physiological punctuation also favored by the saltationist K. Beurlen.26 But where is emergence located in the syndrome of aromorphosis? The organism might have literally emerged, say from water onto land, to be immediately affected by dehydration, gravity, new mechanical stresses, high oxygen tensions, and fluctuating temperatures. But the event was made possible by physiological adaptability that had earlier emerged in the transitional organism, probably before it got to the edge. Maybe components of the unexploited adjacent environment intruded sufficiently to influence the process. Adaptability supported the behavioral choice that led to the environmental change. Then physiogenesis immediately caused internal changes for which adaptability could be adjusted and then advanced. Existing physiological functions, locomotory anatomy, behavior, and inducements of food or escape from predators are the generative conditions for the shift. Physiogenesis, and adjustments to it, then generate conditions for emergence to a higher level of adaptability, such as a critical-point allometric growth of the central nervous system. These all happen to responsive organisms—not to temporary bags of genes. Genetic Assimilation of Physiological Change “Genetic assimilation” is Waddington’s term for what Schmalhausen called “autonomization”: a consequence of change imposed by the environment.27 I will extend the discussion of this topic in a historical context in chapter 5. Although genetic assimilation implies a genetic fixation of an ontogenic characteristic, it is not as Lamarckian as it seems. It combines the direct effect of environment with the syndrome of natural selection: but only if the organism had the original capacity to expose itself to new conditions and be affected by them. Fundamentally it proposes that any phenotypic change imposed by the environment that aids the organism’s survival can be further enhanced by the selection of any gene-based mechanism that reinforces the change— assuming that such a mechanism already exists. Molecular experiments can produce novelties that are appropriate to new conditions. The evolutionary process begins,

156

Chapter 4

either when the environment changes spontaneously, or when the behavior of the organism subjects it to new conditions. The more the adaptable its physiology the more likely is its survival. When an animal is physiologically adaptable enough to experiment behaviorally, physiogenesis may be the next effect: environmentally imposed change or constancy. Then there are internal modifications and adaptations appropriate to the new condition: for example expanded enzyme suites with optimal activity at a given temperature, or in a particular electrolyte medium. At that point homeostatic regulatory novelties that keep the condition constant in the face of change have high fitness. Behavioral exploration, supported by improved physiological adaptability, may effect functional-morphological change. The process therefore combines physiogenesis, behavior, the emergence of adaptabilities, adaptations, and possibly orthogenesis. The provision of genomic change that constitutes genetic accommodation of phenotypic change is not simply a neo-Darwinist mechanism. In chapter 6, I will argue that many such changes are non-random, and some are self-amplifying, so that differential survival and reproduction is in those cases irrelevant. No matter how the mechanisms might be labeled, all of the major physicochemical changes in the internal milieu, though initially imposed by the environment, have been genetically assimilated, so that they persist regardless of further environmental influence. Since genetic assimilation is especially relevant to development it is discussed in more detail in the next chapter. Next I want to consider the buildup of biochemical complexity in primitive organisms. Biochemical Evolution Biochemistry and Molecular Biology Biochemistry used to occupy a comfortable niche between chemistry and biology. Having discovered the general properties of enzymes, and unraveled the metabolic pathways and the physiological actions that they governed, it broadened into comparative biochemistry and assumed an important role in the investigation of evolutionary history through comparing the primary sequences of proteins. Molecular biology then leapfrogged traditional biochemistry to emphasize the informational aspects of macromolecules. It has broad, practical applications in medicine and agriculture, as well as major theoretical implications for evolutionary biology, and a hybrid vigor derived from the cross-fertilization of physics, virology, bacteriology, and genetics. For better or for worse it has also come to dominate traditional organismal biology. Yet, as an erstwhile evolutionary physiologist and a comparative enzymologist, I find traditional biochemistry as pertinent to the larger issues of evolutionary history

The Physiological Arena

157

as molecular biology. The genes, when permitted, dictate the amino acid sequences of the structural proteins, enzymes, and hormones. These products contribute to form, and do the work of the organism, and they cooperate in an organizational hierarchy that transcends molecular properties. Molecular biologists are beginning to realize this to the extent that they declare that now they have tidied up the human genome project they will apply all their resources to the study of proteins—from “genomics” to “proteomics”—a step in the right direction, but something that biochemists were already doing a quarter of a century ago, before they were so rudely interrupted. Physiologists have always had a pragmatic attitude to reduction as a methodology. Analysis at the molecular level simplifies matters and allows the identification of significant functional and evolutionary features that also relate to the whole organism, which is the true value of reduction in the first place. However, molecular functions should not be allowed to become “intolerant abstractions,” as Woodger called them.28 A gene is an intolerant abstraction if its molecular structure is worked out but its interactions and differential expression in the organism are ignored. Reduction with an “ism” attached is often intolerant methodology. Some biochemists and molecular biologists declare “I’m a reductionist, and proud of it,” and do not seem to mind that their graduate students, when asked what their molecules come from, answer “the Sigma Catalogue.” That said, the genome is a dynamic structure with upward causal effects on the organism as well as being subject to change from downwardly acting causes. Molecular biology has a great deal to contribute to the principles that I have been establishing in this chapter. For instance, although I excluded population variability from the definition of adaptability, there are plenty of examples of adaptability at the genome level. And for the sake of clarity, their examination at this level is necessary, not forgetting that the behavior of genes, transposable elements, histone-binding, and methylation are affected by the environment of the cell, the organism at large, and the external environment. Biochemical Origins and Complexification Life proverbially began in the primeval ooze, and trying to understand how it emerged is like being thrown in at the deep end of the swamp. It would be simpler to get on with a discussion of biochemical evolution at the point where recognizable prokaryotes were in place. But the origin of life is the first great biological emergence, in other words the all-or-nothing appearance of a living entity from the inanimate— without the participation of natural selection as formally defined. If the origin of life cannot be left out, how can enough momentum be built in the mental mud to advance our understanding of it? We can try to study model systems logically, but the rest is largely speculation that does not get us very far. And when you listen to what some of the authorities on the subject are saying, they are obviously swimming in

158

Chapter 4

ooze as well, with little more in common than mutual disdain. So there is not much to lose in trying for some coherence and consistency by our own intuitive efforts.29 Organic macromolecules emerged from abiotic substrates before anything like an organized cell existed. What those substrates were, and where on Earth they occurred are two of the controversial questions, and answers have ranged from shallow water, benthic marine clays, to deep rocky fissures around hydrothermal vents. Some of the most ancient prokaryotes still live under pressures and at high temperatures that used to be thought incompatible with survival. Other solutions, focusing on extraterrestrial origins, have suggested the slush of comets that passed near the sun, and some suggestive data have been acquired from experiments that attempt to recreate deep space conditions.30 Since it is conceivable that life originated several times, both on Earth and on Mars, before its life-support conditions deteriorated, it is not out of the question that each planet was seeded by spores from the other, contained in material blasted off by bolide impacts, and that Earthly (or Tellurian) life could have been finally extinguished on Mars, while Martian life lived on to become us on Earth.31 But once biological macromolecules made their appearance in cell-like complexes, most of their remaining evolution, and the extension of primitive autocatalyzing metabolisms, has been organismal. How the biochemical complexity of primitive eukaryotes emerged is partly answered by the endosymbiosis theory discussed in the previous chapter. But how did the molecular complexity of the earliest proto-prokaryotes evolve? Herman Muller’s “naked gene” hypothesis, based on the model of viruses that contain only protein and nucleic acids, was one of the earliest reductionist explanations. But all viruses require a living host cell for their reproduction; as degenerate cell products they are worthless as models for primitive life forms. Nevertheless, variants of the naked gene idea has persisted among molecular biologists. Nucleic acid polymers, randomly and abiotically synthesized in the primeval dilute soup, could have adsorbed amino acids, in a coherent order, according to their degree of attraction to the nucleic acid bases. But unless the abiotic polynucleotides also had an enzymatic ligase function—an ability to link amino acids—this would be unlikely to have resulted in the spontaneous polymerization of proteins. Nevertheless randomly generated free nucleotides and amino acids could have been brought into close physical association by clay adsorption, or in heterogeneous coacervates, and being in the same place at the same time is a necessary condition for the generation of more intimate interactions and symbioses out of loose associations. Von Neumann’s 1966 theory of self-reproducing automata, well known among science-fiction enthusiasts, has slowly been invading biology. It proposes that the minimum requirements are a blueprint, a constructor, a replicator, a switch that activates the copying process, and a box that keeps the bits close enough together to interact, but as an open system. Living processes require the metabolic association of molecules that can access a source of raw materials, and can self-replicate faithfully. But what came first?

The Physiological Arena

159

The primary emergent feature of DNA is its ability to be copied—provided the full resources of a cellular unit are present. Self-copying is bifunctional, in the sense that an entity that reproduces also survives in its offspring, even in destabilizing conditions. Highly stable polymers, such as plastics, can survive much longer than vulnerable nucleic acids or proteins. But to achieve a reasonably faithful copy a living system has to be unstable enough to come apart and go back together again, making it vulnerable to physicochemical influences, error-prone, and thus raw material for natural experiments. Early molecular biologists used the expression “the central dogma” for the assertion that information for synthesis can only flow from DNA to protein, despite the already well-known fact that some viruses consist of only RNA and protein. RNA viruses can synthesize DNA using the enzyme “reverse transcriptase.” Not surprisingly, RNA, of a molecular type similar to messenger RNA, a long-stranded molecule with a genetic code in its primary structure, is now the most popular candidate as the primordial gene.32 But to copy itself, or to make both DNA and protein, RNA would require ancillary support. A more appealing model is an RNA that has a catalytic, or enzymelike function, like that of “ribozyme.” Genetic engineering of this enzymatic RNA has produced forms that polymerize RNA, and some that break peptide bonds in proteins. Other laboratory experiments have randomly produced RNAs that can string together new strands, using high-energy phosphates in the process. And natural experimentation might have brought about the same results at the dawn of life. Familiar ribosomal RNA that participates in the protein synthesis found in presentday living cells has the ability to link amino acids through peptide bonds. The archetypal RNA could have done that, with an additional function of self-replication. The proteins would initially have been randomly structured. These possibilities point to what Leslie Orgel (1994) calls “the RNA world,” from which emerged the “last common ancestor” of all subsequent living organisms. Another option would be a protein world where pre-existing catalytic proteins acted as templates that attracted and polymerized nucleotides or simple bases. Thus, a molecule that already had the spontaneous function of aligning nucleotides could have also been able, indirectly, to replicate itself. Such simple systems could have kept up with entropic wear and tear of their organized complexity, and some of the randomly structured proteins could already have had functions that were meaningful for protobiont integrity. Stuart Kauffman (1995) has taken a different tack, focusing on “catalytic closure” as a property of a chemical system responsible for the emergence of life: Each cell in your body, every free-living cell, is collectively autocatalytic. No DNA molecules replicate nude in free-living organisms. DNA replicates only as part of a complex of collectively autocatalytic network of reactions and enzymes in cells. The cell is a whole, mysterious in its origins perhaps, but not mystical. Except for “food molecules” every molecular species of which a cell is constructed is created by catalysis of reactions, and the catalysts themselves are created

160

Chapter 4

by catalysts created by the cell. To understand the emergence of life, I claim, we must understand the conditions that enabled the first emergence of such autocatalytic molecular systems.33

Kauffman argues from established models that as autocatalyzing systems become more complex there emerges a self-sustaining system with a novel complexity—analogous to a shift from the liquid state of water to solid ice. “A living metabolism crystallizes. Life emerges as a phase transition.”34 Julius Rebeck and his colleagues have worked on non-DNA/RNA self replication of synthetic molecules based on adenine ribose linked with naphthalene.35 He suggests a model for a self-replicating coat of protein. But these, though they show that polynucleic acids are not essential for reproduction, are more rigid than living systems and have no means of capturing or generating energy. Chemist Denis Schwartz has communicated to me his concept of how catalytic closure itself might have happened as an emergent phase shift. He has modeled a pre-RNA autocatalytic, self-energizing chemical system based upon the kinds of molecules found in redox reaction in mitochondria, and as parts of the nitrogenous bases in nucleic acids (adenine again). It bends no physical laws and can probably be tested in the laboratory to determine the limiting conditions of its generation. Moreover, it is a versatile or metabolically multifunctional system that is self-organizing in space and time—much more so than the RNA world. Schwartz believes that the catalytic potential of ribozymes is too small for them to be candidates as ur-molecules. But his system could generate RNA and its own lipid membranes. As to the time frame for the chemical emergence of such a system once the generative molecules had been accumulated, I asked “Seconds or eons?” “Minutes, maybe years,” he replied.36 Thus, both ribozymes and protein enzymes could have been consequences of a Schwartzian abiotic generative complex contained in “protobionts.” Think of the latter as primitive units that have a membrane-bound, heterogeneous chemical structure, but cannot reproduce themselves except by random fission. They were open system/dissipative structures that concentrated the more generative autocatalyzing chemicals, allowed crude reproduction as a first step toward the emergence of a living cell. Randomly generated proteins and ribozymes could have been capable of weakly catalyzing multiple reactions, instead of having the single specific reactive properties of familiar enzymes. This would have increased the probability of spontaneous emergence of a self-sustaining system that could continue to feed on other chemicals in solution in the environment, more life-like in the sense of faithful reproduction. Like Kauffman, Schwartz argues that the emergence of life was inevitable, given the presence of a simple chemical precursor system. And it was fast. The biologist A. I. Oparin used colloidal clumps of coacervates as models of protobionts in his pioneering experiments, reviewed in English in 1938. Coacervates grow by absorbing molecules from solution; they crudely reproduce by disintegration and further absorption, and they can be designed to absorb energy and release it for biochemical reactions, an important step in sustaining a reliable synthetic ability.

The Physiological Arena

161

Proteins that were finally polymerized in a protobiont would have to have shared compatible physical characteristics, so that they could co-exist in an aqueous microenvironment of a particular hydrogen ion concentration, electrical potential, and electrolyte composition without disintegrating or cross-bonding irreversibly.37 Polymerization of proteins and nucleic acids had to go on fast enough to keep up with their rate of degradation. But the necessary high-energy molecules were already present in the constellation of protobiont features. And adenosine triphosphate remains the most common energy primer in living systems. The formation of a simple cell membrane around coacervates is not a major theoretical hurdle, since bipolar molecules such as phospholipids, if mixed with water, tend to orientate themselves as monolayers at the surface If they are stirred, they sort themselves spontaneously into double layers, and can associate with electrically charged molecules like proteins. If the proteins protrude through the membrane they can continue to act as absorbers and concentrators of molecules in solution in the outer environment—forerunners of diffusion channels, ion pumps, cellular adhesion molecules, and cell surface recognition molecules. In the early days of protobiont formation, different types would have coalesced randomly, as well as reproducing by fragmentation, permitting mix-and-match experiments in mainly additive complexification. These speculations do not preclude the possibility that the initial concentration and primitive organization of the molecules that made up the protobionts took place on a heterogeneous, adsorptive surface such as a clay, as suggested by Graham Cairns-Smith in The Life Puzzle (1971). It might have been harder in the aquatic pressure cooker of a thermal vent, since heat-resistant biopolymers would be less flexible, and phospholipid membranes are unstable at high temperatures. Among their emergent properties, protobiotic units would inevitably have had electrical potentials across their lipid membranes because of Donnan equilibria: disproportionate ionic concentrations resulting from the presence of large, non-diffusible ions within the membrane system. Such electropotentials created the opportunity for membrane excitability, even before the emergence of the ion channels, and gates and pumps that we now associate with nervous activity. A necessary complement to the emergence of a living, complex, autocatalyzing system is stable coordination of its components to reduce the chaos of infinite interactive possibilities to a few predictable pathways. This requires a degree of homeostasis that resists change, though in a living system there must always be sufficient inherent plasticity to allow ultimately for reproduction and the evolution of permanent change. Kauffman and his colleagues have shown that model physical systems spontaneously assume such properties—in the absence of any selfreproducing qualities, and in the absence of any influence that might be equated with natural selection.38 In his personal communication to me, Schwartz goes further, to

162

Chapter 4

describe his chemical system as one that not only emerges by autopoesis, but becomes more complex by orthogenesis. As a chemist he finds nothing heretical about selfamplifying systems. Integrity, survival, order and complexification are emergent qualities of complex physicochemical systems, even in the absence of reproduction by means of a DNA-like mechanism. But they had the potential to generate such a mechanism. Kauffman and Schwartz demonstrate that the emergence of ordered systems can be rapid, and geological evidence shows that in the history of our planet life emerged just about as soon as it was cool enough for it to do so. In all probability there were multiple initial experiments in emergent life forms. Some were insufficiently robust to survive environmental contingencies, and some may have pooled their resources symbiotically. By that time, some protobionts would have been at the “chemoton” stage with a rudimentary metabolism and self-replicating coded polymer, or the “progenote” stage with a “rudimentary, imprecise linkage between its genotype and phenotype.”39 The progenote, in which the mechanisms of protein synthesis were still evolving, has been proposed as the “cenancester,” or last common ancestor, of all surviving organisms from Archaebacteria to humans.40 Determination of the exact nature of that cenancester from surviving prokaryotes is difficult because of subsequent multiple exchanges of genes between the lineages. However, when they achieved accurate replication, available environments would have filled up, and then competition and predation would have reduced the array of experiments, influenced the refinement of internal coordination, and the establishment of dynamically stable ecosystems. As unifunctional enzymes, whose products positively affected the integrity of the cell, emerged, along with specifically coded templates for their synthesis, they were subject to adaptation to the local environment of the cell. Michael Conrad (1990) suggests that the structure of protein molecules already provided a variety of buffering mechanisms involving redundancies in the numbers of weak bonds and amino acids, amounting to molecular homeostasis that resisted change in the cellular environment. He also points out that other protein features could result in the amplification of small changes, making them more evolvable. The simplest unicells have signaling systems that pick up external stimuli, and communicate them with the interior; they may only have two or three components. But intercommunication with parallel pathways is sufficient to generate new complexity. These generative conditions for signaling are greatly complexified in nervous systems. For example, there are four major signaling pathways communicating with the post-synaptic membrane of mammalian neurons. Each pathway has multiple components that intercommunicate with parallel pathways, making the total number of possible combinations astronomical. This illustrates that while the generation of complexity in terms of number of interactions is easy to appreciate, the

The Physiological Arena

163

empirical problem is in discovering how the systems are constrained and limited to an efficiency that is not hindered by meaningless interactions. It also illustrates that the contingent mutual impact of independently evolved biochemical pathways has played a role in minor saltatory emergences within simple cells. In their essay “Complexity in biological signaling systems” (1999) Weng, Bhalla, and Lyengar develop this view of biochemical complexity with a computer simulation of a simplified four- pathway system: Such a network exhibits interesting emergent properties, including integration of signals across different time scales, generation of distinct outputs depending on the amplitude and duration of the input signals, and the presence of feedback loops that behave as bistable switches to process information flow through the network. Although this first glimpse of emergent complexity appears to be intriguing, rapidly accumulating experimental evidence suggests that several other considerations need to be taken into account in order to develop a minimally accurate picture of a living cell. Prime considerations among these are compartmentalization and regional organization of signaling components.41

These authors go on to discuss hydrophobic lipid and hydrophilic aqueous compartments and the difficulty of quantifying the micro-quantities of ions therein. They also note the importance of the cytoskeleton for anchoring membrane receptors in position and providing scaffolding for “an assembly line along which a series of enzymes process their substrates in a well-defined sequence and with an efficiency and specificity that are orders of magnitude higher than would be possible in freely diffusing systems.”42 MAP-kinase (= mitogen-activated-phosphorylase-kinase) pathways exemplify this scaffolded order. The emergent property is called “reaction channeling.” Even in vitro, the efficiency of the reactions and specificity of flow is retained. Where biochemical pathways are isolated within subcellular compartments, such as the mitochondria and nucleus, intercommunication is possible through the diffusion of translocation molecules. Subsequent biochemical evolution is a mixture of natural experiment and conservation. (This will be elaborated in chapter 6.) When DNA is brought into the picture, the complications for analysis and understanding the complexities of molecular intercommunication are again astronomical. Compared to these problems the identification of the genetic code and even the completion of the human genome project seem almost simple. The code for structural proteins and enzymes is almost identical throughout the genophores or chromosomes of all living organisms. Any mutations that cause a frame shift in reading the code of an exon, by the deletion or insertion of a base, for example, can throw off protein synthesis so badly as to be lethal. On the other hand a non-synonymous point mutation in a single base, i.e., one that changes the code for one amino acid in the primary structure of the protein, produces an alteration of protein function that may be advantageous, neutral or detrimental. By conventional wisdom all of the potentially beneficial variations have been tried and retained, and

164

Chapter 4

any further mutations will only repeat old mistakes, assuming that internal and external environments remain constant—big assumption. However, Lynn Caporale (2003) marshals a host of instances of non-random hypermutability in specific regions of the genome that constitute gene adaptability. Though most common in prokaryotes, these phenomena also occur in eukaryotes. The construction of antibodies is the example that will be most familiar to her readers. Another conservative feature of biochemical evolution is that similar spectra of genes and proteins, structural, regulatory, and enzymatic, are found throughout the living world. This re-emphasizes the impossibility of explaining evolutionary divergence or progress in terms of the cumulative possession of unique proteins or biochemical reactions. The same old genes and proteins are mix-matched time and again to come up with modified variations and combinations. Repetitive Differentiation Gene amplification is a versatile process that widely affects physiological evolution. Alteration of existing DNA removes its old function from the organism’s repertoire. If that old function was essential the mutation is detrimental, even if the modified protein has an immediate utility. One way to get around the problem is to duplicate genes, and leave some to carry out business as usual, while the duplicates provide the raw material for natural experimentation. In this way biochemical systems can be intensified, complexified, and made more adaptable. This is where “repetitive differentiation” comes in. E. D. Cope’s antique term “repetitive addition” referred to anatomical complexification that resulted from segmentation in worms and arthropods. He and William Bateson realized independently that the functional-anatomical repertoire of an animal could be enhanced by repeating, or segmenting, some of the old parts. These could continue doing what they always did, assuming that those functions remained necessary, and the others could experiment with new functional morphologies. Appendages used for swimming by the old segment could be modified for walking in the new one, repeated nerves and muscles still working in the same basic way initially. The potential for a particular segment or group of segments to produce some kind of variant of a limb resides in the expression of a cluster of developmental Hox genes that have also arisen by repetitive differentiation. But other organismal mechanisms participate as well as the genes. In the middle of the twentieth century, several farsighted authors realized that the duplication of the total number of chromosomes, as well as duplication of portions of chromosomes had a potential for evolutionary experimentation. (See also chapters 6 and 8.) The old anatomical name, “repetitive addition,” is tautological, and it does not imply differentiation as well as multiplication. The concept is filtering into comparative biochemistry and molecular biology. And in epigenetics it has been included in the syndrome of modularity. But much has

The Physiological Arena

165

yet to be made of it in an evolutionary context, so I dare to modify the name to “repetitive differentiation.”43 The concept is very useful at the level of molecular changes, where it refers to a gene or a specific DNA sequence that has repetitively duplicated, and whose duplicates have then mutated into differentiated variants. These are sometimes called “variant repeats” or “varied repeats” and the result of simple duplication without differentiation is “dose amplification” or “dose duplication.” It has long been known that many proteins exist in a variety of molecular forms with slightly different electrical properties, allowing them to be separated by electrophoresis; in the case of catalytic proteins they are called “isoenzymes.” They have functional differences in addition to their electrophoretic ones, and can emerge through simple mutation as well as repetitive differentiation. There may be a number of mutant allelic variants present in a population, any two of which a given diploid individual will possess. However repetitive differentiation provides an individual with an extensive suite of variant repeats. One of the advantages of such a suite of isoenzymes was made clear in the now classical example of salmon muscle kinases, enzymes that release energy for swimming.44 The isoenzymes have different temperature optima, and switch on and off according to the ambient temperature, so that the fish can keep on swimming efficiently in a range of temperatures instead of slowing down when the temperature drops, as it would if there were only a single kinase with a fixed and narrow temperature optimum. A significant portion of amphibian genomes is taken up with variant repeats that ensure normal development under different temperature regimes. Repetitive DNA, “amplified DNA” or “duplicate genes” as the bits are sometimes called, are very common, many of them, such as satellite DNA near the centromeres and telomeres, apparently doing nothing, or at least not contributing to protein synthesis. (But see chapter 6.) Many enzymes come in a number of isoenzymic forms, some simply allelic variants of non-duplicated genes, some variant repeats. Susumo Ohno (1970) observed: “Only the cistron which became redundant was able to escape from the relentless pressure of natural selection, and by escaping, it accumulated formerly forbidden mutations to emerge as a new gene locus.”45 R. J. Britten and E. H. Davidson (1971) supported this opinion, pointing out that repetitive DNA gives the organism the opportunity to experiment with point mutations of the original gene, “out of sight of natural selection.”46 This gets over the problem that if a gene that is doing something useful for the organism mutates, it is no longer available for the original function, and its substitute, while it may be more useful under certain circumstances, might be useless if the conditions of existence revert. So adaptability can be achieved through repetitive differentiation. Suites of mutant genes and transposable elements might be held on standby—potentially useful if circumstances change—while the original keeps on doing its necessary job. These natural experiments need not interfere with the internal

166

Chapter 4

coordinative conditions of the cell or organism unless they are constitutive, i.e., always turned on, and wasteful of energy. Therefore internal selection favors the refinement of feedback control. In tissues where they are not required they are inhibited by histone binding and methylation. Plasticity in natural molecular experimentation can also be achieved by both duplication and differentiation and rearrangements of subgene portions of DNA, the exons that carry the code for the meaningful bits, and the “nonsense” introns providing for the rare emergence of functionally novel genes. This is dealt with more extensively below in the section “New Genes for Old.” It should be mentioned here that changes in methylation, acetylation, and histone binding patterns that repress gene expression can be environmentally induced, and the impact transmitted from generation to generation, as Eva Jablonka and Marion Lamb pointed out in Epigenetic Inheritance and Evolution: The Lamarckian Dimension (1994). This has now become acceptable to some mainstream molecular biologists, who are now speculating about the non-DNA inheritance of acquired characteristics.47 The concept of repetitive differentiation simplifies the problem of how metabolic pathways evolved. In the ancient, anoxic, aquatic environment of Earth many organic molecules were synthesized abiogenically by the energies of solar radiation, volcanic emissions, electrical discharges, and impacts of bolides—the general term for meteors, comets, asteroids and other space debris that impact Earth. The first protocells, with a simple biochemistry, and unable to photosynthesize, depended on abiogenic food molecules in their environment for energy, growth, and reproduction. What happened when the abiotic energies diminished and the oxygenation of the environment began to destroy free organic molecules faster than they were made? One way to make up the loss was to prey upon other organisms that already had essential compounds. Another was to alter precursors that were still available. The enzyme that makes a particular product might be duplicated and differentiated to become one that can convert the first product to something else. Substrate structure is already related to product structure, thus the active sites of enzymes in a metabolic pathway tend to be similar as well. Therefore it is conceivable that some of the enzymes in a given pathway arose from a single original gene by repetitive differentiation. But there remains the problem of how the evolutionary process could respond quickly enough—when the abiotic source of essential molecules was depleted, the final phase would have been sudden, and drastic. It only takes a month or so for an oceanic phytoplankton bloom to crash because it has used up key mineral nutrients, usually phosphate. The random evolution of a series of enzymes, one at a time, by repetitive differentiation, could not have responded in such a timeframe—duplications and mutations are instantaneous, but are not made to order. Instead, some biochemically adventurous cells had to have had a pre-existent suite of enzymes, immediately able to make good the deficiencies.

The Physiological Arena

167

B. E. Wright (2000) proposes that biosynthetic pathways evolve as a result of “depletion and interconversion of related metabolites.” Starvation may result in the absence of a key transcriptional inhibitor metabolite. Subsequently, loss of feedback control is followed by transcriptional derepression of specific enzyme-encoding genes. Given that increased transcription is linked to higher rates of non-transcribed strand mutation, this chain of events would give rise to an increased number of variants that might initiate novel catabolic pathways, enabling the derepressed enzyme to utilize a new substrate. Experimental evidence demonstrates that this type of biochemical evolution is not uncommon, and Wright believes that mutation is not exclusively random, and that the environment plays a significant role in evolution at the molecular level. Simple but biochemically versatile prokaryotes could also acquire additional DNA through transformation, conjugation, and transduction, and such cells could also have provided the foundation of a food pyramid. Before eukaryotes emerged, the elaboration of membrane proteins was well under way. These variously responded to extracellular influences by recognizing external stimuli and absorbing specific nutrients. With the advent of multicellularity their influence was extended for cellular communication. Some of the most complicated examples of repetitive differentiation are to be found in the neural and hormonal systems of multicellular animals. Ion channel proteins for potassium, sodium, and calcium involved in selective activation of impulses in nerve cells evolved by repetitive differentiation. Hormones that affect water transfer across cell membranes mostly belong to the same peptide family. For example, vasopressin, the antidiuretic hormone of mammals that causes urine retention, increases the water permeability of part of the kidney tubules. It is closely related to the molecule that makes frogs absorb salt from the environment, and one in freshwater fish that causes increased urination rather than water retention. Prolactin stimulates lactation in mammals, causes broodiness and crop milk production in birds, and acts as a growth hormone in reptiles. It may have been involved in alterations in gill membrane permeability in primitive fish, and it reduces salt secretion from the gills of modern bony fish. Prolactin is also found in invertebrates. Most of its functions have to do with the transfer of solutes and water through membranes by altering porosity; so major reconstruction is not needed to switch high permeability over to low permeability. Prolactin is itself a derivative, through repetitive differentiation, of the more ancient molecular family of growth hormones. Their genes too have undergone rapid episodes of duplication and variation throughout vertebrate evolutionary history. Among the mammals, alteration of growth hormones coincided with “sustained bursts of rapid evolution” in the artiodactyl and primate lines.48 In humans, three variants are involved in placental function, in addition to the ancestral growth hormone gene.49

168

Chapter 4

Hormone-like steroids and peptides are ancient molecules found in prokaryotes. Mammalian subjects have been intensively investigated because of medical priorities in physiology. The gaps in our knowledge of their functional evolution from bacteria to vertebrates simply reflects a relative lack of interest in other animal groups. However, there is some consensus that a few major events catalyzed vertebrate evolution. The duplication and reduplication of the entire vertebrate genome in the late Cambrian quadrupled the potential for adaptability of the entire lineage through natural experiments in differentiation of the repeated genes. Not only were novel variations of the existent possible, but new major systems, such as the immune system emerged. Furthermore, early in the craniate line, the neural crest emerged as a source of cells that were to be fundamental tools of vertebrate evolution. This improvement in functional anatomical evolvability is a story for a later chapter on developmental evolution. However, it is a matter of historical record that embryonic innovations together with physiological adaptability permitted our aquatic ancestors to survive the major catastrophes that terminated the Ordovician and Devonian, and, once on land, to soldier on through the Permian, Jurassic, Triassic, and Cretaceous extinctions. There was a lot more to it than being in the right place at the right time. Comparative studies have begun to establish hormone lineages both in the protostome and deuterostome lines of animal evolution. Many of them are mediated by ubiquitous second messengers, such as calcium ions, and cyclic adenosine monophosphate (cAMP). The latter is a close relative of adenosine triphosphate (ATP), which universally energizes biochemical reactions. It is also related to the nucleotide constituent of DNA that contains the base adenine. Among the many processes activated by cAMP are the hormonal control of intestinal activity, blood sugar, urine volume, blood calcium levels, reproduction, and the production of a variety of neurotransmitters. It also has an inhibitory effect on other hormonal actions. The question of how a single mediator can elicit such a range of responses, and which of them are effected in specific tissues, is again addressed by the concept of repetitive differentiation. The membranes of differentiated cells are responsive only to specific hormones, although they mostly possess an enzyme, adenylate cyclase, involved in the conversion of Mg-ATP to cAMP. Then, when cyclic AMP is produced, there are specific, differentiated enzymatic response mechanisms that again are characteristic of the cell type. The ability to synthesize cAMP remains a general feature of many different cells, but its production and action depends on repetitive differentiation of cell receptor and effector molecules. A final point relating to homeostasis is that in mediation processes involving hormones and series of enzymes there can be a cascading amplification of the products, and control by rapid enzymatic degradation of the activating hormones is necessary. The dangers of unregulated positive feedback mechanisms are common in myth. The Fantasia version of The Sorcerer’s Apprentice, may strike a chord—Mickey Mouse tries to stop the magic broom from inundating the

The Physiological Arena

169

alchemist’s laboratory; but when he duplicates it by chopping it up, the cascade is amplified. New Genes from Old The most elementary knowledge of DNA action ought to dissuade anyone from the old notion that new physiological abilities are due to the acquisition of brand-new enzymes. The complexity of a protein requires the coded complexity of DNA, which cannot pop into existence out of nowhere. Classical examples of “protein novelty” keep being eliminated. Prolactin is derived from the older growth hormone family. Hemoglobin is a protein that appears to have evolved de novo a number of times. But every mitochondrion has the iron-porphyrin heme that makes hemoglobin interact reversibly with oxygen, and suitable globin proteins are universally distributed. Prior to the emergence of hemoglobin, an ancestral globin gene lost an intron and then duplicated. At some time in the late Pre-Cambrian the duplicates mutated, and from the divergent coding emerged both myoglobin (the oxygen carrier of muscle cells) and hemoglobin (which is usually found in blood). Further gene duplication provided for the emergence of the hemoglobin supermolecule. Sequencing evidence shows that when the original duplication occurred that gave the hemoglobin molecule its two globin strands there was an increase in the rate of base substitutions in the duplicate genes.50 Functions can be modified or even changed by repetitive differentiation. For example, there are suites of hemoglobins that function under different physiological conditions in individual organisms. Adult human hemoglobin oxygenates and releases oxygen at higher levels of environmental oxygen than does fetal hemoglobin. Hemoglobin consists of the conservative iron-containing heme that actually captures the oxygen, and two globin strands that modulate the action of the heme, and the genes for the globins are on different chromosomes. This complements the hypothesis that where there is a duplicate there is greater freedom for natural experiment. The improved adaptability conferred by the new configuration also relaxed the strictures of adaptation. Another example of emergent novelty arising from repetitive differentiation is the protein thrombin, which promotes blood clotting in higher animals. Its gene is a variant repeat of the more ancient gene for the protein-digesting enzyme trypsin found in digestive systems throughout the animal kingdom. Receptors for numerous hormones involved in the regulation of digestion and blood sugar also belong to the same G-protein-linked molecular family.51 In other words, repetitive differentiation of genes can generate developmental and physiological plasticity. “Novel” digestive enzymes such as chymotrypsin and pepsin were produced by repetitive differentiation of the genes for trypsin and cathepsin D respectively. Cathepsins are universally distributed in eukaryote lysosomes—organelles that have an intracellular digestive function in invertebrates and unicells. In arthropods and

170

Chapter 4

mollusks they have been seconded as supplementary extracellular digestive enzymes too. Cathepsins recycle cellular components during development and starvation in vertebrates. For example, their activity is elevated in the tails of tadpoles during metamorphosis, and in the muscles of migrating salmon, whose protein is sacrificed for energy and for re-investment in eggs and sperm. Cathepsins have other biochemical roles as well; one is used to activate a thyroxine precursor. Another protein “novelty” is a component of snake toxin derived from a common molecular ancestor of pancreatic trypsin inhibitor of mammals.52 Ancient intracellular fibrous proteins differentiated into the keratin of skin, scales, feathers, and the mucus of hagfish. Tubulins were put to use for a variety of mechanical functions including the contraction of muscles. An equally interesting emergence of a physiological innovation at the gene level involves both repetitive differentiation and contingent association of the functional domains of previously unrelated proteins. The milk-sugar synthesizing enzyme lactose synthetase has two sequences, one derived from galactosyl transferase, itself an enzyme of the Golgi apparatus, the other arising from milk protein lactalbumin, which in turn differentiated through mutation of a duplicate of the gene for lysozyme.53 Lysozyme is an ancient enzyme that protects against some bacteria, as Fleming and Allison discovered in the early years of antibiotics research. They first found it in human tears, but it turned out to be almost universal.53 Lysozyme has been involved in the evolution of cellulose digestion in mammals, as well as production of lactose. Variant repeats make the mature organism more adaptable in changing environments. Exon duplication, shuffling, and transposition is responsible for the buildup of proteins involved in blood clotting in vertebrates. But a different kind of novelty is illustrated by the crystallins of eye lenses in many animals. All that is required of these proteins is that they be small enough to avoid clumping and be transparent in aqueous solution. These are common properties of a wide variety of enzymes that have been independently recruited as crystallins for the eyes of different animals—an opportunistic molecular convergence.54 Crystallins are all coded by so-called housekeeping genes that need to be constantly available for the production of intermediate metabolic enzymes in most cells. One fascinating example of a new enzyme function arising from a simple point mutation is a change from an insect carboxylesterase to an organophosphorus hydrolase. This confers resistance to certain insecticides. The mutation results in an amino acid substitution that bonds with a water molecule to alter the tertiary structure and function of the enzyme.55 The relative paucity of examples of such novel genes is probably due to lack of investigation, but no matter how strongly “selection pressure” demands that something “ought” to be there, it is not always forthcoming. Since lots of people die of amino acid deficiency without having children, there must be a “strong selection pressure” for a gene for protein storage. Yet most animals cannot

The Physiological Arena

171

physiologically store protein at times of plenty, and release essential amino acids from that store in times of famine. They can only recycle proteins that are in use as muscles and enzymes, to the detriment of normal function. Simple molecules like glucose can be stored as glycogen, and fat can be stored in adipose tissue. But because a particular protein requires a correspondingly coded segment of DNA, there is an information barrier. Amino acids cannot simply be thrown together randomly to make proteins, and genes for protein storage cannot operate without a full complement of the amino acids that the proteins contain. Evolution has no ability to predict what amino acids are going to be available, nor in what erratic combinations, in the future. On the other hand, it would be useful if our lipid deposits were more like egg yolks, with vitelline storage proteins as well as fat. Insects seem to have pulled it off: their “fat body,” which occupies much of the abdominal cavity, contains storage protein. In some insects this is complemented by symbiotic bacteria that can turn excretory uric acid into useable amino acids. In actuality, every cell of our bodies, including the adipose tissue, has genes that code for vitelline protein; they would just have to be switched on. But no amount of selection pressure has done that to compensate for dietary deficiencies in mature vertebrates. All the same, protein starvation in humans is often more of a political or agricultural problem than a biological one, and that is true of other nutritional deficiency diseases. The mammal’s abilities to synthesize vitamins, essential fats and essential amino acids have regressed during its evolution, and if an adequate natural diet is available they do not need to be synthesized. But humans put each other in concentration camps, and used to send emissaries on prolonged sea voyages on a diet of hardtack and salt pork. Selection pressure should not be expected to compensate for such bizarre behavior, even for those who treat it as a real agent of evolution. New Physiological Systems from Old Interactions between previously independent genes and protein domains are common evolutionary experiments. Such changes occasionally happen at the physiological organ-system level. For example, after a period of development in the ocean, the blue swimming crab Callinectes sapidus, enters the brackish water of Chesapeake Bay, exposing itself to dilution physiogenesis. It responds by reducing the osmotic pressure of its cells through the metabolism of amino acids, releasing ammonia as an end product. Ammonium ions are then excreted into the environment in exchange for sodium and chloride uptake. Another advantage is contingent upon this osmoregulation mechanism. The oxygen-binding capacity of the crustacean blood pigment, hemocyanin, is diminished by dilution, creating a potential respiratory problem, since the crabs swim constantly during their migration, and need more oxygen than usual. But conversion of ammonia to ammonium ions increases the blood pH, which coincidentally makes the hemocyanin more efficient.56 This congruity of previously

172

Chapter 4

unrelated functions represents emergent adaptability—the whole is greater than the sum of its parts—but it is independent of homeostasis. Fish did not need such a stroke of luck since their hemoglobin function is not diminished by dilution. A more common example in the vertebrates is the interaction of the ventilation system with locomotion. Switching Biochemical Pathways Once a functional series of biochemical reactions is in place and all of the possible improvements have been incorporated, enzyme pathways tend to be conservative; but they can still be altered. First, an alternative branch might be grown, whose initiating enzyme, again originating by repetitive differentiation, competes with one of the old ones for a common substrate. The adaptability of switching backward and forward from the old to the new pathway is sometimes triggered by an environmental change. For example, in facultative anaerobic respiration an animal has two choices. ATP can be produced by simple glycolysis, in the absence of oxygen, or by using the normal oxygen-demanding mitochondrial pathways of oxidative phosphorylation when oxygen is available. In intertidal clams the normal response to the ebb of the tide is to close the valves and switch to anaerobic respiration. The burrow water stagnates rapidly and there would be nothing gained by circulating it over the gills. Glycolysis, which is the processing of glucose to release its energy in the form of high-energy ATP and reduced redox molecules, produces phosphoenolpyruvate. Here the metabolic pathways branch. When the bivalve is clammed up, the pH drops, because carbon dioxide accumulates. Greater acidity favors the action of phosphoenolpyruvate kinase, and initiates the full anaerobic chain of reactions.57 When the tide comes in, the bivalve extends its siphons hydraulically and squirts stagnant mantle water into the air. These celebratory fountains are typical of a clam beach at the early flood; hence the saying “happy as a clam when the tide is in.” Accumulated carbon dioxide is voided, and fresh air comes down the siphons. The rise in mantle pH turns the phosphoenolpyruvate over once more to pyruvate kinase, and the aerobic pathway is restored. As the tide ebbs, the changed environment effects a behavioral change in the clam, and a physiological alteration of the intermediate metabolism. Some upper shore bivalves simplify matters by using the mantle cavity as a kind of lung when the tide is out. In plants, a parallel switch in the pathways of intermediate metabolism has allowed certain “C3” plants to survive in dry environments. “C4” plants must keep their stomata open to obtain enough carbon dioxide for photosynthesis, thus risking transpirational water loss. The C3 plants use a carbon-fixing pathway that operates at low carbon dioxide levels, and the stomata can be closed. The difference can be seen dramatically in the prevalence of C3 couch grass on unwatered lawns that were originally seeded with thirsty C4 grasses.

The Physiological Arena

173

Regression I have already touched upon how complex organismal homeostatic systems can regress in a way that increases available energy—mostly having to do with temperature homeostasis, as in the camel, the naked mole rat, and the Arctic cormorant. The principle applies at the biochemical level too. A conservative pathway can be circumvented simply by dumping some or all of it. As Jakob von Uexküll remarked, the pianoforte can still bring forth complex concerti if it has lost some of its keys. French physiologist André Lwoff, writing in 1944, emphasized the importance of regression in biochemical evolution. He saw it to be parallel with paedomorphosis, the process of simplifying the life cycle by reverting back to the form and functions of the earlier stages. The example of a branching metabolic pathway, one of whose branches has regressed by epigenetic suppression, or by the lability of the non-essential DNA, illustrates his point. The pathway goes back to an earlier substrate stage where it can take off in a different direction. This process of drawing back in order to make an evolutionary leap in a different direction will be expanded upon in chapter 5. Biochemical regression has been important in the evolution of certain modes of excretion of nitrogenous waste. Animals that have urea as an excretory end product have lost the ability to synthesize the urease that can break urea back down to ammonia. Those that have uric acid as the waste molecule have lost the enzymes uricase and arginase, the latter a necessary agent in the production of urea. Like Caesar burning his boats when he crossed the Rubicon, these organisms have committed themselves entirely to the new excretory venture. In the case of uricotelic animals like reptiles, birds and land snails, the egg must accumulate nitrogenous waste, and uric acid is the least toxic molecule. Metabolic errors that would produce urea, or worse, ammonia, are a liability, and it is safer to get rid of them altogether. Nevertheless some reptiles retained maximum adaptability: crocodiles can excrete ammonia or uric acid, and aquatic turtles can excrete uric acid or urea. Primitive nitrogen cycles have a wonderful complexity. Early eukaryotes could probably synthesize urea and uric acid and break them back down again to ammonia. But that was before those molecules had anything to do with excretion. Uric acid was then part of a recycling mechanism in purine metabolism, which also involves guanine, cytosine and uracil, some of the nitrogenous bases of nucleic acids fundamental to cellular physiology. Urea may have been used as a temporary store of nitrogen, and then co-opted for osmoregulation, in the way that modern marine sharks and rays and coelacanths do, in order to maintain a low body fluid salinity without incurring osmotic imbalance. The regression of this primitively useful flexibility can be understood in cases like uric acid storing eggs; any other nitrogenous waste would rapidly accumulate to toxic levels. Natural selection has no interest in holding on to an adaptable system of low fitness when a specialized one presents itself. The loss of uricase in humans seems like gross carelessness, since it inflicts gout, a

174

Chapter 4

painful buildup of uric acid crystals in the joints. But hyperuricemia may also have been an epigenetic condition of rapid brain growth in the higher primates, and in humans in particular. Endosymbiotic Biochemistry The ultimate biochemical complexification is found in the symbiotic coming together of complementary systems that have already been established in independent organisms—the acquisition of heritable metabolisms. Particulars of this kind of biochemical emergence were discussed in some detail in the previous chapter. The impact of eukaryotic endosymbiosis on cellular biochemistry may be seen by scanning any textbook with metabolic diagrams that have two compartments representing the cytosol and the mitochondrial protoplasm, with a mitochondrial membrane in the middle and arrows going back and forth between the two compartments to indicate the exchange of metabolites. Since the establishment of endosymbiosis there have been many shifts of mitochondrial genes to the nucleus, and exchanges of enzymes and biochemical reactions between cytoplasm and mitochondria. The parts have been tuned to make up a more harmonious whole. The energy-hungry homeostasis of higher animals is fueled by high-energy molecules synthesized by the mitochondria. The dependable delivery of oxygen to them resulted from the evolution of gills and lungs. In the mitochondria, oxygen acts as the terminal acceptor of hydrogen protons and electrons from reduced carbon molecules. Although most people associate the word “respiration” with breathing or ventilation, this is only the preliminary to the absorption of oxygen through respiratory surfaces, the transportation of oxygen by hemoglobin in red blood cells, and the release of oxygen to the cells that need it for mitochondrial function. Reduced carbon molecules, such as sugar, amino acids, and fats, are acquired by animals as food, often in complex forms like proteins and starches. The evolution of feeding mechanisms and digestive and absorption mechanisms has channeled such fuels to cells where they are oxidized for ATP production. Nutrients largely arise from the primary trophic level of the food chain, where the reduced carbon energy comes largely from the capture of sunlight for photosynthesis. Therefore it is tempting to believe that this biological primary production emerged before the evolution of heterotrophy, i.e., feeding on preformed organic molecules. However, the primitive environment was probably rich in reduced carbon molecules that were continuously, abiogenically synthesized, and the most parsimonious evolutionary interpretation is that heterotrophy preceded chemosynthetic and photosynthetic autotrophy. Photolysis, the breakdown of water into hydrogen and oxygen by light, is a spontaneous abiogenic process too, and any hopeful prokaryote that could use hydrogen to reduce carbon dioxide, as some of the earliest Archaebacteria did, could have enjoyed simultaneous environmental expansion and energetic independence by

The Physiological Arena

175

harnessing water photolysis with a chlorophyll-like molecule. One hypothesis suggests that chlorophyll originated in organisms that used infrared energy rather than light for hydrolysis.58 Photosynthesizing prokaryotes were eventually acquired along with protomitochondria to make independent phytoplanktonic eukaryotes. Not only were these cells able to live in diverse aerobic and anaerobic conditions, they turned the dangerous byproduct, oxygen, into an energetic asset, although sometimes energy was bled off as bioluminescence. They were also in a position to survive the inevitable consequence of their own biochemical adaptability, namely the oxygenation of the ancient anoxic environment, and the creation of a new biosphere, within which oxygen-breathing organisms might evolve. In the longer term this made it possible to achieve the high metabolic rates required for the development of complex nervous systems. Plants created the conditions for the emergence of creatures that could initially browse on them—and eventually think about their role in evolution. The evolutionary history of physiology involves biospheric changes caused by organisms, and changes in their internal milieux caused by the environment. These were intensified as greater adaptability permitted more flexible behavior and exploration of more extreme conditions. We need to look at ecological conditions where natural experiments can prepare for emergence into new environments, such as interfaces between oxygenated and anoxic regions, between sea water and fresh water, and between aquatic and terrestrial. During the course of evolutionary history, environments free from competition and predation, rich in food resources but rife with physiogenic challenges, awaited the physiologically and behaviorally adventurous. Rapid diversification could occur until the climaxes of dynamic stability were reached. In the past there were also environments swept clear by catastrophes. Mass extinctions, drastic disruption of food sources, instantaneous changes in temperature, and lingering climatic change, affect every arena of evolutionary causation. Competition, and predation may be removed entirely, and physiological adaptability is immediately at a premium for simple survival. That adaptability has evolved in part through progressive improvements in regulatory systems that stabilize the internal milieu, and let the organism live la vie libre. Physiology, Developmental Biology, and Evolution I remarked earlier that the physiology of the mature organism has to be studied in its own right rather than as an extension of developmental physiology. Yet it should already be obvious that changes in the interactions between mature organismal physiology and developmental processes are crucial. Therefore, in anticipation of the following chapter on development, a few general remarks about the links between the physiological, behavioral and developmental arenas might be useful.

176

Chapter 4

Roy Pearson’s synthesis of the physiological qualities of the developing and mature organism is a good start. He also places it in an environmental context: “. . . the common thread woven through the problems of macroevolution, morphogenesis and cancer is the hormonal-homeostatic complex of the organism. Life history (environment) interlaces with homeostatic and thus homeodynamic forces to create the organism’s evolution, development and demise.”59 Pearson argues that if adaptable responses to new environments have an endocrinal basis, the new environments can affect development and reproduction. The “vertebrate hypothalamic/pituitary/ thyroidal/adrenal/gonadal axis which regulates homeostasis is also intimately involved in metamorphosis (development) and reproduction (e.g., vitellogenesis and spermatogenesis).”60 Eventually, in animal evolution, many epigenetic responses to environment are genetically assimilated. After birth, human babies do not have to be made to stand up and manipulate objects in order to acquire a typical human gross anatomy. That process is already under way in the fetus. Yet bipedalism requires a mechanical striving during childhood, so that there is an ontogenic adjustment of the skeleton and musculature. Brain development also requires ontogenic experience, particularly in learning, to bring out certain potentials. The idea of the genetic internalization of environmental effects was pursued by Schmalhausen, who called it “autonomization.” It is common knowledge that reproductive physiology and behavior, as well as dormancy and other physiological factors in plants and animals are often governed by climate and season. External physical stimuli act through hormones or neurosecretions to influence the genes responsible for synthesizing the other proteins appropriate to the season. Schmalhausen was sure that primitively the environment was a major inducer of developmental change, and that gradually the induction was internalized, thus complementing Bernard’s precept that independence from external change was the precondition of physiological freedom. Schmalhausen’s stabilizing selection included internal co-adaptations that would lead to physiological stability, or homeostasis. As part of the process, “duplicative mutations” of simple regulatory mechanisms could build up an array large enough to take over from the old environmentally induced one—an evolutionary hypothesis equivalent to the concept of repetitive differentiation.61 Finally, disequilibrating influences of stressful environments might be needed to induce evolutionary changes to occur, and to escape from stasis. Epigenetic modification can be initiated by changes in temperature, precipitation, photoperiod, food (or lack of it), and population density, as well as extrinsic biological triggers like pheromones. The morphological, physiological and behavioral consequences seen after metamorphoses are particularly striking. Digestive, locomotory and respiratory systems may be totally reorganized for example. Locusts start out as simple grasshoppers and do not form their devastating flying swarms until an accumulation

The Physiological Arena

177

of environmental factors trigger changes in adult morphology, physiology and behavior.62 Another discovery of how overcrowding affects these factors is the appearance of monstrous carnivores in salamander populations.63 Mary Jane WestEberhard’s Developmental Plasticity And Evolution (2003) is the best current source for these kinds of cases. Conclusions Referring back to the checklist of questions in chapter 2, which were to be kept in mind by explorers of emergent evolution: 1. Molecular and functional-morphological repetitive differentiation is common to the arenas of association, and physiology. As we work through the next chapter we will see that repetitive differentiation at the gene, cell and organ levels are important components of developmental differentiation. Furthermore there will be commonalities found between disequilibration of existing homeostatic stabilities and homeorhetic stabilities as part of emergent evolution. 2. The generative conditions from which physiological emergences spring combine physiological and behavioral adaptability with environmental opportunity at interfaces between ecosystems. Concentration at interfaces is also relevant to the origin of protobionts and to endosymbiosis. 3. Novel emergent qualities of physiology, especially those with a symbiotic content, altered external environments, so that a way was prepared for further natural experiments and ventures 4. Transition to new environments caused disequilibrating physiogenic change that became a key generative condition for further modifications. In higher animals the acquisition of a sophisticated homeostasis potentiated novel behaviors and anatomical specializations. 5. The interaction between organism and environment vindicate the neo-Lamarckist argument that the environment imposes change, and Lamarck’s law that the organism responds to its environment. These are important evolutionary factors, even in the absence of a somatogenic mechanism to genetically fix them. An environmental influence may finally be genetically accommodated if there exists a relevant genetic variation. What was missing from Lamarckism was that pre-existing physiological and behavioral adaptability allow the organism to expose itself to environmental changes and respond successfully.

178

Chapter 4

6. Evolutionarily important novelties in physiological evolution are too numerous to summarize. They have often been features that impose protective barriers between organisms and their environment, such as waterproofing in plant and arthropod cuticles and vertebrate skin and heat insulation in birds and mammals. The invention of cellulose and lignin skeletons in plants and calcium carbonate shells and calcium phosphate skeletons in animals potentiated their anatomical evolution. The placenta, the hypermorphic cerebral hemispheres, and the corpus callosum were immediately adaptable. It must be borne in mind that “key innovations” happen in the context of other generative conditions. 7. Physiological evolution may also be affected by unpredictable contingent factors, such as the interaction of independent genes to produce new enzymes, or the interaction of previously independent organ systems. More predictable are physiogenic contingencies that occur as a result of transition from one environment to another. Major cataclysms are unpredictable but may have a direct effect on biochemical functions through heat shock. They also remove competitors and predators that may have hindered the diversification of organisms with advanced adaptabilities. 8. Physiological emergences have often produced a constellation of multiple functions. Glands associated with hair provide for evaporative cooling, milk production and sexual attraction, as well as maintaining insulative properties of hair. The most outstanding example is the potential for multiple behaviors that accompanied emergence of the neocortex. 9. The course of emergent physiological evolution has led to greater self-organization, independence, and freedom of choice. 10. Most important events in physiological and behavioral evolution occur above the gene level. Evolution is not a process of changes of the distribution of alleles in gene pools. It involves instead the reality of the organism, which represents the past and shapes the evolutionary future. A Note on Plant Physiology Some, but not all of the concluding generalities above apply to plant physiological evolution. I will bring the subject back in chapter 9 under “physiological theory.” I have not neglected the plants out of indifference or incompetence, but because I am constrained to limit the size of the book.

5 Development and Evolution

The major empirical fact about the development of animals—a fact which has no theoretical inevitability, but which is so obtrusive that only the rudest observation is necessary to establish it—is that the end-products which it brings into existence usually vary discontinuously. —C. H. Waddington, 19571 If evolution is emergent, the basis for this is to be found, not in the natural selection of random mutations but in the creative potential of epigenesis. —M.-W. Ho and P. Saunders, 19822 . . . an empirical approach to the problem of novelty has to focus on the organizational principles of developmental systems and their ability to generate new structures. —G. B. Müller and G. P. Wagner, 19913

Before Darwinian theory, it was known that many animals, diverse in their mature forms, shared anatomical homologues that must have arisen by the same basic processes of development. Darwin concluded that they must therefore have evolved from common ancestors. However, efforts to elucidate developmental evolution was more intense among neo-Lamarckists, who did not object to its saltatory implications. Furthermore, long before the molecular structure of the gene was worked out it was realized that the diversity of organisms was not to be understood in terms of different gene complements but in terms of different expressions of the similar sets of genes that all organisms possess. Von Uexküll’s metaphorical piano, with its keyboard of Mendelian genes, did not have to be reconstructed or given new keys to create different music, only played according to different instructions in the scores. In the language of modern biology, the structural genes that code for proteins are similar across the biological spectrum from bacteria to humans. Even regulatory genes are highly conserved. Although the number of protein domains that are coded in DNA is far smaller than the number of genes, they can be arranged and rearranged in different combinations. What has changed during the course of evolution is the set of instruc-

180

Chapter 5

tions for differential gene expression and protein domain combinations. And these, in combination with gene duplication and differentiation produced more and more complexity as time went on. This message has not completely sunk in. Some biologists still believe that the natural selection of random point mutations that improve ecological fitness is the essential or ultimate process of evolution. Everything else is proximate, or contributory, as well as random. For them, complexity is an accidental accretion of adaptational molecular changes. Molecular biologists were taken aback by the genome project’s discovery in 2001 that a human has only 10,000 more genes than a simple roundworm. For von Uexküll, however, it would still have made sense if we had been found to have fewer genes. Most non-biologists—sometime less myopic than the specialists—identify evolutionary diversity with differences in form, or anatomy. These are caused by alterations in developmental patterns in conjunction with changes in gene expression and environment. C. H. Waddington could not bring himself to say that developmental evolution is saltatory. But in his epigraph he admits that epigenetic divergences are discontinuous in their phenotypic manifestations, as Geoffroy had proposed before Darwinian theory came along. Mae-Wan Ho and Peter Saunders help us focus on the emergent nature of developmental evolution. Gerd Müller and Günter Wagner infer that the generation of novel organization is saltatory. Such implications have always made the gradualistic Modern Synthesis wary of incorporating epigenetics. Another troubling prospect is that the environment can have a directing effect on the course of developmental evolution, and that brings in the physiology and behavior of the whole organism too. If that were not enough, I am going to rub salt in the wounds by resurrecting orthogenesis in chapter 7. However, I am not addressing the fainthearted here. They have long since left the field to the travelers who have persevered out of interest in evolutionary origins and diversity, and dissatisfaction with neo-Darwinist gradualism. Now that we are at the developmental ring of the evolutionary circus, with its liveliest of performances, it is salutary to recall where we have already been. In the symbiosis and association ring, the raw material for natural epigenetic experimentation—sexually reproducing, eukaryotic cells in multicellular groupings—had already come into existence. Moreover, evolutionary physiology and consequent behavioral flexibility of the organism operate in an adjacent arena, and these three rings of the evolutionary circus are all linked under the big top of the environment. Therefore, we should not let the razzamatazz in a particular ring mislead us to think that it is the best of the show, and to forget the others. While the circus metaphor serves to simplify the causal complexities of evolution, the reality is the whole organism in its environment, bearing in mind also that the environment consists to some extent of other organisms of the same and of different types.

Development and Evolution

181

Perceptions of Epigenesis and Evolutionary Development A quick primer in some of the modern terminology for embryological or developmental evolution is needed at this point. “Epigenesis” is archaic, being Aristotle’s term for the embryonic ordering of a complex being from disorganized matter. It is now synonymous with the general process of embryonic development. With regard to the importance of embryology for evolutionary theory, J. H. Woodger (1929) wrote: “What we do observe . . . is a gradual rise in the level of organization of the developing organism, i.e., this appears to be beyond reasonable doubt an epigenetic process.”4 “Epigenetics,” as it was used by C. H. Waddington in 1942 and 1957, meant the study of developmental processes that involved differential gene expression, and its regulation—how the genotype gave rise to the phenotype. In Epigenetics (1974), Søren Løvtrup defines it as the study of anything that affects normal or evolutionary epigenesis, including the role of non-DNA influences, both organismal and environmental. The definition of epigenetics is the subject of some ongoing etymological debate, which has even run as a Socratic dialogue in the pages of Science.5 Some definitions limit epigenetics to heritable mechanisms, pertaining to DNA and its regulators. But that is too narrow, since development can be affected consistently and persistently by non-heritable environmental factors. And some environmental influences are heritable though no DNA changes are involved. Evolutionary epigenetics considers how changes in epigenesis might result in anatomical and physiological change. Like Løvtrup (1974), I take epigenetics to encompass all of the mechanisms of epigenesis, including its evolution. Its scope includes alterations of gene expression, the effects of the exterior environment, the internal milieu— including the influence of symbionts, the interactions of embryonic cells, the cytoplasmic environment, and the process of genetic assimilation of environmental effects. Some modern authors prefer the term “evolutionary developmental biology,” which they feel better captures the entirety of embryology in relation to evolution.6 They are the ones who call themselves “evo-devos.” Theoretical connections between development and evolution had been established before the publication of The Origin of Species. The quasi-evolutionary paradigm called “nature philosophy,” which proposed a unity of plan among organisms, inferred the relatedness of organisms based on similarities of anatomy. Out of this school of thought came Goethe’s ideas on plant and animal homology, and Owen’s concept of the specially created archetypal organism, whose anatomy diversified to produce different homologous structures. Nature philosophy co-existed with Lamarckian evolutionary theory, and both were integrated with developmental evolution by Lamarck’s colleague Étienne Geoffroy de St. Hilaire, usually referred to simply as Geoffroy.

182

Chapter 5

Geoffroy promoted the importance of understanding epigenetic processes of evolution. He believed that species diverged from common ancestors through alterations that the environment induced in their developmental pathways. The earlier in the developmental process that the alteration took effect, the greater would be the change in the mature organism. Those insights should have guaranteed a strong epigenetic component to all subsequent evolutionary thought. But he overgeneralized—as T. H. Huxley remarked, he spoiled his case by stating it. For example, he proposed that all that was needed for a vertebrate to emerge from a protostome like a squid was reversal of the dorsoventral axis. Nowadays, the study of deep homology at the gene level brings us much closer to Geoffroy’s position, although the reversal of the dorsoventral axis would have occurred in very early ancestors of the protostome and deuterostome lines, rather than in a squid or a fish. However, his contemporary Charles Darwin shied away from any such saltation, because it would reduce the causal power of natural selection. Embryology only mattered to him for evidence to support the evolutionary relationships of different animal families. It should already be evident that much of evolutionary development remains a tangled bank of speculative suggestions, well-described processes, some molecular mechanisms, metaphysics, and wishful thinking—including a dysfunctional relationship with selectionism. To bring a degree of order to my discourse, I will first take the route of describing individual developmental (ontogenic) changes that indicate how evolution might have occurred. Then comes the problem of understanding what kind of ontogenic plasticity allows change without disintegrity. This, in turn, provides some answers to the question of how radical phylogenetic or evolutionary changes are accommodated. Good Sports The most familiar examples of epigenetic stability and developmental change are to be found in plants and animals in our homes, gardens, and farms. Also, when we look in a mirror in a reflective frame of mind we recognize uniqueness, similarity to close relatives, and a connection with our distant primate relatives. Individuals who are abnormal provoke thoughts about how they might have come to be that way. As Aristotle said, nature is true to type—dogs beget dogs and cats beget cats—but there are odd exceptions that breeders call “sports,” which are quite distinct from their parents in some way. The conditions of confinement for breeding animals, as well as deliberate selection for tameness, may all be epigenetically destabilizing, inducing the production of developmental novelties, even when the breeding stock is not closely inbred.7 Good sports are chosen as the founders of new pedigrees. Broadly they remain true to type, but not to form, as Darwin very well knew. He was fascinated by pigeons with feathered feet and webbed toes, fantails with multiplied tail feathers, and others with extra vertebrae and elongated beaks and bones. He did not even stop to think of

Development and Evolution

183

them as saltations, although he accepted Sir John Sebright’s word that “he would produce any given feather in three years, but it would take him six years to obtain head and beak.”8 One of the most frequently cited sports in the Darwinian era was the Ancon ram, which had very short legs, and was selected by American breeders because its progeny were easier to fence in than the normal long-legged form. Note well that in farming, as in nature, the developmental saltation comes first, followed by the act of selection, whether conscious, natural, or metaphorical. The dachshund is another familiar shortlegged sport that was selected for its ability to hunt badgers in their setts without knocking its head against the roof. Both the Ancon ram and the dachshund result from the premature completion of limb bone ossification, similar to the condition in human achondroplasic dwarves, who have almost normal trunks, but short arms and legs. Such phenotypic conditions may be caused by a variety of epigenetic hormonal changes. Apart from dwarves there are other human growth enormities, such as pituitary midgets and giants, and pygmies. Although the word “enormity” suffers from usage slippage, its proper sense of “out of the norm” seems to be more appropriate than “abnormality”; African pygmies regard the rest of us as abnormal. Acromegaly is a burst of growth of the extremities near the end of development, which increases the dimensions of the jaw and hands and feet. It is another common expression of human developmental hormonal effects, based on alterations in gene expression, but with some environmental influences such as diet as well. Geoffroy’s evolutionary thinking was influenced by his study of abnormal human embryos and fetuses, for which he coined “teratology.”9 He mistakenly believed that the origin of the change was in the developing organism. Or was he in error? By modernist lights, if it does not occur in the germ plasm of one of the parents it is not heritable, and therefore meaningless in an evolutionary sense. But Geoffroy also thought that the environment could directly cause such changes, and keep on causing them from generation to generation, and I will not exclude this as a contributory evolutionary factor. For Geoffroy, birds might have arisen from reptiles by such an embryonic saltation. And he did not hesitate to conclude that such was the mechanism of “the origin of species,” which is how he phrased it in the title of his 1833 memoir. In Geoffroy there is an element of truth that has to do with the genetic fixation of phenotypic changes, or genetic assimilation, all of which will be dealt with shortly. With Geoffroy came the first musings about processes and mechanisms of developmental evolution. These influenced German embryologists, such as von Baer, who were to lead the discipline for the best part of a century. In 1864, von Kölliker argued that Darwin’s gradualism was inadequate to explain progressive evolution. His alternative was the emergence of novel forms through sudden embryonic metamorphoses, like those that gave rise to medusae from polyps and adult echinoderms from

184

Chapter 5

their planktonic swimming larvae. In response to von Kölliker, T. H. Huxley repeated the notorious remark that he had made to Darwin that The Origin of Species was hampered by unrelenting gradualism and that saltations might be possible.10 The modernist view is that metamorphosis does not represent evolutionary saltations, but are compactions of previously gradual developmental processes. This is not contradicted by epigeneticists, but they see that metamorphoses have another evolutionary significance, since they are life cycle stages where the developing animal might have a variety of evolutionary epigenetic options. J. J. Murphy (1869) argued that evolution must be a matter of differentiation and integration, and phylogeny followed ontogeny. This inspired Darwin’s critic St. George Jackson Mivart to propose that saltatory evolution, manifested during development, would help to solve one of the problems of gradualism: how the incipient stages of a gradual process could have selective value.11 There was no such gradual phase, just a creative saltation. His other explanation, an autonomous drive, will be discussed in chapter 7. Mivart also took neoteny to be a kind of developmental saltation, albeit a leap backwards. And he puzzled over parallel evolutionary patterns that appeared to have more to do with inherent developmental trends than with adaptation. The American neo-Lamarckist Edward D. Cope made an important theoretical contribution to developmental evolution by emphasizing the importance of changes in relative developmental timing, such as acceleration and retardation. These could cause more complex allometric shifts if the relative rates of development in different cell and organ lines were affected. Acceleration of the growth of the front end of an okapi would produce the long neck, heavy pectoral girdle and strong front legs of a giraffe, the development of the rear end being retarded. Relative changes in timing and rate now come under the heading of “heterochrony.” Cope also theorized that by his Law of Repetitive Addition, a simple organism could become complex through multiplication and then differentiation of its body segments. This concept, which I have renamed “repetitive differentiation,” applies just as well at the molecular level, and I have already discussed its importance in physiological evolution.12 Before he became an emergentist, Conwy Lloyd Morgan appreciated the need for a generative theory of evolution to complement the Darwinian description of what happened after change had been generated. His Animal Life and Intelligence (1891) focused on the importance of variation. In addition to superficial variations that involved color and external form, there were “organic” variations of physiology, and behavior, and “reproductive” and “developmental” variations, which included heterochronic alterations in epigenesis. Natural experiments in development might result in saltatory evolution; but increased organismal complexity was a response to increased environmental complexity, as Herbert Spencer had proposed. Later he would shift that stance to become an emergentist committed to discontinuous endogenous change in complexity, albeit with environmental influences.

Development and Evolution

185

William Bateson is now remembered for popularizing Mendelian heredity, and coining the name “genetics.” Although he studied at Cambridge under a strictly Darwinistic, comparative embryologist, F. M. Balfour, he leaned in the direction of neo-Lamarckism and saltationism. The director of his study of Balanoglossus in the United States was W. K. Brooks, whose Theory of Correlated Variation prompted Bateson to explain how complex innovations arose.13 Minor changes in the internal embryonic environment could trigger correlated changes in associated systems. If some such “accommodatory mechanism” did not exist, developmental saltations would have been too disruptive to ever make a contribution to evolution.14 Bateson set out exhaustive research results in Materials for the Study of Variation with Especial Regard to Discontinuity in the Origin of Species in 1894. In it he proposed that there had to be natural saltatory tendencies of progressive variation, but by that time he had rejected the notion that the environment had a direct influence on evolution.15 As the title and publication date of Bateson’s book suggest, he was already ahead of Hugo De Vries’s Mutation Theory of evolution, which he promoted on its publication just after the turn of the century.16 Just after T. H. Morgan’s Drosophila group began to focus the attention of evolutionists on genes and chromosomes, Hans Spemann and his contemporaries discovered that while nuclei affected differentiation, determinants from the cytoplasm of surrounding cells could have a feedback effect on them.17 Spemann himself did not give any great weight to the evolutionary significance of the organizer effect, being more concerned with experimental facts than theoretical speculations.18 Indeed the history of developmental biology indicates two categories of investigator. The first was pragmatically interested in a comprehensive factual description of embryogenesis and the details of its processes and mechanisms. The second was more concerned with theoretical interpretations in an evolutionary context. A synthesis of epigenetics and evolution was retarded by the indifference of the pragmatists, and lack of evidence to support speculation. Moreover, the theorists alienated gradualistic evolutionists by concluding that developmental evolution had to be saltatory.19 Good sports need to be cosseted by breeders, and guarded from genetic mixing as well as being selectively bred. Isolation of some kind is assumed to be an essential part of natural speciation. But, could they take care of themselves in the struggle for existence in the wild? Gradualists argued that developmental saltations were too disintegrative for them to survive at all, far less to contribute to progressive evolution. The neo-Darwinist R. A. Fisher wrote in 1930 that macromutational “leaps of Nature” could only produce damaged goods. Because they had eluded the benign attention of natural selection, saltatory deviants could not survive. For example, the bithorax condition in Drosophila—an extra pair of wings—could not be functional, because its nervous system was not properly coordinated for flight with four wings. This is true of bithoracic wings in fruit flies, but most other insects have two coordinated pairs of

186

Chapter 5

wings, and primitive flying insects managed to accommodate without detriment the development of multiple simple wings in many body segments, through some change in developmental regulation.20 And this was followed by other phenotypic saltations resulting from the repression of wing development from all segments except a pair in the thorax. Echoing Fisher’s objection to a role for macromutatory saltations in evolution, Rupert Riedl, in Order in Living Systems (1978), uses the metaphor of a high-rise building under construction. If the builder misreads the basement plan and puts the elevator well in the wrong position, and if the plans for the other floors are then read properly, the elevator will never be able to leave the basement, and the building would be unfit for habitation. But just as a builder could modify the old plans and finish construction so that the elevators would work, so could an organism accommodate an early error. For instance, veins in the wings of young insects guide the path of developing nerves. Yet, in the absence of crucial veins in some mutant insects, the neurons manage to find new and effective developmental routes, and induce the appropriate sensory organs when they reach their goal.21 Moreover, many insects make major morphological changes in their larval basements, such as multiple, atavistic legs in caterpillars, and yet come back to the original form of the adult. The regulation of epigenesis in ways that keep the organism true to type involves back up systems that allow the usual goal to be reached despite potentially debilitating mutational losses or modifications of gene products. One example in human development is syndactyly: what starts out as a six-fingered embryonic hand becomes the usual five-fingered appendage, because two of the fingers fuse into one. Bateson was right that there are accommodatory mechanisms, and we are coming to understand them better. So we know that an epigenetic elevator well that has been misplaced in the basement can accommodate itself by inducing architectural changes in the upper floors. There are some clearly established ways in which the old plans can be junked, a divergent design followed, or a new design superimposed, which we will explore later in this chapter. Some epigeneticists are not aware that there ever was an accommodation problem, which shows that ignorance of history can occasionally mean liberation from “big issues” that were never really significant. The crucial point is that accommodatory mechanisms not only keep development true to type, but might compensate for detrimental consequences when a radical epigenetic change leads off in a new evolutionary direction. Hopeful Monsters After Mivart focused on the good sport as a model for epigenetic evolution, the next development of the idea came in the form of “the hopeful monster” (Goldschmidt 1940). Its creator had good grounds for seeking alternatives to selection theory, since his extensive field studies on Lymantria, the gypsy moth, showed that Darwinism

Development and Evolution

187

encompassed only minor elements of evolution, affording no clue to progressive change. Before the hopeful monster was conceived in 1933, for a lecture at the Chicago World’s Fair, Goldschmidt had already considered heterochrony in development, and mused that the accommodation mechanism that would buffer radical embryonic alterations must be a coordinated shift, like the operation of an automatic transmission in an automobile. The major criticism leveled at Goldschmidt nowadays is that he did not try to explain how any kind of monster would be fit enough to survive, far less come to dominate a population. He did admit that an experiment like the tailless Manx cat was “just a monster.” But Archaeopteryx, a feathered reptile, was a hopeful monster, because it could put its feathers to some use for gliding, if not true flight. He should have asked himself how just-a-monster could persist long enough for him to characterize it. The answer is that it retains organismal integrity, and its monstrosity is neutral or of minor relevance to competition and predation until perhaps a catch-it-by-the-tail predator comes along. The “Goldschmidt toad” reported by David Rollo (1994) as “doing very well” in an Ontario garden, had functional eyes that had developed in the buccal cavity from the roof of its mouth.22 No doubt there would be a selective pressure for that one—seeing its way better to its food perhaps? But it was a saltation nonetheless. Roy Pearson (1999) points out that among wild starlings there are sports with long bills reminiscent of nectar-feeders. One was observed to feed normally, without any obvious disadvantage. Whether an innovation, or an atavism to a nectar-feeding ancestor, the phenomenon is probably a saltation involving unusual neural crest cell action and morphogenetic protein production.23 The just-a-monster features of some human terata have afforded not only hope, but success to some other vertebrates: giantism and achondroplasic dwarfism are natural in otters. The armless phocomelic condition, where the hands emerge from the shoulders, brought to public attention by the thalidomide scandal, is not monstrous in a seal; and complete limblessness is typical of snakes, as well as some lizards and urodele amphibians. Although some of these examples can be interpreted after the fact as being adaptational to particular habits and habitats, the initial emergence is by developmental saltation, and the primary requirement is organismal wholeness, which is causally prior to the tests of natural selection. One of the monster’s hopes is to have other monsters to mate with, so that its new qualities will be passed on to the next generation. In organisms that can clone themselves or self-fertilize there is no such difficulty. Plants and primitive animals that produce vast numbers of gametes could establish instant populations of the new type. For the others, the condition that arose in their parent’s germ cells might emerge in only a few siblings, and require close inbreeding to establish a distinct new lineage within the species. Recessive alleles associated with monstrosity in the homozygous condition could accumulate by genetic drift and so eventually produce a larger broodstock.

188

Chapter 5

If it is not better at what its parents did, the emergent will be obstructed by established patterns of competition and predation. If it is physiologically adaptable, alternative environments or habits proffer great hope. The proto-bird can take to the air, and the proto-otter to the water. Monsters with greater adaptability than their parents may still have to wait until a radical change in the existing environment has occurred, or head for the fringes. So much for the fate of epigenetic novelties once they have occurred. Questions remain: how do they get over the normal developmental constraints, and when they do, how is it that their natural epigenetic organization doesn’t fall apart? I will begin to answer this by looking at ontogenic plasticity, or the scope for deviation that exists in the development of an individual organism. In the higher plants, extensive phenotypic plasticity helps to offset the occasional disadvantage of having to remain rooted in the same spot for life.24 Animals are much more predictable in their final mature form, but they have more flexible habits. In both cases, epigenetic and physiological adaptability provides some room for maneuver. Epigenetic Plasticity In the physiological arena we saw two major types of adaptability. The first accommodates to change by switching to mechanisms that keep it going, e.g. when the temperature changes, the enzyme with the matching temperature optimum is turned on. The internal milieu may change; but the operational efficiency of the organism is not affected. Alternatively, external change is buffered by homeostasis and internal stability is maintained. In development there can also be an accommodatory switch called “facultative phenotypic expression,” an either-or choice described by Mary Jane West-Eberhard, where “condition-sensitive expression of alternative phenotypes means that in a variable environment a novel phenotype (such as worker behavior, or a new trophic specialization) can evolve alongside an established specialization without being expressed (competing) in the same situations.”25 This describes circumstances where the same genotype can give rise to two distinctly different phenotypes, the switch being activated by an environmental cue. In termites, and presumably other social insects, the timing of the supply of food and the effects of juvenile hormone modulate the caste system. The relative numbers of workers and soldiers are adjusted according to the needs of the nest.26 Boris Uvarov detailed striking examples of novel phenotypes that appear in response to population density in his encyclopedic work, Grasshoppers and Locusts (1966, 1977). At low population levels the adults are drably colored and solitary in their behavior. With increasing population the egglaying behavior of females changes so that they lay eggs en masse, so that the deme density is increased, and the adults of several species become vividly colored and striped. The crowded wingless hopper stages of locusts form “marching bands.” Eventually as adults they may take to the air in flying swarms that can cover an area as large as 780 square kilometers. Such behaviors are stimulated by acoustic and visual

Development and Evolution

189

cues, and modulated by juvenile and sexual hormones. These kinds of plasticity involve an all-or-none commitment for the individual, and lacks the modifiability associated with physiological and behavioral adaptability. Nevertheless, it illustrates that mechanisms that keep development predictably on track can, under some circumstances, be derailed without lethal disintegrity. Whether an accommodated change makes any difference in selective value or not, the new epigenetic experimental organism will be ready to go about its life in a novel way if its integrity is preserved. The semi-independence of developmental pathways has to be sufficient that change in one does not act to the detriment of others. Yet there has to be enough feedback between the pathways to make integrating adjustments. West-Eberhard (1989) notes that epigenetic shifts are further accommodated by the plasticity of behavior and correlated functional anatomical shifts. She gives the example of Slijper’s goat, which was born with very short front legs and normal hind legs.27 But this good little sport learned to walk upright, and its anatomical development responded with accommodatory changes in the spine, thorax, neck and muscle insertions. There is no argument that the epigenesis of an individual is plastic, which being so, the difficulty of adjustment to change is mitigated and monsters might not be altogether hopeless. Historically, however, accommodation presented a major theoretical problem. The Canalization Bind: Does Lack of Plasticity Obstruct Evolvability? Aristotle’s aphorism that nature is true to type is so consistent that it presents evolutionists with the perennial dilemma of understanding how it works, and how it might be overridden. Although he knew of the existence of abnormalities in human development that survived to maturity, he argued that the soul had a quality called “entelechy” that ensured that cats begat cats. The dilemma was emphasized for materialist Ivan Schmalhausen, because he could see that developmental stabilization became more rigid with the passage of time, as external environmental triggers became internalized in the organism. But he had a good general idea of how to resolve the problem. Entelechy was then re-invented by Waddington (1957) as “homeorhesis,” which signifies the stabilized flow of epigenesis. “The thing that is being held constant is not a single parameter but is a time-extended course of change, that is to say, a trajectory.”28 Thus, the organism’s development is kept on track by compensations for minor misguiding influences. Waddington distinguished the processes of homeorhesis and physiological homeostasis. He did not, however, completely round the circle by comparing their evolutionary roles. Homeostasis allows the mature organism to persist in its being while exploring new environments, and acquiring habits that might otherwise tend to disequilibrate its internal milieu. It places no obstacle in the way of functional anatomical change that would complement new behaviors in new environments,

190

Chapter 5

quite the opposite. Homeorhesis does, however, obstruct anatomical change by conserving the integrity of the original form. Therefore homeorhesis cannot be equated with evolutionary adaptability in the same way as physiological homeostasis. Nevertheless, anatomical changes have occurred, and, as Peter Saunders points out, “Homeorhesis is a necessary property of an epigenetic system. But a system which possesses this property will also have the capacity for heterorhesis, i.e., for large, organized change.”29 Developmental adaptability is inversely correlated with Waddington’s “canalization,” the degree of fixation of epigenesis along particular developmental pathways. The concept of canalization arose from the well-known metaphor of the epigenetic landscape. Imagine a water-eroded slope, where minor channels would run into large channels as they flowed downhill. If a round stone is randomly pushed from the ridge at the top of the slope it will eventually find its way into a minor channel, whence it is canalized by a deeper channel in a determined downward direction. At first, the kinetic energy of the stone, and the shallowness of the channel, might make it bounce away from its likely course. This would be loose canalization. Eventually, however, the stone would roll into one of the lower, and deeper channels, from which its escape would be difficult, and its final goal all but certain. The behavior of the rolling stone offers an heuristic metaphor for the migration of embryonic organizer cells as well. Both of them slow down, are obstructed, and eventually “gather moss” which fixes their position firmly. Tight canalization prompted the crucial question of how it could be loosened sufficiently for epigenetic evolution to occur. Waddington’s model suggested disequilibria in the topography of the landscape itself, and his equation included disruptive alterations in the real external environment. It is often argued that the body plans of animals that emerged around the time of the Cambrian explosion have become so tightly canalized over geological time that further evolution is impossible. But the vertebrates are exceptions that prove the rule. They modified the basic plan, made novel additions, and kept on doing it. In hominid evolution the process has been rapid by conventional calculations. These changes are epigenetic, and therefore decanalization and accommodation can be quite radical, despite the passage of time and the growing defenses of homeorhesis. The question remains: what are the circumstances that bring such changes about? Or how, as Waddington asked, does nature get away with it? Schmalhausen had already suggested that physiological adaptability in the developing embryo, and in the mature hopeful monster, provided some accommodation. Even stable environments can cause phenotypic changes, especially where there are distinct climatic and photoperiodic cycles. There is also room for adaptational specialization in microenvironments within stable environments. Finally an ecosystem will contain a diversity of specialists. The more specialized an organism becomes in a stable environment, the more it is locked into a line of evolution that can only go in one

Development and Evolution

191

direction—more of the same, whether by directional selection or allometry. Exploration of alternatives is increasingly prohibited by competition; and with specialization there comes a more rigid epigenesis, loss of plasticity, and a diminishing ability to find new options. The Specialization Bind Long before Waddington’s attempt to formalize the problems of evolutionary development, zoologists were aware of incongruities in the line of animal evolution leading to the vertebrates. Sedentary echinoderms and sessile sea squirts, which have stayed the same for 500 million years, seem to be in the “specialization bind” at which Cuvier originally hinted, with form and function so well integrated and adapted to their conditions of life as to be impossible to alter. How could vertebrates have arisen from such creatures? The answer lies in their larvae, which have potentially greater behavioral and morphological plasticity than the specialized, sedentary adults. If they fail to metamorphose to the inflexible adult form, but still become reproductively mature, they have more scope for alternative behavioral and morphological changes. It might be argued that these particular specialization locks are altogether spurious, if the larvae represent the mature morphology of common ancestors, and echinoderms and seasquirts side-branches that quickly reached dead-ends. But the larvacean urochordates, which are tadpole-like swimmers, unlike their sessile seasquirt relatives, clearly illustrate the possibility of drawing back to juvenile plasticity. And their ancient relatives could then have leapt in new directions. Paedomorphic repression of multiple limbs was a particularly important step in the line of evolution that led to the insects, although the typical adult form that was finally established is curiously inflexible. Radical evolutionary change in body pattern through paedomorphosis was a distinct possibility that met no serious objections from any quarter. Nor was there any opposition to the idea that there could be add-ons—evolutionary changes later in life that would not interfere with normal development. The question of how novel body plans could evolve from the plastic larval forms was open. Geoffroy and later saltationists had suggested that the earlier an alteration in development occurred, the more radical would be the ultimate change in body form. However, this option was strongly resisted, for the reasons that Fisher gave, and because it threatened neo-Darwinist gradualism: saltations would undermine the necessity for natural selection to have acted persistently on a graded series of slight changes in order to produce its evolutionary effect. The hypothetical mechanism could only be saved by rejecting the possibility of saltations. In The Ghost in the Machine (1967), Arthur Koestler referred to evasion of the specialization bind as “reculer pour mieux sauter.” If the embryo first stepped back it might make a better leap to evolutionary novelty. The concept was known almost a century

192

Chapter 5

earlier when it was brought to the fore by Walter Garstang. A well-known example is that of the axolotl, a neotenous tiger salamander, whose sexually mature tadpole does not metamorphose because of an epigenetic change involving prolactin and thyroxine. Among salamanders there is a full spectrum of these kinds of neotenic changes. The least canalized may be reversed by temperature change; the intermediate condition, in the axolotl, can be altered with thyroxine treatment, and the fully committed condition cannot be experimentally altered. Considering the earlier diversity of large, heavily armored terrestrial extinct Amphibia, this “return to childhood” could be a character of many extant representatives of the class. Drawing from many animal phyla, Ryuichi Matsuda (1987) gives examples of paedomorphic simplification of the life cycle and adult morphology, focusing on the effects of the environment, especially where they affect egg size, and cause rapid reorganizations of metamorphoses. This is getting ahead of the story, but it might help you to persevere if you know that the historical speculations are justified by subsequent data. Time for a brief synopsis of evolutionary life cycle changes that help to loosen canalization and evade or transcend homeorhesis. For the full credit course, start with Gavin De Beer’s Embryos and Ancestors (1940), advance to Stephen Jay Gould’s Ontogeny and Phylogeny (1977), and then consult some of the later works on heterochrony, i.e., the process of changing body forms by alterations in the timing and rates of developmental processes—ideas set in motion by E. D. Cope. Here I draw largely on Gould’s treatment, which is straightforward and familiar to modern students. Adding-on to the old adult form is called “peramorphosis,” and the extension or exaggeration of particular adult characteristics “hypermorphosis.” For orthogenetic hypermorphosis that might hypothetically lead to functional-morphological instability and possible extinction I use “ultramorphosis.” Paedomorphosis is persistence of a juvenile form in the adult. Neoteny is the variant of paedomorphosis where the animal grows to sexual maturity and adult size at the normal rate, but retains the juvenile form instead of undergoing gradual or metamorphic change to the old adult form. This applies in Matsuda’s examples of talitrid amphipods and salamanders. The other kind of simplification in form is “progenesis,” usually heritable, in which the simple early juvenile form becomes sexually mature precociously, remains very small, and fails to complete development of the former adult anatomy. This is common in phyla such as arthropods and mollusks, and, as Gould deduces, such uncoupling of growth and development is a significant evolutionary phenomenon. Since the time taken to reach maturity is reduced, several generations can fit into the time formerly taken for one, although success in this epigenetic arena is congruent with the larger environment. The life history of an organism can be altered by “cenogenesis,” in which a novel phase is inserted in early development to produce a larva that might be quite different in its anatomy and behavior from the mature organism. The best known examples are

Development and Evolution

193

in insects whose larvae are maggots or caterpillars, which obtain most of the energy needed for growth, development, and gametogenesis. Finally, the life history metamorphoses back to the insect norm, and the biological functions of the “imago,” or adult, may be largely limited to reproduction. The most drastic change in life history is “deviation,” which occurs early and alters the functional morphology, and ultimately the behavior of the adult. This is no evasion of hemeorhesis, but a frontal assault. Ultimately, deviation becomes genetically fixed and characteristic of the type. But the ontogenic plasticity that I have just described suggests how deviation is accommodated when it first appears. This kind of evolutionary novelty is anathema to neo-Darwinists because it negates the explanatory power of natural selection. But if it can be demonstrated that radical deviations can be spontaneously accommodated by subsequent embryonic development, without genetic change, their argument is jejune. Gavin De Beer subsumed radical embryonic change under neo-Darwinism, but marine biologist Alister Hardy saw it differently. His father-in-law, Walter Garstang, had already proposed that ontogenic change in marine animal larvae was a means of reintroducing evolutionary plasticity, especially in the deuterostome/echinoderm/ chordate lineage.30 Hardy went further, to associate such evolution with behavioral change and genetic assimilation, in his book The Living Stream (1965).31 Hardy also thought that radical deviations were part of evolution, and in this context he resuscitated the classical evolutionary problem of vertebrate limb placement. E. S. Goodrich, who had originally explored the matter in 1913, called the shifts “transpositions.” They arise from axially arranged segments or somites, several of which contribute to a particular limb. But they do not derive from the same somite sets in different animals. The problem how limb development could be transposed can now be reduced in part to the triggering action of homeotic genes. These are the foundation of a homology that goes much deeper than the form and placement of limbs in a particular class of vertebrates. Over time, the genes themselves have duplicated and varied, and where they are expressed, limbs appear. Like the variants of all genes, they exist in every cell in the body, but are repressed where particular organs are not wanted. Limbs can potentially arise almost anywhere at the appropriate phase of development. They are placed where they function most efficiently in relation to the other body structures and behavior. But that does not explain the cause of transposition—how could an epigenetic algorithm know where and when to shift the legs? The orthodox explanation is that trial-and-error transpositions were sorted by natural selection. More heterodox is the idea that an epigenetically harmonious novelty emerges from the egg and goes about some new business in a fit manner. The best hypothesis lies somewhere between the two. These kinds of accommodations could be originally have resulted from epigenetic experiments made with limb

194

Chapter 5

structure and placement, without the necessity for improbable co-incidental correlated mutations in developmental genes. From his study of the evolution of ears, Keith Thomson (1966) concludes that it required “correlated progression,” which reinvents Brooks’s and Bateson’s version of Darwin’s “correlated variation.” Later data from molecular epigenetics have not qualitatively improved the idea. In contrast, neoDarwinism proposes that key innovations, if adaptive, change the selection pressures so that a cluster of supporting variations are finally accumulated. Variant repeats of homeotic genes that stimulate the development of antennae, mouth appendages and legs in Drosophila are expressed chronologically from anterior to posterior.32 A similar head to tail progression occurs in mammals. Therefore, as a general principle, natural experimental changes in the anterior limbs might result in compensatory changes in the posterior limbs, both in placement and size. It can be done to fruit flies in the laboratory; why not naturally? Anterior reduction of the forelimbs might result in a transposition or size increase in the posterior limbs. Behavioral plasticity is one of the key accommodations of such changes. A monster analogous to Slijper’s goat, with large hind legs transposed “too far back,” might attempt a similar balancing act of bipedalism, using the forelimbs as stabilizers or claspers. Such creatures would find their feet more easily if spared the competition and predation of natural selection. We only have to go to Australia to find such good sports—kangaroos! Limbs might go altogether. A proto-snake, for example, with increased segmentation, and small limbs too far apart to hold up the elongated body, could locomote without much need for limbs, through emphasizing the archaic sinuous movement still seen in lizards, and which derived originally from lobefin ancestors. Ontogenic responses such as scale or rib modifications could have then been genetically assimilated. Accommodation of Change in Epigenesis I have already raised the problem of how radical epigenetic change can be accommodated, and have just emphasized behavioral plasticity with regard to limb transposition. Earlier I mentioned the accommodation of epigenetic change in individual development, and dropped a few broad hints about how similar processes occurred during evolution. Comparative embryology has always suggested relationships between ontogeny and phylogeny. And this has been reinforced by current developments in the study of genes involved in epigenetic regulation. We need a more complete picture of how such large embryological changes occurred, how they were accommodated, and how the resulting hopeful monsters realized their aspirations. Schmalhausen’s synthesis of Darwinist and epigenetic thought had no difficulty with accommodation. He asserted that new differentiations were effected as the organism responded to different environments. What follows is an almost-verbatim

Development and Evolution

195

passage from Factors of Evolution which I have slightly edited (and hence italicized) to bring out the distinction between adaptation and adaptability, both epigenetic and physiological: These became integrated into a harmoniously constructed and functioning entity through regulative developmental systems and physiological adaptability. A highly plastic organization was produced, so that both advantageous hereditary variations and functional adaptability harmoniously transformed the entire organization. In the course of individual development, adaptabilities were supplemented by suitable co-adaptation of organ functions through stabilizing selection. In this way, the organism always evolved as a integrated whole and novelties did not destroy the harmony of organization. Indeed, new functional differentiations increased the heritable complexity of organization.33

That Schmalhausen adduced the latter, in addition to the effects of the environment, as parts of the whole evolutionary picture is typical of his holistic view of the natural world. The synthesis is however incomplete, since it does not specify the morphological plasticity of plants, and the behavioral plasticity of animals. Arguments to the effect that radical embryonic change will always be deleterious, raised by Ronald Fisher, Ernst Mayr (1977), and other populationists, were advanced to justify the gradualistic worldview, rather than to assess epigenetic evolution. Without question, the mutation of a structural gene responsible for an epigenetic hormone or enzyme is likely to be harmful, unless, as is often the case, there is a compensatory suite of back up genes already in situ. Yet, classical examples of accommodatory development were familiar to C. H. Waddington when he published The Strategy of the Genes in 1957. For instance, embryonic implants such as eye primordia not only induced an appropriate array of muscles and nerves, they also affected the developing brain of the animal relative to the size of the implant. He called this coordination and integration “ontogenic buffering.” Although Waddington did not discuss the specifics of loose canalization and accommodatory mechanisms, he realized that the subsidiary developmental pathways, which he called “creodes,” have some independence of homeorhesis. Some are quite plastic, for example the one that deals with the development of the endoderm, the fundamental tissue layer that forms the gut. Because it has some body space in which to maneuver, its morphospace is similarly larger, and almost any type of gut can become epigenetically modified and specialized without affecting the overall development of the organism. Other creodes of organogenesis acquired some capacity for differential development long after overall body plans had been laid down. The evolutionary explorations of migratory neural crest cells in the vertebrates that affected cranial, ear, eye, jaw, and limb structure continued to have a strong impact in hominid evolution. These give us further clues to evolutionary plasticity and accommodation in development. Nevertheless, there is still a large gap between what we think biologically possible and what has actually been demonstrated. Also the fact remains that we do not

196

Chapter 5

repeatedly see natural experiments that result in hopeful monsters raising their lovely heads in every generation. I have dug tons of clams and looked at thousands of their juveniles through a microscope without seeing a single deviant specimen, except for those that have been ontogenically distorted by an unusual substrate. But I was looking at a type that has probably been generatively entrenched for 200 million years. If I were a botanist my subjective view would be different. Some of my house and garden plants show some very distinct changes in leaf and flower color through somatic mutation. My pink-flowering cactus has just developed flame-colored blooms from a new shoot. Moreover, if, at any time, you want to see the product of a series of epigenetic changes that periodically occurred over the last 5 million years, take a look in a mirror. Whatever molecular and morphogenetic accommodating mechanisms there might be, it is obvious that the later in development an epigenetic novelty occurs, the more likely it is to fit harmoniously into the organismic whole. In fact, peramorphoses, involving hypermorphosis of established patterns of growth and maturation, and novel additions to them, are the easiest kinds of natural developmental experiments to accommodate. Going back to Riedl’s architectural metaphor; if the original plans for the high rise have been realized, more floors might be added to a different design, with hanging gardens, penthouses, satellite dishes, and helicopter pads tacked on as afterthoughts. These peramorphoses and hypermorphoses might have the potential to change the fundamental function of the building. The late development of feathers and hairs from the outermost surface of birds and mammals depends initially on simple hypermorphic evaginations and invaginations of the epidermis. But the physiological and behavioral consequences of feathers and hairs made them essential components of bird and mammal emergence. Ontogenic buffering is now almost axiomatic for evo-devos. For example, Raff and Kaufman (1983) say that plasticity of epigenesis and adaptability of regulatory mechanisms are always flexible enough to accommodate epigenetic diversification once it is initiated. Epigeneticists such as Matsuda, and Balon, also have less trouble with the accommodation of deviation than Riedl. For them, metamorphoses or developmental thresholds are nexuses of plastic rearrangement, offering various changes of direction along new pathways. Gerd Müller (1991), although he nods to the ultimate causes of neo-Darwinist theory, argues that emergent novelties arise at the transgression of threshold points immanent to developmental systems, where sequences of developmental interaction are disrupted or new interactions become established, and the resulting structure will depend on the reaction norms of the system at that point. This may directly translate into the adult phenotype or may initially produce a transitory structure on ontogeny not immediately expressed in the adult. Such transitory structures can later become expressed in the adult stage of descendants through processes of heterochrony or they can become future modified and provide a developmental basis for other morphological innovations.

Development and Evolution

197

Particular opportunities for the modification of developmental patterns exist during the transition phases of changing mechanisms in the stepwise formation of organs. Generally it can be expected that new structures arising are easily integrated into the organism through epigenetic adjustments of the associated systems.34 [Epigenetic thresholds are points of discontinuity (metamorphosis being an extreme example), where the emergences of developing structures are triggered by a variety of stimuli, some endogenous, some environmental. —RR] Novelty can thus arise as a side effect of evolutionary changes of size and proportion, with the specific result depending on the reaction of the affected systems. In this scenario the emerging structure becomes only secondarily a target of selection which will determine its maintenance and persistence throughout the population; the disruption of a morphogenetic sequence lies at its origin.35 [emphasis added]

Müller illustrates these generalizations with three categories of morphogenetic threshold change. The first of Müller’s categories arises from modification of body size. In the early stage of vertebrate skeletal development the total number of prechondrogenic, or precartilage-forming cells, and the overall size of the limb buds, affect the formation of the adult limb. The trigger for condensing the centers of cartilage formation may be as simple as a carbon dioxide/oxygen gradient.36 The number of these cells has been experimentally manipulated to show these effects. And in nature large animals are more likely to form extra digits than small animals of the same type. The second is the epigenetic effect of the relative position of organizer cells, or organ primordia in the embryo, a factor known since Spemann’s classical experiments, which imitated what in nature had been going on for 500 million years. That the site of action of morphological change is affected by earlier differential development is simply illustrated by the development of cheek pouches in rodents. These are either internally derived, lined with buccal epithelium, or externally derived, and lined with fur. A very minute position change in the point of pouch invagination makes the allor-none difference between the two types.37 This kind of late-development hypermorphosis is common and non-disruptive. But A. C. Burke (1989) argues that a similarly small modification of epithelial-mesenchymal interactions in early development resulted in the saltatory emergence of the turtle carapace, a radical anatomical structure that appeared suddenly in the fossil record.38 Simultaneous accommodation of the limb girdles within the rib cage is necessary for the new arrangement. Incipient intermediate forms necessary to satisfy the theoretical demands of selectionism may never actually have occurred, and are redundant for a theory of developmental emergence. I have already mentioned changes in cranial, facial and tooth anatomy involving neural crest cell migrations that could be saltatory in their initial phenotypic expression and also show continuing orthogenetic trends. Brian Hall (1998) reviews historical and current advances in this area of study. An organizer role for the neural crest has been demonstrated in the evolution of part of the

198

Chapter 5

autonomic nervous system, inner ear structure, and a variety of endocrine glands, as well as the better-known instances of cranial and facial anatomy. John Fondon and Harold Garner (2004) have established that at the gene level craniofacial changes involve various tandem repetitions of codons in cis-regulatory genes. The causal connection between such molecular changes and the activities of neural crest cells has not been established. One of the best examples of early deviation that was successfully accommodated by subsequent epigenesis was discovered by Rudolf Raff and his associates.39 Early heterochronic changes in the segregation of cell lineages in sea urchin blastomeres cause novel larval forms in which some primitive features are not manifested, and others appear precociously. “These and other derived features of direct developers such as changes in cleavage pattern and mitotic rates are dependent on the heterochronic changes in developmental mode and not on adaptations in the traditional sense.”40 Müller’s third category of causes of novel developmental emergences is biomechanical. As early as 1874, the German embryologist His had proposed that folds in early embryos were caused by the mechanical pressure of differential growth. Gravity physiogenesis certainly had an important impact on proto-tetrapods emerging from water onto the land. And patterns of cartilage and bone formation are strongly affected by mechanical stimulation. Tension, pressure, shear and locomotion can all bring about developmental change. In bird embryos, which begin active movement within the egg, biomechanical stimuli have a major impact on wing development and are essential to the shinbones.41 Hydrostatic pressure forces the expansion of some developing eyes and brains.42 Other triggers of threshold phenomena include hormones, morphogens, and adhesion molecules, as well as egg size and the kinds of environmental effects proposed by Ryuichi Matsuda (1982, 1987). Müller also details how transitory intermediate structures, or “interphenes,” can be introduced into development. In the nineteenth century, Haeckel divided these into “palingenetic” recapitulatory structures and “cenogenetic” embryonic novelties. As an example of the former Müller cites the extra joint in the upper jaw of bolyeriid snakes. Paedomorphic suppression prevents the fusion of two ossification centers that in earlier vertebrate evolution rose from two separate bones. In the snake this throwback morphogenesis provides a double-jointed maxilla allows easier ingestion of large prey. Cenogenetic novelties are represented by the fibular crest in the hind limbs of theropod dinosaurs and their putative descendants the birds. This structure appeared de novo through the growth of a stress-induced cartilage sesamoid. The wishbone (or furcula) in birds may also have arisen from a transitory stress-induced dermal cartilaginous structure. The same is true of the of pandas’ thumbs, extra sesamoid digits that Stephen Jay Gould (1980a) made familiar to a wide audience. Such functional skeletal elements had to be integrated with the neuromuscular system, and this process was also mediated in part by biomechanical factors involved in how the animals put them

Development and Evolution

199

to use. Søren Løvtrup argued in 1974 that the notochord was a similar saltatory innovation, and that in combination with the absence of mucopolysaccharide from the body cavities, the notochord had a major impact on the evolvability of the chordates. Müller proposes that heterochronic modification of interphenes is responsible for some epigenetic evolutionary discontinuities. Where they crop up they could be amplified, and development of the old adult phenotypic characters could be repressed or lost. The kind of switching involved in reculer pour mieux sauter effects is also addressed by Müller. Each point during development where there is a switch in the epigenetic mechanism, affords an opportunity to change emphasis or qualitatively alter direction. The vertebrate skeleton, while conservative in its fundamental morphogenetic patterning—a feature that made vertebrate homologies so evident to nineteenth century zoologists—has a wide variety of mature forms. Despite the conservatism of the system the diversification of these forms must have occurred at early morphogenetic thresholds, such as between cartilage formation and bone formation.43 Furthermore, radical morphogenetic changes need not involve molecular epigenetic novelties, but simply the redistribution of their sites of expression, sometimes through topological rearrangements of cells with crucial inductive functions. David Wake and G. Roth (1989) call this “repatterning”—your advancement may depend on how good your re-connections are. I will incorporate Müller’s other theoretical contributions to developmental evolution into “Re-inventing Emergence” and “A Theory of Emergence” (chapters 8 and 10). Still, it is appropriate here to recall the accommodation question, since he, along with Stuart Newman, has proposed that major evolutionary processes that gave rise to animal body plans were largely mechanical responses to environmental epigenetic stimuli. Genes had only a small part to play, and their significant role in developmental regulation came after the body plans had emerged.44 This implies that accommodation was unnecessary at that stage. Any experiment could be tried, and its survival depended only on simple bodily integrity, including an ability to feed. Another way of tackling the accommodation phenomenon is by computer modeling. In Making Sense of Life (2002), Evelyn Fox Keller has focused on the theories of Garrett Odell and his colleagues, who work with robust epigenetic networks in insects. Although the initiating mechanisms of segment development vary among the insects, the segment polarity gene networks are highly conserved. This suggests how there is consistent segment homology among the insects. Initial models that attempted to get all the physicochemical and simple biological parameters right did not match the natural epigenetic processes. They concluded that the robustness of a system that could accommodate parameter changes derived from the entire system: “. . . not only does the network topology embody many different solutions, but most solutions are highly robust to variation in individual parameter values.” Their next

200

Chapter 5

words are very much in tune with an emergentist view of epigenesis: “The simplest model that works at all emerged complete with unexpected robustness to variation in parameters and initial conditions.”45 The whole is greater than the sum of its parts. But although these investigations say a lot about how nature is true to type, we still have to get to the next stage of modeling a system that circumscribes how insects underwent divergent evolution despite their conserved and robust genetic segment polarity networks. We know part of the solution to this particular problem: leave the robust system in place and work around it. Let particular segments go on producing eyes, antennae, wings, legs, and reproductive appendages. The end products can themselves be enlarged or elongated or regressed without upsetting the epigenetic applecart. The lifecycle can be modified allometrically, heterochronically, physiologically and behaviorally. And segmental conservatism remains strong enough for the least experienced eye to see that it’s still some kind of insect. Not all phyla or orders share the robust conservatism of insects. Whether the mollusks escaped from segmental regimentation, or never had it at all, they are to be characterized by evolutionary plasticity instead of morphological conservatism. If the somites of vertebrates can be compared to insect segments they have undergone much greater evolutionary changes. A last word on accommodation comes from Marc Kirschner and John Gerhart’s 2005 book The Plausibility of Life: Resolving Darwin’s Dilemma. Assuming cellular molecular processes are adaptable enough to adjust to genomic mutations, subsequent epigenetic events provide further modification through “exploratory behavior” at the cellular level. For example, alterations in regulatory genes that result in potential anatomical changes in a pair of limbs can be responded to by the exploratory processes of nerve, muscle, and tendon development. Regardless of the nature of the genomic change, if not fatal, subsequent adjustments can be made at all developmental levels. There is no developmental accommodation problem if there is adaptability to changes at the molecular level. As can be deduced from the earlier portion of this synopsis that deals with Müller’s practical and theoretical observations, Kirschner and Gerhart’s treatment of the accommodation problem does not have chronological priority, but it is the most accessible of current publications. If we step back to get a broader view, we might try comparing homeorhesis and epigenetic evolution with homeostasis and physiological evolution, and discover one common occurrence. They have both undergone some disequilibration that jolted them free of their stability, enough for change to occur: not so much that the results are dysfunctional, but not so little that they return the experimental organism to its original state. That is why it is so important to study emergences that surmounted the barriers. The fixation or tightening of dynamic stabilities by the natural selection of slight adaptational adjustments (or internal selection) within the systems leads to evolutionary inflexibility. Subsystems that are epigenetically adaptable may be culled in

Development and Evolution

201

favor of those specialized to respond to the most common stimulus in only one way. Life that carries on under the scrutiny of natural selection is very different from life that has been freed from it to emerge as new forms into fresh woods and pastures new. Direct Environmental Effects and Genetic Assimilation Earlier structuralistic treatments of developmental evolution were biased in favor of autonomous changes in functional morphology that would occur regardless of environmental conditions. Epigenetic self-amplification, or orthogenesis, is an important example of such endogenous evolution. However, any organism lives in an environment that affects its life and influences the future of its descendents. In my previous chapter on evolutionary physiology I noted how homeostasis in animals was periodically reset by environmental change and physiogenesis. Schmalhausen and Waddington realized that these factors were important for developmental physiology as well, and that this provides a key to the escape of the insects and the vertebrates from the homeorhetic lockup. Did the sameness of the old marine environment contribute to the evolutionary stagnation of its denizens, while the differences of the terrestrial environment opened the experimental laboratory for new emergences? Neo-Darwinism insists that new environments provide different strong selection pressures that simply elicit appropriate innovations. But I have gone to some trouble to show how the interplay between physiology and behavior can produce internal physicochemical changes, prior to any genetic assimilation and subsequent adaptational adjustment. Just how important are the direct effects of the external environment, and the internal milieu in developmental or epigenetic evolution? Ryuichi Matsuda catalogues seventeen phyla in which there are cases of metamorphosis being subject to environmental influences in Animal Evolution in Changing Environments with Special Reference to Abnormal Metamorphosis (1987). The biochemical epigenetic processes of salamanders provide the best established illustrations of environmental effects and genetic assimilation. But arthropods offer almost as complete an inventory of known pathways. Further examples of major phenotypic changes induced by the environment in animals that undergo metamorphosis are offered by Scott Gilbert and his associates in Metamorphosis: Postembryonic Reprogramming of Gene Expression in Amphibian and Insect Cells (1996). In chapter 3, I pointed to epigenetic changes brought about by microorganisms such as those that induce the broad thallus of the sea-lettuce; nitrogen-fixing bacteria that stimulate root nodule growth in terrestrial plants; and gut commensals that influence the postnatal development of mouse guts, as well as firing up the immune and digestive systems. I also speculated that symbionts could have epigenetically effected allometric shifts in giant tridacnid clams, and bivalves with sulfur-oxidizing symbiosis. A localized internal increase in nutrients could have initiated these processes. Some of these associations may have been in existence for

202

Chapter 5

hundreds of millions of years, consistently affecting development without being determined by the genes of the developing organism. It is possible that some geneswapping has occurred, as it did between endosymbiotic organelles and their hosts. What of it? The environment induces changes. Nobody is going to argue, unless it is also proposed that these changes can somehow become heritable. Recollect that in his discussion of homeodynamics Roy Pearson notes that the endocrine system that affects epigenesis in vertebrates is already present and heritable, requiring environmental retuning to have an evolutionary impact (q.v. coda of chapter 4). But several modern epigeneticists propose a more radical neo-Lamarckist approach. How then could environmentally induced changes in the organism become genetically fixed, beyond gene acquisition from symbionts and retroviruses? It does not require the inheritance of acquired characteristics, but modification of genes, and of a variety of organismal resources that already exist. The subject has been treated most thoroughly by Eva Jablonka and Marion Lamb in Epigenetic Inheritance and Evolution (1995). To date, Mary Jane West-Eberhard (2003) offers the broadest exploration of the general topic of environmentally-induced change and organismal response. One of the processes triggered by environmental induction was called “genetic assimilation” by Waddington, but the central idea was first published independently by three different authors, C. L. Morgan, J. Baldwin, and H. F. Osborn, in 1896. The different shades of meaning and intentions of the originators, as well as those of Waddington and Matsuda are tabulated by Brian Hall (2001). His citation of the King and Stansfield (1984) definition of genetic assimilation is worth repeating here: [Genetic assimilation is] the process by which a phenotypic character initially produced only in response to some environmental influence becomes, through a process of selection, taken over by the genotype, so that it is formed even in the absence of the environmental influence that at first had been necessary.46

The originators of the concept were all intrigued by how phenotypic changes, induced by environmental changes, through behavioral changes, became heritable. For example, calluses are caused by skin abrasion in land vertebrates, yet in ostriches, and in some humans, calluses appear without any preliminary abrasion. Birds that inherit their songs seem to have followed a lineage that originally had to learn them from their parents. How did this genetic fixation take place? Darwinists rationalized such environmentally induced changes as manifestations of a pre-existent genetic propensity that was then enhanced by selection. The callus is a protective structure, so any mutation in the direction of callusing would be advantageous. But calluses do not appear randomly all over the body, to be sorted out by selection according to lifestyle needs. It is environmental abrasion that determines the appropriate pattern in the first place. King and Stansfield’s 1984 definition was phrased almost as neatly by H. F. Osborn in 1896:

Development and Evolution

203

During the enormously long period of time in which habits induced ontogenic variations, it is possible for natural selection to work very slowly and gradually upon predispositions to useful correlated variations, and thus what are primarily ontogenic variations become slowly apparent as phylogenic variations or congenital characters of the race.47

Osborn gave the hypothetical example of a human infant raised in a tree, where its propensity for clutching and turning in its feet might be enhanced, and ultimately become genetically fixed. As it happens, the males of some tribes in S. America are virtually raised in trees, where they do all of their hunting, and their ability to clutch the trunks with their turned-in feet and “distorted” big toes is quite striking. If female tribe members and infants who do not hunt were also found to have unusual feet, it would indicate that the feature has been genetically assimilated.48 In The Descent of Man (1871), Charles Darwin cited Alcide D’Orbigny’s account of exceptionally large lung capacities among the natives of the high plateau of Peru. D’Orbigny had also somehow discovered that the alveoli of their lungs were larger and more numerous than those of the dwellers of the plains. Then Edward Forbes had studied the Aymara, who lived at altitudes above 3,000 meters in the Andean altoplano. They had significantly larger trunks, but shorter femur and humerus limb bones, as well as diminutive heel projections, in comparison with lowlanders. Subsequently Forbes encountered Aymara who had lived at low altitudes, working as gold miners for two generations. Their anatomies were still distinct, though less exaggerated than those of their great grandparents. Both Forbes and Darwin attributed the bone shortening to a “compensation effect” for the larger trunk size. Darwin concluded: From these valuable observations, there can, I think, be little doubt that residence during many generations at a great elevation tends, both directly and indirectly, to induce inherited modifications in the proportions of the body.49

There is no question that living at high altitudes has ontogenic effects such as increased hemoglobin levels as well as expanded lung capacity. Has the effect of living in such an environment for many generations fixed these modifications genetically? Nowadays, possible subjects for research into human genetic assimilation are Himalayan people, such as Tibetans or Nepalese who have been displaced to the lowlands for several generations. In more recent studies of the ontogenic effects of living at high altitudes, J. L. Rupert and Peter Hochachka (2001) note that while visitors to the high Andes can accommodate to the mountain air, the local inhabitants have a substantial advantage. And although they write that the relative contribution of ontogenic variation and genetic adaptation has not been firmly allocated, L. P. Greksa (1996) had previously presented “Evidence of a genetic basis to the enhanced total lung capacities of Andean Highlanders.”

204

Chapter 5

Genetic assimilation was explored experimentally in fruit flies by C. H. Waddington and his co-workers.50 If the pupae are heat-stressed, some of the adults have a distinctive broken wing vein characteristic. A combination of continued heat treatment and selective breeding of the flies with that phenotypic characteristic results in a population that uniformly shows the change in venation. But this seems to be a repression or activation of a regulatory gene involved in the epigenetic expression of wing development. In other words, the regulatory systems were already there, some more temperature sensitive than others. The deliberate act of selection simply sorted out the most extreme cases. The adaptational significance of broken cross veins is irrelevant here except in terms of the investigator’s requirements. In some instances of ontogenic change in heat-stressed fruit flies, Waddington thought that the heat must have been mutagenic, instead of latent genetic characteristics being elicited. That comes close to defining the limit of genetic assimilation: it depends on activation of a dormant gene or mutation of a pre-existing gene or a combination of the two. However, a contingent novelty of gene or protein shuffling might also match the environmentally induced change. Since natural selection does not recognize the difference between a phenocopy caused by the environment and a genocopy established by genetic fixation, the latter has to have a little more selective value before it becomes part of dynamic stability. The universal existence of “heat-shock genes,” and their translation to proteins that have protective and regulatory effects, is now widely known, and they may affect evolvability. “Stress proteins” is their alternative name, since a variety of physicochemical factors in addition to heat may be involved. Under normal conditions they contribute to the formation of the three-dimensional structure of proteins, but under stress they are otherwise engaged, and mutability increases. (See my next chapter.) Some of the processes discussed in the previous paragraph would come under WestEberhard’s term “genetic accommodation.” Her commentary on the general properties of genetic accommodation is imbued with selectionspeak, but I will endeavor to translate. She notes that “genetic accommodation occurs whether a novel trait is mutationally or environmentally induced” and that simple statement, without the accompanying “selective regime,” is a useful starting point.51 I agree with the idea that the environment induces a phenotypic change to which the genotype accommodates; and that the genotype is in a sufficient state of experimental flux that it can generate developmental and perhaps behavioral changes. Her concept also includes the coadaptation or “internal selection” that follows emergent novelties and produces internal organismal equilibria with low energy demands, a phenomenon that I raised in chapter 4. The attempt to synthesize a variety of related phenomena is admirable, provided that the relationships are real. However, it is not always easy to remember all the special implications of a common word such as “accommodation.” And I have already used the word with different shades of meaning for both physiological and

Development and Evolution

205

epigenetic responses. The invocation of molecular adaptability as it relates to genomic change and environmental influence strengthens the overall concept (Kirschner and Gerhart 2005). For Matsuda the “proximate” event was the most interesting part of accommodation, i.e., ontogenic change caused by the environmental condition. Temperature and light are common influences on the production of developmental hormones. Nutrition is also indirectly involved. In many cases animals that enter a stressful environment, such as marine animals invading fresh water, or fresh water animals invading the land, make large eggs. This allows “embryonization” (i.e., the first stages of the life cycle occur in the egg), and the vulnerable larvae are protected from environmental stress. Ovoviviparity, where the egg is protected within the mother, and the young are born as miniature adults, is another means of avoiding stress. And the next stage is viviparity, where the developing young derive nutrition from the mother. This has emerged independently in fish, amphibians, reptiles and mammals.52 To return to the epigenetic consequences of large eggs, “secondary vitellogenesis” is often a factor. This increased yolk formation can be induced by temperature and photoperiod changes that stimulate more synthesis of the hormone “vitellogenin.” In crustaceans the hormone is mediated by light-sensitive organs in their eye stalks. Matsuda observed that when sand-hoppers—intertidal, talitrid, amphipod crustaceans—first invaded dry land they sought a damp environment and found it in leaf mould, which is dark as well as damp. Darkness affects the production of light sensitive hormones in the eye stalks, which in turn have two effects: de-inhibition of androgenic hormone production, which results in accelerated spermatogenesis and the development of the male secondary sexual characteristics, and inhibition of ecdysone production. The latter suppresses the development of normal terminal moults, resulting in a neotenous condition that has a variety of anatomical consequences in several species of these talitrids, including the loss of the swimming legs, and alteration of leg, antennal, and gill structure. That these changes can be brought about by experimental manipulation of extant intertidal amphipod species supports the hypothesis. Some of the changes in the terrestrial talitrids have, however, been genetically fixed. The behavior of transitional talitrids seeking damp, dark places is part of the proximate process. How they got further up the shore in the first place was put down by Matsuda as the result of sudden changes in the level of the sea or land or possibly mass migration. In my experience intertidal amphipods, when removed from their normal environment, usually try to head downhill toward the sea. I was once called by a bewildered neighbor who could not understand why “shrimps” were spontaneously generating in his cat’s water dish. On investigation, I found a pile of seaweed in his compost heap, from which a host of the amphipods had headed downhill, to be channeled into a cul-de-sac at his back door, where the water dish must have presented a refuge of last resort for the frustrated crustaceans. Death by

206

Chapter 5

dilution had turned them shrimp-pink. On the other hand, my colleague Tom Reimchen has observed that in Haida Gwaii (Queen Charlotte Archipelago) in British Columbia, some talitrids migrate en masse up from the shore and overwinter in grasses above the high tide mark.53 This area was also subject to large-scale regressions and transgressions of seawater levels when the last glacial period came to a close in the late Pleistocene, and continuing into the early Holocene, about 12,000–8,000 years ago. Environmental stimuli may have ontogenic consequences that are coincidentally advantageous. The polymorphism of shell markings and color in land snails provided a classical problem for neo-Darwinism. How could such striking differences in banding patterns and colors all be adaptational? In the common land snail Cepaea nemoralis, which in France is a popular escargot, the background color can be green, yellow or pink, and the black banding may be absent, or very narrow, or so wide that the shell appears almost uniformly black. The original investigations by Arthur Cain and Philip Sheppard (1950, 1954) inferred that the different color morphs were adapted to local microenvironments. (The term “morph” can refer to color, color pattern, anatomical structure and size.) They came at the reverse of the problem by analyzing the remains around thrush anvils—stones that the gastronomic birds used to break open their snails. The thrushes most commonly caught the type that was most visually conspicuous in the local microenvironment. So the investigators concluded that the mosaic of different environmental conditions, such as dark woods, open areas, long grass, and short grass, had disrupted the selection process and produced polymorphism. But how is it you find non-adaptational coloration? In Donegal, in northwest Ireland, I have noticed that the pink color of Cepaea makes it stand out in the seashore grass from many yards away—plenty of potential anvils, but no thrushes perhaps? And to what extent is or was the microenvironment directly responsible for these differences? If a similar snail, Helix aspersa, is raised in the dark, it has a dark shell; if raised in occasional light, it has a paler colored shell. The ontogenic effect of the light stimulus coincidentally proffers advantageous protective coloration in both cases, and natural selection, even as a post hoc cause is redundant. To what extent does Cepaea respond ontogenically to its microenvironment? No one will bother to look if they find the disruptive selection story adequate. Several unrelated species of marine snails show similar polymorphism. I have on my desk a jar of the shells of marine gastropod Nerita chamaeleon. They were all collected from the same square meter of sandy shore on Kat O Chow in China. Yet the shells vary far more in color and banding patterns than do those of Cepaea. And from the Jervis Inlet area of New South Wales I have a vial of top-shells that are as varied as mixed humbugs in their colors and stripes. Thus, striking polymorphisms of no adaptational significance may arise de luxe in uniform environments. Cor van der Weele cites a number of cases of environmental induction in her Images of Development (1999). Caterpillars of the moth Nemoria arizona change color

Development and Evolution

207

according to their diet. In the spring they eat oak catkins and are catkin colored. Later hatching caterpillars eat oak leaves and take on the color of the twigs of the tree; it all depends on the amount of tannin in the diet. Changing photoperiod determines whether a light or dark morph is produced in the comma moth. Araschnia levana is a butterfly that comes in different color and pattern morphs, one of which, to add to adaptationist confusion, looks like a map of Europe. Which morph develops in Araschnia depends on the wetness of the climate. Ambient temperature at the end of the larval stage determines the presence or absence of eye spots on the wings of Bicyclus butterflies in Malawi.54 The same author gives examples of other groups including mollusks and vertebrates where diet and physical stimuli cause developmental change. Some of these effects come down to a few facultative phenotypic options that were genetically fixed in the past, so they exemplify the action of environmental switches affecting existing genotypes. Others are like the color of leaf-eating caterpillars. To take an example from my own research field, marine biology, dog whelks may be white if they eat barnacles, brown if they eat mussels, or banded brown and white if they eat both. No doubt they have been assigned selection coefficients that “explained” the differences. The fact remains that the morphs were caused by the environment, would have happened regardless of selective value, and only secondarily were demographically distributed according to degrees of predation. Stuart Newman and Gerd Müller (2000) add the following examples of environmental character determination. Candida albicans, known to sufferers of this fungal infection as “thrush,” varies according to its environment from single cells, or budding strings like yeasts, to septated filaments. And they have no “default” morphology.55 Incubation temperatures determine sex in reptiles: high temperatures produce male lizards and crocodiles, but female tortoises.56 Embryos of a strain of mice that normally have five lumbar vertebrae develop six vertebrae if transplanted to the uterus of a surrogate mother with six.57 Evolutionary epigenetics are often treated as if their emergences were autonomous, or intrinsic, all to do with progressive, internal complexification without regard to the larger environment. Quests for unifying physical principles that will explain complexification at all levels fall into this category, as do structuralistic analyses. Cor van der Weele believes that the internalist/structuralist position in evolutionary developmental studies abandons too much ground to ecological adaptationism, and favors enlarging epigenetics to emphasize environmental factors. But it had already been attempted by Matsuda’s synthesis of a neo-Lamarckist antithesis with a neo-Darwinist thesis. He put the environment as a causal agent back where it belongs: in and around the organism. Eugene Balon also transcends simplistic structuralism to view ontogenic emergences as optional responses to new environmental exigencies.58 He points out that there are critical stages in the life cycles of animals where there is more than one direction to choose from. Out of these choices come environmentally cued, facultative

208

Chapter 5

phenotypic expressions. The final result may be a change of direction in development, with the old phenotype lost through disuse. The radical reculer pour mieux sauter regression to larva followed by a new line of hypermorphosis has taken such a route, and in a number of cases either the paedomorphic type or the old mature adult form can be induced. Balon’s work will be addressed more fully in chapter 8. Another aspect of environmental influence on epigenetics requires attention. MaeWan Ho (1984) has taken a particular interest in the significance of phenocopying, whereby the environment can bring about ontogenic, or phenotypic changes that mimic those produced by particular genotypes. The phenomenon was detailed by Richard Goldschmidt (1940), and subsequently Peter Medawar (1951) came up with the complementary term “genocopying” to signify the genetic assimilation of an environmentally induced phenotypic condition. Ho is convinced that the universality of phenocopying indicates that it is a common prelude to evolutionary change. Reinvestigating some of the earlier genetic assimilation experiments she concludes that the continued amplification of phenotypic effects from one generation to the next cannot be fully explained by the modification of regulatory genes nor the mutation of structural genes. She therefore proposes that phenocopying is mediated by the ultrastructure of the cytoplasm, and that environmentally induced cytoplasmic changes can persist through the egg to the next generation, to be amplified if the environmental conditions persist. Susan Oyama also understands the need to accommodate non-DNA influences both within the cell and whole organism, and to bring environmental influences into the causal interactions. In The Ontogeny of Information (1985), she writes: What we are moving towards is a conception of a developmental system, not as the reading off of a preexisting code, but as a complex of interacting influences, some inside the organism’s skin, some external to it, and including its ecological niche in all its spatial and temporal aspects, many of which are typically passed on in reproduction because they are in some way tied to the organism’s (or its conspecifics’) activities or characteristics or because they are stable features of the general environment. It is in this ontogenetic crucible that form appears and is transformed, not because it is immanent in some interactants and nourished by others, or because some interactants select from a range of forms present in others, but because any form is created by the precise activity of the system.59

Oyama goes on to conclude that the genome can explain neither epigenesis nor the causation of epigenetic change, any more than physics can explain the nature or change of the genome. Since Oyama’s book first appeared in 1985, molecular biology has made major inroads into epigenetics. The actions of Hox genes have created enormous interest, and given epigenetics a “scientific” respectability that it previously lacked. At the same time, molecular biology has pulled epigenetics into a tight genocentric orbit. Only in the last few years have the kinds of “eco-evo-devo” studies and assessments that support Oyama’s original position been forthcoming—see, for example, Hall, Pearson, and Müller 2003.

Development and Evolution

209

In my next chapter, I will enlarge on the significance of other studies of non-DNA epigenetic events, and also summarize the advances that have been made in molecular epigenetics and their significance for evolutionary developmental biology. I omit most of them from this general chapter to keep the broader issues simple. But when we turn to epigenetic mechanisms it will be more obvious that Ho and Oyama belong to a historical tradition of dissent that has quite strong ties with neo-Lamarckism. As Walter Garstang remarked, “Ontogeny does not recapitulate phylogeny, it creates it.”60 And the title of neo-Lamarckist F. Wood Jones’s 1943 book Habit and Heritage speaks for itself. When I began this work I was well aware of how the environment influences physiological change. But I was still biased in favor of autonomous emergences that gave organisms the opportunity to undertake new experiments in behavior through improved adaptability. In the conventional sense they allowed the organism to better “persist in its own being” in a greater variety of environments. Gradually it dawned on me how the organism’s actions determine what is adaptive in the conventional sense. Although they don’t call it genetic assimilation, Stuart Newman and Gerd Müller (2000) improve on the idea in their essay “Epigenetic mechanisms of character origination.” They put it as follows: The close mapping between genotype and morphological phenotype in many contemporary metazoans has led to the general notion that the evolution of organismal form is a direct consequence of evolving genetic programs. In contrast to this view, we propose that the present relationship between genes and form is a highly derived condition, a product of evolution rather than its precondition. Prior to the biochemical canalization of developmental pathways, and the stabilization of phenotypes, interaction of multicellular organisms with their physicochemical environments dictated a many-to-many mapping between genomes and forms. These forms would have been generated by epigenetic mechanisms: initially physical processes characteristic of condensed, chemically active materials, and later conditional, inductive interactions among the organism’s constituent tissues. This concept, that epigenetic mechanisms are the generative agents of morphological character origination, helps to explain findings that are difficult to reconcile with the standard neo-Darwinian model, e.g., the burst of body plans in the early Cambrian, the origins of morphological innovation, homology, and rapid change of form. Our concept entails a new interpretation of the relationship between genes and biological form.61

Thus, Newman and Müller take us out of the genocentric interpretation of autonomous emergence. There the organism simply gets an interesting innovation through genetic mutation or shuffling, and then figures out something useful to do with it—in other words, function following form. In contrast Newman and Müller place us in a realm where extrinsic causes modulate intrinsic self-assembly mechanisms. Behavior and the direct impact of the environment determine what is adaptive, and what specialized adaptations will then arise. They do not however present it as an either-or choice of intrinsic or extrinsic. And before we get too excited I have to remind myself, and you, that these evolutionary processes depend on the

210

Chapter 5

organism having the “constituent tissues” capable of responding to “later conditional, inductive interactions.” Newman and Müller do not miss this point. But, to them, it is based on “an inevitable feature of the viscoelastic cell aggregates that constituted the first multicellular organisms.”62 Superficially, that might seem hylozoic, and it underrates the physiological and behavioral adaptabilities that took metazoa into new environments in the first place. In addition to focusing on the “pre-Mendelian” evolution of body plans in simple plastic metazoans, Newman and Müller also scrutinize the final consequence: morphological adaptation, which is popularly identified with evolution. We are mesmerized by the evolution of whales from hippopotamus-like creatures, by the massive predatory dinosaur Tyrannosaurus rex, and by Cretaceous crocodiles that were big enough to have him for lunch. Clearly, their evolution is squarely in the epigenetics arena. But we might forget that progressive evolution of physiological and behavioral adaptabilities made them possible. And in emphasizing that, we should not lose track of how physiogenesis and epigenetic changes contributed to further physiological evolution. (Physiogenesis, as I have to keep reminding my physiology students, even at the end of the course, is the imposition of physicochemical change by the environment on the organism—the expression came originally from E. D. Cope. And while I am temporarily digressing to deal with definitions, I should point out that, like Løvtrup, Balon, and me, Newman and Müller use the broadest definition of epigenetics as anything that affects development and its evolution. They not only include non-heritable mechanisms and processes, but give them primary importance.) We will take the origination of body plans and adaptational forms into account in the next section of this chapter. But, before we continue, Newman and Müller’s theme is worth re-emphasis: simple, primitive, multicellular organisms were more plastic, and responsive to epigenetic influences than complex organisms with mechanisms that buffer morphogenesis and homeostasis. Epigenetic causes were not gene determined, but were physiogenic, and thus contingent—they may or may not have acted; it depended on the circumstances. Consistent behaviors and contingencies would have led to consistently altered morphogenesis, and only then would the linkage between phenotype and genotype be established. Then the genome would have been able to co-opt the morphological outcome of development. Ontogeny does not recapitulate phylogeny, it creates it. Thus, they conclude that evolvability, at least in terms of large innovations like the emergence of different body forms decreases with time. Genetic determination obstructs it. It’s a pattern that keeps repeating itself. For example, from a primitive behavioral plasticity, insects and birds have evolved distinctive, genetically-fixed behavioral patterns. The argument is logical and can be demonstrated with tangible examples. It parallels, but gives a stronger evolutionary tone to Waddington’s image of the epigenetic landscape. The ball rolls down the hillside, bouncing from rut to rut, but

Development and Evolution

211

finally canalized in a deep channel at the bottom. But this reduction of primitive evolvability can be accompanied by the kind of evolvability that comes with increased complexity and adaptability. That is a general feature of progressive vertebrate evolution, something that Waddington recognized as an escape from canalization. Birds may have become “less evolvable” behaviorally, but physiological and behavioral adaptability in lineages of the placental mammals allowed radical experiments in morphogenesis. Moreover, the jack-of-all-trades qualifications of hominids allowed them to escape genetic co-option of behavior. Reaching a Better Understanding of Developmental Evolution In this chapter I have taken a roughly chronological approach to the history of general concepts of developmental evolution, with occasional sallies into some of the better known mechanisms. We know little enough about epigenetics in living organisms, far less fossil organisms, or unrecorded missing links, if such ever existed. Therefore it is difficult to begin with aboriginal unicellular organisms and provide a comprehensive history of epigenetic evolution. To seek to construct such a history is a worthy goal. Yet some of my students don’t agree. They are indifferent to historical musings about the course of epigenetic evolution, and to their errors and digressions. And they are skeptical about speculation and theorizing based on inadequate data. “Just give us the facts—and the textbook page numbers!” One approach would completely satisfy them. First, work out the developmental processes and epigenetic mechanisms of a single organism. Then, to understand evolutionary change, construct hypotheses that could be empirically tested by modifying the molecular and cellular mechanisms of epigenesis in the fully understood organism. Hence I recommend to them Walter Gehring’s Master Control Genes in Development and Evolution: The Homeobox Story (1998), which also provides a bonus educative insight into the nature of scientific discovery. Despite my doubts about genocentrism, homeotic gene function has provided one of the greatest success stories in modern evolutionary studies. (See the following chapter.) Yet it also illustrates the limitations of gene analysis for understanding evolution. More to the point, Gehring makes frequent reference to one of the best exemples of my students’ ideal, the work of Sidney Brenner and his associates on the tiny nematode worm Caenorhabditis elegans.63 It has about 20,000 genes, and 1,000 cells at maturity, and the invariant course of development of every cell from the fertilized egg has been accounted for. But as Gehring readily points out, the fate of every cell depends on the fates of other cells, and only a few of the determining factors in the cytoplasm of particular cells have been identified. In the development of the worm’s vulva the role of every cell—induction, inhibition, backup, or other supporting roles— has been discovered. The number of inducing molecular signals has been

212

Chapter 5

determined.64 But at the time of Gehring’s review only two genes and their protein products involved in epigenesis had been identified. And those genes no doubt have a number of other essential intergenic effects. The very success of research into cell fates in C. elegans underlines the difficulty of understanding the role of genetic and cellular interactions in development. Gehring’s remarks concerning the determination of eyes in Drosophila also reinforces the nature of the problem. One of the key genes for eye development in the fruit fly is paradoxically called “eyeless”—it originally referred to a mutant condition. The gene has been found to be a homologue of others that help to determine the eyes of mammals, mollusks, and flatworms. “Discovery of the Gene For the Eye!” is the kind of newspaper headline that greets findings of this nature. But there are other genes for eyes—about 2,500 of them in Drosophila, by Gehring’s account.65 The surprising conclusion of this research is that eyes, or at least the fundamental mechanisms of eyedevelopment, may have emerged only once, regardless of whether they stopped at the simple eye spot stage, or diverged and progressed to camera eyes, or compound eyes. Previously, many biologists, including myself, had agreed with Ernst Mayr that eyes could have evolved independently about 40 times. Hence, if it were all that likely, it could hardly be all that difficult, and Darwin’s worry about the evolution of such complex structures could be assuaged. Now all we have to do is to figure out the smallest number of exons or whole genes that could interact with each other, together with the smallest number of cells that need to be organized in a functional sequence to affect differential gene expression and determine the final complex product. The epistemological lesson, dear students, is that to progress in such understanding we have to stagger from speculation to hypothesis, and to experiment. Then, when we find that we cannot go there from here, we have get lucky, or lurch back to more speculation about alternative routes. Development and the Progressive Evolution of Complexity Explanations of the earliest origins of organisms that undergo differentiation and integration during their development lie partly in the causal arena of symbiosis. I have already touched upon the emergence of eukaryotes, and then sexual and multicellular associations, in chapter 4. At that point epigenesis became seriously involved with evolution, although some signatures of the next level of emergence are discernable in eukaryotic unicells, as will be seen in the next chapter. Multicellularity Let us return briefly to early fundamental concepts of differentiation. Erasmus Darwin and Herbert Spencer had intuitions that a primordial filament or a unicell was subject to differential physicochemical stimuli. Before they assembled themselves in multicellular units, the unicellular progenitors of metazoa already had the contractile

Development and Evolution

213

microtubular filaments needed for locomotion. They also had cell surface molecules that could hang on to a physical substrate as well as other cells. Even simple protein stickiness, modulated by the ionic content of the environment or the cytoplasm, is enough to account for the earliest stages of multicellularity.66 The differentiating effects of form were increased as soon as a three-dimensional structure was achieved. Some of the processes that form gastrular stages in metazoan developmental evolution could simply have been differential adhesion.67 Thus, the lack of rigid genetic control of surface protein adhesiveness provided for plasticity of form that was initially determined by non-heritable epigenetic conditions. If the first multicellular organisms were merely one-cell-thick mats, they could have had some differential adhesion, and they were subject to different influences on top and underneath, and except at the margins each cell was surrounded for the first time by other cells. Or the first multicells may be represented by the ediacarans of the Vendian period, which in all probability had phototrophic, chemotrophic or mixed symbioses. Mark McMenamin (1998) has proposed that the vendiobionts were derived from protoctists and consisted of multicellular forms dictated by the number of cell types in the founding organism, which would range from one to four. Each embryonic stem cell gave rise to semi-autonomous “cell families.” Thus, there were one to four lineages which cohered to one another but lacked gap junctions or a fluid-filled body cavity that would have allowed sufficient intercellular communication for the activities of the cell families to be coordinated. A collagenous glycoprotein coat around the cells could have provided enough of a sticky, durable material for organismal integrity, and to allow their ultimate fossilization. McMenamin infers an asexual budding process, each embryo containing each of the different cell types. His hypothesis provides grounds for explaining not only the different forms, but also their taxonomy. In a broader theoretical sense it also explores how nature first tentatively experimented with multicellularity, and how hopeful monsters could diversify in the absence of competition and predation, and what limitations are imposed by lack of intercellular communication and sharing of resources. But predation was about to come, from the emergent wave of Eumetazoa with hungry habits. The first eumetazoan may simply have been a sessile mat that had finally overcome the limitations of the vendiobionts by acquiring gap junctions between the cells. Those would have immediately have improved intercellular cooperation. Conceivably such an organism could have had the same kinds of symbiotic nutrition as the ediacarans. It could have sat still, absorbing particulate matter that settled on its upper surface, or it might have crept over bottom substrates, browsing on biofilms, its ventral surface specialized for phagocytosis. A blastula-like ball of cells might be the next step after a sessile mat, being a physically stable configuration for a single layer of cells released from a flat surface. But how would such an organism make a living?

214

Chapter 5

Blastulae have an independent existence as the early embryonic stages of many marine larvae, but they are non-feeding, and possibly cenogenetic novelties that never existed as primitive, mature organisms. Functional spherical forms are found in colonial flagellates like Volvox, but photosynthesis provides most of their needs and they do not require the differentiation and integration that the simplest heterotrophs possess. Nevertheless, a blastular structure, whose cells cohered with adhesion molecules, and which then developed tight junctions to isolate the internal environment, and gap junctions to allow intercellular communication, would have had differential impact on the constituent cells. Willmer (1960) showed that some protists may have ciliated or amoeboid forms depending on the ionic conditions of their environment. And the ubiquitous presence of ion pumps, already possessed by unicells, resulted in ionic differences between the interior space and the outside environment of a simple blastula. Goodwin (1989) raises these points with reference to the “basic conundrum” of how organized multicellular organisms ever came into existence: One of the most significant steps in the emergence of greater organismic complexity was the evolutionary origin of gastrulation. The transformation of a hollow ball of cells into a multilayered structure, with the consequent combinatorial potential for reciprocal inductive interactions leading to divers patterns of cell differentiation, stands out as a major event in the evolution of the metazoa.68

To make a living, the blastuloid organism could have flattened out over the substrate, continuing to feed as did the ancestral mat, with only the ventral layer specialized for deposit feeding. Adhesion in the constituent cells is polarized, i.e., parts of them stick easily to other cells and other parts remain free to be motile. The polarization makes it easier to form internal compartments.69 Cells that received the largest share of food might have grown differentially, causing the invagination of part of the ventral surface to create a simple gut, the endoderm of the future, with dorsal cells migrating to the interior as the mesoderm of the future. The planktonic blastula could have been an emergent cenogenetic addition to the life cycle that would enhance its dispersal as well as providing a new threshold for epigenetic divergence. A potential for epigenetic differentiation already existed in the cytoplasmic heterogeneity of unicells. A significant role for gradients of concentration of morphogens—bicoid proteins—in the zygote and early embryo was pinpointed by Christiane Nüsslein-Volhard and Eric Wieschaus in 1980, and effectively developed by Nüsslein-Volhard’s 1996 Scientific American essay “Gradients That Organize Embryo Development.” Her generalizations on the importance of gradients are as follows: Some morphogenetic gradients apparently yield but a single effect: if the concentration of the morphogen in a particular place is above a critical threshold, a target gene is activated: otherwise it is not. In other cases different concentrations of morphogen elicit different responses, and it is

Development and Evolution

215

this type of gradient that is most important for providing an increase in the complexity of the developing organism. Although each morphogenetic gradient seems to control only a few target genes directly, interactions between co-factor moleculars that affect transcription can radically change responses to the gradients. These mechanisms of combinatorial regulation open the way to the formation of patterns of great complexity from an initially simple system.70

Nüsslein-Volhard catapults us forward in time to where her research organism Drosophila has a morphological complexity far in advance of the blastula stage that we are examining. But between the heterogeneous unicell and the complex differentiated metazoan such gradients existed, and became important factors in epigenetic evolution. Appropriate conditions for multicellular morphogenesis came from the blastuloid structure, which originally formed a dynamically stable configuration mechanically, rather than by naturally selected genetic instructions. This does not entirely get over the problem raised by Buss (1987) to the effect that a ball of ciliated cells would have had the problem of not being able to undergo cell division, so could not reproduce. But, any cells that intruded into the interior of the blastuloid, and were subjected to a different ionic environment, might not have developed cilia, but would have been available as stem cells for regeneration, reproduction and new experiments in evolution. And when the blastuloid emerged spontaneously, inner migration of cells could have been instantaneous. The Ediacaran fauna were possibly the first wave of relatively simple blastuloids, although, according to Seilacher and his co-workers (1998), there is fossil evidence for more advanced triploblastic embryos during the Vendian. I agree with them that the diversification of triploblastic body plans could have been rapid, however, it does not follow that it proceeded immediately. Newman and Müller (2000) clarify the contrast between Darwinist and emergentist interpretations: A novel implication of this interpretation of the burst of forms during the early history of metazoan life is that the disparate organismal forms would have been achieved with no requirement for competition or differential fitness. Since function would follow form, rather than the other way around, the pre-Mendelian world would thus also have been, in this sense alone, a “pre-Darwinian” one.71

However, the presence of competition, even from Vendian quiltazoids, could seriously obstruct their diversification. It is certainly unnecessary to invoke the nonexplanation of selection-pressure-for-gastrulation to satisfy the need to maintain locomotion and reproduction. Also, as an erstwhile digestive physiologist, I find it odd that theoreticians so easily ignore the maxim that an army (of cells in this case) marches on its stomach. Brian Goodwin had already attributed a large part of the process of gastrulation to its inherent physical stability which parallels the attractors of the non-living systems

216

Chapter 5

that chaos theoreticians study. Although non-living systems do not have the biological ion pumps necessary to Willmer’s explanation, the carrier proteins of such mechanisms have a self-assembly component as well as essential biological components, including the primary structure of the proteins, the enzyme that energizes the system and the biosynthesized fuel molecules. That the blastular internal milieu is ionically distinct from the outside environment is due to a combination of mechanical and biological properties. Goodwin saves his case by noting that while he is trying to find reductive explanations for complex systems, all levels of organization have to be taken into account in understanding ontogeny. Body Plans The evolution of body plans is an essential part of the history of developmental evolution, but I do not have the morphospace to explore it thoroughly here. I will take it up in a future work that will go more thoroughly into evolutionary history as seen from an emergentist perspective. In the meantime, Rudolf Raff’s The Shape of Life (1996) and Wallace Arthur’s Origin of Animal Body Plans (1997) provide adequate accounts. Although the distinctive body plans of the marine animal phyla may have appeared very rapidly in the early Cambrian, their tenacious stability has depended largely on the establishment of strongly canalized homeorhesis. But along with it came some reduction of evolvability. Even the simple anatomies of polyps and flatworms have been canalized to the point of intransigence. Paradoxically, although it took longer for the basic fishy vertebrate plan to emerge, vertebrate developmental evolution kept on progressing in fits and starts for more than 400 million years, while most Cambrian animals, with the other obvious exception of arthropods and the lesser example of mollusks, stayed stuck in the mud. This supports Brian Hall’s (1999b) contention that the neural crest is a major generative condition for the emergence of vertebrates and their continued evolution. This distinctive embryonic ectodermal structure appeared early in the craniate lineage. Before it appeared, some migratory cells, especially the neuroblasts, helped to modify body plan. But the evolutionary versatility of the neural crest justifies Hall’s claim that it is an emergent, fourth germ layer. Along with duplications of the whole genome in the early vertebrates, and further duplication and differentiation of genes that regulate development and physiology, the neural crest provided powerful experimental tools for emergent evolution. During the evolution of fish, amphibians and reptiles there were anatomical experiments, often involving numbers of vertebrae, the limb transposition early noted by Goodrich, and the arrangement of fin-rays and digits. Among the reptiles, for example, contrast the forms of turtles, plesiosaurs, ichthyosaurs, dinosaurs, pterosaurs, and snakes. These could be generated by changes in homeotic gene expression. They could also be effected by changes in cellular interactions, such

Development and Evolution

217

as the migration of neural crest cells, and by heterochronic shifts. But behavior must also have been important in determining what changes would be relevant to the animals’ way of life. This version of vertebrate evolution receives some support from Newman and Müller’s view of epigenetics in the later “Darwinian phase,” when genetic assimilation had resulted in consistent trueness to type. But again they point to a continuing role for physical, particularly biomechanical epigenetic influences, even after the genome had co-opted anatomical determination. Morphogenesis is basically a process of organizing the embryo’s differentiated cells and their secretions into a variety of shapes, such as rods, balls, and tubes with various degrees of rigidity or flexibility. These then have a feedback effect on gene expression. For example, a migrating neural crest cell might encounter a new obstacle. At that point it could stop and change its usual role. Or it could go round the obstacle and induce a new set of cells to change their differentiation. It doesn’t stop there, because changes in its surroundings will again alter the final morphogenic goal of the neural crest cell. But neural crest cells, despite their striking epigenetic role, are not the be-all and end-all of vertebrate epigenesis. Skeletons can be built because of the ability of mesenchymal cells to group together and make cartilage. In a developing limb the shape of the skeleton is determined by various interacting factors. These include spatial constraints, and differential cellular adhesions that are modulated by the differential expression of genes. Diffusible chemical growth stimulants and inhibitors can induce periodicities that generate the pattern of repeated cartilages and bones found in digits and spinal columns.72 When the foundation of a skeleton has been laid, biomechanical stimuli are produced by the contraction of the rudimentary muscles. They consistently affect the shape and operation of joints, and influence innervation, and vascularization. In addition to participating in “normal” skeletogenesis, the ability of connective tissues and tendons to form cartilage and bone in response to mechanical stimuli can generate novelty. Mesenchyme cells in tissue culture are known to arrange themselves along stress fields. Depending on the density of its cells, mesenchyme will make cartilage under the influence of stretch and compression. Emergent structures that come from such interactions include a variety of sesamoid bones. There are genes for the structural proteins, growth factors and enzymes that participate in bone and cartilage synthesis. But there is no gene-for-sesamoids. While they are ultimately integrated into the skeleton their generation is initiated by the mechanical stimuli of embryonic movement. Several examples are known from bird embryology.73 The novel “sail battens” found in the wings of the fossil flying reptile Coelurosauravus jaekeli probably resulted from spontaneous bone formation in the pattern of folds generated by the movements of the embryos before they hatched from their egg.74 Although these prominent wing-stiffeners are only known from a single species, there

218

Chapter 5

is a universal bone unique to pterosaurs, the putatively sesamoid pteroid bone. It extends from the wrist region toward the shoulder, but is not a digit. Most of the leading edge of the wing is made rigid by the extended fourth finger, but between the shoulder and the wrist there is only the wing membrane. Thus, the pteroid bone probably enhanced the aerodynamic quality of the wing by eliminating flutter in a region of the wing especially crucial in the giants Pteranodon and Quetzalocatlus that had wing spans of up to 12 meters. The existence of analogous structures that do the same thing in organisms with no close phylogenetic relationship can arise from the fundamental properties of mesenchymal tissues to respond in the same way to the same mechanical stimuli. Anatomical homologies that do indicate phylogenetic relationship were recognized as evidence for evolution by the nature philosophers of the eighteenth century, long before Darwin incorporated the concept into his historical theory. The perennial example is the vertebrate forelimb, which diversified into a variety of legs, wings, flukes, spades, and hands, all with the same underlying anatomical pattern. In combination with the deep homology of homeotic genes, generic epigenetic processes, (i.e. those that respond to physicochemical effects in the same way, regardless of phylogenetic closeness) originally produced both analogues and homologues. Various recent theoreticians, especially in their contributions to Homology (1994), edited by Brian Hall, have recognized that here lies one of the fundamental secrets of evolution. Again, Newman and Müller have run the furthest with the idea. They propose that once the extrinsic causes of morphogenesis were genetically assimilated these “Mendelian” organisms became more determinate in their epigenesis. The goals of what Waddington called “creodes” became more fixed, and reliable dynamically structural modules led to the diversification of homologues.75 These gene-based functional-morphological patterns became more important than the original generic extrinsic causes. Newman and Müller write: This means that although homologues may first arise by the same epigenetic processes that produce homoplasies [analogues], they eventually become independent of their underlying molecular, epigenetic, and generic constituents and increasingly play an organizational role in morphological evolution. They take on a life of their own and are thus inherited as structural units of morphological organization, not tied to any particular generative process. Homoplasies reflect the origin of morphological innovation in the generic material properties of tissues—they are an echo of the pre-Mendelian world. Homologues, in contrast, act as formal “attractors” of design, around which more design is added.76

I have italicized a sentence here to bring out its relevance to emergence. Homologues are new patterns with new rules of action that are consistent with the underlying rules of genetic determination and response to physicochemical changes. But they have a new emergent property: “they take on a life of their own.” The epigenesis of vertebrate forelimb homologues does not follow an identical algorithm in all instances. The same goal can be achieved by a diversity of routes, even in phylogenetically close organisms.

Development and Evolution

219

Closing Note on Evolvability While genetic determination may have supplanted in large part the former exclusive role of generic epigenetic causes, homologues had a life of their own. An increasingly sophisticated physiological adaptability both buffered the experiments that they could try, and permitted new behaviors that would encourage them to succeed. With the emergence of mammals, the numbers of vertebrae had been reduced, their proportional allocation to the spinal regions made consistent, and limb positions relatively fixed. Like most other mammals, humans and giraffes have the same number of neck bones. But this anatomical stability did not limit evolvability, which could continue to experiment with allometric growth shifts and accommodations. In their early evolution, mammals were also exploring how far their new emergent physiological adaptabilities could take them. This had a bootstrapping evolutionary relationship with their development, allowing them to try out habitats and new ways of living that had previously been off limits. It was humans who were the most adventurous. The Story So Far I conclude this discussion of development and evolution with a summary of the significance of embryonic change and the problems of developmental escape from homeorhesis or epigenetic stasis, together with some recapitulatory remarks about the physiological and associative arenas. This, along with the following chapter’s summary and the completion of the field-trip checklist in chapter 7, is a preamble to chapter 8. 1. Before the developmental evolution of multicellular organisms, cellular complexification was achieved by endosymbiosis and the sexual association of eukaryotic unicells, followed by primitive experiments in multicellularity. 2. Once the foundational multicellular association had formed and emerged to the level of complexity of a gastruloid, developmental evolution became possible. It initially involved differential adhesion, and experiments with body spaces, and simple structural patterns that were responses to extrinsic and intrinsic epigenetic stimuli. Ultimately epigenesis came to involve genes and their regulation. This was then made more variable by repetitive differentiation of molecules and cell types, coupled with reorganization, integration and regression. 3. “Epigenetic algorithm” is easier to say than to understand. “If this, then that” depends on the environmental circumstances (“this”) as much as the differential responses of the genome (“that”). Moreover, the genome cannot contain the algorithmic program. Epigenesis depends on interactions between this and that.

220

Chapter 5

4. Developmental changes result in diversity of form, and affect the physiology of the mature organism by the alteration of existing cell lineages or organ systems, or the emergence of new ones. 5. Developmental/physiological emergences are often multifunctional, which may be overlooked if only “key innovations” are sought. Furthermore, in additional to several immediately useful emergent properties there may be features that will increase the evolvability of the lineage. 6. Modification of existing organ systems involves heterochrony and allometric growth shifts that may have a component of genetic drive that causes change in a particular direction, regardless of fitness. This being the case, then the heresy of orthogenesis needs to be resurrected as a legitimate embryogenic concept. 7. When basic body plans of the animal phyla were constructed in the early Cambrian, homeorhesis was weak. They did not all necessarily originate simultaneously, but they all “took off” at the same time, due to the weakness of competition and predation. Natural experiments that resulted in successful emergent body plans are best placed on a time scale commensurate with the duration of the life cycles of simple marine animals, rather than on a geological time scale. 8. Internal selection or physiological coadaptation over geological time has established developmental dynamic stability (= homeorhesis). This constraint on body plans has occasionally been surmounted by animals that remained confined to the marine environment, such as crustaceans, mollusks and primitive fish. 9. Emergence from the sea to fresh water and terrestrial environments initiated progress in the developmental evolution of plants, and provided direct environmental effects on the epigenesis, and mature physiology of animals. These were then internalized in varying degrees by genetic assimilation. More flexible behavior increased the feedback between environment, physiology and development. 10. Although homeorhesis affects terrestrial animals and plants, there have been enough environmental disequilibrating effects to induce or encourage experiments in evolutionary progress to greater complexity, and enough related disruptions of ecostasis to help establish novel emergents. 11. Bursts of diversification of successful emergents accompany the relaxation of natural selection: a. in the wake of catastrophe.

Development and Evolution

221

b. in the invasion of pristine environments. c. on the re-invasion of old environments after a period of progressive evolution in new ones (e.g. teleosts returning to the sea). 12. Lamarck was right about the importance of the individual, and the way it can affect evolution by its behavioral choices. As evolution progresses, the freedom of choice increases exponentially, which further affects epigenetic and physiological evolution. 13. The neo-Lamarckists were right about the importance of the direct effect of the environment on the individual, and evolutionary consequences. 14. There are numerous ways in which an organism can break away from firmly established developmental patterns: a. Simple regression to a juvenile form that often works well in association with symbionts, or hosts, in the case of parasites. b. Regression to an early stage, and development from there in a non-traditional direction (= reculer pour mieux sauter). c. Cenogenesis, i.e., insertion of a diversionary life cycle stage such as a butterfly caterpillar, while coming back on line to terminate in the typical adult form. d. Hypermorphosis—adding on to the end of epigenesis—affects both anatomy and physiology. e. Early diversion of normal development by a particular creode or cell line that is internally integrated and also accommodated (ontogenically buffered) by other developing lines. Sometimes this is caused by the insertion of temporary interphenes that may be atavistic or novel (= key epigenetic innovations). This is the process most likely to create the unlikely, or to produce saltatory novelty. f. Metamorphic phases or thresholds in life cycles that are affected by environmental conditions increase the likelihood of developmental change. 15. The conservative nature of the genes that contribute to foundational epigenetic functional units (or holons, or modules), such as homeotic gene clusters and key organizing embryonic cells, partly explains the tenacity of homology, in addition to parallel evolution and aspects of convergent evolution. Their repetitive differentiation also allows them to be repatterned with highly diverse phenotypic consequences. 16. But there exists an emergent property of homology that cannot be reduced to genes and their differential expression. And long after a developmental role had been established for the genes, there persisted responses to intrinsic physicochemical causes, and extrinsic environmental causes.

222

Chapter 5

17. These effects resulted from the behavioral activity of animals. Physiological adaptability. not only permitted behavioral changes, it also helped to buffer developmental novelties. 18. Some extrinsic biotic causes (symbionts sensu lato) have had persistent and prolonged effects on development that have very little to do with genetic determination.

6 Epigenetic Mechanisms

Some genetic structures do not adapt the organism to the environment. Instead they have evolved to promote and direct the process of evolution. —J. H. Campbell, 19851 In pan-environmentalism, environment consists of both morphogenetic and selective factors. It is envisaged that the former induces, by response of the genotype, variation upon which the selective factor(s) work. It follows, then, that there will be appreciable evolution with environmental changes. (Conversely, there will be no appreciable evolution without environmental change.) —Ryuichi Matsuda, 19872 . . . the present relationship between genes and form is a highly derived condition, a product of evolution rather than its precondition. . . . [The concept] that epigenetic mechanisms are the generative agents of morphological character origination, helps to explain findings that are difficult to reconcile with the standard neo-Darwinian model, e.g., the burst of body plans in the early Cambrian, the origins of morphological innovation, homology, and rapid change of form. —Stuart Newman and Gerd Müller, 20003

Since much of the information in the structural genome is shared by all organisms, from bacteria to humans, the explanation of differences among organisms has to be sought, in part, in “genetic structures . . . that promote and direct evolution.” These include mechanisms that affect mutability, and regulate differential gene expression. To operate, some mechanisms of evolution at the genome level require spare parts, such as the “junk” DNA components that constitute about 75 percent of some eukaryotic genomes. Geneticists classify anything involving DNA as genetic, and define epigenetic processes as those heritable regulatory mechanisms that also involve DNA. But this is too genocentric. The DNA keys of the genetic keyboard are necessary if the music is to be played, but they are neither player nor score. Suppose that conventional wisdom were true: the developmental instructions for any organism are in its DNA. Let’s further suppose that the DNA can initiate protein synthesis all by itself, and

224

Chapter 6

everything goes from there. We are already in the Looking Glass World at this point; but even if what we saw in there were real, we still would not understand how it could have evolved to make different organisms. “Isn’t mutation of the genes, and natural selection enough?” cries a student who wants it to be simple. I tell him that reduction to the genes leaves out the organism, its actions, its environment, and its evolutionary history. Environment was a large part of that history, and although it is a large part of selection theory as well, we should be more interested in the environment as a generator of evolutionary novelties, than as a selective eliminator of them. Furthermore, although organismal and environmental factors can be reduced to the molecular and physical levels, they cannot be induced from those, since they involve physiology, behavior, environmental changes, and all of their interactions—he was sorry he asked. Because of the phenomenal and causal complexities, I do not limit epigenetics to heritable features, but prefer Søren Løvtrup’s broad definition of epigenetics as encompassing anything that affects the course of development, along with his inference that anything that changes the course of development causes evolution (Løvtrup 1974). Two of the epigraphs for this chapter were chosen to emphasize how environment initiates epigenetic change; because it is a general principle that must not be forgotten. However, this chapter focuses largely on the molecular and cellular levels. In 1913, Jakob von Uexküll proposed that Mendelian elements must be regulated by “supergenes”: a hierarchy of control must be involved in evolutionary change. Richard Goldschmidt also realized that changes in the genes could neither explain speciation, nor the large epigenetic changes that might generate hopeful monsters. The Material Basis of Evolution (1940) proposed that chromosome mutations and their consequent position effects altered epigenesis. Stimulated by Goldschmidt’s theory, Barbara McClintock discovered position effects in corn genetics in the 1950s. They did not result from chromosome mutations, but were the effect of small, transposable elements that caused changes in gene expression when they jumped from one chromosome to another. They are not necessarily beneficial, but provided that their effects are harmonious, and that the integrity of the developing organism is not compromised, such natural experiments may continue until a hopeful monster emerges. Position effects are important at the ultrastructural chromatid level since interacting segments of DNA have to be in the same place at the same time. We could also say that position effects are caused at the cellular level by the influence of adjacent cells. While McClintock’s research was getting under way, François Jacob and Jacques Monod were investigating the bacterial lac operon, the first mechanism of gene regulation to be worked out at the molecular level. Their discoveries excited the same Cold Spring Harbor molecular biologists who had treated McClintock to blank looks. But although the lac operon became well known, its theoretical relevance to evolu-

Epigenetic Mechanisms

225

tionary epigenetics, though suspected by its authors, was largely ignored as a special case of bacterial function. Epigenetics and its associated vocabulary were not elaborated until C. H. Waddington published The Strategy of the Genes (1957). Though well received, it lacked mechanistic detail that might explain how gene strategies were effected or altered. Subsequently, Løvtrup’s Epigenetics (1974) was a valiant attempt to marshal the biochemical evidence for evolutionary epigenetic change. Conceptually persuasive to some readers, its molecular database was insufficient to fuel a paradigm shift. Due to technical difficulties, and the hostility of the Modern Synthesis to epigenetics and other relatives of the hopeful monster, it was not tried and found wanting, but found difficult and not tried.4 In Phenotypes (1994), David Rollo remarks that although persuasive molecular evidence for developmental evolution did not accumulate appreciably until the 1980s it was time for the focus of evolutionary thinking to be shifted to epigenetics, and that is how things have gone in the last decade. Wallace Arthur comes at the subject from the same background of evolutionary ecology, and his The Origin of Animal Body Plans (1997), makes a similar recommendation. A rich lode of epigenetic molecular biological evidence from current literature is Epigenetic Inheritance and Evolution (1995) by Eva Jablonka and Marion Lamb, who admit to a Lamarckist bias that I will come back to consider at the end of the chapter. Fundamentals of Molecular Evolution (1991; second edition, 1999), by Li Wen-Hsiung and Dan Graur is a good general source. Cells, Embryos, and Evolution by John Gerhart and Marc Kirschner (1997) has the same fin-de-siècle resonance as William Bateson’s 1894 Materials for the Study of Variation. For anyone who wants the full degree program, Scott Gilbert’s Developmental Biology (1997) is not only the most encyclopedic presentation of the information that I examine in this chapter, it also pays some attention to its history and its more provocative implications. Brian Hall’s Evolutionary Developmental Biology (second edition, 1999) is another comprehensive treatment of epigenetics in the context of the Modern Synthesis, and also explores its historical roots. Mary Jane West-Eberhard’s Developmental Plasticity and Evolution (2003) takes the subject beyond mere evolutionary epigenetics to the point where it stretches the Modern Synthesis to the limits of its plasticity. Lynn Caporale’s Darwin in the Genome (2002), does not focus on epigenetics, but has serious implications for it, and is a rich source of information regarding genomic mechanisms that “have evolved to promote and direct the process of evolution” as Campbell’s epigraph puts it. Marc Kirschner and John Gerhart’s 2005 book The Plausibility of Life: Resolving Darwin’s Dilemma provides the most recent and comprehensible synopsis of epigenetic ideas, but it entangles them almost inextricably with neo-Darwinism. Eva Jablonka and Marion Lamb’s Evolution in Four Dimensions: Genetic, Epigenetic, Behavioral and Symbolic Variation in the History of Life is more accessible to the general reader than

226

Chapter 6

their 1995 opus, but although they are familiar with some of the escape routes they are still trapped in the hand of Darwin. To grasp the difficulty of understanding molecular evolutionary epigenetics, consider one of the advertised goals of the human genome project. At the outset, it claimed it would be able to provide a clearer understanding of human evolution, as if this would automatically result from accurate knowledge of all the base sequences of the genome. However, to understand evolution we would also need to know what the genes do, discover all the proteins for which they code, and how the proteins function, and how all of these interact. And all this would not solve the more refractory problems of physiology and behavior and associated environmental effects. Moreover, to understand the evolutionary significance of human base sequences alone, we would also need similar information about, say, the genome of the chimpanzee, and a number of other organismal types going back through the line of evolution of the vertebrates and their ancestors. Even so, the well-known genomes of bacteria, yeast, roundworms, fruit flies, and a hundred other species are proving to be almost as relevant to evolution as those of our close relatives. They are an invaluable repository of information necessary for the thorough establishment of phylogenetic relationships, and they relate how base sequences have changed in relation to evolutionary diversification. However, information about the regulation of gene expression is more valuable to evolutionary theory than a genomic encyclopedia. In scale, the largest genetic differences between us and our closest cousin, Pan, the chimpanzee, are chromosomal inversions and translocations. These repositionings and novel combinations of chromosomal segments may be partly responsible for some of the phenotypic differences that we can see, but may simply be random changes that contributed to mutual sterility before or after the divergence of our lines. The genomes, at the level of structural gene content, are virtually identical. The putative 1.3 percent difference is just as great within the human species as it is between us and the chimpanzees. In fact there is a huge overlap between the genetic constitutions of all living species, in terms of their abilities to make structural proteins and enzymes. All organisms have the same building blocks of DNA and protein. It is how they arrange and use those building blocks—to make a simple cell, or to make a complex, multicellular human—that is important. Think of the basic genome as a pantry full of ingredients.5 A Chinese chef and a French chef might go inside, take out very similar arrays of items, and produce two totally different meals. Similarly, epigenetic recipes produce two distinct organisms by combining the same basic ingredients in different ways and by allocating different seasonings and cooking times. If we have comparative information about how genes are expressed during embryonic development, in quantity, in quality, in rate, and in concert, we have an important key to understanding the fundamental evolutionary mechanisms that make organisms so diverse. To contribute some kind of answer to

Epigenetic Mechanisms

227

that problem, this chapter will survey molecular mechanisms that relate to morphogenesis and epigenetic evolution. Neo-Darwinism was satisfied that random point mutations of the structural DNA that codes for proteins were sufficient raw material for the operation of natural selection. To a degree this is indeed sufficient for simple, though inflexible, adaptation to environment. Yet epigenetic evolution could progress without the accumulation of such point mutations, through alterations in mechanisms of gene expression. A conventional response is this: “It doesn’t really matter whether it is mutations of structural DNA or regulator genes. So long as they happen at random, natural selection will sort them and so gradually direct both simple adaptation and the increase of complexity.” I will establish in this chapter that all change at the DNA/RNA and chromosomal levels is discontinuous, and that many such changes can be considered saltatory, with dramatic epigenetic consequences. Through the redundancy of repetitive DNA, experiments can be conducted in non-adaptive change, “out of sight of natural selection.” Moreover, many genomic changes are non-random, and can have a self-amplifying effect correlated with allometric morphological change, i.e., shifts in the anatomical proportions of organisms. Furthermore, it is possible for the environment to effect phenotypic changes that are heritable and that persist when the environmental influence has been removed. Some epigenetic evolution has to do with the ultrastructure and biochemical composition of the cytoplasm of eggs. Finally, and most heretically, as the Newman and Müller epigraph suggests, it is possible that most early evolution was caused primarily by the interaction of the organism and its environment, and only secondarily worked into the genome. The introduction of technical language that will be unfamiliar to some readers is unavoidable. Even for professional biologists the terminology is difficult because it has grown haphazardly. Sometimes several terms are used for the same mechanisms, and sometimes terms retained for their historical priority are thoroughly misleading. I will keep them as simple as possible, as much for my own sake as anyone else’s. Any specialist biologists who are still with me will have to tolerate the catalogue of elementary explanations of mechanisms and structures that are needed to make sense of molecular epigenetics. A plain-language summary of established molecular mechanisms of epigenesis that bear upon evolution is provided at the end of the chapter. Modifications of Gene Structure The best-known process conventionally associated with evolution is non-synonymous base substitution through point mutation. This automatically results in an amino-acid substitution in the protein for which the DNA codes, once the transcription and translation of the code through RNA intermediates have been completed. Ever since

228

Chapter 6

the point mutation was first understood, it was taken to be a random process. This is now questioned by Lynn Caporale (2003). She cites, for example, mechanisms that increase hypermutability in particular parts of genes. In their essay “Large phenotype jumps in biomolecular evolution” (2004), F. Bardou and L. Jaeger model the effects of mutation on biopolymer folding (i.e. tertiary structure). They find three categories of point mutation: 1. Those largely lethal mutations that cause extreme changes in structure; 2. Those that cause moderate changes with phenotypically significant and non-lethal effects; 3. Those that have minimal effect that might allow for “fine tuning” of an existing system. Their model fitted the effects of 157 point mutations of a ribozyme. As to randomness, the majority of actual point mutations were in category 2, and so, there is some quality of the genome that makes them non-random. Bardou and Jaeger also postulate that large phenotype jumps are possible in this system, especially involving category 1 under environmental stress. None of these events are cumulative adaptational processes directed by natural selection. Eukaryotic genes contain exons, which have the structural genetic code, and introns, which are “spacer” components that are not translated into protein structure. However, they facilitate shuffling of the sequence of exons, and that does have an effect on protein structure. Gene structure can also be changed by intragenic recombination, that may reduce part of the DNA structure of a chromatid, or add to it, or simply mix parental chromosomal characteristics. I will refer to some of these processes in further detail below. DNA Modifiers of Gene Expression Prokaryotic Mechanisms The “lac operon” system has no direct bearing on eukaryotic epigenetics, but it is familiar to most biology students and is a useful introduction to the vocabulary needed to address the genetic components, and their products and interactions. It also illustrates how distinct strains of prokaryotes can be established by cytoplasmic inheritance, which is important for epigenetics in multicellular organisms. The system is an array of genes that controls the output of the enzyme β-galactosidase in Escherichia coli, a common bacterium of the human gut. The “structural gene,” which codes for the relevant enzymatic protein, is part of the lac operon complex. Two other structural genes in the complex make permease, which gets the lactose into the bacterial cell, and transacetylase, which acetylates unusable galactosides before their elimination from the cell. The three structural genes are physically together on the same strand of DNA, accompanied by a DNA switch called an “operator.” The operator is adjacent to a DNA “promoter” sequence, which allows the enzyme RNA polymerase to begin to transcribe the structural genes to make messenger RNA. If the operator

Epigenetic Mechanisms

229

switch is off the transcription does not take place and the enzymes are not synthesized. The enzymes are not often needed anyway, so the switch is usually off. Alongside the operator there is a “regulator” gene that codes for a “repressor” protein, whose presence in the cell locks the operator in the off position. For the lac operon to be activated fully, two conditions have to be met: lactose has to be available, and glucose, the energy source most commonly used by E. coli, has to be absent. If lactose is in the cell, some of it spontaneously converts into the isomer allolactose, which combines with the repressor molecule, inactivates it, and thereby unlocks the operon, allowing the ultimate synthesis of the necessary enzymes. However, if glucose is also abundant in the bacterial cell there is a scarcity of cyclic AMP, which is needed to make the lac operon promoter more receptive to RNA polymerase. The cyclic AMP (cAMP) combines with “catabolite activator protein” (CAP) to stimulate the expression of the lac operon structural genes. Therefore, the production of the lac operon enzymes can be inhibited and stimulated by separate systems. Two different strains of E. coli can arise in media containing lactose. Its uptake is stimulated in cells that have some permease. In others that lack permease to begin with, there is no uptake of lactose to trigger the production of the galactosidase. When cells that have permease reproduce, their daughter cells have enough of the enzyme already in the cytoplasm for them to continue to use lactose. Thus, cytoplasmic inheritance of the appropriate protein molecule determines the action of the strain in the next generation, although both the strains have identical DNA.6 The bacterial control of gene expression shows how well tuned and responsive cellular homeostasis is in the most primitive kind of organism. Eukaryotic Mechanisms Knowing that such a regulatory system existed in bacteria led to a search for similar systems in eukaryotes, but because gene structure and regulatory systems in eukaryotes proved to be much more elusive and complex than the operons of E. coli there was no rush to embrace their implications for epigenetic evolution. One problem arose with the discovery that the genes that need to be called into use simultaneously for a coordinated biochemical pathway are often found on non-homologous chromosomes, and each has its own promoter. This required some mechanism, such as a common sequence responsive to a common triggering substrate, capable of making all their promoters say “go” at the same time. Moreover, some promotor sequences are complex algorithmic switches that probably evolved by the repetition of simple DNA holons, which then differentiated to become responsive to a variety of activating molecules. The nomenclature for genes that influence the expression of the structural genes varies somewhat throughout the literature. Among the choices available, “modifier genes” and “regulators” signify the most general category of DNA sequences that affect

230

Chapter 6

gene expression, and related molecular mechanisms. These fall into several categories. “Replicators” mark the spots that begin and end DNA replication; “recombinators” mark recognition sites for recombination enzymes involved in phenomena such as crossing over, transposition events, increases in exon shuffling, and gene duplications. “Segregators” are DNA sites where chromosomes attach during mitosis and meiosis, before they are pulled apart to enter the new daughter cells. Promoters, along with sequences called “enhancers,” are attachment sites for inhibitor or activator molecules, and switches for DNA transcription. The promoters respond to proteins synthesized in the nucleus. In some cases they bind proteins synthesized in the cytoplasm. These are activated by steroid hormones, upon which they enter the nucleus and affect specific promoter sequences. The activation of promoters demonstrates how nuclear proteins, cytoplasmic proteins, steroids from the extracellular milieu, whole organism physiological cycles, and environmental cycles may all be influential in regulating structural gene activity. In addition to structural genes that code for structural proteins and enzymes, there are many structural genes that code for regulatory proteins, including “selectors” such as Hox and Pax-6 genes involved in appendage and eye development respectively. Others code for enzymes involved in DNA and RNA transcription; RNA modification; DNA repair; DNA acetylation and methylation; recombination, and the synthesis of non-protein activators or inhibitors such as cyclic AMP and steroid hormones. Products of this kind are also important in regulating organ function in multicellular plants and animals. Finally there are the genes that code for DNA packaging, and binding proteins such as histones. Codon, exon and gene duplications, transpositions and point mutations of any of these genes could have an epigenetic and hence evolutionary impact. Some authors use “developmental gene” as a catch-all term for any modifier or structural gene involved in epigenesis. Regulatory Role of RNA RNA interference can silence gene expression, either via direct interaction with DNA (transcriptional gene silencing—TGS), or by interactions with other RNAs (post-transcriptional gene silencing—PTGS). Predominantly a homology-driven means of associating with other nucleic acids, “interference RNAs” (iRNAs) can recruit methylation machinery to genomic sites. RNA can prevent the expression of genes by interfering with the translation of mRNA into protein. This involves single-strand or double-strand RNA molecules binding to a gene’s RNA transcript, in essence silencing that transcript. Occurring universally in plants and animals, “micro-RNA molecules” (miRNAs) interfere with messenger RNAs (mRNAs) by directing their cleavage by RNAase. Because of their small size, about 22 bases in length, miRNAs long evaded detection.7 Thus, a gene’s ability to speak can be muzzled, preventing translators from making sense of the words, or the spoken word can be snatched before it is heard—certain

Epigenetic Mechanisms

231

RNA molecules have the ability to completely alter gene expression without any change in the genome sequence.8 Role of Steroids Steroids are molecules whose synthesis occurs by the activation of synthetases, proteins whose own synthesis requires a DNA code. Therefore changes in the code, and hence the structure of the synthetases can effect alterations in the steroids that are produced. In addition to regulatory proteins, steroids affect enhancement sites and activate mRNA transcription. In multicellular animals, steroids originate in endocrine glands and are passed by simple diffusion or in the blood circulation. Their presence is often activated by environmental conditions. More than 40 years ago, the insect steroid ecdysone was known to change the chromosome puffs in Drosophila polytene chromosomes, because these could be seen under a light microscope. The presence of messenger RNA could be detected histochemically by radioisotope labeling.9 Some vertebrate steroid hormones and other hormones inactivate a protein repressor and release the positive action of the enhancement sites for structural gene transcription. Thus, the duplication and variation of the genes for steroids and other nuclear receptors contribute to cellular differentiation and growth during development. Their repetitive differentiation has also played an important part in reproductive physiology and the evolution of homeostasis. Hormones involved in gametogenesis, yolk deposition, metabolic rate, excretion, salt and water balance, calcium regulation, peroxisome proliferation, the elaboration of visual pigments and other functions all require differentiated nuclear receptors. The emergence of the vertebrates has been linked with the doubling and redoubling of the entire genome. This had multifunctional potential, because now four copies of the original primitive steroid receptors, and all other genes, were then available for differentiation.10 Some regulatory genes have several promoters, making it possible for them to be transcribed by various activator molecules a number of times during cell differentiation, an arrangement that may modify the products of the regulatory gene as well. It is also likely that regulatory proteins interact with each other as well as with DNA recognition sites involved in gene transcription. The co-existence of all of these regulatory mechanisms presents a wide range of possible combinations that are needed for a sophisticated differentiation hierarchy of regulatory levels. From these, large-scale or small-scale evolutionary changes can emerge. Mechanisms That Introduce Genomic Variability Point mutation substitutions, part of the process of protein alteration, were once taken to have the miraculous power of providing any molecular variations demanded by selection pressure. Ever since it became possible to work out the primary amino acid sequences of proteins, it has been known that many substitutions have occurred

232

Chapter 6

during evolutionary history, and it was inferred by selectionists that these were all adaptive, though there is now wide support for the view that many are simply neutral and persisted through genetic drift. Many beneficial point mutations involve duplicated genes, which contribute to an adaptability that gives the organism choices appropriate to conditions. Beyond change in the DNA of the genes, there are the kinds of epigenetic controls of gene expression that have just been outlined. But we have not yet done with events that increase variability at the gene level. Exons and Introns Exons are the bits of the eukaryote gene that are ultimately translated into protein, but in eukaryotes they are separated by intron sequences that are initially transcribed into messenger RNA and then pruned out before translation, so that their DNA code normally has no influence on the structure of the protein that is synthesized. Selfsplicing introns are found in bacteriophages, and protozoa, as well as chloroplasts, and mitochondria. From the genes of the latter they were introduced into the eukaryotic nuclear genome, and spread by the actions of transposable elements called “retroposons,” which are described below. The self-splicing function of these introns was replaced in the nuclear translation apparatus by “spliceosomes.” These are complexes of small RNA molecules and proteins capable of editing newly synthesized RNAs, and altering these templates for protein synthesis via translational start and stop sites. Alternative splicing permits the generation of numerous proteins from a single gene transcript.11 Histone genes have no introns, but some structural genes consist mostly of introns that may widely separate exons from one another. All of the DNA in exons and introns may be subject to repetition through non-conservative DNA replication and repair mechanisms. After initial transcription there are steps, over and above silencing mechanisms, that can still affect gene expression before it is finally translated into polypeptides, and afterwards, during the construction of the final protein product from the polypeptides. Large proteins may be assembled from polypeptides produced by several separate coding sequences. Since each of the coding sequences consists of exons interrupted by introns there is a potential for gene repatterning through intragenic recombination—the crossover occurs in mid-gene, resulting in the production of modified polypeptides and therefore different proteins. With the discovery of introns it was quickly realized that gene exons could be shuffled to make “new” genes, which would enhance genetic variability. Exon shuffling could provide for evolutionary experiments through random differentiation of the resultant proteins, and there is evidence for such rearrangements.12 For example, the differentiation of genes for cellular adhesion proteins results in the wider differentiation of tissues. Exons can be deleted, duplicated and translocated. Furthermore, novel genes may be generated by involving introns in protein synthesis instead of having them excised

Epigenetic Mechanisms

233

from the messenger RNA. This involves mutation in the splicing instructions to cause the intron to be read as an exon, which is known to occur in Drosophila epigenesis.13 A more dramatic differentiating effect may be achieved by “overprinting,” whereby existing base sequences are read in a new frame, with the result that a “new gene” arises from the overlapping of two adjacent old genes that still retain their original functions.14 Introns also have a role in the regulation of gene expression.15 That might in part be why eukaryotes, which have introns, have greater evolvability than prokaryotes, which lack them. Protein Domains The significance of exon shuffling points to the next higher level in the hierarchy of the cell, namely the “proteomic” level. Large proteins consist of a number of “domains” (distinct units with structural or functional roles). In different combinations they account for the larger variety of proteins. An annotation of the draft human genome sequence included a comparison of protein domain types from several organisms. Protein collections obtained for yeast, nematode, fruit fly, and human contained 3,136, 9,254, 8,896, and 15,312 proteins, respectively. This indicated 973, 1,183, 1,218, and 1,865 domain types.16 But the number of genes, in the tens of thousands for most higher eukaryotes (approximately 30,000 in humans), is far greater than the number of different protein domains. Some domains may involve several amino acid sequences brought into functional relationship by the three-dimensional structure of the protein. In other words, such “functional domains” are not determined by a particular gene, but by several genes, or by particular exons of several genes. We cannot understand them in terms of genomic information alone, since the epigenetics of protein complexification adds greater meaning. Thus, it seems likely that the more fundamental relationship is not between the numbers of genes and proteins, but between the number of possible protein domains and the original number of primitive exons. It is exon shuffling, duplication, and insertion, in addition to intron co-opting, and intron insertion and deletion, that have built up the more familiar variety of genes and proteins—and made repetitive differentiation possible at the molecular level. Another active participant in genomic shuffling is the intron that is not merely a meaningless partitioning segment, but can code for endonucleases, splicing directions, and other elements that affect RNA structure and protein synthesis.17 Nevertheless, the genome can still inform us about protein “structural domains” whose primary amino acid sequence conforms to the sequence of exons in a particular gene. A good example of a functional change brought about within the gene by exon duplication is cited by Graur and Li (2000). In the guts of most animals, chymotrypsinogen is activated to become protein-digesting chymotrypsin. In the Antarctic cod, Dissostichus mawsoni, the gene for this endopeptidase was duplicated.

234

Chapter 6

An original chymotrypsinogen gene was retained for digestive functions. But in the other copy, exon duplication and mutation, as well as incorporation of some intron sequences, turned out an antifreeze glycoprotein. There is a self-amplifying component in glycoprotein production as well since often a large protein with repeated tripeptide segments is cleaved to produce many antifreeze molecules. Logsdon and Doolittle (1997) attribute this to a strong positive selection pressure brought about by climatic change when Antarctica froze between 5 and 14 MYA. This textbook example of the power of the metaphor of selection pressure should be a cautionary tale for those who combine selectionism with molecular biology. An alternative explanation does not draw upon a metaphorical force, and it puts the molecular information in a wider context. Repetitive differentiation, which had previously produced trypsinogen and chymotrypsinogen from a common gene ancestor, now resulted in another molecule that added to the adaptability of the fish— it could penetrate colder water with new food resources and fewer predators. Not only were competition and predation diminished in that environment, the cod population there would be largely composed of the strain with the antifreeze. It was the behavior of the cod that determined both the demographic and genetic consequences. What the organism does determines what is advantageous. The drop in temperature need not in this case have epigenetically triggered an increase in antifreeze synthesis, but that too deserves consideration. The more adaptable the organism, the greater the freedom to conduct and test evolutionary experiments. And those experiments depend upon the availability of components that are not adaptive but redundant. Repair Mechanisms Since the operation of repair mechanisms has more to do with cellular homeostasis than evolutionary change, it is their failure that is involved in epigenetic experiments. It has long been common knowledge that high-energy radiation (including x rays and ultraviolet light), and many chemical environmental factors, are mutagenic and carcinogenic, and organisms are in constant danger of these influences to themselves and their offspring. But there are ways of repairing damaged DNA. In his extended metaphor of natural genetic engineering, James Shapiro (1992) calls this “quality control.” A wide range of enzymes exists for this purpose. For example, radiation has the common effect of covalently bonding adjacent thymine bases into dimers that cannot participate in DNA transcription. The affected triplets can be excised by a repair mechanism, and the gap filled by the action of DNA polymerase and ligase. The accidental chemical conversion of cytosine to uracil is corrected by a specific enzyme, and there are also enzymes that proofread and correct DNA strands during replication, such as 3'-5' exonuclease and a component of DNA polymerase III. Mutation of the DNA sequence that codes for this component causes a great increase in the mutability

Epigenetic Mechanisms

235

of the affected organism. There is also an inducible SOS system (as in the Morse code for “help”) that responds to radiation injury, both repairing the faults and stopping mitosis until the corrections are complete. However, the operation of the SOS process can also increase mutability.18 When breaks in DNA are detected, the cell usually shuts down until such time as the break is either repaired, or the cell enters into programmed cell death (apoptosis), a key feature of multi-cellular organisms that maintains overall organismal integrity. There are two general pathways by which DNA breaks can be repaired: “homologous recombination” (HR) and “non-homologous end joining” (NHEJ).19 The first involves copying of homologous template sequences, to ensure the proper replacement of any lost sequences, but can lead to errors in crossing over or gene conversion events (discussed below). The second class of repair (NHEJ) is via direct religation of broken ends, which usually leads to little or no change in DNA sequence but can produce translocations or gross chromosomal rearrangements. Left unrepaired, these kinds of damage would most likely have a negative effect on epigenesis. However, there are other kinds of DNA alteration that could change the course of epigenesis, and repair mechanisms are part of the apparatus of canalization. Therefore any decrease in the effectiveness of repair could contribute to epigenetic change by loosening canalization. Heat-Shock Proteins At the molecular level, all organisms are capable of mounting a response to physicochemical and biotic stresses. A highly conserved group of “heat-shock proteins” (HSPs) normally assist other proteins in folding (conformational change—in some cases the HSPs are referred to as “chaperonins”), subcellular localization, and other cellular homeostatic functions. Under stress, they protect the cell and repair other proteins. Under temperature increase, exposure to oxidants, or infection, the heat-shock transcription factors (HSFs) undergo conformational changes, whereby their domains are disrupted and new sites are exposed, enabling them to bind DNA and induce the transcription of heat-shock genes. These proteins act to repair other damaged proteins, and defer the entry of the cell into apoptosis. Proteins capable of changing conformation and activating genes in response to changes in the cellular environment certainly introduce another level of control with respect to differential expression of existing genetic information.20 Gene Duplication I have made much of the general process of repetitive differentiation. If a body segment, or an organ such as a limb, or a molecule, such as an enzyme is repeated, or duplicated, the spare copy can be modified or differentiated to complement the existent unit, thus increasing overall adaptability. Therefore the variety of ways in

236

Chapter 6

which multiple copies of DNA and RNA can contribute to complementary changes in gene expression deserves scrutiny. Some kinds of repetitive sequences called “satellite DNA” are repeats of short nucleotide sequences usually found in condensed chromatin, and are clustered near the centromeres and the ends of the chromosomes. They may play a structuralfunctional role as transition zones between euchromatin and heterochromatin. There are perhaps millions of copies of some of these sequences and they may add up to 25 percent of the total DNA of the genome in some organisms. Their somewhat misleading name is derived from the way in which they separate in an ultracentrifuge. They are also lumped under the heading “junk DNA,” which is also an unfortunate misnomer for a pool of material that probably has an important role in evolution. Even so, the puffer fish, Fugu rubripes, has a genome that has been largely stripped of its junk DNA, and it survives and reproduces perfectly well.21 Whether or not it has sacrificed evolvability in so doing, it now provides a useful contrast for the interpretation of junk DNA in other species. A kind of repetitive DNA that has attracted much attention constitutes “telomere repeats” at the ends of chromosomes, which are replenished through the action of “telomerase,” a specific reverse transcriptase. Telomeres are now understood to extend the longevity of cell lines. Thus, their gradual degradation is equated with aging, since telomerase activity seems to diminish with increasing numbers of cell divisions. Sometimes, amplified DNA is physically separate from the chromosomes. The toad Xenopus laevis has many copies of the gene responsible for the synthesis of ribosomal RNA in one chromosome. Many additional copies float free in the cytoplasm of the developing ovum. Other amphibians can get by with much fewer ribosomal RNA gene repeats. This “dose repetition” or “amplification” guarantees the rapid synthesis of many ribosomal RNA units needed during development, after which the extra copies are degraded. Gene duplication that causes only dose repetition results in more of the same product being produced faster, and is sufficient to alter development. Other examples of dose repetition through somatic gene duplication are found in widely different organisms. For example, groups of genes in the fruit fly responsible for laying down the outer layer of the egg are multiplied during oogenesis. The most controversial aspect of gene duplication is the possibility that it might be initiated in direct response to a particular exigency. Mammalian cells in culture, exposed to the drug methotrexate, replicate by several hundredfold the gene for the enzyme dehydrofolate reductase, which breaks it down.22 However, there is a slight tendency for this gene to be duplicated even in the absence of the drug. Specific repetitions of genes for the enzyme aspartate transcarbamylase and for the metaldetoxifying protein metallothionein-I have also been found in mammalian cell cultures.23 These are somatic duplications that would not normally be heritable; but there is also an example in a human family with a history of exposure to organophos-

Epigenetic Mechanisms

237

phate insecticides. They have a hundred repetitions of a mutant gene for cholinesterase, which confers resistance to the toxins and probably arose during spermatogenesis when the gene is active. In some plants, genes that proffer resistance to herbicides are similarly multiplied.24 Could chronic exposure to such stimuli initiate appropriate duplications in germ line cells at vulnerable developmental stages and pass such duplications on to subsequent generations? Activity of retrotransposons seems to peak in germ cells, suggesting that during the global rearrangements in epigenetic patterns there is room for copying and repositioning genetic sequences within the genome. In a few types of animal, including nematodes and arthropods, there is “chromosome diminution” or “chromosome deletion” (differential chromosomal loss as cell lineages develop). This may be why August Weismann (1893) generalized incorrectly in his Biophore Theory that cells differentiate by retaining only a small, distinctive portion of the original genome of the zygote. Some differentiated cells lose all of their chromosomes; for example the red blood cells of mammals are anucleate, but they have a large enough supply of the appropriate mRNA to keep on synthesizing hemoglobin for the several weeks of cell life. The kind of gene duplication that is particularly relevant to repetitive differentiation occurs where there are multiple copies of the same structural or modifier gene, either clustered on a particular chromosome, or dispersed on non-homologous chromosomes. Duplication of these genes can occur through uneven crossing over during the first division of meiosis, when gametes are being produced. This is one of the least disruptive possibilities, although the gamete that loses its genes to a sibling cell will be sterile. Particular genes in particular lineages have a tendency to be duplicated, suggesting a built-in susceptibility for uneven crossing over at specific chromosomal loci. This duplicative trend occurring in the same part of the genome, rather than randomly at unpredictable points, could contribute to allometry. Replication slippage during the synthesis of new DNA strands before mitosis and meiosis can also duplicate DNA sequences. This is most likely to occur in DNA strands where there is already a degree of duplication of coding sequences, so that the slippage does not totally disrupt replication. Only short DNA sequences are affected. Repetitive microsatellite DNA has a variety of functions in bacteria. The ability of Neisseria gonorrhoeae to infect epithelial cells of the host is made possible by a cell surface protein whose stickiness depends on multiple copies of a CTCTT base sequence in its gene.25 These result from frequent replication slippage during cell reproduction. The same “errors” produce CAAT base repeats in three different genes of the meningitis bacterium Hemophilus influenzae. If present, the microsatellite repetitive DNA seems to make the bacterium more resistant to the hosts immune system. If absent, the bacteria are better able to invade the host. Alterations of the CAAT repetitions allow infection first, and then resistance to immunity. Note that a single individual cannot do the

238

Chapter 6

switching. It must occur with division. However, the resulting adaptability is at the population level. Such contingency genes have also been found to affect cell movement, reception of environmental chemical stimuli, and resistance to antibiotics. In eukaryotes, a positive role for repetitive microsatellite DNA has been harder to find. “Anticipation” is a condition caused by replication slippage; Huntington’s disease and myotonic dystrophy are examples. A particular gene codon is repeatedly amplified to the point where the gene is no longer functional. This self-amplification process is discussed in chapter 7. David King (1994) has already pointed out that this phenomenon could have a non-detrimental evolutionary effect of altering gene regulation.26 And he notes that it would not only result in a rapid generation-bygeneration amplification but would also be highly mutable. Subsequently, King’s prediction has been in part borne out by a study by John Fondon III and Harold Gardner (2004), which is also discussed in the following chapter. Walter Schaffner’s laboratory has demonstrated that microsatellite DNA for multiple glutamines or prolines, if inserted at the beginning of a gene, would increase protein production.27 One way of getting satellite DNA or other sequences into genes is by means of the repair process for double strand breakages that occur quite commonly during mitosis. Recent studies in plants, yeast, and mouse cells show that doublestrand breaks (DSBs) in DNA can infrequently be patched with other fragments of DNA. Presumably, the same proteins which bind to broken DNA ends in the genome will also recognize extrachromosomal DNA fragments as DSBs, and recruit them to the repair site. This process is found in humans.28 Moreover, a report of an inherited chromosomal translocation in mouse cells showed the insertion of a mitochondrial DNA sequence into the breakpoint. This not only suggests that insertional repair patching can not only affect the germ line but also effect organelle-to-nucleus gene transfer.29 Another possible duplication mechanism is reverse transcription of messenger RNA. DNA repeated in this manner lacks promoter sequences, but may come under the care of a pre-existing promoter. This is another way that some of the genes of proto-mitochondria could have been transferred to the nuclear genome. According to Lewin (1988) about 25 percent of the mammalian genome arose in this manner. And the recent human genome analysis suggests that more than 50 percent of the human genome arose from the insertion of repetitive elements.30 Reverse transcription contradicts the “central dogma” that information can only flow one way, from DNA to protein. A number of cases have been documented in which a non-transposon mRNA had somehow recruited a reverse transcriptase enzyme and produced a cDNA copy of itself, which has inserted elsewhere in the genome, resulting in a “pseudogene.” Errors in the initial expression of retrotransposon element can also result in a “read-through” event, whereby additional sequences downstream of the

Epigenetic Mechanisms

239

element are included in the mRNA transcript. When reverse transcriptase produces a cDNA copy of the transcript for insertion elsewhere in the genome, the original flanking genomic sequence is carried along for the ride. Other involvements of transposable elements in a separate process of gene duplication are discussed below. Extra copies of genes are also acquired through the duplication of an entire chromosome, or of the full assembly of chromosomes through polyploidy. Variation of Duplicate Genes Some duplicate genes become “varied” (or “variant”) “repeats” through mutation or exon shuffling. This is the official name for repetitive differentiation at the molecular level. These changes can produce variations on the same protein theme, or molecules with novel functions. They are less likely to be detrimental to the organism if there are back up copies to conduct business as usual while natural experiments on genes are taking place. Examples of repetition and variation of the genes for hemoglobin, protein hormones, muscle kinases and hormones involved in digestive tract and blood sugar regulation were given in chapter 4. Of particular epigenetic interest is repetitive differentiation of genes for hormones involved in development. Growth hormone underwent sustained bursts of rapid modification during early stages of elasmobranch, teleost, and amphibian evolution. Then during the diversification of placental mammals there were similar spurts in early artiodactyl and primate diversification. An increase in growth hormone in mice, induced by transgenesis of rat genes, results not only in giantism but also behavioral change.31 Prolactin was derived from the same gene family as growth hormone by repetitive differentiation. Its variant repeats regulate salt and water in fish and amphibians, broodiness and the production of crop milk in birds, and its most familiar function of mammalian milk production. Duplication and differentiation of steroid hormones has been important in development and reproduction. In addition, the systems that they regulate can be further complexified through repetitive differentiation of their receptors. Baker (1997) suggests that “receptors for androgens, glucocorticoids, mineralocorticoids, and progesterone evolved from an ancestral steroid receptor gene by two successive duplications over a brief time that could have coincided with the origins of vertebrates.” He continues: Moreover, the duplications of these steroid receptors may be additional evidence for the two duplications on a genome-scale that have been proposed to be important in the evolution of vertebrates. The two successive duplications of steroid receptor genes and their subsequent sequence divergence leading to steroid-specific receptors that regulate growth, development, reproduction and homeostasis in vertebrates may have been one of the events important in vertebrate survival after the Cambrian during global extinctions that occurred about 440 and 370 million years ago.32

240

Chapter 6

The range of receptors in the repetitively differentiated molecular superfamily that directly affects the nucleus includes those for thyroid hormones, steroids, retinoids, peroxisomes, vitamin D, and a variety of others. Swapping of domains from different origins was also involved.33 In addition to the simple tendency for duplications to keep affecting the same chromosomal loci rather than being purely random, there is a mechanism called “concerted evolution” that emphasizes the multiplication of particular variants. Within a gene family that had arisen by tandem duplication, i.e., side by side on the same chromosome, an apparent directiveness to mutational variation in the genes was discovered. Moreover, in related species where the same gene family existed, presumably because the common ancestral species initiated the duplication trend, there could be a divergence of this mutational variation in a different direction.34 This kind of mechanism could relate to divergent evolution, but if concerted evolution continued the same trend in related clades it could contribute to parallel evolution. Concerted evolution spreads particular mutations within gene families. The phenomenon was first called “molecular drive” by Dover (1982), who saw that its operation was independent of natural selection: The widespread fixation of variants by molecular drive is different in that it is an outcome of a variety of sequence exchanges within and between chromosomes that give rise to persistent nonmendelian patterns of inheritance. Significantly, there are circumstances in which the activities of the genomic mechanisms, in spreading sequence information between chromosomes, would lead to the progressive increase of a variant through a family more or less simultaneously in each individual of a sexual population. This concerted pattern of fixation by molecular drive may provide an explanation for the origins of species discontinuities and biological novelty.35

Li and Graur add: “Concerted evolution essentially means that an individual member of a gene family does not evolve independently of the other members, a multigene family evolves together in a concerted fashion, as a unit.”36 The mechanism that brings this about is much the same as that which causes the duplications in the first place: unequal crossing over, and a related phenomenon called “gene conversion.” The processes cause reduplication within multiplied gene families containing varied repeats. Particular variants become duplicated as a result of concerted evolution, and the more copies of a particular variant that there are, the greater the probability of it becoming reduplicated again and again. This trend can go on without reference to any adaptive value of the initial duplicate genes or their concerted evolution. Gene conversion does not increase the total number of duplicates, only the proportion of particular variants within the gene family, so it avoids the possible detrimental effect of “dosage imbalance.”37 This results from the disequilibration of dynamic stability by competition from multiple duplicates for scarce energetic or molecular resources.

Epigenetic Mechanisms

241

In contrast, if, as neo-Darwinism originally assumed, mutational experiments are random, there should be no identifiable trend except that caused by the natural selection of adaptive variants, and both related species should show a heterogeneous scattering of mutations, related only where selection pressures are identical. By neoDarwinist logic, two distinctively different kinds of selection pressure would be needed, for such trends to diverge in two species of the same genus. In addition to the duplication of whole genes and chromosomes, exon duplication within a gene also occurs, and some large proteins with repetitive domains (multiple functional regions) have originated in this way. For example, the trypsin inhibitor of egg white is a protein with three functional domains for inhibiting trypsin and similar serine-endopeptidases, such as chymotrypsin and bacterial subtilisin.38 This gives egg white an antibiotic quality. The synthesis of fats involves the most multi-domain “mosaic” genes that are known. Curiously, while this applies both in fungi and mammals, the components of the genes are different, and have undergone different evolutionary routes to converge with the same function.39 Transposable Elements Most of this segment is drawn from Li and Graur 1991. Transposable elements can move from their original position in a chromosome to different sites in the genome, usually affecting gene expression in some way. The synonym “jumping gene” has been bumped from the official lexicon because it is not always a gene that jumps—it may be a bit of one or more than one. And a “jump” is too close to a saltation for traditionalists. A transposition may be conservative, in which case the element takes itself to a new region of the genome, leaving a gap behind, or it may be duplicative or replicative, leaving a copy of itself behind when it jumps. A transposable element may also make the leap by proxy through “retroposition.” It is transcribed to RNA that is then reverse-transcribed to DNA, which is inserted into a new chromosomal locus. Transposition usually leaves a signature in the form of a “direct repeat,” a short sequence duplication of the adjacent DNA. Some transposable elements are only active in specific tissues, for example germ cells. This is consistent with the differential availability of DNA in different tissues, determined by methylation and histone binding. Some transposable elements always jump to the same site, some prefer a particular chromosome, some prefer a region rich in certain bases, and some jump randomly. Transposable elements fall into three categories. “Insertion sequences” contain only enough genetic information to effect transposition—vehicles without passengers. They are known to occur in prokaryotes, viruses, and eukaryotes. “Transposons” are larger, containing genes that code for a variety of products in addition to the information that lets them transpose. One kind of bacterial transposon carries the genetic information for an entire bacteriophage, including the genes that code for the protein coat and invasive enzymes of the virus.

242

Chapter 6

“Retroelements” or “retroposons” contain DNA or RNA sequences for reverse transcriptase, the enzyme that causes DNA to be synthesized from an RNA template. In mammals there are multiple copies of these, such as the “long interspersed nuclear elements” (LINEs), consisting of up to 800,000 copies in humans, though the vast majority are no longer full-length, and hence incapable of producing the machinery necessary for further transposition. There are also short interspersed nuclear elements (SINEs), as well as other transposons and endogenous retroviruses. Altogether, these make up more than half of the human genome. However, it is misleading to class the parts of the genome that do not code for proteins or constitute modifiers of gene expression as “junk DNA.” Many genes have been found to contain parts of repetitive elements, such that the most common protein domain in the human proteome is one derived from reverse transcriptase, and the introduction of not just exons but also introns from repetitive elements can introduce novel splice sites. The most complex kind of retroelement is the retrovirus. The retrotranscription of its RNA to DNA in the host cell initiates the synthesis of new viral protein and RNA that contains the genetic information for the next generation of viruses Retroelements produce “retrogenes,” which are open to copying errors and may not be functional when inserted back into a DNA strand. In mammals there are multiple copies of functionless “retropseudogenes,” the constant eruption of which has been called “Vesuvian evolution.”40 Although “evolution” here seems somewhat of an exaggeration, the possibility of a significant novelty arising randomly by this method should not be overlooked. The following are potential evolutionary epigenetic effects of jumping genes: Transposition of genetic defenses and novelties in prokaryotes. This is in addition to plasmid exchange through conjugation. Enhancement of downstream gene expression on the same DNA strand. Dormant genes may be awakened (known in yeast). Gene activation, and inhibition. General increase in position effects. Potential for novel exon shuffling through the transposition of exons, introns, control sequences, etc. Increase in unequal crossing over, which results in gene duplication. Gene duplication, with the potential for doubling rates of protein synthesis, with links to allometry (or orthogenesis).

Epigenetic Mechanisms

243

Increased mutation rate (known in E. coli and Drosophila). Gerhart and Kirschner (1997) comment that suppressors of transposable elements could be lost through stress or inbreeding, resulting in “an active period of gene rearrangement and insertional mutagenesis.”41 (Did Hugo De Vries’s “condition of mutability” not imply something of the sort?) Hybrid sterility. “Hybrid dysgenesis” is known from strains of Drosophila where a group of transposable elements produce high mutation rates, chromosome breakage and failure of gonadal development.42 A similar dysgenesis arising from “feminization” of males is caused by the parasitic bacterium Wolbachia in a variety of insects and terrestrial isopods.43 To top that, Wolbachia infection can “rescue” Drosophila oogenesis made defective by mutation.44 Paedomorphosis. A simple transposition could also be responsible for paedomorphosis through the inhibition of a gene coding for the synthesis of a key developmental hormone. Transposition that causes repeated gene duplication in the same locus is also one of the likely causes of gene duplication leading to allometric change and orthogenesis. Speciation. Li and Graur (1991) speculate that transposon-mediated alterations in gene regulation could be responsible for rapid sympatric speciation. Transgenesis. Retroviruses, the most complex form of retroelements, have long been suggested as mechanisms for one of the most radical kinds of natural genetic engineering experiment: transphyletic exchange of genes. One early known example involves a virogene that was transferred from baboons to cats about 10 million years ago. Although the number of known examples of such exchange are limited, exhaustive searches were not made until the human genome project was completed. Even when a transgene pops up from the genome there remains the problem of crossreferring it to its origin, if it does not match up with existing genome data banks. However, numerous bacterial genes have been identified in the human genome. Also, the synthesis of salivary amylase is promoted by an exogenous retroviral segment of DNA.45 Position effects. If operator and activator segments of DNA can be repositioned, the variability of gene expression increases. Thus, the evolutionary epigenetic potential of position effects, whether through chromosomal mutation or transposable elements, is now better understood. Goldschmidt (1940) came close to the modern interpretation when he argued that repatterning of the genes was more important than the mutation of the genes. Many of the phenomena that have been raised here involve heritable

244

Chapter 6

change, and must have occurred in the germ line of developing organisms. Many others occur only in specific lineages other than the germ line and are not therefore heritable. Plants are much more likely to pass on the effects of somatic gene multiplication since their cells retain totipotency and can enter the gametogenic cell lineage at a late stage in development. Selector Genes “Selector gene” is a catch-all expression for all the genes involved in the development of all the body compartments in the antero-posterior axis. Best known from Drosophila research, they include segment polarity genes, terminus genes, and homeotic genes such as Hox, Pax-6, and the Wnt group. The word “selector” is not intended to imply that the genes are agents of natural selection, but rather that they are involved in the determination of the particular pattern of development that the compartment will undergo. Noting that a comprehensive treatment is impossible here, and that there have been rapid advances in the last ten years, I organize this section in the manner of Gerhart and Kirschner (1997).46 Not all molecular biologists and journals follow the rules for gene nomenclature consistently. Specific gene names are always italicized. Sometimes the general group name is not italicized although the first letter is capitalized. Like many molecular biologists (and all biological taxonomists) I prefer to italicize the “genus” or group name as well as the specific gene name. Segment Polarity and Terminus Genes In insects there are 17, 18, or 19 segments, most of which are recognizable in the adults. Twelve segment polarity genes are involved in the development of parasegmental compartments, which give rise to the mature segments, each of which retain the same general cellular and organ architecture. Terminus genes are associated with the anterior and posterior ends of arthropods, parts whose development is regulated independently of segmentation. In primitive arthropod larvae the anterior end, which is involved in the formation of the mouth and gut opening, develops very early, allowing the microscopic planktonic animals to feed. The posterior end is specialized as a swimming organ which is essential for planktonic existence, and also employed very early in the life cycle. Most of the segmental development takes place after the termini have reached a functional condition in the larvae. Homeotic Genes Hox One of the most interesting discoveries in the fruit fly is the existence of highly conserved, linearly arranged clusters of homeotic Hox selector genes, with distinctive marker sequences called “homeoboxes.”47 Sometimes the physical sequence of the

Epigenetic Mechanisms

245

genes on the chromosomes conforms to the chronological sequence of their expression during development. They have been found in many animals, including cnidarians, mollusks, and vertebrates, and are responsible for the anteroposterior axial regulation of segments, or somites, and their appendages. Thus, it is in the fruit fly’s Hox genes, or somewhere in the epigenetic chain of command from Hox to the final construction of the limbs, that changes produce the classical conditions antennapedia and bithorax, among others. In the “antennapedia” condition, a leg appears in place of an antenna. In the “bithorax” condition, an extra thoracic segment with an extra pair of wings appears. The polycomb gene is known to perpetuate the repressed state of Hox genes where their action would be inappropriate, and its mutation is one of the reasons for the homeotic macromutations just mentioned. Hox genes are not responsible for the “nuts and bolts” of limb construction. They regulate segmental construction through their protein activators and are in turn activated or locked up by other epigenetic mechanisms. The earliest flying insects had multiple segmentally arranged wings. There are two theories of how they arose: as extensions of the dorsal exoskeletal plates, or as modified legs that first functioned as gill/swimmerets in aquatic insect larvae.48 If the winglessness of Apterygota, such as springtails and silver fish, is a primitive feature, wings emerged later than the basic body plan. The familiar arrangement of two pairs of wings was not primitive, but what was left after regression of an extensive segmentally arranged set.49 Although Hox genes alone did not dictate the emergence of wings, they are believed to be involved in the ultimate suppression of all but two pairs of thoracic wings. Moreover, the homeotic Ubx gene suppresses the posterior pair of wings and contributes to their modification as halteres in Diptera, such as Drosophila.50 What is fascinating about insects is the way that homeotic genes allow so much morphological plasticity within the rigidity of an adult body plan that basically conforms to two antennae, four wings, and six legs arising from a similar array of body segments. There is a consensus that what came before the insect body plan was an arthropod with many similar segments with legs arising from most of them, as in myriapods. Therefore, Hox genes were expressed in most of the segments. Since the gene clusters share deep homology with others found in the simplest unsegmented animals, they must have remained active when segmentation emerged, and then undergone repetitive differentiation coupled with changes in downstream phenotypic expression to produce the varied forms of the antennae and palps and legs (and swimmerets in crustaceans as well as gills in aquatic insect larvae).51 William McGinnis and his colleagues have demonstrated how laboratory experiments with Hox gene expression can reduce the number of legs in the brine shrimp Artemia.52 In terms of evolutionary tempo, there is no other description for this kind of experimental result other than “saltatory.” Once the regression of abdominal limbs in the early insect lineage

246

Chapter 6

occurred, expression of Hox was locked, resulting in adult forms that are all recognizable as insects. As De Beer (1940) emphasized, the evolutionary rigidity in the mature insect body plan is offset by the evolutionary plasticity of the larvae. Insects of species with almost identical adults can have quite different larvae. The most sophisticated holometabolic insects, like the Lepidoptera, undergo a metamorphosis that effects marked changes in form, physiology and behavior from the larvae to the adult form—the “imago.” “Cenogenetic” phenomena involve epigenetic changes in larvae—usually referred to as “feeding adaptations” or specializations—that deviate from the ancestral type. Hox expression in the caterpillar or maggot is different from that of the adult. Metamorphosis in insects is not the condensed hypermorphosis found in the development of frogs from tadpoles. Instead, once the parasegments have formed in the holometabolic insect embryo, some clumps of cells are set aside as “imaginal discs.” These grow considerably as the voraciously feeding larva develops, and then they differentiate during pupation, obtaining additional nutrients from recycled larval tissues. In a sense, two different organisms grow and develop out of the same genome, providing more nutritional and behavioral flexibility during the life cycle of the whole organism. An intuition of the evolutionary significance of segmentation of body parts, and their subsequent differentiation, was what gave E. D. Cope his concept of repetitive addition. William Bateson came close behind with “meristic variation,” and it is intellectually satisfying to see how the explanation of all of these organismal phenomena ties in so closely with the way they are repetitively differentiated at the molecular level. Although fruit flies with four wings, or with legs growing out of their heads, may be hopeless monsters, the universality of Hox genes indicates that some homologies are very ancient, conserved metazoan features. For every hopeless antennaped fruit fly, there are all the successful monsters that emerged in the past to give rise to new phyla and orders. Not only did Hox genes affect the arthropods, they were likely involved in the explosive diversifications of the Cambrian as well. Calling this universal phenomenon “deep homology” conveys the idea nicely. In the vertebrates Hox genes participate in craniofacial development, hindbrain morphogenesis, axial structure, including the anatomy of the vertebral bones, limb arrangements and other aspects of epigenesis.53 Many of these functions are controlled at the cellular level by the organizing effects of neural crest cells, as described in the previous chapter. The fish-like lancet, Branchiostoma lanceolatum (formerly known as Amphioxus), a non-vertebrate chordate, has a Hox cluster whose gene sequence anticipates the lineup of the multiple Hox clusters of jawed fish. And primitive jawless fish, still represented among the living by the lamprey and the hagfish, have two Hox clusters. This signifies evolution by homeotic gene duplication, once before the emergence of the Agnatha, and another time before the emergence of the jawed fishes,

Epigenetic Mechanisms

247

giving the latter four clusters. (This kind of evolution might also be considered to be orthogenetic.) In fish evolution the pectoral and pelvic fins may have emerged from continuous lateral fins by suppression of Hox genes in most of the intervening body somites.54 Further differentiations of the Hox genes that initiate pelvic and pectoral fins then led first to lobe fins in the sarcopterygian ancestor, limb formation in its tetrapod descendants, and finally regression of the number of digits to the familiar pentamerous pattern.55 Hox participation in limb growth is independent of somite, or vertebral number. Wnt The name Wnt originally signified “wingless” in Drosophila, which has seven variants that, like Hox, are involved in segmental organization. Also like Hox, Wnt genes were primitively involved in the differentiation of the antero-posterior axis, especially the posterior end. There is one Wnt in the freshwater cnidarian Hydra. It participates in the organization of the oral region, which is regarded as morphologically posterior. In the deuterostome lineage they influence the development of the blastopore region. Vertebrates may have 19 different Wnt genes, and one later evolutionary role is the morphogenesis of the mid and hind brains. They are also partly responsible for the early stages of organogenesis by influencing the separation of the mesoderm into the organ rudiments called “anlagen.” In addition, Wnt genes interact with migrating neural crest cells and help to induce tissue formation in the early embryo. Their interaction with other selector genes is summarized and extensively referenced in Schubert and Holland 2005. Pax Another deeply homologous homeotic gene type is the Pax-6 cluster responsible for initiating the morphogenesis of eyes. The implantation into Drosophila of a squid Pax6 gene cluster causes a multiplicity of eyes to appear all over the bodies of the fly’s offspring—fruit fly compound eyes, not squid camera eyes, mind you.56 This riotous result shows what can happen when genes are transferred to organisms that have no stable regulatory mechanism for them. But from the evolutionary point of view the more interesting point is that the lineages of squid and insects have been separate since they diverged from their last, common, unsegmented worm ancestor in the early Cambrian. Lacking a common ancestral organ, their eyes are non-homologous at the organ level but deeply homologous at the level of the homeotic epigenes that trigger their final construction. Walter Gehring (1998) discusses the earlier bridging of an even greater phyletic gap, namely between vertebrates and insects. Mouse eyeless is a molecular relative of Pax-6. When transferred to the imaginal discs of larval fruit flies, they induce ectopic eyes (i.e. eyes that aren’t where they ought to be) on the antennae, wings and legs. These

248

Chapter 6

transgenic eyes have normal photoreceptor cells that respond to light stimulation by emitting an electric pulse. Gehring also makes the point that such “master control genes” are not genes for eyes but are involved in eye induction along with another 2,500 genes.57 Although we now know that in animals that already have the capacity to develop eyes, certain genes can trigger their appearance in odd places, we still have to understand how all those genes cooperate, and how the capacity to develop eyes arose in the first place. On the ability of nature to experiment with eyes Gerhart and Kirschner have this to say: There are many more mysteries in the convergent evolution of eyes in different animals, but perhaps none so eerie as the complex anatomy of the cubomedusan jellyfish eye in a species without a brain. . . . The 24 eyes of this primitive jellyfish each have a lens, a cornea, a pigment layer, and a sensory layer. . . . The existence of an eye with anatomy similar to that of vertebrates in an animal without a central nervous system shows how far complexity of one part of function can go without the other. If jellyfish were ever to develop a brain, they would be ready to appreciate the world around them.58

The jellyfish does have a nervous system, and George Mackie (1999) shows that it possesses a considerable degree of organized complexity despite its non-centralized nature. Cubomedusan eyes have considerable resolving power and species that attach to substrates can possibly select suitable objects for settlement. These sea wasps orient toward light, swimming in response to the signals from the most strongly illuminated eyes, and some can avoid objects that cast shadows or create turbulence. Higher animals with image-forming eyes can, however, put them to a wider variety of uses such as finding mates and prey and avoiding predators. But if there is no “need” or “selection pressure” for these functions, can there be a selectionist “explanation”? The phenomenon may instead illustrate how far natural experiment in the epigenetic evolution of organized complexity can be taken prior to acquiring selective value. Brian Goodwin (1994) argues that there are universal generative conditions for the emergence of eyes in animals that include translucent epithelia, light-sensitive neurons, and the presence in the diet of light-sensitive pigments that came from plants that used them in photosynthesis. A range of potential crystallin proteins for lenses occur in every cell. Pax-6 genes and their relatives that stimulate lens formation and brain growth are common throughout the animal kingdom. And cells have cytoskeletons whose shapes and sizes can be changed in reaction with calcium to invaginate or bulge. Eye morphogenesis is in part triggered by simple physical forces. Goodwin writes: The processes involved are robust, high-probability spatial transformations of developing tissues, not highly improbable states that depend on a precise specification of parameter values (a specific genetic program). The latter is described by a fitness landscape with a narrow peak, corresponding to a functional eye in a large space of non-functional (low-fitness) forms. Such a system is

Epigenetic Mechanisms

249

not robust: the fitness peak will tend to melt under random genetic mutation, natural selection being too weak a force to stabilize a genetic program that guides morphogenesis to an improbable functional goal.59

The conventional focus on natural selection as the generator of improbable ends is what makes Michael Behe (1996) conclude that complex structures like eyes are irreducible, and what makes Richard Dawkins demand strict gradualism to accommodate reducibility. For example, as cited by Behe, Dawkins’s rationale is this: Evolution is very possibly not, in actual fact, always gradual. But it must be gradual when it is being used to explain the coming into existence of complicated, apparently designed objects, like eyes. For if it is not gradual in these cases, it ceases to have any explanatory power at all. Without gradualness in these cases, we are back to miracle, which is simply a synonym for the total absence of explanation.60

Since evolution cannot be the explanation of itself, Dawkins must be really arguing that evolution has to be gradual for natural selection to be have an adequate causal role. In contrast, Behe’s irreducible structure is one whose subunits by themselves confer no selective advantage, so they cannot be built as accumulations of adaptations. He infers that a mutation at any step in a specific genetic program may throw the whole process into confusion. Also, it is highly improbable that all of the mutations necessary to produce a novel complex structure could coincide. But Behe’s view is clouded by the conventional assumption that the accumulation of point mutations is essential. The default mode need not be the miracle of “intelligent design.” Goodwin’s mechanistic spatial-transformation model illustrates how eyes, as highly probable robust structures, have emerged independently a number of times and have persisted despite genetic accidents. They do not require an accumulation of point mutations, but self-amplifying processes that have been outlined above, plus the influence of epigenetic mechanisms involving Pax6 regulators. Autonomous complexifying processes that generate eyes arise from ubiquitous generative conditions. Their emergent transition into a more organized, and more energetically stable phase requires no directing selection pressure. This is how Kauffman as well approaches the emergence of complexity in The Origins of Order (1993). Furthermore, as Goodwin argues, the robustness of the process can opportunistically take alternative routes to reach the required form in different homologues. Epigenesis beyond Gene Determination In this chapter I have paid quite a lot of attention to DNA-epigenetic factors, and relevant RNA and proteomic factors that are necessary components of epigenesis. I too inhabit a genocentric universe, and I have not reached the escape velocity implicit in the idea of Stuart Newman and Gerd Müller (2000) that gene modification might only

250

Chapter 6

consolidate prior epigenetic changes. But I never intended that genes be taken as the controllers of specific epigenetic programs. To have any role they must interact with physicochemical, ultrastructural, organismal, and environmental factors. Some of these combinations have intriguing consequences—Marilyn Monk (1995) suggests that throwbacks and some apparent cases of inheritance of acquired characteristics might be due to the expression of deep homology through demethylation—hence derepression—and mutational modification of ancient ancestral “motifs.” Stephen Jay Gould’s 1983 essay “Hen’s teeth and horses’ toes” anticipated Monk’s conclusions, and shows that Darwin had already thought that deep within organisms crowds of “invisible characters . . . separated by hundreds or even thousands of generations from the present time . . . lie ready to be evolved whenever the organization is disturbed by certain known or unknown conditions.”61 The question of hens’ teeth, synonymous with scarceness, arose from an experiment by Kollar and Fisher (1980) that caused mouse mesenchyme to produce teeth, when chick gill arch epithelium was used as an inducer. By itself the mesenchyme only synthesized spongy bone. Chicken epithelium still has the robustness to induce tooth formation, after 60 million years of toothlessness in birds, although bird mesenchyme has lost the ability to react to the induction. That kind of plasticity is controversial, since tradition has carved the Use-It-Or-Lose-It Principle in stone. Epigenetic Inheritance Systems The term “epigenetic inheritance system” includes any process whereby an organism in general, or its cell lineages in particular, follow characteristic, established lines of development. The genome alone is not responsible for differentiation—it requires a variety of epigenetic interactions that achieve developmental consistency. Neural crest cells reveal how such systems can change and evolve. They are modified by the tissues through which they migrate, and in turn induce changes in some of those tissues. Some of the effects are physical, some chemical, and some involve gene regulation. A most satisfactory discussion is provided by Eva Jablonka and Marion Lamb in Epigenetic Inheritance and Evolution (1995), where the emphasis is on non-DNA systems, transcellular inheritance, and the possibility of transgenerational inheritance. Non-DNA Epigenetic Inheritance Systems As outlined above, the cytoplasmic inheritance of permease can give rise to two biochemical lineages in E. coli. Jablonka and Lamb provide a model that suggests how this could have been established by an ephemeral environmental stimulus. This has a general theoretical relevance for the kinds of internalization of environmental stimuli that fascinated Ivan Schmalhausen and C. H. Waddington. “Cytotaxis” is the inheritance of induced changes in the cortical architecture of unicellular ciliates, without reference to their nuclear DNA. Its existence has been

Epigenetic Mechanisms

251

known for the last 30 years. Also known as “directed assembly,” it has since been found in mammalian cells in culture, and in the epidermal cells of caterpillars.62 That the cytoplasmic architecture and non-genetic chemical make up of the ova of multicellular organism is responsible for a number of the earlier stages of epigenesis has also been accepted for a long time. “Chromatin marks” are defined by Jablonka and Lamb as “the non-DNA parts of the chromosomes, for example binding proteins or additional chemical groups attached to DNA bases that affect the nature and stability of gene expression.”63 Chromosomes in different cell lineages have different chromatin marks, which may change during the course of development, reflecting the activation status of the genes concerned. The chromatin marks that characterize particular developmental lineages of differentiating cells are largely passed on to the daughter cells. When DNA is being replicated, the methylation of the parent strand remains, but initially the daughter strand is not methylated until methyl transferase responds to the asymmetry by methylating the appropriate cytosine bases of the new strand. Similarly, the new strand lacks histones, and these are synthesized and assembled under the direction of protein transferase. Some errors in methylation are random, others seem to be directed by a modifier molecule or an environmental stimulus that acts through one. Almost as stable as point mutations of DNA, they are much more frequent. Jablonka and Lamb argue that chromatin mark systems are a key to understanding epigenetic evolution: Since the chromosome-marking maintenance mechanisms are independent of the functional state of the gene, and the gene product is not required to have a specific regulatory role, the system is potentially much more flexible. It is able to change both during development and during evolution.64

They point out that the genotype, including modifier genes and their products that participate in epigenesis, is relatively constant, but the non-DNA components of epigenesis exist in numbers and variations that provide for an almost unlimited combinatorial creativity. In addition, an acquired change to the non-DNA epiphenotype is more easily inherited than one that requires incorporation into the DNA itself. Several clear examples of such inheritance persisting through meiosis are known in protozoa. Beneficial qualities are more likely to persist, and, if heritable, likely to strengthen homeorhesis. Methylation For transcription to messenger RNA, DNA must be chemically as well as physically available. Its cytosine bases must be free of “methylation,” which is common in inactive DNA, and takes it out of the running for transcription even when it is uncondensed. Methylation involves the post-synthetic addition of methyl groups to cytosine bases, often found in CpG dinucleotides and associated with promoter

252

Chapter 6

regions. It changes the fit between the major groove of the DNA double helix and DNA-binding proteins. Methylation of histone tails also changes the interplay between acetylation and methylation of histones that affects chromatin structure. Specific types of proteins are capable of recognizing acetylated histones or methylated DNA, such as the methyl-CpG-binding proteins (MeCPs or MBDs). Whole complexes of chromatin remodeling factors are known to restructure local chromatin architecture, capable of converting areas between transcriptionally open or closed conformations. Methylation, which is caused by the enzyme methyl transferase, is not random, but a regulated procedure that ensures that differentiating lineages of embryonic cells only produce the proteins appropriate to their special requirements. Since methylation is absent from yeast and rare in Drosophila, there must be other ways of silencing DNA. Imprinting Mendelian genetics assumed that a gene is a gene, and the action of a gene in an organism was not affected by the sex of the parent who contributed it. But the phenomenon of imprinting has such implications. Two apparently different genetic “diseases” of humans are caused by the same mutation, but the phenotypic expression of the gene as one or the other of the two conditions depends on whether it came from the mother or the father. The sex of the parent somehow has an imprinting effect of the offspring. Moreover, whatever parent the imprint comes from in the previous generation, the imprinting is re-set according to the sex of the individual who will be a parent of the next one. The case for methylation as one of the causes of imprinting is promoted by Marilyn Monk’s 1995 essay “Epigenetic programming of differential gene expression in development and evolution.” Methylation patterns that may have developed ontogenically in a parent are undone during gametogenesis, but the ovum is slightly methylated, and the sperm slightly more so, to allow imprinting by each parent. The methylation mechanism regulates the expression of the gene, but is not itself coded in the DNA. The mechanism “falls outside the genic paradigm,” but has to be encompassed by the epigenetic paradigm.65 In Monk’s mice, the degree of methylation is at its lowest overall in the blastocyst stage of development, perhaps because the methyl transferase laid down in the cytoplasm of the egg by the mother has been gradually degraded and diluted by cell proliferation, and the embryo has not yet begun to synthesize its own enzyme. The germ line cells continue to be minimally methylated, but extra-embryonic tissues are progressively methylated, and the rise is even steeper in the somatic cell lineages. Methylation patterns of the latter are responsible in part for their differential features. Although there are a number of hypotheses regarding the value of imprinting in relation to organismic integrity, such as avoiding competition between the mother

Epigenetic Mechanisms

253

and the implanted embryo, Monk wonders if imprinting per se is simply a carry-over of methylation that had been primarily correlated with gametogenesis. More recent work has suggested the presence of cytoplasmic factors in the oocyte in addition to DNA methyl transferase (DNMT) can perpetuate imprints, including heterochromatin protein 1 (HP1) which recognizes methylated histones, and other transcriptional repressors. In mammals, imprinted genes are clustered into differentially methylated regions (DMRs), and this may protect imprints from the global methylations and demethylations that occur in embryos. Mammalian cloning, which made Dolly the sheep an international media star, involves the technique of somatic cell nuclear transfer. The nucleus of a differentiated cell is transplanted into the rich environment of an oocyte which has been stripped of its own genetic complement. This shows how epigenetic patterns can be reversed, since differentiated nuclei can revert to totipotency and (albeit rarely) provide for the whole of development. The failure of most cloning attempts of this type proves how important epigenetic patterns are, and that a complete DNA sequence complement is not sufficient to direct the cell cycle.66 How might these waves of demethylation and remethylation have evolved? Some argue that they provide the advantage of erasing any potentially damaging epigenetic modifications incurred in the parent. The existence of methyl groups on cytosine bases greatly increases the risk of a point mutation, in that these 5-methyl-cytosines are often spontaneously deaminated to thymine bases. Another correlation could be the relative activity of transposable elements, known to be more highly expressed in germ line—possibly as a result of the global demethylation. However, there is growing evidence for the persistence of methylation and histone-binding patterns between generations that amount to a neo-Lamarckist process of the inheritance of acquired characteristics. This will be discussed below. Found in a variety of animals, methylation reaches its most extreme form in some insects, where the entire chromosomal complement derived from a particular parent might be silenced as heterochromatin, or lost entirely. Removal of the previous generation’s methylation patterns in the early embryo is conventionally interpreted as the result of “strong selection against carrying the remnants of an individual’s epigenetic history into the next generation.”67 This is not a barrier to any emergent novelty of epigenesis that is based on DNA point mutation, exon shuffling and transposable element effects in the parental germ line. But Jablonka and Lamb wonder if there is any exceptional transgenerational inheritance of adaptive chromatin marking, particularly methylation patterns. In the fruit fly, methylation of “transgenes,” i.e., foreign genes experimentally transplanted into host cells, persists from one generation to the next. Persistent methylation of transgenes also occurs in protists, plants, fungi, flatworms, nematodes, crustaceans, insects, and mammals, including humans. Plants offer the bulk of the examples, and this is because somatic cells that may have been exposed to environmentally induced epigenetic changes enter the germ line quite late

254

Chapter 6

in the plant’s development. Barbara McClintock found that inactivation of the jumping genes of corn during epigenesis was heritable; methylation is responsible. Examples of heritable chromatin mark alterations are relatively sparse. Since methylation of the cytosines of nucleic acids is a relatively recent discovery there was no early guiding hypothesis for phenomena that did not obey Mendelian rules, except some kind of aberrant change in DNA. Chromatin mark changes may be random, as well as directed by environmental change. Early neo-Lamarckists were fascinated by highly visible somatic phenotypic responses to environment, but did not have experimental or observational access to changes that might occur in the germ line in early embryogenesis. Looking in the right places with modern molecular techniques and a working hypothesis based on methylation effects is now expanding the body of evidence pertaining to non-DNA inheritance systems. Leslie Pray (2004) reviews some current research on the inheritance of methylation patterns, and histone binding, under the title “Epigenetics: genome, meet your environment,” commenting that “as the evidence accumulates for epigenetics, researchers reacquire a taste for Lamarckism.”68 This article shows how the mainstream comes closer now to the Jablonka and Lamb direction.69 These are indeed heritable changes, but strictly speaking, they are neo-Lamarckist effects, and some are detrimental conditions produced under stress. Epigenetic Effects of Environmental Stress Compounding methodological problems is the scarcity of information on what environmental changes affect chromatin structure in such a way as to alter gene expression, and activate enzymes, repair mechanisms, and changes in transposon action. The role of heat-stress proteins in response to physicochemical stresses has already been discussed. Recombination is increased by the effect of heat stress on highly condensed chromosome regions. The same stimulus increases the vulnerability of DNA to the action of transposable elements, whose activity is also modified by changes in their chromatin structure. Such examples show that the environment can have an efficient causal effect. The potential of a genome to respond to stress was called “catastrophe insurance” by Koch (1993). (Recollect that I have defined stress as a condition that requires abnormal and energy-demanding compensatory responses.) But the genome responds in a larger context of behavior, physiology, and environment. For example, stress arises from the domestication of wild stock, where behavior is restricted and new environmental regimes are imposed. Hormonal effects on domestically bred silver foxes, and the parallel case of Djungarian hamsters, have already been mentioned. Another example is the Laysan duck, whose entire population arose from a low of ten individuals 90 years ago. A few birds that were subsequently domesticated began to produce novel plumage colors within a few generations. Since the ducks had already

Epigenetic Mechanisms

255

been inbred for most of their history, alteration of the color of the feathers is less likely to have arisen from continued inbreeding than from the domestic breeding environment.70 A natural parallel may be the rapid speciation that occurs in isolated and limited environments such as islands, and in new lakes, as a result of stressful changes in behavior and social structure, as well as physicochemical differences. As ever, note that the typical agents of natural selection, competition and predation, are absent under those circumstances. Stress responses can come from molecular mechanisms other than heat stress proteins.71 Mutability is known to increase in bacteria under stress, which in a natural setting may be most of the time. Moreover, some eukaryotes that normally reproduce asexually switch to sexual reproduction, thus increasing variability through recombination. In a range of effects of environmental change, the least stressful simply alters gene expression through slight alterations in chromatin marks. If the lineage has experienced periodic, specific stress, such as the lack of an essential nutrient, preexistent mechanisms may be activated. One example is “directed” or “adaptive” mutagenesis in bacteria. General stress, such as change in oxygen tension or temperature change, may increase gene duplication and mutation, resulting in increased physiological adaptability. Acute stress may result in the activation of errorprone DNA repair mechanisms, increase in the activity of transposable elements, and crossing over during meiosis. Catastrophic Stress If stress has such a prominent epigenetic impact, the expression “catastrophe insurance” is doubly significant. Canalization might be loosened as a direct result of catastrophic environmental change. I have noted that major emergent forms often arose before major bolide or geophysical disasters, and then diversified explosively in their wake. Catastrophes cleared the bench of the usual agents of natural selection, allowing earlier natural experiments, especially those with emergent adaptability, to multiply and diverge. But this does not exclude the possibility that some new emergences might have occurred as a direct result of sublethal heat shocks occurring at the fringes of impact or volcanic blast zones. Because water buffers heat, slow sustained rise in temperature would be more likely for aquatic organisms than sudden heat shock. But that might be even more effective than a transient stimulus. Canalization in hopeful emergents might have been loosened as a direct effect, thereby stimulating rapid diversification. In addition, small, highly stressed populations of survivors would have been involved. The combination of an imposed founder effect with physical and biological environmental changes, especially altered interactions between individuals, would have had some effect on epigenetic variability.

256

Chapter 6

Origin of Epigenetic Inheritance Systems Epigenetic inheritance systems are of major importance in multicellular organisms where true-to-type lineages must originate and reach their predictable goals as parts of complex mature forms. But how and where did the epigenetic inheritance systems originate? John Bonner (1974) calculates that the multicellular condition has emerged independently at least seventeen times, the major products being plants, fungi, and animals. (All animals are presently considered to have arisen from a common, probably unicellular, ancestor.72) Simple multicellular organisms had the emergent qualities of being hard to eat and slow to starve. Without much differentiation, they also had an easily realizable potential to eat larger things, store more food, locomote more efficiently, and commit themselves more effectively to reproduction. Looking at the progressive evolution of ever higher emergent levels, Bonner (1988) saw that while there is no strong correlation between genome size and phylogenetic status, adaptability and the number of different cell types are correlated. Higher plants have about 30, higher invertebrates have about 55, vertebrates more than 120. It could be argued that epigenetic inheritance systems participate in the cell cycles of unicells, and are not therefore novel emergent qualities of multicells. Yet there is novelty in the emergent qualities of chromosome packaging and methylation mechanisms. And it is in multicellular organisms that differentiation is so dramatically complexifying. As Erasmus Darwin and Herbert Spencer speculated long ago, the differential effect of the physicochemical environment on a multicellular eukaryote would alone be sufficient to prime cellular differentiation in every generation, without DNA-epigenetic inheritance systems. These effects are seen in some quasi-colonial unicells, provided that the appropriate environmental stimuli are constantly present. DNA-epigenetic systems are necessary for the complexity of multicellular organisms, but they may all have arisen through interaction with environmental effects. And they also require complementary factors such as adhesion/integrin molecules, cell junctions, and the establishment of intercellular induction mechanisms. The elaboration of skeletal matrices that would physically affect tissue assembly would also enhance structural organismic integrity. Cellular cytoskeletons, integrins, the basement membranes of epithelia, and cell walls, as well as the more visible structures found in sponge skeletons, coral concretions, mollusk shells, chitinous exoskeletons, echinoderm spines and plates, and vertebrate cartilages and bones all contribute to the self-assembly of tissues and organs. Jablonka and Lamb note that epigenetic inheritance systems had a double role in the transition to complex multicellular organisms: First, they enabled the emergence of a new unit of structure and function, the phenotypically distinct cell lineage. Second, they allowed the formation of the stable interdependences between

Epigenetic Mechanisms

257

epigenetically distinct cell lineages, which resulted in the evolution of integrated organism from loose groups of cells.73

They also anticipated Newman and Müller (2000) with their argument that the phenotypic epigenetic changes produced by extrinsic and intrinsic environmental effects were prior to their canalization by genetic assimilation and natural selection. Sexual Reproduction Bell (1988a) asserts that the absence of sexual reproduction kept living organisms in a primitive unicellular state for about 2 billion years. Furthermore, natural experiments in multicellularity may often have been tried during that time, but failed because of the lack of the repair facilities that emerged with meiosis. Thus, the tradeoff between faithful reproduction and experimental flexibility held early progressive evolution back until a saltatory boost arrived in the form of sex. For most of the time in question prokaryotic unicells could complexify themselves by gene acquisition through a variety of routes, and could reproduce asexually. Although some engaged in conjugation, which falls within a loose definition of sexual mating, sex in eukaryotes involves chromosomes. Chromosome packaging was an immediately advantageous feature for mitotic asexual reproduction, and it potentiated sexual reproduction as well. Membrane adhesion molecules, and the pre-existent experience of conjugation were other generative features. The natural experiment of sexual reproduction succeeded because of the prior lack of reconditioning mechanisms—there was nothing to prevent it at that stage. Once in place, repair mechanisms could be refined, and a new, efficient level of change-resistant dynamic stability (i.e. homeorhesis) established. In sexually reproducing organisms it is common for there to be impregnation of a haploid egg by a haploid sperm, returning the number of chromosomes to the diploid or double number characteristic of the type. This is sexual reproduction, but did not initially require the differentiation of sex chromosomes and phenotypically distinct (dimorphic) genders. Those emerged later and independently in different plant and animal lineages. The familiar Y chromosome of human males probably appeared early in the mammalian lineage. It retains sufficient of the collinear genes of the X chromosome to demonstrate that it arose from such an ancestor. The major sexdetermining trigger is the protein synthesized by the SRY gene (sex-determining region Y), although there remain a number of “housekeeping” genes that operate in every cell. Once maleness has been catalyzed, a number of genes from other chromosomes are involved in the final determination of gender in the mature mammal. Testosterone production in the testes is an important part of this process. The Y chromosome retains the ability to pair with the X chromosome at synapsis during Meiosis I of gametogenesis, but for most of its length it is incapable of crossing over with the X chromosome. As Karin Jegalian and Bruce Lahn (2001) write in “Why

258

Chapter 6

the Y is so weird,” the Y chromosome retains much of its evolutionary history because of the lack of recombination: . . . the Y lost the ability to swap DNA with the X in an unexpected, stepwise fashion—first involving a swath of DNA surrounding the SRY gene and then spreading, in several discrete blocks, down almost the full length of the chromosome. Only the Y deteriorated in response to the loss of X-Y recombination, however; the X continued to undergo recombination when two copies met during meiosis in females.74

Now, I have already inferred that a number of molecular processes from DNA mutations, jumping genes, and chromosomal mutations are all-or-none saltations. As these authors express it we here have an example of saltation through chromosomal mutation when the first chunk of the proto-Y containing the SRY gene inverted. The leap at that stage was neither advantageous nor detrimental; but it might have resulted in reproductive isolation of the organisms in which it happened. This was around the time when the mammalian lineage separated from the reptiles. Further inversions seem to have occurred at or close to the bifurcations of the mammalian line that gave rise to the placentals, and anthropoid primates (monkeys, apes and humans). Use-itor-lose-it is invoked to explain the deterioration and shrinkage of much of the Y chromosome. The emergence of chromosomally different sexes was accompanied by sexual dimorphism This has evolutionary implications since the fate of future generations now depends on the reliable interaction of two sets of gametes in two organisms that have not only developed as differentiated genders, but may not only look different as mature organisms, but in the case of animals may behave differently, use different resources, and thereby complexify their ecological niches. Furthermore, the necessity of sexual interaction is what potentiated the familial interactions of the higher mammals. Summary of Epigenetic Mechanisms 1. Point mutations of the genes that code for structural proteins and enzymes have a role in adaptational and adaptability evolution. More important for epigenetics are numerous genes that can modify structural gene activity by coding for regulatory proteins and enzymes, or by interacting with the regulatory proteins to act as switches. Alteration of these controlling mechanisms may contribute to epigenetic evolution. 2. In the egg and the early embryo there are gradients of regulatory RNA and protein molecules whose alteration may affect thresholds of activity and have an evolutionary effect, prior to involvement of the genome of the new organism.

Epigenetic Mechanisms

259

3. Modifier genes and the actions of their products are hierarchically controlled, so that a single alteration can have a multiplier or cascade effect. 4. The action of transposable elements, in amplifying or changing established regulatory patterns, can be a mechanism for radical epigenetic change. In many cases this is a non-random process. While epigenetically potent, jumping genes lack the risks of point mutation. 5. Non-DNA influences such as chromatin marks, involving methylation and histone binding, can be initiated by environmental stimuli. These can in some limited instances be inherited. 6. Consistent reduplication of genes to produce gene families and repetitive differentiation is assured by the action of transposable elements and by mutations involving recombination enzymes. 7. Genetic drives, involving codon amplification, gene duplication, concerted evolution, and variant repeats (repetitive differentiation), may be responsible for allometric growth during development and for the phenotypic phenomenon of orthogenesis. 8. Concerted evolution ensures that all of the members of a clade possess the same pattern of repetitive differentiations. 9. Novelty in structural genes is not only caused by point mutation, but also by new cistron combinations, intragenic duplication, exon shuffling, and the conversion of introns to exons. Mutability is increased through methylation and the weakening of repair mechanisms under stress. 10. There are deep homologies among selector genes, e.g. Hox and Pax, that contribute to atavisms, saltatory changes, and parallel and convergent evolution. 11. The role of retroviruses in transphyletic exchange of genes is real and known in a few cases to be advantageous, but its general significance is enigmatic. 12. Genes participate in epigenesis, and genes are regulated by other genes. These activities are, however, controlled by non-genetic factors. Within the organism, these include chemical heterogeneity and changes in the ultrastructure of the cytoplasm, the shape and size of cells and epithelia, and the proximity of chemical and physical inducers in the internal milieu, and the behavior of organizer cells. Development and

260

Chapter 6

epigenetic evolution are expressions of the whole organism and its female parent operating in the larger environment, as well as expressions of genes. 13. Some epigenetic inheritance systems, such as cytoplasmic composition and architecture of the egg are not coded in DNA but can alter the phenotype of the offspring. Within a given species the size of the egg alone can be enough to drastically alter the form of the developing embryo. Non-DNA maternal effects can carry the embryo through its early stages and influence the establishment of its cell lineages before its own gene-based, epigenetic organizing mechanisms kick in. Moreover, maternal effects that are altered by prolonged environmental change may ultimately be genetically accommodated. 14. The semi-independence of embryonic cell lineages, and their internal hierarchical arrangement makes it possible for radical epigenetic change to be integrally accommodated in emergent organisms. It also allows the orderly shutdown of multiple cell lineages during paedomorphosis. 15. Subsequent to emergent epigenetic changes, the fine tuning of homeorhesis becomes a general quality of a clade (or of the cell lineage). The more dynamically stable (or generatively entrenched) these systems are, the tighter canalization becomes. Development is more likely to keep true to type. 16. Intron dissemination, repetitive differentiation, molecular drive, self assembly, anticipation, and other aspects of complexification can occur without causal reference to adaptiveness—in other words, “out of the sight of natural selection.” 17. The greatest challenge for developmental evolutionary theory is the difficulty of modeling epigenetic programs that would contain all of the real variables that have affected evolution. Afterwords The Dover TRAM Although I compiled the above material independently, from a variety of sources, Gabriel Dover’s book Dear Mr. Darwin (2000) brings together similar mechanisms with evolutionary potential under the heading of TRAM systems, an acronym for “Turnover, Redundancy, and Modularity.”75 Although I agree with the comprehensive principles that Dover is trying to establish I will not borrow his acronym. “Turnover” could apply to mitosis, meiosis, recombination, gene transposition, and gene

Epigenetic Mechanisms

261

conversion, all of which could undergo evolutionary experiment through unequal crossing over, DNA slippage and molecular drive. So it is difficult to remember all of its implications. “Redundancy” is redundant, since it is a subset of modularity. Nevertheless, Dover makes the important point that co-evolution of compensatory molecular systems could allow sublethal experimental results to persist. At the molecular level, “modularity” applies to codons, exons, genes, protein domains. Such modules, or holons, make it possible to produce “new” genes and proteins, both regulatory and structural. Then, there are cellular modules involved in development and hence in epigenetic evolution. The interaction of modules is a necessary aspect of self-organization. It is therefore essential to remember that the organism is greater than the sum of its modules, regardless of the level at which analysis is focused. Beyond that, do not trams run in directions determined by their environment (in the form of rails?). Environmental Direction Environmentally induced epigenetic inheritance system changes, usually involving methylation patterns, are common in unicells and plants, where the germ lines are not isolated from early development. The evidence from animals is sparse, perhaps because animal epigeneticists have not thought to look for it. Even if animals have not been generally influenced by these kinds of epigenetic effects, there would still be an important distinction between plant and animal evolution that needs elaboration. Genetic assimilation, or the larger category genetic accommodation falls between non-DNA epigenetic inheritance and conventional natural selection of alleles, but the process could just as well be called “environmental direction.” Unquestionably the environment can induce phenotypic change that involves ontogenic modification of existing heritable genetic factors. Their “adaptiveness” is also altered before any change in the genes themselves. Where there is a coincidental competitive advantage arising from these phenotypic modifications, any directional genetic shift will reinforce anatomical change. Internally, as Schmalhausen proposed, coordinative conditions that are energetically economic, and autonomization of environmental stimuli will lead to greater stability, and tighter canalization. And as Newman and Müller say, gene change is after the fact of epigenetic change. Natural Genetic Engineering Before Darwin voiced any thoughts on the matter, the English evolutionist W. C. Wells commented that what breeders do “by art seems to be done with equal efficiency, though more slowly by nature.”76 At the conclusion of my book Evolutionary Theory: The Unfinished Synthesis (1985) I paraphrased this to ask “Is what genetic engineers do by science done with equal efficiency, though more slowly by nature?” James Shapiro had already anticipated that question in 1977, with the multifold actions of transpos-

262

Chapter 6

able elements in mind. The tools of genetic engineers—plasmids and various enzymes for snipping and splicing—are all borrowed from nature. And transgenes are natural structures slipped into a foreign genome, an operation not all that different from natural bacterial transformation and retroviral transgenesis. Shapiro extends the simile in his 1992 essay “Natural genetic engineering in evolution”: . . . evolutionary novelty often does not reside in the invention of new biochemical processes by the continual modification and selection of individual proteins. Instead evolution appears to proceed by the utilization of basic biochemical routines in different combinations in different organisms. With few exceptions, the structural proteins of all mammals, for example, are probably interchangeable; what makes a mouse different from an elephant is when and how those molecules are synthesized and assembled during development. . . . Much of genome change in evolution results from a genetic engineering process utilizing the biochemical systems for mobilizing and reorganizing DNA structures present in living cells.77

From the large catalogue of information about the causes and effects of molecular modification, all that the proponents of the Modern Synthesis have chosen to explain evolution is the accumulation of adaptive random mutations by an infallible metaphorical force. This is not only hopelessly narrow, but scarcely even relevant to Shapiro’s radical way of thinking. He properly demands that any known mechanisms that can bring about change through an orderly hierarchical rearrangement of gene expression must be integrated into evolutionary theory. He points to some cases of large-scale genomic change, such as the chromosomal fragmentation and reintegration that occurs in ciliates during reproduction. The cascade of DNA effects that result in the enormous array of antibodies in vertebrates belong in this category too. I have already described the changes caused by transpositions that lead to hybrid dysgenesis in some Drosophila. Some of the offspring of any animal that had multiple ova and sperm altered by transposable elements, would be ready to form a distinct new lineage. We know that genomic and organismic integrity is retained in such cases, so we are further along in answering Bateson’s perennial question about the accommodatory mechanism. J. H. Campbell, who is quoted at the heading of this chapter, takes the view that mechanisms of genome regulation have evolved to meet the goal of providing novelties of variation rather than for ensuring the continuance of the status quo. Some jumping genes could certainly fit the role of active agents of potential evolutionary change. Others are provided by Lynn Caporale (2003). To continue the engineering metaphor, epigenetic mechanisms provide a variety of machine tools, which do not produce particular products, but make the specialized tools that do. However, the inference that epigenesis is genetic in the narrow sense of being all to do with genes and their expression comes from the genocentric universe, where other factors are completely ignored. The major achievement of Jablonka and Lamb’s Epigenetic Inheritance and Evolution (1995), nicely dedicated to “parents who gave us more than

Epigenetic Mechanisms

263

genes,” is to move us toward a system centered on organismal holism, which in turn wheels around in the larger environmental sphere. This is complemented by Susan Oyama’s philosophical propositions concerning epigenetic evolution (Oyama 1985 and 2000). Epigenetic Algorithms Mechanical metaphors have appealed to many philosophers who sought materialist explanations of life. The definitive work on this subject is T. S. Hall’s Ideas of Life and Matter (1969). Descartes, though a dualist, thought of animal bodies as automata that obeyed mechanical rules. Julien de la Mettrie applied stricter mechanistic principles to humans in L’Homme machine (1748). Clockwork and heat engine models were popular during the Industrial Revolution. Lamarck proposed hydraulic processes as causes of variation. In the late nineteenth century, the embryologists Wilhelm His and Wilhelm Roux theorized about developmental mechanics. However, as biochemical and then molecular biological information expanded, popular machine models were refuted, but it is not surprising that computers should have filled the gap. Algorithms that systematically provide instructions for a progressive sequence of events seem to be suitable analogues for epigenetic procedures. A common error in applying this analogy is the belief that the genetic code, or at least the total complement of an organism’s DNA contains the program for its own differential expression. In the computer age it is easy to fall into that metaphysical trap. However, in the computer age we should also know that algorithms are the creations of programmers. As Charles Babbage (1838) and Robert Chambers (1844) tried to tell us, the analogy is more relevant to creationism than evolutionism. At the risk of offending the sophisticates who have indulged me so far, I want to state the problems in the most simple terms. To me, that is a major goal of theoretical biology, rather than the conversion of life to mathematics. DNA contains information that can be transcribed into RNA sequences, and then translated into proteins. It also contains promoter sequences that constitute on and off switches for transcription. Some of the RNA and proteins that are synthesized can flip those switches; and some of the regulator molecules are the products of enzymatic reactions. So the big question is, where is the program of instruction, or epigenetic algorithm, that can tell those mechanisms where and when to operate? Although promoters in bacteria are quite well understood, the number of binding sites on the promoters of eukaryotic chromosomes are considerable. If combinations of bound sites are involved in downstream gene regulation the potential number of activations is astronomical. For example, there is a multifunctional protein involved in sea urchin development whose primary sequence is coded by a structural gene endo16. Its promoter sequence is known to have 50 binding sites, for 20 of which there is a known binding factor.78 This promoter molecule, and hypothetically all other modifying

264

Chapter 6

molecules, might be thought to contain an algorithm that demands precise outputs in response to molecular inputs, though it would be stretching the point to describe the system as a “computer.” Nor is it an algorithm for development. It is an algorithmic switching mechanism for the regulation of endo16 protein production. Presumably many structural genes are regulated in similar ways, making for a highly complex array of switches waiting to be differentially activated by proteins and steroids during development. The cells of the developing organism can also be considered to be algorithmic, since they mediate between regulatory messengers coming from other cells in the embryonic milieu. If there is such a thing as a biological computer it is the whole organism, which also happens to be the computer operator and service technician. Hypothetically, in a simple unicell, all systems could go all out until the cell reaches critical mass and then divides. But this ignores all kinds of problems about acquiring nutrients, making choices of biochemical pathways, and generally maintaining cellular homeostasis, especially in a changing environment. Without there being any change in the DNA or RNA of that unicell, physiological strains can diverge and persist. In a multicellular organism, the first few cell divisions may occur through simple mitosis, without any differential expression of nuclear DNA. But the cells can differentiate as a result of the heterogeneity of the fertilized egg, such as is provided by the distribution gradient of the messenger RNA that codes for bicoid protein in Drosophila.79 The concentration of the protein is an epigenetic factor in the sequential differentiation of the anterior series of embryonic structures. The heterogeneity of the egg usually constitutes a heritable pattern, but is it constructed by a genetic program, or by the totality of the female organism? In this case, probably the former, but there are other cases in which the female’s nutritional condition, and therefore her behavior in the face of ecological encounters, are certainly influential. Matsuda (1987) demonstrated how the amount of yolk laid down in crustacean eggs would affect development. In early epigenesis, DNA regulation soon becomes involved in cellular divergences, and the activating mechanisms are often the end products of protein synthesis and their enzymatic actions. But often those are affected by the physicochemical condition of the developing organism, and from the larger environmental conditions that lie beyond. Epigenesis does not start with the genes nor finish with the genes. Nor does it start and end with the environment nor any hierarchical level between it and the gene. But the system could not work without the participation of all those levels— which is essentially the “interactionism” of Susan Oyama (1985). Pinning down bits of epigenetic algorithms at the gene level is possible, but understanding the operation of the whole organism is not made much easier by seeing it as a computer.

Epigenetic Mechanisms

265

Is It Lamarckism? Lamarck applied four laws to progressive and adaptational evolution. The two that refer to a trend to increase in overall size, and allometric growth in proportion to utility can be understood in broad epigenetic terms. Another states that the organism responds to “needs.” Logically, the metaphor is equivalent to the idea that the organism responds to selection pressures. But there is a deeper reality to Lamarck’s version: Any organism has the real and essential feature of responsiveness. If it is an animal, it also has greater freedom to behave in a variety of ways. And if it keeps behaving in the same way it may be constantly subjected to the same environmental influences. As I have emphasized in previous chapters, this will result in some ontogenic, phenotypic changes that might eventually be genetically accommodated. The speed at which the genotype co-opts the phenotype will depend on how varied and numerous the genetic molecular experiments are, and how lucky the correspondence between them and the current actions of the organism are. But in the meantime the animal is not restrained from continuing to behave in a particular way, nor from changing its habits. The more adaptable it is the more freedom it has, although it risks regression if it specializes too much. Unlike the neo-Lamarckists, Lamarck did not believe that multicellular organisms acquired characteristics imposed directly by the environment. But in its essentials Lamarckism involved an autonomous evolutionary progress modified by response to the environment. For him, the inheritance of acquired characteristics was not the crucial law, and he believed it could get in the way of gradation to more complex forms. As Richard Burkhardt Jr. convincingly argued in The Spirit of System (1977), modern biologists and philosophers have lost sight of what mattered to Lamarck, and taken the inheritance of acquired characteristics to characterize his thought. Suppose we were to continue to insist that the crux of Lamarckism comes down to the inheritance of acquired characteristics; to what extent might it be possible? Early in this chapter I gave an example of a cytoplasmic feature, the presence of permease in bacteria, that is passed on to future generations, and so establishes distinct physiological strains, without DNA involvement. This is literally the inheritance of acquired characteristics, and would probably have satisfied Lamarck. Transgenerational methylation patterns and other chromatin marks are in the same category. The acquisition of foreign genes, or paraheredity, can occur in bacteria through transduction and conjugation. The emergence of the endosymbiotic eukaryote was a major feat of whole genome acquisition. Eukaryotes can also pick up exotic genes through the action of parasitic bacteria like Agrobacterium tumefasciens and retroviruses. The genome projects are shedding some light on this, and such effects are more common than previously thought. But the acquisition of heritable characteristics is not the inheritance of acquired characteristics.

266

Chapter 6

Lamarck understood that adaptation, however it came about, could get in the way of evolutionary progress. In contrast, Darwin, who finally imported the inheritance of acquired characteristics into his own theory, thought that adaptation through natural selection was the mechanism of evolution. And the ultra-Darwinists have taken that interpretation to extremes. When Lamarck inferred that the organism could respond to environmental change he was right, in the sense that if the organism was already adaptable it could make the appropriate changes in its behavior and might finally acquire appropriate anatomical modifications. What about neo-Lamarckism, a synthesis even more elastic than neo-Darwinism? There are two specifics that neo-Lamarckism added to Lamarckism: the environment can directly cause changes in the organism; the changes are heritable and, though random, some are advantageous. Therefore it is within neo-Lamarckism rather than Lamarckism that heritable changes in methylation and histone binding patterns should be placed. Neo-Lamarckists also adopted physiological and behavioral adaptability. Physiogenic change imposed by the environment has been an important part of physiological evolutionary history, and some environmental stimuli, such as heat shocks, can loosen canalization and increase transposon activity with heritable consequences. However, if these are changed by the environment, they should be interpreted as neoLamarckist rather than Lamarckist. Lamarck properly emphasized the importance of choices made by individual organisms for subsequent evolution, something that was lost by the “population thinking” of the Modern Synthesis. In animals especially, specific behavioral responses to environmental changes are important, and the higher the emergent level of physiology and behavior the more choice the animal has. Neo-Lamarckists properly emphasized the effect of the environment on the individual, with the qualification that it was not always beneficial. Beneficial or not, there seem to have been many incidents of emergent evolution that either would not have occurred or would not have succeeded, without disequilibration of internal or external environment. I don’t accept the inheritance of acquired characteristics, in the sense of environmentally directed adaptive gene construction, but I do appreciate that Lamarck was closer to a generative theory of evolution than Darwin.

7 Orthogenesis

In many other cases, modifications are probably the direct result of the laws of variation or of growth, independently of any good having been thus gained. But even such structures have often, as we may feel assured, been subsequently taken advantage of, and still further modified, for the good of species under new conditions of life. —Charles Darwin, 18721 Organisms develop in definite directions without the least regard for utility, through purely physiological causes, as the result of organic growth. —Theodor Eimer, 18982

Darwin wrote that “laws of growth” were independent of natural selection, giving the example of ubiquitous hooks on the stems of bamboos, which gave no advantage to normal plants, but which were turned to good use by creeping and climbing bamboos. As his epigraph implies, he believed that trends of variation could be emerge without initial selective value, though the affected lineage might find them advantageous at some stage. St. George Jackson Mivart (1871) had argued that “innate tendencies” of growth were responsible for the parallel evolution in tooth structure between marsupials and placental mammals, although the parallel features had been absent from their common ancestors. The quotation from Eimer gives some of the sense of the law of growth named “orthogenesis”—it implies an evolutionary process that goes in a definite direction through an autonomous drive. Many examples of directional exaggerations of particular anatomical characters, which come under the category of allometry, are to be found in the fossil record. Enlargement of the canine teeth in evolutionary lines of “saber-toothed” cats, and the increasing proportions of the antlers of the Irish elk are well known. Since those lines are extinct, orthogenesis was thought capable of uncontrolled ultramorphosis to a form that could not easily survive. Although orthogenesis cries out for re-invention, and belongs with developmental evolution, I have saved this discussion until its molecular context could be set out in the previous chapter on epigenetics.

268

Chapter 7

Once a respectable evolutionary idea, orthogenesis was prominent in textbooks until the 1940s. The particular concept of “straight-line” and “definitely directed” evolution was introduced by Wilhelm Waagen in 1867. Wilhelm Haake, who coined the word “orthogenesis,” and Theodor Eimer, who popularized the concept, regarded it as the effect of an autonomous force or drive.3 Some modern evolutionists believe that developmental constraints might be responsible for phyletic trends, something Richard Goldschmidt had considered in 1940. They ignore his further proposal: What is called in a general way the mechanics of development will decide the direction of possible evolutionary changes. In many cases there will be only one direction. This is orthogenesis without Lamarckism, without mysticism.4

At one time it was proposed that human evolution had an orthogenetic component. The Australopithecus and Homo series both show progressive phenotypic paedomorphosis. Bolk (1926), who called it “fetalization,” argued that it was caused by an innate directional tendency that lacked any immediate adaptive value. This got him into hot water with critics who otherwise agreed with the sequence of neotenic events in human evolution that he suggested. To satisfy neo-Darwinism, every separate anatomical modification must have an immediate adaptive value, scrutinized by natural selection. It is either that, or a phenotypic manifestation of a genetic unit that pleiotropically produces a different but adaptive trait, subject to selective approval—a common fallback proposal of mid-twentieth-century population biologists. However, it is an argument from ignorance unless genetic proof exists. According to Bolk, some adaptational fine-tuning was ultimately involved, but only after the primary “consecutive” or orthogenetic features had been built up in harmony with the whole, without necessarily conferring an adaptational superiority. In small groups such innovations would spread by inbreeding. As late as 1940, Gavin De Beer, in his book Embryos and Ancestors, had no trouble equating orthogenesis with allometric growth shifts that harmoniously alter the anatomical proportions of organisms during development. And although it would be safer for me to do the same, I prefer to recognize the intuitions of the past and retain the word “orthogenesis” to represent a larger phenomenon. Julian Huxley referred to it positively in the first edition of Evolution: The Modern Synthesis (1942), citing paleontological observations of general orthogenetic trends. For example, the trend in fossil amphibian labyrinthodonts involved flattening and broadening of the head and anterior body. The skull was shortened, and the bone structure of the anterior cranium was extended downward, and simplified. These all occurred synchronously in disparate lineages.5 At the time Huxley warned against excluding adaptationist interpretations of such phenomena, and in his 1962 second edition he came much closer to Bernhard Rensch and Ernst Mayr. They had taken orthogenesis as no more than the directional selection of random mutations that affect development.

Orthogenesis

269

“Orthoselection” was a term coined by Ludwig Plate (1913) for what neo-Darwinists now call “directional selection.” According to a personal communication from the biologist-historian Igor Popov, Plate was a born-in-the-bone orthogenesist who regarded orthoselection not as a substitute for orthogenesis but as a complement to it. The story is elaborated by Georgy Levit and Uwe Hossfeld (2006). I argue that some stages of orthogenesis have no adaptiveness, but there might come a point in the exaggeration of a trait where it might be distinctly advantageous. Then it would increase demographically, no longer restricted to the originating lineage. I emphasize that the increased adaptiveness of the orthogenetic trait results from orthogenesis, and directional selection is a redundant concept. However, neo-Darwinists argue that directional selection is all that is required to make a variant that will become a general population characteristic. If ecological conditions are stable, further exaggerations of that variation in the same direction will also spread throughout the population. Take the example of plumage: if the colorful display of a proto-peacock is attractive to peahens his offspring are likely to be more numerous. Further exaggerations of male plumage are likely to have a similar effect, so selection of the variation in the direction of more and more colorful display will go on and on, perhaps to the point of handicapping the individual in other respects, such as difficulty in locomotion, and attracting predators instead of mates. Exaggerations of the lobate lines in the extinct ammonites can be explained according to the same logic. These creatures were shelled cephalopod mollusks similar to the modern chambered nautilus. Their lobate or suture lines mark the septa between the chambers of the shell. The outermost septum is an interface between the living organism, in the outer chamber, and the interior, water and gas-filled chambers that confer buoyancy. The larger the surface area of the septum, the more rapidly salt and water can be pumped out of the inner chamber, to be replaced with gas. Thus flotation is increased, and adaptationists are provided with a functional explanation for the elaboration of the lobate lines. Historically what started out as simple curves in ammonites became convoluted to foliaceous structures and the clade became extinct. Then a simple lobate line reappeared late in the era, only to be extinguished along with all the ammonites by the catastrophe that terminated the Triassic. But no explanation of ow the lobate line exaggeration was caused was proffered. Thanks largely to Bernhard Rensch’s Evolution Above the Species Level (1959), orthogenesis was displaced by directional selection. Yet it was cast it away for ideological reasons, before the molecular basis of heredity was understood. Now, since few molecular biologists have any sense of history, they have not thought to revisit orthogenesis. Modern authors only raise orthogenesis to immediately dismiss it as a historical fallacy and display how selectionism has progressed beyond such a mistake. Orthogenesis “is one of the many examples demonstrating that, in science, it is fallacious to suppose (as many people do) that truth must reside in a compromise, or

270

Chapter 7

middle ground, among opposing views.”6 This attempts to kill two distasteful birds— orthogenesis and the dialectical synthesis—with one stone. Often the errors of the old horse evolution exhibit at the American Natural History Museum are trotted out to discredit orthogenesis. Four specimens had been chosen to demonstrate an orthoevolutionary trend involving the reduction of toes, increasing leg length, increase in overall size, and molar tooth specialization. But “the pull of the present” was more responsible than the push of the past; i.e., the assumed reality of direct linear trends resulted in the choice of specimens that would fit, and the actual irregularity of horse evolution was ignored.7 As G. G. Simpson convincingly argued in Tempo and Mode in Evolution (1944), the evolution of the Equidae was not linear but bushy. Simpson disliked “orthoevolution” and “orthoselection” for their “inelegant etymology”—he preferred “rectilinear evolution.” But he did not deny the existence of such trends.8 Nor did he rule out the possibility that patterns of mutation that were more likely to go in one direction than any other were a possible mechanism of rectilinear evolution that could operate independently of natural selection. But he qualified this by noting that known examples, such as the mutation from wild type to white eyes in Drosophila usually went “against the evolution of the wild population.”9 The existence of a directional “law of growth” independent of natural selection, which had been acceptable to Darwin was anathema to neo-Darwinism. Now orthoevolution is taken to be spurious, and orthogenesis a historical fantasy. The “pull of the present” is now anti-orthogenetic. If their premises are accepted, the logic of Simpson and Rensch is impeccable— so much so that selectionists never bothered to suggest any way in which genes could randomly mutate so as to keep on producing particular trends of phenotypic exaggeration that directional selection could encourage. Even if the ultramorphosisto-extinction argument is dropped, the possibility that a kind of orthogenetic process—a mechanistic, directed allometric growth shift—underlies directional selection cannot be ruled out. If so, all such cases would require reconsideration, and that includes most of organismal diversification. Where do directional exaggerations come from? Are they due to the selective accumulation of random advantageous mutations, or could the appearance of directional change through selection pressure mask a more fundamental underlying drive. When I began to consider the possibility that orthogenesis was a real process, I thought it irrelevant to evolutionary progress, leading only to more of the same old thing. Intent on emphasizing emergence to higher levels of organization, I viewed adaptive radiation, or diversification, as a mere working out of the potential provided by a major emergence. If you can, like me, but a little bit faster, I hope, detach yourself from the anti-orthogenetic bias of the Modern Synthesis, start with the notion that most of adaptive radiation amounts to allometric growth shifts. And then you might realize that many of them could have been caused by autonomous, self-amplifying epigenetic mechanisms.

Orthogenesis

271

I should have paid more attention to D’Arcy Thompson. His book On Growth and Form had this to say, in both editions (1917 and 1942): . . . in particular cases, the evolution of a race has actually involved gradual increase or decrease in some one or more numerical factors, magnitude itself included—that is to say increase or decrease in some one or more of the actual and relative velocities of growth. When we do meet with a clear and unmistakable series of such progressive magnitudes or ratios, manifesting themselves in a progressive series of “allied” forms, then we have the phenomenon of “orthogenesis.” For orthogenesis is simply that phenomenon of continuous lines of series of form (and also of functional or physiological capacity), which was the foundation of the Theory of Evolution, alike to Lamarck and to Darwin and Wallace; and which we see to exist whatever be our ideas of the “origin of species,” or of the nature and origin of “functional adaptations.” And to my mind, the mathematical (as distinguished from the purely physical) study of morphology bids fair to help us to recognise this phenomenon of orthogenesis in many cases where it is not at once patent to the eye; and on the other hand, to warn us in many other cases that even strong and apparently complex resemblances in form may be capable of arising independently, and may sometimes signify no more than the equally accidental numerical coincidences which are manifested in identity of length or weight or any other simple magnitudes.10

This paragraph would have been more than enough for the Modern Synthesis to exclude the Thompson transformation theory, since such changes would occur without reference to natural selection. But Thompson made it worse from the conventional point of view by proposing that there had to be breaks in continuous transformations, thresholds at which new formulations would apply. And he was quite satisfied from the evidence of quantum jumps in physicochemical phenomena, as well as phylogenetic breaks, that saltation was a common mode of evolution: Our geometrical analogies weigh heavily against Darwin’s conception of endless small continuous variations; they help to show that discontinous variations are a natural thing, that “mutations”—or sudden changes, greater or less—are bound to have taken place, and new ‘types’ to have arisen, now and then.11

In Making Sense of Life (2002), Evelyn Fox Keller asks why Thompson continues to be mentioned with approval, although his concepts have scarcely helped to advance biological theory. I suspect that all who look at one of his coordinate geometrical transformation grids sense that they are missing something important without being able to put a finger on it. What they are seeing is evidence that doesn’t exist for them because it doesn’t fit selection theory. Yet there it is, and Thompson put his finger on it with complete assurance. His invocation of orthogenesis and saltatory emergences, and his virtual denial of natural selection, still create the paradox, and guarantee that he will continue to frustrate all his neo-Darwinist readers. Straight-line evolution is not inconsistent with divergence. As John Grehan and Ruth Ainsworth point out in their 1984 essay “Orthogenesis and evolution,” straight lines can stop, deviate, and bifurcate. Thus, lines of orthoevolution—real, or

272

Chapter 7

apparent—can diverge, converge, or run parallel to each other. In Evolutionary Theory: The Unfinished Synthesis (1985), I included orthogenesis in a comparative table of several processes that Darwin called “laws of growth.”12 Now I feel that many of them could be subsumed by orthogenesis. They are expressions of critical-point emergence, built upon the interaction of the complementary mechanisms of heterochrony, organizer topography, and perhaps amplification by some kind of gene drive. Self-amplifying processes are known to exist, in the form of non-random codon repetition, non-random tandem gene duplication, and polytene chromosome production. Some are known to diminish organismal integrity over several generations, and result in death; but some are also known to be reversible. There is no reason to suppose that orthogenesis invariably continues until it is detrimental to the point of lineage extinction—most molecular mechanisms have “off” switches. And there are clearly many living organisms whose distinctive forms arose by non-random, repeated allometric shifts. Nevertheless, at any stage in its activity, orthogenesis could be disintegrative, and kill off the organisms that do it. At the risk of flogging a dead horse, I reiterate that natural selection is superfluous to self-generated destruction. If adaptively neutral, a novelty will not spread to become a universal population trait, but will persist in its own descendent lineage, whose numbers will depend on many other factors. Where it coincidentally confers adaptational advantage the proportion of its lineage in the species will then increase. Orthogenesis could help to generate hopeful monsters, which in turn could be the founders of new populations, in new environments, all with the active orthogenetic mechanism. I will shortly detail molecular processes relevant to orthogenesis. They might all contribute to a genetic drive that affects morphogenesis. The implication that these could be a “recipe” for orthogenesis was twigged by Jack Heslop-Harrison (1983) and by John Grehan (1984) before me. But the author of molecular drive, Gabriel Dover, repudiated the inference, requiring separate evidence for the existence of orthogenesis before he would accept that his mechanism had anything to do with it.13 One caveat: While orthoevolutionary trends are evident in fossil series over geological time, we do not discern major orthogenetic trends happening in existing wild species. Yet, the molecular mechanisms are hypothetically rapid enough to produce change from one generation to the next. The leaps of transposable elements are one example. Another is the pathological phenomenon called “anticipation,” which involves orthogenetic amplification of particular codons through replication slippage, and finally results in ultramorphosis to the point of extinction in the lineage. However there might be parallel cases with advantageous qualities. In any case, orthogenesis might bring about a critical-point emergence and then stop. Or it could occur early in the lineage of a new clade, reach its characteristic degree of development and then stop, leaving the new type in a relatively unchanging condition. In other words it would occur with the punctuation, or even be the punctuation, but not continue to amplify itself through the equilibrium phase.

Orthogenesis

273

Perhaps we simply cannot see the fossil record with fine resolution, and perhaps orthogenesis of the anticipation type is too fleeting to be noticed in wild populations. All the same, orthogenesis can be found at the chromosome level in particular cell lineages. The polytene chromosomes of Drosophila were a workhorse of research into the epigenetic hormonal control of gene expression. The question of how the chromosomes are duplicated and reduplicated without an accompaniment of cell division has been addressed, but nobody has pondered the other implications. Here is a cellular lineage, confined to the salivary glands of these creatures, that has clearly undergone orthogenesis. Related flies with similar life styles do not have the giant chromosomes, so their relative adaptiveness is questionable, both in incipient and mature stages. Some insects have polytene chromosomes in their excretory Malpighian tubules, again without any discernible improvement over their relatives that lack them. In these examples allometry has been largely confined to particular organs, has offered minimal advantage, but has nevertheless stopped before becoming analogous to a cancer. Moreover, we only need to look through a seed catalogue to find exaggerations of color and form in cultivated flowering plants. These are not invariably caused by the “supernatural” selective action of the horticulturist, i.e., by manipulative hybridization. They could be the products of ongoing orthogenetic trends that are encouraged by inbreeding, while protected from any negative side effects. Indeed, when we consider domestic animals, such as dogs, allometric shifts are easily detectable on the scale of a century. Sometimes, in pedigreed lines of animals, the exaggerated traits are not orthogenetic, but regressive and sometimes dysfunctional. Anyone familiar with pug dogs, for example, knows how difficult it is for them to draw breath when exerted, and how susceptible they are to respiratory diseases. Without the ministrations of their human caretakers they would quickly die out. Neo-Darwinists might argue that selective inbreeding has flattened the face of a pug dog, in a manner that is quite clear from paintings of the breed over the last three centuries. But the exaggerations had to occur before they were recognized. What does “exaggeration” mean at the DNA level, and how does it keep on happening? Convention assumes that non-synonymous point mutation of a structural gene is involved in directional selection, and that it can keep mutating in a way that magnifies its original function. As Wallace Arthur (1997) says, “the exact changes in DNA are irrelevant to the formal population model,” and, by extension, irrelevant to neo-Darwinism.14 Yet they are relevant to evolution, and it is difficult to imagine that point mutations or the exon shuffling mechanisms that Arthur proffers would keep on producing more of the same instead of something different. Point mutation of a regulator or a growth hormone gene can indeed affect growth; but could another such mutation amplify the effect? Most mutations are detrimental, and it is highly unlikely that any non-detrimental random point mutations would keep producing viable

274

Chapter 7

results consistently in the same direction, as opposed to varying their phenotypic expressions. Such a mechanism would require a heap more “separate evidence” or “mystical force” than orthogenesis ever needed. It is much more likely that duplication of a structural gene, or an epigenetic acceleration resulting from an alteration in expression of a modifier through dosage amplification, is at the heart of the matter. At the cellular level of epigenetic induction of a pug face, a topographical shift of organizer cells produced differential effects, resulting in a phenotypic saltation distinct enough to attract the attention of breeders. When they then inbred them to intensify the desired feature, they were allowing the trend to continue by preventing dilution by cross-breeding. I reiterate that intentional selection works the opposite way to natural selection, which makes it all the more desirable to examine how its effects are manifested under conditions where natural selection is absent. There is no lack of raw data for such a program of analysis. The non-random nature of orthogenesis at the phenotypic level parallels the nonrandom nature of molecular drive, or “concerted evolution,” associated with gene duplication. Initial duplications through unequal recombination or gene jumping may be completely random. But it is known that reduplication affects particular genes, not any gene anywhere at any time, and some transposable elements have preferred landing loci for their jumps. Only certain portions of the chromosomes are affected, and within a species the affected regions are consistent. The differences in DNA repetition between the domestic mouse, which may have a million copies of a particular sequence, and related species of mice that only have one, was one of the first known comparative examples among similar types.15 Finally, amphibian and fruit-fly studies have found that if repeated sequences are lost, or diminished, they are replaced in subsequent generations.16 Simple duplications have the immediate potential advantage of functional amplification. Whatever proteins those genes code for, every duplication means that the organism can make twice as much in the same time. In the longer term, repetitive genes may differentiate, to produce adaptability, or they might cause epigenetic acceleration, or the exaggeration of an adaptive feature. Then they might be perceived as an example of directional selection. Non-random duplications of structural or modifier genes are not the only candidates for orthogenesis. But these repetitions are the most obvious way to get large increases in the raw materials necessary for construction during developmental accelerations brought about by organization changes at the cellular level. Most diversification of archetypal body plans is expressed through allometric growth shifts. Somehow they stabilized before they led to widespread extinction. Darwin thought that there was a compensation effect, whereby the resources of the uterus were taken from one region of an embryo and given to another. But a more likely mechanism is dosage increase by gene duplication, differentially expressed at

Orthogenesis

275

the locus of allometric increase. Since all cells in the body are genetically identical, molecular governors of DNA operate in the regions where rapid growth is inappropriate, presumably through repression, methylation, or histone regulation. Above the DNA level there are growth-stimulating and growth-inhibiting factors that affect cell membrane receptors. Amplification of the intracellular growth cascade system is another way to intensify and prolong the duration of allometric alterations. Søren Løvtrup noted in Epigenetics (1974) that allometry is most likely an effect of growth hormone activity. And although growth hormone gene duplication has been associated with episodes of rapid diversification in the vertebrates, there are many other genes involved in growth. The trick is to get the growth localized sufficiently to produce the allometric shift instead of an overall size increase. Nevertheless, overall size increase does itself result in allometric shifts, as Stephen Jay Gould pointed out for the Irish elk. Elisabeth Vrba has numerous examples of relatively sudden universal size increases in the fossil record of Southern Africa, consistently associated with rapid climatic change to colder, drier conditions.17 Thus, autonomous drives can be modulated by the environment. A hierarchy of epigenetic commands must be invoked before the phenotypic effects of orthogenesis are discernible. Examples of allometric trends such as anterior proportions and skull modifications in the ancient labyrinthodont amphibians, cited by Julian Huxley, may have been associated with environmental changes. But the more efficient causes were probably alterations in the behavior of neural crest organizer cells in conjunction with the effects of mutated homeotic genes. Questions remain: how are these cells freed from tight canalization, and, once free, do they have a tendency to keep on exaggerating particular novelties in the same direction? We do know that these things happen, but is it by drive or adaptation? Regardless of whether a breeder seems to be selecting pugs for their short jaws, or nature seems to be selecting saber-tooths for their competitive predatory skills, the selection focuses on organisms in which an epigenetic trend is already under way, and which started and might continue in the absence of selection. Paleontologists used to associate orthogenesis with extinction. The allometric trend became so exaggerated that the final products were unable to survive. In ammonites, the enhanced buoyancy correlated with increase in the surface area of the foliaceous lobate lines could have finally been lost because different parts of the septum were counteractive to the point where the buoyancy mechanism was compromised. Does their extinction and replacement with others with simple lobate lines indicate an ultramorphogenetic momentum that natural selection could not stabilize? The male of the extinct Irish elk Megaceros had massive antlers that were supposedly sexually selected by susceptible does. But those antlers laid Megaceros stags open to chronic neck pain, if not increased predation, and offset their sexual advantage. Did orthogenesis carry them to their extinction, or was a change of climate responsible?18 In this

276

Chapter 7

case, since the Irish elk survived to leave many fossil remains, the trend to orthogenetic doom had probably already been switched off, but although it did not drag it to extinction, it left it vulnerable to environmental disequilibration. A similar kind of massive head ornamentation in Rhinoceros has not made it extinct—yet—that will come for entirely different reasons. The dinosaur Triceratops had rococo flanges and horns that probably conferred some protection.19 Their extinction almost certainly had an external, catastrophic cause, the K-T bolide impact. But there remains the possibility that orthogenesis is not only a real, material, heritable phenomenon, but that occasionally there have been no governors upon it, and it has gone beyond terminal viability. Lineages that show orthogenetic trends may diverge because one has briefly resumed the trend and others not. During horse evolution and diversification, allometric shifts resulted in the emphasis of a single digit and the disappearance of the others in the legs, changes in skull and tooth structure, and an increase in overall size. These affected diverse branches of horse evolution, and taken all together do not constitute a single continuous orthoevolution of a single trait. Attempts that have been made to identify such a single line, whether from the adaptationist or orthogenetic point of view, are oversimplifications. But the fact remains that the lineage of modern horses did somehow acquire all of those changes, even if their emergent punctuations were followed by long, periods of orthogenetic stasis.20 The novel variation in an epigenetic algorithm that starts it moving in a particular direction is by definition a saltatory emergence. Its subsequent harmonious development depends on both allometric, genetic and epigenetic coordination, and at critical points functions might change. These processes are independent of natural selection, but if they keep on in the same direction they might finally reach a condition that is “better adapted,” especially if the orthogenetic animal changes its behavior to take advantage of its changing anatomy. Pierre-Paul Grassé (1977) suggested that orthogenesis could be equated with parallel evolution. He noted how the evolution of hippomorphs, or horse-like forms, diversified in remarkably similar trends in the Old World and the New World, and suggested that their common ancestors already had a propensity to undergo orthogenesis along similar lines. Some of these problems can now be addressed with reference to deep homology, the expression of duplicated and differentiated regulatory genes such as Hox, and the activity of neural crest organizers. But orthogenesis is closer to the actual process than directional selection. To earlier generations of biologists the thought of welcoming back the prodigal principle of orthogenesis would have been intolerable. Bernhard Rensch’s book Evolution above the Species Level (1959) was devoted to the thesis that the origin of varieties within populations was the same process as that leading to speciation, and all points beyond. However, the replacement of orthogenesis by directional selection was

Orthogenesis

277

not a matter of a demonstrably faulty explanation being supplanted by one based on more rigorous evidence. Here, a majority opinion prevailed without reference to any rational evidence at all. Michael Ghiselin, in a recent personal communication, counters this with the opinion that “the internal self-amplifying drive was rejected not for ideological reasons, because it would make natural selection redundant, but because there was no reason to invoke it.” I would think that the proven existence of selfamplifying drives more than sufficient reason to invoke them. It was such a dialectically successful attack on orthogenesis that, malgré Grassé, modern molecular biologists and some students of body plan evolution now deal with the process without being aware of either the term or its morphogenetic implications. I must note an exception in Keith Thomson who mentions it as orthogenesis, before substituting the term “trends.” In Morphogenesis and Evolution (1988) he writes: Their most obvious explanation would be causation through strong and historically consistent external directionality of selection. However, it is also obvious that many evolutionary trends also have a strong internal generative aspect. An obvious example would be any trend in shape that was driven by allometric size change. Some aspects of horse skull evolution are the direct consequence of increase in size. Size increase then is the real trend, and skull shape and proportion merely follows. But, as MacFadden (1986; cf. Radinsky, 1984) has shown, the early part of the history of horse evolution was accomplished in the absence of major size increase. Therefore, it is worth asking whether evolutionary trends may have some other deeper, internal cause.21

Although Thomson then argues that size increase per se could fit the directional selection model, he has to assume that there is a generating mechanism in place to give selection something to work with. As a variation on the theme he adduces developmental constraints. Such constraints might apply asymmetrically, so that some proportions of a limb, for example might elongate, while others did not. Then Thomson suggests that “many different genetic variations, introduced at different levels in the morphogenetic cascade, could combine under this ‘integrative’ influence to cause phenotypic changes in the form of a ‘trend’.’”22 Call it trend or call it orthogenesis; however many diversifications have occurred, each of its products had its own lineage. Its track might be lost among the bushes, but it is there, and more likely to have been produced by orthogenesis than directional selection. I have already pointed out how difficult it would be for a series of random point mutations of a single gene to produce an orthoevolutionary series of phenotypic exaggerations. But it can always be argued from ignorance that every trait is determined by multiple gene interactions, and that hypothetically there could be an astronomical number of combinations from which directional selection could produce an apparent trend, every stage of which is progressively fitter. It still doesn’t get around the fundamental problem of the non-fitness of incipient stages. And recombination is always a difficulty for precarious combinations. Thomson is sufficiently troubled by

278

Chapter 7

the phenomena of parallel and convergent evolution in disparate lines of clovenhoofed and single-hoofed ungulates to insist on a generative developmental mechanism, rather than leaving it to directional selection alone. Such mechanisms must represent “the inverse of developmental constraints. They are not predictable from a study of individual genes and gene expression and, except at a phenotypically trivial level, neither predicted nor explained in terms of population genetics.”23 I think he is ready to run away and join the circus. In The Shape of Life (1996), Rudolf Raff just dismisses the idea that orthogenesis exists, using the oft-cited example of bushy rather than orthoevolutionary trends in horse evolution, but he echoes Keith Thomson when he writes: Although the idea of internally directed evolution is not tenable, the potential role of internal factors or constraints remains a viable one even in a selectionist context. Existing genetic and developmental systems are not neutral features, nor are they necessarily readily dissociable. Existing developmental systems must produce constraints on the degree of freedom with which selection can operate. Thus, a major principle of evolution exists beyond an all-powerful selection working on randomly generated variation, and the inner workings of genetic regulatory systems and developmental processes as well as their histories must be considered as key elements of evolution.24

Right—as far as it goes! If an organism’s development is so tightly canalized that only one alternative avenue of development remains unconstrained, that will be the avenue taken, and the result will be orthoevolution. It will continue for as long as the lineage is exposed to the triggering stimulus, and natural selection will have no role in its generation. If they were to entertain the untenable a little more, epigeneticists such as Raff might want to take a closer look at molecular mechanisms that produce the same kinds of effects, regardless of developmental constraints and natural selection. Orthogenesis merits a significant place in the ranks of evolutionary causes, and it would be worth re-examining every case of directional selection for an orthogenetic component. Genetic Drive, Allometry, Anticipation, and Orthogenesis For the phenomenon of self-amplifying repetitive differentiation at the molecular level, I have suggested the term “genetic drive,” which involves any non-random selfamplifying genetic change, such as gene duplication, mutational repetition and concerted evolution. As Britten and Davidson argued in 1969, duplication of particular genes seems to keep on happening regardless of adaptiveness or natural selection. Selection would only have a limited metaphorical role in culling any that show dosage imbalance, which could waste energy and adversely affect the availability of amino acids and nucleotides for essential syntheses. Selection does not cause the buildup of multiple genes, but where products of gene drive coincidentally show a pronounced competitive superiority, the proportion of organisms that possess the drive will, of course, further increase in the population.

Orthogenesis

279

I have applied the concept of repetitive differentiation to the development of biochemical adaptability, where a suite of proteins can meet the demands of a varying environment. Here highly concerted molecular evolution would be counterproductive, since the range of variation within the individual organism has to be preserved to afford adaptability. To be effective there would also have to be further adjustments to the regulation of the duplicated genes, so that in the case of salmonid kinases the activation of the appropriate genes would be sensitively set to narrow temperature ranges. The distinct trends in allometric growth shifts that have been associated with orthogenesis in the past cannot simply be explained by exon or gene duplication per se, although this kind of DNA repetition does suggest how growth might be accelerated. There has to be an additional regulatory mechanism, involving, for example, a growth hormone. During morphogenesis it might react with larger numbers of receptors in specific organs, or stimulate a greater amplifying cascade of second messengers in the affected cell lineages. Or there might be a more profuse synthesis of the effector enzymes that respond to the growth stimulus. Tissues that are being allometrically enhanced need to grow faster, and their cells need to divide faster, but in a manner that can be accommodated within organismal integrity, rather than like a cancer. Recall also that allometric growth may require more than simple tissue growth acceleration. Taking the example of the giraffe’s neck once more, not only does there have to be an increase in the rate of growth and elongation of the cervical vertebrate and associated tissues, this also has to be achieved in the entire front quarters, including the pectoral girdle, legs, heart, blood vessels, and nerves. A simpler example relates to the allometric growth of feathers. Minimally this requires the interaction of the genes sonic hedgehog and bone morphological protein 2 in promoting cellular proliferation and differentiation. But these genes participate in many anatomical processes, their role in bony facial structure, tooth and beak structure among others. For them to work effectively in feather epigenesis they have to be expressed in the right place at the right time and the right rate, in conjunction with regulatory factors, such as modifier genes, and neural crest cells. To work, allometric shifts have to be not only harmonious with other patterns of growth, but also directional. Self-amplification—orthogenesis—is an effective way to achieve this, provided there is also an “off” switch. My question for selectionists is this: Does any directional evolution rely upon natural selection acting on random mutation? Hypothetically it could result from genetic drive, local proliferations enhanced by gene duplication, or induced by epigenetic mechanisms at a cellular level that coincidentally give a competitive edge. Individuals with dosage imbalance would fall by the wayside in any case, through failure of organismic integrity. This seriously challenges the one kind of natural selection that even dissidents regard as evolutionary. I am perhaps overcomplicating allometry, since it can sometimes arise from size changes alone. Aside from these,

280

Chapter 7

embryonic canalization might only allow alterations in a single developmental direction. In that case, allometric shift must always be directional, without the necessity of directional selection. As an analogue of orthogenesis, the pathogenic condition known as “anticipation” is instructive. It was first known from myotonic dystrophy, a muscular disfunction.25 More recently its genetics have been investigated for Huntington’s disease, and for mental retardation caused by fragile-X.26 The most general observation made about these diseases is that they were genetic, and that it could be predicted that they would worsen from generation to generation. The word “anticipation” suggests predictability, but actually means that in subsequent generations the onset of the condition becomes progressively earlier, as well as more severe. Myotonic dystrophy was initially detected from cataracts in elderly but mildly affected sufferers. The condition would worsen in subsequent generations to the point where cataracts would affect the middle-aged, and then muscular symptoms in the young. The history of the major lineage of myotonic dystrophy has been traced to a single triplet duplication about 300 years ago.27 Since only one codon is concerned, and it occurs during mitosis and meiosis, replication slippage is believed to be the initiating molecular process, although unequal crossing over could be significant in extreme cases. Fragile-X is named for its cytological phenotypic expression: breakage of the X chromosome due to heavy methylation at the locus of FMR1. At the genotypic level there are duplications of the base triplet coding for arginine. In the population at large up to 54 duplications can occur, without any phenotypic manifestation of the condition, nor any detriment to offspring. Between 52 and about 90 repetitions of the triplet is the “normal transmitting” condition where there is no initial phenotypic expression but the parent can be considered a carrier. From here it can be predicted that fragile-X will be phenotypically expressed, with an average of 10 duplications added per generation. However, regression is also known. Transcription from the strongly affected gene to messenger RNA is inhibited. There is nothing speculative about the fact that anticipation at the level of molecular processes is exaggerated by an autonomous, self-amplifying driving mechanism. It is orthogenesis by another name. Natural selection qua differential reproduction has no part to play in its progress, which in the case of anticipation leads the lineage to a self-eliminating, detrimental condition. As I have argued before, it is redundant to adduce a metaphorical undertaker to reach a fate that was implicit in the generative mutation. If a pathogenic condition can proceed orthogenetically at such a speed, one that starts out in a selectively neutral state might progress even longer through genetic drive, and ultimately display beneficial qualities under the prevailing conditions, with the option of stopping, or going on to eliminate itself through detrimental hypermorphosis. The action of natural selection is irrelevant to these events.

Orthogenesis

281

Gabriel Dover (2000) writes that he regrets coining the expression “molecular drive,” since it suggests a deterministic quality—despite his acceptance of natural selection as “a deterministic, non-random product of selection pressures arising in the environment.”28 If he were simply to equate his “drive” with any autonomous selfamplifying molecular process he would not have a problem, especially since he goes out of his way to demonstrate how such processes occur without reference to ultimate reproductive success. Two examples that he gives are the proliferation of segments in snakes and centipedes through the repetition of identical Hox influences. Segment proliferation is a “lavatory-roll model of evolution”—identical bits keep coming and coming. Although Dover (1982, 1986, 2000) does not mention orthogenesis within molecular drive or TRAM systems (q.v. previous chapter), he asserts that these kinds of processes are directed only in certain gene families and are not random, in contrast to the effects of natural selection and genetic drift. He is really writing about orthogenesis.29 Dover (2000) also describes another self-amplifying “genetic drive” phenomenon involving a jumping gene called P-element. Dover’s P-element was accidentally passed from Drosophila willistoni to D. melanogaster by a parasitic mite. In the new host, the effect was sterility through chromosomal disintegration and gonad destruction. Once established in a chromosome of the new host it was replicated during mitosis, so that the homologous chromosome in the cell lineage came to possess it as well. Part of the single transposable element codes for a transposase that triggers the jump. When it jumps, the gap it leaves is filled by a new copy induced by the non-transposing element in the other homologous chromosome. The self-amplifying DNA is then spread through the population by sexual reproduction. Coincidentally, in populations that were not driven to extinction through sterility by P-element, several compensating mechanisms had appeared. One of them inhibited the effect of the transposase, and so prevented the jump. Another was a repressor protein that enhanced the reproduction of the individuals that possessed it. Any compensating mechanism that appears will save the day. Therefore it is redundant to say as Dover does, that it is “seized on by selection to overcome the debilitating effects of hybrid dysgenesis.”30 “Spiegelman’s Monster” is also relevant to the orthogenesis v. directional selection debate. This is the name sometimes given to the subject of “A Darwinian Experiment with the Replicating RNA Molecule.” It refers to a viral RNA strand that over a period of time showed reduction in size and increase in replication rate, and the research is reviewed by Spiegelman (1967). Denis Schwartz communicated his view that it was an orthogenetic process, noting that no one had challenged the “Darwinian” label. Paul Davies (1999), however, treats it, along with Eigen’s work on the self-assembly of RNA strands from a solution of simple ribonucleotides, as a model for Darwinian evolution at the molecular and possibly prebiotic level.31 The starting components of Spiegelman’s experiments were RNA strands from a bacteriophage virus, an RNA replicase that could recognize them, and ribonucleotides

282

Chapter 7

that the system could use to build the replicates. Over a series of 75 reactions, he regressed the time allowed for the process, and transferred a portion of the reaction products to the next stage. He was deliberately selecting the fastest replicating and smallest products. After the fifth transfer the viral strands had lost their biological competence, i.e., they could no longer act as infectious viruses. By the final transfer the strands were reduced to 17 percent of their original length and the growth rate was fifteen times faster than that of the original viral molecules, and still recognizable to the replicase. This is unequivocally an exercise in artificial selection. Factory farming tries for the same rapid production of stock, though attempts to increase size are usually more common. In any case, competition from slower growing RNA strands was removed, and the selected molecules were given unlimited resources and protected from the biological consequences of over-simplification. As to orthogenesis: the RNA strands seem to have had a self-reducing characteristic that the experiment allowed to continue beyond the biological point of no return. This had to be present before it could be selected. In nature such a process would be masked by the complexities of the virus and its host, and also be self-eliminating, unless the self-reducing strands took up a permanent parasitic existence in the host bacterial cell. Dwarfs and Giants The discovery of the fossil remains of Homo floresiensis, putatively a dwarf strain of H. erectus, has reawakened interest in evolutionary trends toward dwarfing and giantism among island fauna. Examples from Flores Island in Indonesia include komodo dragons (giant Varanus lizards), pygmy elephants, giant rodents, and dwarf hominins.32 Overall size increase could be explained as a saltation due to a growth hormone gene mutation, or an orthogenetic trend involving the amplification of growth hormone and its receptors. Dwarfing is unlikely to be due to a negative orthogenesis, simply a heterochronic epigenetic change. Here a saltatory phenotypic effect is more likely to be the effect of an epigenetic change that cuts growth at an early stage. The same may apply to the example of the reduction of the snout in pugs. Traditionalists struggle to explain these effects in terms of the gradual natural selection of random point mutations. But when an organism finds itself on an island with any kind of useable resources the agents of natural selection, i.e., competition and predation, are diminished or absent. Other things being equal, it doesn’t matter if the organism gets huge or tiny provided it maintains its integrity. There may be no particular advantage to either until the hypostasis of natural selection is imposed by increased competition. In conclusion, Mivart’s supposition that the laws of growth were due to “innate tendencies” were certainly true in as far as divergent lineages would still have the majority of their epigenetic inheritance systems in common, and that the same rules

Orthogenesis

283

of constraint would likely apply after divergence. Beyond that, self-amplifying systems also represent innate tendencies that operate within the rules of constraint, but produce novel allometric shifts, regardless of the “relentless demands of selection pressures.” Postscript While I was preparing the final draft of this chapter, there appeared a most interesting and challenging publication by John Fondon III and Harold Garner: “Molecular origins of rapid and continuous morphological evolution” (2004). It deserves particular attention, since it bears out David King’s suggestion (q.v. chapter 6 above) that tandem repetition of codons need not be pathological, but could be a useful source of variability. Since it also suggests a more selectionist-friendly interpretation of phenomena that I would treat as orthogenetic, I review it here. First, I repeat their remarks that are most supportive of my position. “Rapid and continuous evolution” speaks for itself. They report a consensus among molecular geneticists that random point mutation is insufficient as a source of variation for natural selection to be effective, and that mutations in cis-regulatory elements are the “predominant source of the genetic diversity that underlies morphological variation and evolution.”33 In their study of the evolution of domesticated dogs, the “mutations” are tandem codon repeats within such modifier genes as Alx-4, Runz-2, Twist, and Dix-2. They modify the production of proteins with runs of glutamine and alanine, for example. These proteins participate in the epigenesis of craniofacial features, limb length, and digit number. Tandem repeats within regulatory genes are always involved in the epigenetic changes under discussion. Fondon and Garner note a correlation between allele length (as determined by the number of repeats) and the size of a phenotypic character. Point mutation rates are far too low to provide dog breeders with enough variation from which to select. In contrast, increased codon repetition, as well as deletions, provide for “gross morphological novelty, reminiscent of the saltatory genetic events Goldschmidt envisioned for his “hopeful monsters.”34 Especially interesting is their conclusion that what in the case of anticipation is pathological, is in far more cases in nature a potential for viable evolutionary change. More controversial for me is their inference that the evolution of dog breeds is driven by the breeders’ selection of particular features; that tandem repetition and in some cases deletion of repeats occurs rapidly enough to allow for the historically rapid changes in dog features. Extending their study to other canids and a variety of mammals including otters, walruses, rabbits, bats and humans, they find similar repetitions in the same kinds of epigenesis-modifying genes, and conclude that in these cases the divergences are driven by natural selection. I will only comment that

284

Chapter 7

these are in part products of orthogenetic self-amplification. They are initiated neither by artificial nor natural selection, and they proceed until the breeders notice that there is something new from which to breed, or until an advantageous quality for their conditions of life helps the mechanism to spread. As Grehan and Ainsworth (1985) point out, direct lines of evolution can branch and go in different directions, so orthogenesis can be a significant part of diversifying evolution, whether within the domestic dog species or within the entire taxon of placental mammals. Where Are We? In the prologue to chapter 2, I submitted a field-trip checklist for the visit to the causal arenas of emergent evolution. This was updated at the end of chapter 3, at the end of chapter 4, and at the end of chapter 5. We are now in a position to add epigenetics and orthogenesis. 1. Mechanisms that perform in all the arenas of emergent causation. The prominent mechanism is repetitive differentiation, applying to molecules, segments and organs. To stretch the point it could also be applied to social insect caste systems, and human manufacturing activities. Simple gene duplication or dose amplification may precede differentiation at the molecular level and contribute to allometric growth and orthoevolution. Modular shuffling of exons, combined with mutations of the receptor sites in promoters, and the consequent increase in their algorithmic complexity universally affect developmental changes. Transpositions of genes have also been important in epigenetic evolution. All of these have effected not only developmental evolution, but also, as a consequence, physiological and behavioral evolution. Major adjustments to symbioses have also been effected by gene jumping. Environmental stimuli have directed specific epigenetic changes. Physicochemical physiogenesis has also been generally affective. Physiogenic changes have often been internalized to consolidate homeostasis and homeorhesis. Retroviral and bacterial transgenesis potentially affects all molecular causal mechanisms. Margulis’s principle of mixing and matching applies not only to the generative conditions for symbiosis, and biochemical rearrangements between host and symbiont. At the molecular level, the same applies also to processes of topological and sequential gene and protein rearrangements, and shifts in reading sequences. While much of adaptational evolution may be driven by particular and persistent organismal actions, the molecular laboratory in the basement, even operating at random, has a huge capacity to come up with experiments that will accommodate those actions.

Orthogenesis

285

2. Generative conditions from which emergences spring and common features of effective generative conditions. The holonic, or modular, nature of life, which involves the reproduction of hierarchically arranged units is a major generative condition of developmental, biochemical, physiological, anatomical and behavioral complexification and diversification. Especially important was the establishment of cellular modularity in multi-cell organisms. The concurrence of efforts is easier if the separate offices are close to each other. The generative conditions for some complex organs like eyes and appendages are universal, making their emergence highly probable, under appropriate organismal conditions, as illustrated by the number of times they have independently emerged. Yet these independent emergences are also dependent upon the deep homologies provided by homeotic genes. Conditions conducive to epigenetic evolution often involve developmental nodes or thresholds, as well as changes in gradients leading toward those thresholds. Environmental gradients and thresholds (or interfaces) are among the generative conditions for symbiosis. At interfaces, the prolonged concentration of different types of organisms with a potential for sharing their talents is a good way for the unlikely to become the likely. Such gradients and interfaces also affect physiological and behavioral evolution. They sometimes allow respite from the agents of natural selection. But while environmental gradients and interfaces physiogenically altered the internal milieu of simple organisms, the provision of a stable internal milieu was to become a generative condition for many emergences. Cellular environments had to be stable enough to allow eukaryotic endosymbiosis. The emergence of viviparity provided constancy for the fetus prior to birth, and in placentals was a generative condition for the success of offspring whose own homeostatic mechanisms were late-developing. The more stable the internal milieu, the greater the capacity of the type to diversify (anatomically, by orthogenesis) after passing through environmental thresholds. The emergence of any kind of adaptability potentiates further evolution through extending the ability of the organism to try new behaviors and environments whose influences reverberate through all of the causal arenas. 3. Key innovations that catalyze emergence, provided that other appropriate generative conditions are present. Major ones are early biochemical acquisitions such as photosynthesis, and sulfuroxidation; structures such as chromosomes. waterproof insect integument, and neuron myelination; cleidoic eggs, feathers, and parabronchi in birds; and hair,

286

Chapter 7

placentae, and neopallia in mammals. The multiplicity of examples makes generalization daunting, but they often involved epigenetic change and enhanced physiological and behavioral adaptability. 4. Particular contingencies, predictable or unpredictable that have affected generative conditions. Disequilibration of homeorhesis and homeostasis by environmental effects was a necessary generator of emergence in early epigenetic and physiological evolution. The availability of environments that could have these effects is predictable for any Earth-like planet with seas and continents and plate tectonics. Concentration of adaptable plants and animals, metaphorically at the edge of chaos, or actually at the interfaces between environments where competition and predation is reduced is also predictable. Physiogenesis resulting from environmental shifts is a predictable matter of physicochemistry. Catastrophic extinctions and physiogenic effects of bolide impacts, and volcanic action remove the agents of natural selection and lead if not to direct emergences of adaptable organisms, then to the demographic success of those already in place, and their rapid diversification. 5. The constellation of multiple functions characteristic of new emergences. While they differ from one example to another, multifunctionality is a common attribute of emergent systems. Simple epigenetic changes such as alteration of egg size or reduction of photoperiod can cause paedomorphosis with multiple effects. Simple increase in size causes allometric shifts. Changes that increase adaptability are virtually synonymous with multiple functionality, and they occur in all the causal arenas. The problem with the selectionist “explanation” is that it sees only one feature among many that it ignores. 6. The course of emergent evolution that has progressively led to greater self-organization, independence and freedom of choice. At different times in the history of evolution symbiosis, association, epigenetic differentiation, physiological adaptability, behavioral freedom, have interacted to send out waves of progressive change that generated new major emergences. “Bootstrapping” says it more succinctly. Orthogenetic/allometric shifts in the central nervous system have been an integral part of several independent animal lineages, with the most dramatic found in the primates. While chronological priority is easy to establish, it is

Orthogenesis

287

impossible to rank the causes of emergent evolution in order of importance for evolutionary progress. 7. Emergent evolution and adaptive radiation. Although the emphasis of this book has been on progressive evolution, this synopsis would be incomplete without reference to the diversifying evolution that results from major emergences. This chapter has demonstrated the feasibility of self-amplifying processes responsible for allometric shifts that are key to anatomical diversification. Indeed the question has been raised if the concept of directional selection is necessary for anything but the simplest adaptational changes. Allometric shifts, which may be subject to strong epigenetic constraints, exemplify critical-point emergences, and are subject to behavioral adaptabilities Can the organism work with the shift? Or is it capable of changing its behavior to suit? It should be remembered that regressions occur on the way to specialization, again limiting the organism’s functional options. Yet some developmental regressions such as paedomorphosis can be keys to escape from specialization.

8 The Re-invention of Emergence

The distinction between sums-of-parts (or aggregate) and structural characters is crucial to biology: if new and nonaggregate structure had not emerged at various times in the history of life, neither the tiers of life’s hierarchy nor selection at these successively higher levels of organization would ever have evolved. —Elisabeth Vrba, 19891 The vast mystery of biology is that life should have emerged at all, that the order we see should have come to pass. A theory of emergence would account for the creation of [that] stunning order. . .as a natural expression of some underlying laws. It would tell us if we are at home in the universe, expected in it, rather than present despite overwhelming odds. —Stuart Kauffman, 19952 [The big questions are about] causes, strengths of causes, levels of causes, and contingency. That’s not a bad formulation. . . . The emergent property is the emergent property, and that’s all you can ever say about it. —Stephen Jay Gould, 19963

Emergence wants re-inventing for several reasons. First, the old concepts have been largely forgotten, along with their emphasis on spiritual emergents that transcended biological realities. Second, its re-invention might relieve the current sense of unease among some evolutionists. Third, biologists already treat emergent phenomena with easy familiarity, and are likely to be receptive to a formal treatment. But there remains a need to explain how emergences are generated and why they matter. Elisabeth Vrba says that as non-aggregative “structures” they have been essential for evolution. Stuart Kauffman expects an emergence theory to state the “underlying laws” of the creation of order—thereby explaining progressive evolution. Stephen Jay Gould challenges us to find more to say about the emergent property than that it simply exists. Kauffman, like most complexity theorists, is on a grail quest—for a simple formula that would apply equally well to physicochemical phenomena, the emergence of life from heterogeneous mixes of autocatalytic molecules, or to the emergence of new levels of organization in multicellular organisms. Fittingly, he uses the plural,

290

Chapter 8

“underlying laws,” because there are more than one, and they do not lie only at the physical foundations, because each new emergent level manifests a new set of overlying laws. Therefore emergence implies discontinuity. And although the current literature is still prejudiced against the word “saltation,” it would, by any other name, be just as jumpy, even in cases where the generative conditions have varied continuously. For example, when dealing with change in non-living, dynamically stable systems, complexity theorists use “bifurcation” to signify the emergence of “sudden and dramatic changes in trajectories and attractors.”4 Another fundamental aspect of emergent evolution is that it often arises from organizational changes in hierarchical systems, something that its early proponents perceived but did not analyze. To explain it in that context is a major challenge to modern emergentists. In chapters 4–7 I brought together some of the pertinent biological information, detailing the programs of performances staged in the various causal arenas. Now we need to transcend the circus metaphor to bring in hierarchical structures. We can approach the problem of understanding evolution at any level in a biological hierarchy, provided that we are willing to move up to the highest levels as well as down to the lowest, without judging any one of them as more significant than any other. Physiologists do that without even thinking about it. Unfortunately, much of conventional biology that doesn’t stubbornly stick at the population level plummets to the genetic basement. The epigenetics department is attracting greater numbers, but the physiology mezzanine is largely deserted. I will begin this chapter with a more comprehensive statement of the definition of emergence that appeared in the introduction and in chapter 2: Emergence is the spontaneous appearance of novel qualities through the interactions and constraints of generative conditions, consisting of the dynamic structure of the original, and properties of its environment. Thus stated, emergence includes a wide range of physical events from the Big Bang to the physicochemical reactions that produce liquid water from hydrogen and oxygen at appropriate temperatures and pressures. It also allows for the introduction of a catalytic factor. And it assumes physicochemical and biological constraints on natural experimentation. Evolutionary emergences normally depend upon a genetic foundation and its reproduction to see them through to the next generation. But there are intermediate biological emergences that are effected by environmental conditions that may persist for many generations. Until such time as they are genetically assimilated they are not biologically reproduced. They are simply always present as part of the generative conditions. In that sense they are “innate” in the broad usage of Susan Oyama (1985) and William Wimsatt (1998). Now I will pick up the patchy history of mechanistic attempts to grasp emergence, and survey the thoughts of biologists who have portrayed it as an independent generative process of evolution. Some Modern Synthetists have acknowledged its importance, while subordinating it to the “ultimate” mechanism of natural selection.

The Re-invention of Emergence

291

Following a round up of others’ ideas, I will delineate intrinsic and extrinsic events and their saltatory or gradual gaits. Emergence after Morgan “Emergence” is to me the obvious choice of name for a process competent to leap the barriers of ecological, organismal, and theoretical stability. Thus stated, the concept challenges the causal gradualism proposed by neo-Darwinism. It modifies and builds upon C. L. Morgan’s Emergent Evolution, which has an honorable if largely forgotten past. In Evolutionary Theory: The Unfinished Synthesis (1985), I discuss Morgan’s work extensively, and therefore kept my chapter 2 outline brief. The concept is unprejudiced as to the sources of evolutionary changes, whether they be genetic, epigenetic, associative, physiological, behavioral, or environmental. It has been perennially perceived as innovation in any of those arenas. My emphasis has, however, been on changes that establish multifunctional, adaptable properties. Such novelties can arise in a variety of ways from any level in the hierarchy of life: at critical points (or thresholds) in continuities of change from a complementarity of previously unassociated systems by radical autonomous re-organization though the complexification of existing regulatory dynamic structures by simple mutations that advance adaptability from the organism’s actions from the environment’s effects and the organism’s reactions. This, however, is getting ahead of the game, since Morgan did not attempt a serious analysis of emergent processes. Nor did Arthur Lovejoy, pioneer of “the history of ideas.” However, his essay “The meaning of emergence and its modes” (1927) set out the range of qualities that characterized emergences, excluding transcendental processes such as vital sparks. I have here rearranged his list of the qualitatively different kinds of emergences that might occur in evolution, in descending order of importance5: l. New events irreducibly different from old ones.

292

2. 3. 4. 5.

Chapter 8

New types which have some new qualities and may lack some old ones. A new quality in a pre-existing organism. An improved method of doing the same general thing. Proportionate numerical increase.

Lovejoy did not intend this is to be an ideal table of mutually exclusive categories, and it covered the bases so widely that almost any biological change could be labeled an emergence. But he gave us the first materialistic, analytical approach to emergent evolution. May we take it from the bottom? The least category, number 5, includes natural selection as differential survival and reproduction. Since, at the molecular level, structural DNA mutations are all-ornothing saltations, they would also be considered emergences, but their consequences might be merely neutral or adaptational—improved methods of doing the same general thing, as in category number 4. Progressive steps in the evolution of adaptability, or self-organized integrity, could also come in at this level. They might be just as important as some processes at the top of the table. Number 3—new qualities in old organisms—would include the physiogenic changes in fish migrating back and forth between the sea and fresh water, and the development of lungs in stagnant-water fish, or wings in reptiles, or all of the familiar divergent homologies of placental mammals, as well as mechanisms of allometry and orthogenesis. Under item 2 there are any number of archetypal novelties of plants and animals, poised to invade new environments or dominate the old, and to diversify widely as a result of their emergence. Number 1—new events—are exemplified at the primordial functional level by sexual reproduction, or at the organismal level by the symbiotic emergence of eukaryotes. The major emergences of life and mind would top the list. Emergence through the Looking Glass Ernst Mayr Adaptation and numerical increase, the least of Lovejoy’s categories, have drawn almost all the attention that selectionists have given to specific cases of evolution for the last century. However, Ernst Mayr’s 1960 essay “The emergence of evolutionary novelties” took a broader view. Since some modern biologists still cite it as the authoritative word, it deserves further scrutiny. To keep emergence in safe hands, Mayr proposes that Darwin had already anticipated it, although it had surely been “greatly neglected during the past two or three decades, in spite of its importance in the theory of evolution.”6 Mayr subse-

The Re-invention of Emergence

293

quently denounced all of the saltationists, essentialists, typologists, and hopeful monstrologists who had not neglected emergence. Notwithstanding, his earlier essay established the selectionist view of emergence, and shows insights that were not blinkered by its conventions. “Evolutionary novelty” gets the working definition of “any newly acquired structure or property which permits the assumption of a new function.”7 Although Mayr remarks that biochemical novelties based on a single gene mutation are more likely to confer the quality of “general adaptation” (which I would call “adaptability”), he mainly discusses structural novelties that represent “adaptation to a more specialized situation.”8 Here follows a brief summary of what Mayr takes to be the causes of the emergence of novel structures, along with my comments in brackets: l. Pleiotropic by-product: a secondary, “neutral” phenotypic feature of an existing gene acquires selective value. [Or, a hitchhiker takes the driver’s seat. Where the expression of a gene contributes to several phenotypic features it is said to be pleiotropic. To an epigeneticist, the earlier a gene participates in development, the more pleiotropy would be involved.] 2. Intensification of function. An example is the intensification of the locomotory function of a limb that leads to specialization for high-speed running. In artiodactyles—cloven-hoofed animals—there is no radical innovation but an improvement in the mechanical efficiency of a tarsal joint. A small structural change makes drastic reorganization of the phenotype possible. Moreover, such improvements can lead to an “evolutionary avalanche.” [Mayr does not, however, address the cause of the allometric shift in the growth of the limb, or the trend toward epigenetic emphasis of a single digit, or even ontogenic anatomical modifications that result from behavior. And although these are not radical innovations either, they would seem to be more central to this intensification of function than a joint modification. This category comes under critical-point emergence.] 3. Change of function. This includes duplication of an organ, and variation of function. The same effect can come from extension of an organ and its subdivision into two areas with different functions, as in a digestive system. There is always “a transitional stage during which both structures function simultaneously,” e.g. the changes from ovipositor to sting, jawless fish endostyle to thyroid, scales to teeth, and leaves to petals. A final type of functional change is one where an “existing structure is preadapted to assume a new function without interference with the original function.”9 [This brings in multifunctionality, a common emergent property; for example, body surface vascularization serves variously for respiratory gaseous exchange, heat absorption, and cooling. But these come into play simultaneously at the point of emergence.]

294

Chapter 8

So far I find little to disagree with, although biochemical adaptability has been given short shrift, multifunctional adaptability is put down to preadaptation, and perish the thought that orthogenesis could be a process of intensified function. Mayr also emphasizes the role of the environment and the behavior of the organism. Ongoing minor changes in large environments were unlikely to be of major importance, though “broad adaptation” [ = adaptability] was likely responsible for success in the face of such change. I take the contrasting position that specialization, regression of adaptability, and ecostasis are the most likely products of dynamically stable environments. A small pool of adaptable generalists will persist if periodic changes are sufficient to give them an occasional break from the usual restraints of selection, although their fitness will usually be low relative to more specialized variants. Mayr gives some credence to this point of view by citing certain songbirds of the Oscines genus. Among them, the omnivorous generalists have relatively large brains, while specialists are more typically “bird-brained.” In his work on the plasticity of the song control system of birds, Eliot Brenowitz is more specific. The brain nuclei in birds that have a greater adaptability to learn new songs into adulthood, are capable of seasonal expansion.10 According to Mayr, a phenomenon of even greater significance occurs where “the active shift of an organism into a novel niche or entirely new adaptive zone will set up a powerful array of new selection pressures. An organism must have a special set of characteristics to cope with the demands of the new environment. It must be “preadapted” for the new world in which it will henceforth live.”11 Mayr goes on to remark: “Perhaps most astonishing is the relative slightness of reconstruction that seems to be necessary for successful adaptation to rather drastic shifts of adaptive zones.”12 It is not astonishing to a physiologist, who might be more aware that physiological adaptability gives the occupants of interfacial environments the potential to cross over, and to experiment with various behaviors. Morphological change is initially a minor consideration. New selection pressures are supernumerary since the adaptability is already there. Mayr insists that unspecialized types are less able to emerge into new environments than specialists, and that to conclude otherwise is a relict of typology. Archetypes “could never have existed in nature.”13 It is true that a “combination of genetic and developmental homeostasis may give the phenotype such uniformity and stability that it may not be able to respond phenotypically to a change in the environment.”14 But exceptions have already been noted, and physiological homeostasis allows the organism to go on doing the same thing when the environment changes, and do different things when the environment remains the same, which amount to a whole hill of phenotypic responses. Finally, Mayr writes that “the more drastic the change in environment, the more rapid will be the evolutionary change and the more far-reaching, in general, the structural reorganization.”15 I agree with this conclusion not because of the “ultimate

The Re-invention of Emergence

295

causal” hypothesis of new, strong selection pressures, but because the emergent is already adaptable, and to a neo-Darwinist there will be an appearance of evolutionary change due to its sudden proliferation and diversification. Natural experimentation has been freed from selection pressure, to diversify, and finally to be entrapped by it again into specialization and regression. Mayr properly emphasizes the importance of behavioral change in initiating adaptive shifts that result in changes of functional anatomy. However, underlying the behavioral shift is the physiological adaptability that sustains it. Mayr also puts his finger on the kind of emergence that occurs as the result of the coincidental integration of previously unrelated organismic features: pre-existing building blocks, which when pieced together, give rise to an improbable new character complex of high selective value: The role of natural selection in these cases is apparently not the bringing-together of the individual units; this is done by forces independent of the prospective new structure. Natural selection enters the scene as soon as the pieces have been combined into a new complex which can function as a unit and can respond to natural selection as a unit.16 [my emphasis, and my inference to follow]

These independent forces would include physiological emergents arising from the contingent co-operation of independently evolved systems, illustrated by the origin of lactose synthetase in mammals from two proteins with previously different and independent functions. Another example is the complementarity of osmoregulatory pH change and respiratory hemocyanin function that allows the blue swimming crab of Chesapeake Bay to migrate 100 kilometers in brackish water. The principle would also apply to the coming together of previously independent symbionts. Repetitive Differentiation Mayr does adduce this at the anatomical level, and tries to save it for selectionism by giving Darwin priority. Although Darwin discussed it as “serial homology” and “vegetative repetition,” he gave the credit to Isidore St. Hilaire (son of Geoffroy) and Richard Owen, adding that differentiation of the parts was due to cell multiplication acted on by natural selection.17 In The Origin of Species, Darwin also referred to Milne Edwards’s (1834) physiological division of labor, which is associated with repetitive differentiation. If we were to be pedantic about it, we would also remember that in Theoria generationis (1759) Caspar Wolff described the duplication and differentiation of cells that originated as identical units in plant meristematic tissues—thereby anticipating Cell Theory. Mayr gives Severtsov (1931) and Gregory (1934) honorable mentions for dealing with repetitive differentiation, but omits Cope and Bateson. I have already shown how the latter duo put the concept in a non-Darwinist context. In 1929, L. J. Stadler

296

Chapter 8

discovered polyploidy and “gene reduplication” in barley and wheat species treated with x rays. Severtsov’s student Schmalhausen referred to chromosomal duplications as a kind of polymerization: “The most important transformations that have occurred in various types of plants and animals are based upon an increase in the number of similar parts and the divergent differentiation of serial homologues and homonomes.”18 Julian Huxley (1942) realized that any duplication of the chromosomal material had evolutionary potential. Entire karyotypes could be doubled by “autopolyploidy” due to mitotic error, or augmented from a foreign source as “allopolyploidy.” C. W. Metz (1947) brought out the significance of “duplication [and differentiation] of chromosome parts as a factor in evolution,” synthesizing current work in fruit-fly genetics as well as Barbara McClintock’s early research on corn.19 G. Ledyard Stebbins The neo-Darwinist G. Ledyard Stebbins (1974) lists about 70 evolutionary emergences that would fit into Lovejoy’s highest three categories. Then he comments that the relative rarity of these saltatory emergences—not that he calls them that—makes them less important than the much more frequent selection of minor variations that resulted in allelic distribution changes in certain environmental conditions. John Maynard Smith John Maynard Smith, who is unequivocally committed to the ultra-Darwinist sect of the Modern Synthesis, addresses similar issues in “Evolutionary progress and levels of selection” (1988). He constructs a hierarchical table of levels of complexity: l. Replicating molecules 2. Populations of molecules in compartments 3. Eukaryotic cells 4. Multicellular organisms 5. Demes, social groups 6. Species 7. Groups with cultural inheritance.20

That such assemblages exist and represent an evolutionary progression is not at issue. Maynard Smith’s crucial questions are as follows: i. What is the nature of the genetic information that is passed from generation to generation at each stage? ii. How is the integrity of that information protected against selection at lower levels? iii. How did natural selection bring about the transition from one stage to another, since at each transition, selection for ‘selfishness’ between entities at the lower level would tend to counteract the change?21

Full and honest answers would solve some of the outstanding problems in evolutionary biology, but to confine the analysis to “genetic information” is to exclude the

The Re-invention of Emergence

297

organism and its actions yet again. The simple part of his third question has already been explained. Natural selection does not bring about transitions, and the transitions/emergences initially thrive best in the absence of natural selection. Question two already hints that protection against selection is part of the general answer. In a sense that selectionists find acceptable, though it is redundant to me, natural selection participates in the stabilization of new emergent levels by fine-tuning their connectedness. But it prevents the success of more radical natural experiments. The integrity of information resides in the integrity of the whole organism, and the organism is all that natural selection has to deal with. The extreme case of two molecules or tissues or cell lineages within the same organism competing with one another to the point of disintegration hardly enters the equation since the organism would not survive, much less reproduce, and natural selection has no explanatory relevance. Given a choice among similar individuals, those whose wholes are slightly greater than the sum of their parts will outcompete those whose wholes are slightly less. In The Major Transitions in Evolution (1995), Maynard Smith, along with Eors Szathmáry, goes further by compiling the processes involved in large-scale emergences. However, their program is to explain the generation of emergences by breaking down the major transitions into steps for which selective value can be identified. Those steps may be easier for gradualists to take than high jumps to hopeful monstrosity. On examination, however, they are revealed as smaller saltations of unexplained origin. My own take on transitions to higher levels will be addressed in more detail shortly, along with the outstanding problem of the generative conditions that promote emergences, once we get out from behind the Looking Glass. I will also respond to the units of selection problem in chapter 11. Before going on to recent currents of emergentistic thought that lack selectionist bias, I want to re-emphasize a cautionary note regarding oversimplifications that are made on both sides of the mirror. Key Innovations While the potential for multifunctionality and adaptability is the measure of the emergent property, evolutionary progress often seems to be realized by simple natural experiments that produce a single crucial emergent feature. Neo-Darwinists, cladists, and some earlier emergentists agree that specific, simple, catalytic novelties have led to new emergent levels that enjoyed immediate high fitness and subsequent, successful diversification. However we should be careful of seeking particulars and ignoring the larger context. The emergent condition of flight in insects may have been the allometric extension of dorsal flanges into discrete wings. Or was it the joints that allowed the wings to beat with aerodynamic efficiency? Flight needed a greater complex of generative conditions, including the arthropod exoskeleton, combined with waterproof wax layers, and an efficient tracheal system that provided enough

298

Chapter 8

oxygen for flight metabolism. Altogether they increased insect integrity so that everything functioned better simultaneously. Reptiles already had multiple features that could have led to warm-bloodedness, and some dinosaurs may have attained it. The avian lineage that perfected homeothermy also acquired feathers, which not only insulated them but also turned forearms into stabilizers and wings. Nevertheless, wings could only reach their avian versatility through allometric amplification, and in association with the parabronchial respiratory system and an increase in behavioral versatility. Was hair the catalytic innovation for emergent proto-mammals? It insulated these small primitive explorers during ventures into higher latitudes, mountainous heights, or cool nocturnal foraging. However, it was only significant as one of a set of generative conditions that included a nervous system sensitive and responsive to temperature change, connected with shivering and non-shivering thermogenesis. Hair not only contributed to heat preservation at the new emergent level, it also brought epitrichal glands, whose secretions were natural hair conditioners, air conditioners, and soft drinks for babies. And then they could diversify to produce pheromones, or expand to become mammary glands. In other grand emergences that have paraded in the evolutionary circus it is possible to point to specific, inconspicuous experiments that catalyzed the main event. Parasitic invasion, regression of defense mechanisms, where they existed, or resistance to digestion, could be interpreted as key innovations of proto-symbionts. Condensation of DNA and histones into chromosomes was a major innovation in eukaryotes. Multicells needed cellular adhesion molecules. As parts of integrins that could anchor to the intracellular skeletal matrix, and vary in their external adhesive and receptor qualities, they were more adaptable, allowing greater differentiality in developmental evolution. But it takes single vision to interpret these simple novelties, however produced, as the causes of emergent transitions: e.g., “Hair and feathers are adaptations to cold weather”; “Wax layers are adaptations to dry environments”; “Chromosomes are adaptations to chromatid confusion.” It is even harder to identify single beneficial “adaptive” features of cell communication and symbiosis. They add to organismic integrity, and adaptability: the “innate flexibility of constitution” that Darwin regarded as the foundation of persistence in being. And each one could only effect the emergence in the context of the array of other functions of a whole organism. To restructure the whole, all the other parts were required. A simplistic search for key emergent properties could be as misleading for emergentism as selection pressure is for selectionism. Though rare, the ability to fly and the ability to maintain a high body temperature are not unique to particular lineages. There have been multiple experiments in each. Can the conditions and requirements and potentials for emergences in all the arenas of epigenetics, physiology, behavior, associations, and environment be merged into a

The Re-invention of Emergence

299

single set that generates a formal universal law? For the moment, just say No. At this point a discussion of emergentists who have come out from behind the mirror would fit a logical sequence. However, they often emphasize emergence to new hierarchical levels. Therefore I digress to introduce the language of hierarchies and to demonstrate the parallels between hierarchical and emergent evolution. Emergent Levels and Hierarchies Molecules per se do not evolve; but their changes affect the development, physiology and behavior of whole organisms. And some molecular changes, beyond the gene level, are heritable. Organisms as individuals do not evolve either, but if they change, for example by invading and responding to new environments, they can influence evolution. To be appreciated in evolutionary terms, hierarchies need a temporal dimension, and, in that context, organisms do contribute to the evolution of their lineages through development, physiology, behavior, and association. Maynard Smith’s tabulation of “levels of selection” demonstrates how, when we come out from behind the Looking Glass, we must deal with hierarchical emergent levels, albeit with natural selection relegated to a stabilizing rather than an innovative role. One thing to keep in mind from the beginning is that emergence to higher levels does not always involve the addition of a new layer on top of the older ones. In the case of the vertebrate neocortex and the expansion of the cerebral hemispheres it does appear that way superficially, but they are functionally internested with foundational systems, such as the limbic and hypothalamic portions of the older brain, and its ancient medulla that regulates some important functions of homeostasis. In the prologue to emergence in chapter 2, I noted how similarly Henry Drummond (1894), Samuel Alexander (1920) and John Holland (1998) expressed how the new emergent level has a new relatedness whose qualities allow it to perform innovatively, without disobeying the laws that constrain the lower levels. Alfred North Whitehead’s Theory of Organic Mechanism, set out in Science and the Modern World (1925), illustrated this concept with an example that I here expand. He said that the behavior of an electron in a nerve axon is unlike that of a free electron in a disorganized, non-living system, because it is subordinate to myelin insulation, the passage of the electrical impulses in the nerve, and the physical restraints of the axon. His point is well taken, though we should be looking at ions rather than electrons. Nerve impulses are made possible by the electrolyte composition of the intracellular and extracellular environment, which is governed by biological ion pumps and gated channels switched on and off by neurotransmitter molecules. Initiation of an autonomic nerve impulse can be stimulated by the local environment, and then overridden by the central nervous system. For example, secretion of the various hormones and enzymes of the upper alimentary tract, which are initially under neural

300

Chapter 8

stimulation, may be terminated by “somatostatin,” the “body stopper” hormone, in response to an emergency that requires that blood be diverted from the intestinal mucosa. In the language of hierarchy theory, at the “focal level” of the nerve cell (the level under scrutiny), an electrical potential can be built up according to physicochemical, and biological conditions. At the lower, “generative” level, there exist operational constraints imposed by the availability of chemical energy, the state of membrane pumps and channels, and condition of receptors for neurotransmitters. The neuron’s action is triggered or inhibited by the influence of a higher constraining or organizing level. Behavioral and cognitive levels can activate or override the lower levels. To flee or hide from a predator is a common automatic reflex in lower animals. But in higher animals, intelligence can overrule automatic reactions and suggest different behavioral options, such as pausing to check for fire below, before leaping out of frying pans. Logic and language impose yet another regulatory level—the challenge “Friend or Foe?” can determine the degree of danger of the unknown and suggest options before anyone gets shot. At no point has the physical nature of the electron or ion been compromised, nor the physicochemical action of the nerve cell, nor the local hormonal and vascular qualities of an organ system, nor the automatic or reflex functions of the nervous system. Each of these represents a hierarchical level with its own array of holons and distinctive emergent features. The intrinsic emergent evolution of self-organization involves hierarchical relationships between building blocks that include molecules, genes, tissues, and organs. Organisms are also members of intraspecific and interspecific hierarchies, whose organization may wax and wane with environmental change. There are two kinds of biological hierarchies: “control hierarchies” and “compositional hierarchies.” Here I am taking a leaf from Stanley Salthe’s Evolving Hierarchical Systems (1985). The chain of command in a nervous system exemplifies a control hierarchy. Conscious choice orders the central nervous system to activate muscles and produce behavior. Some of the neural control hierarchy, as well as the hormonal system, is involuntary. Although we have seen how it can be affected by the larger environment, including other organisms, the intraorganismal control hierarchy is a very distinctly independent unit. A compositional hierarchy can depict the biosphere as the largest component (or even include the entire planet or the entire cosmos). Within it are internested biocenoses, ecosystems, communities, populations, demes, organisms, organs, cells, molecules. The compositional hierarchy is usually depicted with a Venn diagram of circles within circles. (By the way, the figure that I have used for the three-ring circus within the big top borrows in part from a Venn diagram but owes nearly as much to electron-orbital depictions and the Celtic triskelion. In other words the image is a pictographic simile, not a formal attempt at marrying causality, function, and structure.) Causal arenas could conceivably be forced into a Venn diagram, and thereby be made

The Re-invention of Emergence

301

more convincing to hierarchy theorists. I will demonstrate how misleading such simplicity can be. But for the moment I return to the nuts and bolts of hierarchical systems. “Holon” is a useful general term for the units that are usually associated into cooperative groups arranged in a hierarchical order of interaction. From the physiologist’s point of view an array of molecular holons, such as hormones and enzymes, may be characteristic of an organ-level holon like the pancreas or liver. Such organs are under the regulative control of higher-ranking holon organs,. such as the pituitary and the brain. The word “holon” was used by Arthur Koestler in The Ghost in the Machine (1967), building on the foundational ideas of J. H. Simon’s 1962 article “The architecture of complexity.” For Koestler, “holon” meant simply a subunit or module of a hierarchical level, which is the most effective usage. A holon could be part of a hierarchical anatomical structure, i.e., part of a compositional hierarchy, and part of a control structure, i.e., under the command of a higher level as well as interacting with other holons at the same level. In more recent literature, the treatment of this subject comes under the heading of modularity, where holons are called “modules.” Furthermore, repetitive differentiation would be taken as a subset of modularity. I prefer “holon” because of its holistic implication, but I will use “holon” and “module” synonymously. Waddington’s (1966) version of modularity did little to advance the principle. The repetitive differentiation component of modularity was brought to the fore in Susumo Ohno’s Evolution by Gene Duplication (1970). Rupert Riedl works the repetitive differentiation of modules into the organization of hierarchical patterns, in Order in Living Organisms (1978). Although he emphases its morphological aspects, his general principles apply at all hierarchical levels. Developmental evolutionists such as Rudolf Raff (1996) and Scott Gilbert (1997) have popularized “modularity,” and Modularity, edited by Werner Callebaut and Diego Raskin-Gutman (2005) is the latest and most comprehensive work. Koestler’s illustration of the hierarchical structure of an army emphasizes the chain of command. Individual soldiers are holons ranked into higher holons—platoons, under a junior officer. Companies are holons that unite several platoons under a more senior officer, and so on up through regiments and brigades. Thus, any rank, or level of organization, consists of a number of separate but coordinated holons, each of which have a self-assertive tendency and an integrative tendency; in other words, independence and cooperation. An array of holons may be identical, as in the case of multiplied genes that turn out large amounts of a given protein. Simple duplication of modules evolves first, and may immediately serve the function of increasing a crucial process. With or without such a function an array may then differentiate. An example would be the ensemble of genes for enzymes with identical functions but different temperature optima. These are what W. R. Ashby (1952) called “step mechanisms.” They are usually turned off until appropriate conditions for their operation arise. An

302

Chapter 8

array produced by repetitive differentiation may alternatively form a biochemical pathway, taking a substrate through several changes until the final product appears. Its organization may be made more complex by branching, the final route being decided by prevailing physicochemical conditions. Holons, or modules, at the anatomical level evolve according to similar principles, first multiplying and then differentiating. If the holon begins as a multifunctional organ, the repeated units may specialize for particular functions, involving both anatomy and the organization of biochemical syntheses. The progressive evolution of the gill pouches of primitive vertebrates illustrates how far such physiological specializations can diverge. Just as independent biochemical pathways might contingently associate, so also do formerly independent anatomical units cooperate by progressive packaging. The close physical association of the pituitary gland with the hypothalamus, via a vascular portal system, is an example. The evolution of higher levels of organization is what emergence theory is largely about, so when they first appear it is appropriate to call them “new emergent levels.” Recall that the level of organization or emergence under consideration is the focal level, and the level below it is the generative level. I have already hinted at the importance of holistic analysis of the generative conditions for a given focal level, in assessing how emergence to a new level occurs. David Rollo, whom I regard as an honorary emergentist, comments on how hierarchical structure is apposite to emergence: Hierarchical organization may be crucial for compartmentalizing interactions and allowing large, complex systems to retain stability. Not only does hierarchical organization provide reliability and stability but modularized structure also allows modification of subcomponents without global disruption. Thus, hierarchical structure allows systems of great complexity, which also retain the ability to evolve. Both are key attributes of life.22

Now what do my other fellow biological emergentists have to say about hierarchies? Order in Living Organisms (1978) merits Rupert Riedl’s conscription as an emergentist. For example, Riedl refers to “quick breakthroughs to new forms of organization.”23 He frequently alludes to how the reorganization of modules brings new adaptabilities that are then lost as a result of canalization. And, in his conclusion he notes that the highest states of order involve “the greatest possible differentiation, complexity, and individualization.”24 Regarding hierarchies, Riedl writes: Hierarchical order is characterized by features (or concepts) whose fields of validity do not overlap but are contained within each other, so that several lower concepts of equal rank are usually included in a higher concept. The higher concept specifies the significance of its lower concepts, and the latter specify its contents.25

The Re-invention of Emergence

303

Since the evolution of hierarchies “depends on the definitive fixation of additional features,” repetitive differentiation is important.26 Riedl’s synthesis is thoroughly Aristotelian, marrying a hierarchical system’s “degree of complexity” with its “degree of integration.” The former is defined as no more than the sum of holons. The latter depends, first, on the “degree of coordination.” Some systems might operate effectively if its holons vary a little, but not if they vary a lot; i.e., there is some limit of tolerance to variation. Integration also works according to a “degree of dependence.” That refers to how essential a particular holon might be in a dynamic system. Riedl’s example is an automobile that can be driven indefinitely without a hub cap, but not with a bent axle.27 He places strong emphasis on the importance of hierarchical position in determining how much of the rest of the system will be adversely affected if it changes.28 Numerous biological examples are provided of dramatic phenotypic variations that arise late in the epigenesis of orchids, insects, birds, and antelopes when Riedl discusses systems of maximal freedom from genetic determinism.29 But he also points to organs that have been strongly conserved throughout evolutionary history due to their “epigenetic burden,” presumably because they have key hierarchical positions, and their alteration would be disintegrative. William Wimsatt (1986) calls this “generative entrenchment.” Epigenetic burden can be referred back to the accommodation problem of evolutionary developmental change, which worried Bateson, and made Fisher reject the hopeful monster. Yet, late phenotypic developments or not, Riedl is showcasing not only hopeful, but successful monsters. And some modern theorists, Riedl’s students among them, have done an end run around the accommodation difficulty, as I will point out in the next chapter. For the time being it is sufficient to acknowledge that hierarchical position is indeed important, but that there need be no adverse consequences arising from early deviation if there is harmonious reintegration during the chronological processes of epigenesis. Riedl was the first biologist to provide a comprehensive theoretical analysis of the evolution of hierarchical organization, and an inspiration to those who were to follow, including myself. Also among them are Elisabeth Vrba and Niles Eldredge (1984): Hierarchy is a central phenomenon of life. Yet it does not feature as such in traditional biological theory. The genealogical hierarchy is a nested organization of entities at ascending levels. There are phenomena common to all levels: (1) Entities such as genomic constituents, organisms, demes, and species are individuals. (2) They have aggregate characters. . .but also emergent characters (arising from organization among subparts). Character variation changes by (3) introduction of novelty and (4) sorting by differential birth and death. Causation of introduction and sorting of variation at each level may be (5) upward from lower levels, (6) downward from higher levels, or (7) lodged at the focal level. The term “selection” applies to only one of the possible processes which cause sorting at a focal level. Neo-Darwinian explanations are too narrow, both in the levels (of genotypes and phenotypes) and in the directive process (selection) which are

304

Chapter 8

stressed. The acknowledgement of additional, hierarchical phenomena does not usually extend beyond lip service. We urge that interlevel causation should feature centrally in explanatory hypotheses of evolution. For instance, a ready explanation for divergence is “selection of random mutants.” But upward causation from genome dynamics (or downward causation from the hierarchical organism) to the directed introduction of mutants may be more important in a given case. . . . A general theory of biology is a theory of hierarchical levels—how they arise and interact.30

If that were so, an emergence theory would be little more than a theory of hierarchical levels. How they arise would certainly be the crucial question, and how they interact its elaboration. “Sorting” would be seen as the consequence of the quality of novelties as they emerged, not as an explanation of their generation. And sometimes it is a barrier to evolutionary progress. Moreover, throwing demes and species and genes into the pot distracts us from the organism, and its immediate interplay with its environment. Vrba and Eldredge do not come out and say it in so many words, but you don’t have to read too deeply between the lines to see that “a general theory of biology” would have to rest on a new theory of evolution. In “What are the biotic hierarchies of integration and linkage?” (1989), Vrba calls for an “expanded evolutionary theory” and baldly states that it is dishonest to claim that that the Modern Synthesis can be stretched and modified to infinity. At this point I have no wish to engage in semantic quibbling, but should explain that she uses the word “structure” in the special sense of a biological emergent phenomenon. Structuralists would use the expression “dynamic structure” which includes developmental, and physiological functions, as well as anatomy. The generative conditions at any hierarchical level is usually based on some kind of structure—molecule, cell, organ etc. Although Vrba knows that behavior has a downward causal effect on both physiology and anatomical development it stretches the word “structure” to include behavior, the physiogenic interpenetration of organism and environment, and the interaction between the organism with its own kind, or other organisms in its environment. Although simplistic hierarchal analyses tend to be empty of such interactions and interpenetrations, Vrba does not omit them. Therefore, as I quote from her essay I will substitute (in brackets) “emergent” for “structural” and “emergence” for “introduction of structure.” “Aggregate characters” will be left alone as the additive parts that do not result in greater wholes: Introduction (or origin) of new variants. De novo introduction of both aggregate and [emergent] characters occurs at all levels. I use the term introduction in the simple vernacular sense of a first appearance to refer solely to the first origin of a new variant, in distinction to sorting among variants. Thus, [emergences] at the gene level included various forms of gene mutation. [Emergences] at the organismal level refers to the first appearances of “phenotypic mutants,” including minor phenotypic innovations and the one-step origins of complex ontogenetic products that may diverge considerably from the parental phenotype. At higher levels heritable [emergences] occur less frequently than at lower and require changes at all lower levels.31

The Re-invention of Emergence

305

The emphasis in the second last sentence is mine. Hopeful monsters, whatever they are called, deserve the attention. Vrba’s rules of emergence that “apply in common to the evolution of all levels” follow, with some comments of my own at the end of each in brackets. 1. Each new hierarchical step was initiated by the origin of recognition and interaction among entities already present and differentiated at the existing level. Thus, recognition between cells was required for the origin of sex (i.e. species sensu stricto) and of metazoan differentiation, just as it still is during each reproductive process and each ontogeny. [Being in the same place at the same time: I think this rule justifies my lumping of associative emergences in chapter 3.] 2. Complex structures and processes at existing levels presented a pool of available materials and, at the same time, strong constraints for integration in the new evolution of higher levels. Since then, in evolution within each level, the same principles have applied again and again. [Repetitive differentiation—separation of offices and concurrence of efforts.] 3. During the earliest evolution of each higher level, as emergent structure and therefore selection were initiated, there must have been “selection wars” between higher and lower selection regimes (Buss 1987, and below). These wars were concluded long ago in net favor of the higher individuals that exist. (Where the higher level lost, it is not there for us to see.) [Some experiments in the separation of offices are simply disintegrative, and don’t need to be explained in metaphorical militaristic terms.] 4. Sorting and change at lower levels need not affect a higher level in more than a sums-of-parts or aggregate way (such as cellular dynamics within the same organism, and organismal turnover and change within persisting species). Yet any selection at higher levels must entail sorting among individuals at all lower levels. [Co-adaptational fine-tuning follows the emergence of physiological novelty, but does not generate further emergences. However, fine-tuning that gives the organism a slight edge will likely result in differential reproduction among organisms that share the emergent property.] 5. A related observation is that hierarchical evolution has involved a progressive loss of autonomy at lower levels as autonomy increased at higher levels.32 [For example, we don’t regenerate as well as earthworms when we are cut in two—imperfection of adaptation?] Much of what I have developed in earlier chapters integrates with Vrba’s assessment of the relationship between hierarchies and emergence. At this stage of development

306

Chapter 8

of an emergence synthesis we need rules that apply at all levels, within levels and between levels. Otherwise we will never reduce to a model that is easier to grasp than the confusing reality. Yet, in addition to rules common to all hierarchical levels, we still must find newly emergent rules that are not to be found at lower levels. Coeval with Vrba’s thoughts on the importance of hierarchical evolution are those expressed by Stanley Salthe in Evolving Hierarchical Systems (1985). He too strays from his neo-Darwinist roots, treats Koestler’s ideas with some respect, and recognizes the relationship between hierarchical evolution and emergence. For example, he observes that the predictability of emergence is low where there are many possible generative conditions. But when they are few and repetitive, the resulting emergences are obvious and predictable. Therefore there is a degree of observer bias.33 Salthe also accepts that differentiation produces more complex new holons whose interactions provide new generative conditions. And, he notes, “the organic evolutionary process obviously is easily registered in the organism, contemplation of which was the original impetus to imagine that process. Traditionally, organic evolution is measured and tabulated by organismic alterations in time.”34 This traditional measurement also applies strongly to emergent evolution. But Salthe then extends the assessment to population and ecosystem levels. How he does it makes good sense to me, since it involves interactions within communities and between demes that have a downward causal effect on their constituent organisms. But he runs the risk of pandering to “population thinking” as a superior approach to the measurement of evolution—which brings me back to the oversimplification that can come from the compositional hierarchical concept. It is often used as a filing system for entities based upon their degree of magnitude. For example the cosmos contains biospheres, made up of ecosystems, communities, demes, and organisms. Then the organisms contain organs, cells, and molecules. However, emergent levels cannot easily be equated with compositional hierarchical levels. Within the organism, the evolution of the two have identifiable similarities. For example, this hierarchical system—molecular systems < simple cells < complex cells < simple multicells < multicells with differentiation and integration < complex organisms—is equivalent to a series of emergences. Moreover, it is recapitulated in part during the epigenesis of the individual organism. However, emergences, and new rules that they generate, have not obeyed this kind of simple progression. At many times in evolutionary history the activity of the higher levels has had a downward directional effect on the lower levels. Emergent adaptability may come at the molecular level long after emergent complexity has been completed at the gross anatomical level. Furthermore, the compositional hierarchical system that has just been laid out is often continued as organism < deme < population < species. In a Venn diagram these are just circles within circles; but in reality there is a major break between the organism and its environment, physically, biologically, and philosophically. Demes and populations are entities, but they are not higher, regulatory, emergent novelties. They always

The Re-invention of Emergence

307

existed, regardless of the emergent level that had been reached, whether it was a group of simple proto-cells, or a gaggle of geese. To be sure, the coordinated behavior of a flock of birds is an emergent phenomenon. And social behavior has a feedback effect on the physiological state of the individuals that make up the group. But emergent evolution does not continue on a larger scale into demes and populations. There is a break point. We know how important families and societies are for the intellectual development of young humans. But in most other organisms there is a gulf in evolutionary causation between intraorganismal and interorganismal phenomena. Such caveats regarding the comparison between emergent evolution and the evolution of hierarchal systems were behind my original decision to organize my thoughts in the context of causal arenas. Ultimately, however, the latter also need to be analyzed in terms of the former. Recent Currents of Emergentistic Thought In the early 1970s, when I first read C. L. Morgan and began to take an interest in evolutionary emergence, a movement called “biological structuralism” was developing. It addressed the hierarchy of biological organization, and the emergence of higher levels of from simpler ones. Structuralism gives a limited interpretation of emergent evolution, since it largely ignores the significance of environmental contingencies, including the potential for organismal associations. The structuralistic concept of “autoevolution” developed by Lima-de-Faria (1988) hints at emergence, in the loose sense of the appearance of a new phenomenon, but for him everything new is old again: innovations primarily arise from the ancient physicochemical nature of the universe. The significance of physical autocomplexification cannot be gainsaid when the origin of life is being considered. But subsequent emergences in biological evolution have largely been generated by biological causes. Neil Campbell’s popular introductory textbook Biology presents “hierarchy of organization” as the first topic of the first chapter, with “emergent properties” close on its heels: With each step upward in the hierarchy of biological order, novel properties emerge that were not present at the simpler levels of organization. These emergent properties result from interactions between components. A molecule such as a protein has attributes not exhibited by many of its component atoms, and a cell is certainly much more than a bag of molecules. . . .35

Campbell goes on to explain that it is unnecessary to adduce vitalism to explain emergence, and then points out that DNA function was far better understood once its interactions with other cellular functions were discovered. In other words, a holistic approach was necessary: “Biology balances the pragmatic reductionist strategy with the longer-range objective of understanding how the parts of cells and organisms are functionally integrated.”36

308

Chapter 8

That would be fine if it were true. Fifty years ago, when DNA structure was elucidated, a few wilderness voices cried about the need to understand the control of gene expression. Jakob von Uexküll had foreseen the need for a hierarchical genetic regulatory system as soon as the first results from Drosophila research began to be published. L. L. Whyte (1949) had already anticipated the importance of the organization of interacting genes and enzymes, admonishing reductionistic biologists for thinking of the cell as a bag of molecules. In 1957, C. H. Waddington’s “The Strategy of the Genes” included not only the adaptational significance of their mutations, but epigenetic variation of their expression and integration. What took the rest of us so long? And when it comes to working the concept of emergence into evolutionary theory as well as physiological organization—well, that is why you and I need to have this tête-à-tête. In “Life’s irreducible structure” (1968), Michael Polanyi wrote about how emergent levels in hierarchies imposed boundary conditions on the lower levels. He had already remarked in Personal Knowledge (1958) that it was difficult to get such ideas accepted unless the recipients are willing to learn new words and definitions. If they have a premonitory distrust or fear of what might be in store they will resist the neologisms and the framework of the ideas. “Proponents of a new system can convince their audience only by first winning their intellectual sympathy for a doctrine they have not yet grasped.”37 It may not now be as insurmountable as Polanyi suggested, due to a slow, subconscious osmosis of “the emergent property,” “holons,” “modularity,” and “hierarchies”—to the point where “emergence” is commonplace in the titles of current books. The neurobiologist Roger Sperry was an important latter-day emergentist, particularly interested in research into “split-brain” phenomena and the role of the corpus callosum in the integrity of mind. I have already mentioned his emergentist epiphany in the 1960s. Although he concentrated largely on mind as emergence, Sperry continued to have a significant, if controversial role, in promulgating an “emergent interactionism” that merges with views that Susan Oyama has elegantly expressed. My own tentative ideas on the need to develop a theory of emergent evolution were published in Evolutionary Theory: The Unfinished Synthesis (1985). They are re-disseminated throughout the present work, so I mention them only to map my own entry point on the road to emergentism. Eugene Balon was also developing emergentistic ideas before his publication of a tentative theory of saltatory ontogeny in 1986. Its major premise is that a series of stable developmental phases is punctuated by thresholds, at which bifurcations of behavior, physiology and anatomy with evolutionary potential are possible. This intuitively leads to the notion that saltatory ontogeny recapitulates saltatory phylogeny, which I would call phylogenetic emergence. We agree that ontogeny is not predetermined by a genetic template, but that the two are in a state of dynamic interaction. The following authors merit their own sections for their recent contributions to emergentist thought.

The Re-invention of Emergence

309

Gerd Müller and Günter Wagner Gerd Müller and Günter Wagner discuss emergentism in all but name in their 1991 essay “Novelty in evolution: restructuring the concept.” They begin by pointing out that the problem of the origin of novelty was acknowledged by Darwin as well as by his critics. But advances in genetics provided only the general answer of mutational change, and the problem was subordinated to the species question and dissipated by adaptationism. They acknowledge that Ernst Mayr’s hypothesis of function change, in terms of repetitive differentiation of structure, is part of the picture. But a major unanswered question is “Can new structures arise without a change of function?”38 Taking the level-of-organization approach, they attend first to populations, ecosystems and natural selection. Giving these short shrift, they go down to the gene level. Here they acknowledge the evolutionary significance of structural and regulatory gene mutations, chromosome mutations and polyploidy, noting that none of these dominate as candidates for the emergence of novelty. But for the most part their focal level is morphogenesis. Here I will abstract their broader evolutionary generalizations, returning to the details of developmental heterochrony in the following chapter. Müller and Wagner are primarily concerned with any morphological novelty that can be defined as “a structure that is neither homologous to any structure in the ancestral species nor homonomous to any other structure of the same organism.”39 They recognize that the appearance of novelties derived by differentiation from repeated elements is important, as are innovations derived by regression. But they concentrate on Lovejoy’s primary category of “new events irreducibly different from old ones.” In placental mammals, the novel corpus callosum in the forebrain crossconnects the expanded cerebral hemispheres. Unique marsupial bones emerged in the mammals, and in marsupials they support the pouch, but placental mammals have lost them. Epitrichal sebaceous glands that gave rise to sweat and mammary glands are regarded as true novelties. Innovations at lower taxonomic levels include the thumb of the giant panda, and horns and antlers in hoofed mammals. I take issue with the assertion that non-trivial novelties, including those that arise by differentiation of repeated elements, and truly novel elements such as new bones or fiber tracts must have “profound adaptive value,” and hence can be fixed in populations by natural selection.40 For one thing, “nontrivial” and “profound” amount to the same thing—selectionist redundancy. For another, if they emerged suddenly as full-blown adaptive features, the process was likely saltatory. And for a third, they could first have emerged as trivial features beneath the attention of natural selection. But the more important questions about morphogenic emergences proposed by Müller and Wagner deserve serious attention: 1. What is the generative potential of the developmental mechanisms to the members of the ancestral taxon?

310

Chapter 8

2. What are the critical changes in generative mechanisms of development that allowed the realization of the derived feature, i.e., the novelty? 3. Which genetic changes were the reason for the heritability of morphological novelties?41 These can largely be analyzed by comparative studies of extant organisms, even although the direct ancestors of emergents with novel properties are extinct. However, Müller and Wagner caution us that the organismal-developmental context needs to be known before question three can be answered. To initiate a study of the generative modes of morphological emergence, they organize them into the following categories whose contents I paraphrase, again with my personal comments in brackets. Hierarchical Organization Levels of organization from molecules to ecosystems are hierarchically ordered. In development hierarchical organization is built in a chronological sequence. Therefore [as Geoffroy long ago realized] changes in early stages effect large changes in later stages. Since concrete examples can be provided of the effects of early changes in cell lineage, this is no longer a merely speculative supposition. Interactivity and Dissociability During development lineages that diverge from one another can still form a network of interaction. [These can settle into the dynamic stability of “homeorhesis,” as Waddington called it. Or, as Wimsatt (1986) would say, the degree of interaction determines the degree of generative entrenchment.] Thus, development is canalized [systems are entrenched], and change requires dissociation of networks. Nontrivial novelties emerge despite epigenetic constraints that prevailed at the generative level, i.e., in the ancestors. Equilibria and Thresholds Existing conditions of homeorhesis are most vulnerable to change at critical stages or thresholds in development. There is therefore a potential for saltatory, emergent novelty if physicochemical influences are altered, or heterochrony occurs. [Like Lovejoy (1927), Müller and Wagner allow that their categories of the generative modes of emergence are not mutually exclusive. Also, they seem to view emergent novelty in the context of Spinozan/Darwinian persistence in being—no criticism implied.] The origin of new body plans required the origin of morphological novelties, but it also requires the integration of this new character with the other parts of the organism. In this context it is irrelevant whether integration is due to functional necessities or due to epigenetic interdependencies. What counts is that some

The Re-invention of Emergence

311

characters acquire an indispensable biological role that is conserved in spite of changing adaptive pressures.42 Brian Goodwin and Ricard Solé I have frequently referred to Brian Goodwin’s publications that deal with emergent phenomena. With its call for a “Science of Qualities,” How the Leopard Changed Its Spots (1994) influenced my own thoughts on the matter. Such a science would emphasize the generation of change and the nature of novelty. During the development of the present work, he and Ricard Solé published Signs of Life (2000). They offer a selection of physicochemical and biological emergences whose properties can be mathematically captured, but which, as non-linear dynamic processes, are unpredictable. This approach complements that taken by John Holland, and together the two books contribute to a solid quantitative foundation for emergentism. Jack Cohen and Ian Stewart Jack Cohen and Ian Stewart, in The Collapse of Chaos (1994), come close to a formal description of emergence when they say that a “collection of interacting components can spontaneously develop collective properties that seem not to be implied in any way in the individual pieces.”43 Emergences collapse chaos in the sense that they bring order to chaotic confusion. Cohen and Stewart also accept the importance of both intrinsic and extrinsic events. Like Stuart Kauffman, and Ivan Schmalhausen before him, they see intrinsic events as responses to a sort of internal emergence pressure, an inherent tendency of simple systems to complexify themselves, but beyond which there are processes “in which totally different rules converge to produce similar features, and so exhibit large-scale structural patterns.” These events depend on the extrinsic contingencies of other independent systems, occurring “where the interaction of several spaces of the possible leads to an explosion of the combined space and the emergence of features that can’t in any sense be traced back to the components.”44 Cohen’s and Stewart’s enthusiasm is encouraging, and their broad analysis is on the right track. But before these hunches are followed, more exhaustive legwork is required, and it will find no support for the mutual exclusivity of the intrinsic and extrinsic. The same is true of Kauffman’s quest, although he does not use the word “emergence” formally, associating it with the evolution of self-organization. Stuart Kauffman Since The Origins of Order: Self Organization and Selection in Evolution (1993) gets close to a theory of emergence, I am obliged to compare what Kauffman has said with my own opinion. Although he is not quite out from behind the Looking Glass, Kauffman ventures that “we must integrate the fact that selection is not the sole source of order

312

Chapter 8

in organisms.” Then he acknowledges that “biologists are secretly aware that selection must be working on systems which to one degree or another exhibit order by themselves. D’Arcy Thompson (1942) told us so with eloquence years ago, but we have not troubled to think through the implications. How strange, yet therefore how inviting, that we may one day bring ourselves to see life in a new light.”45 Coming closer to the heart of the matter, Kauffman adds: We shall in fact find critical limits to the power of selection. As the entities under selection become progressively more complex, selection becomes less able to avoid the typical features of those systems. Consequently, should such complex systems exhibit spontaneous order, that order can shine through not because of selection, but despite it. Some of the order in organisms may reflect not selection’s success, but its failure.46

Kauffman quests for rules of autonomous self-complexification that ought to shine with grail-light through the murk. But a rule that can be derived for one level of emergence may not offer predictions for emergence to the next level up. An explanation of the origin of life might be “surprisingly easy”—“an expected emergent collective property of a modestly complex mixture of catalytic polymers, such as proteins or catalytic RNA.”47 What is unsurprisingly easy about this assessment is that it depends on a priori knowledge of the existence of life and its qualities. Can such an emergence be post-predicted: inferred from a knowledge of the generative level of organic molecules? John Holland Holland deals largely with emergent order in very simple non-living systems whose structures and functions can be described with total mathematical rigor. But his models and metaphors touch on biological systems sufficiently to allow our ideas to interdigitate. Therefore I prefer to introduce some of his terminology rather than to paraphrase him in the vernacular.48 He begins with a set of elements and defined laws that govern their combination, taking the model of a board game involving pieces that respond to the stimulus of the play according to the constraints of space and permissible action. Thus, he has a set of generators functioning according to if [such be so] then [do this] clauses or stimulus-response actions—in other words, reactions to simple algorithms or guiding programs. The state of the system is equivalent to the disposition of the game pieces on the board at any one time, or the various conditions of neurons in a neural network. The “transition function” brings in the effect of an input or stimulus on the state of the system that results in a new state. As the system changes state, different strategies can come into play, if it is possible to respond in more than one way, and if the responses are recursive, i.e., can be evaluated before being put into action. The use of strategy delineates a “trajectory” or unique sequence of states as the game progresses. This is relevant to orthoevolution, regardless of its causal interpretation. The path of

The Re-invention of Emergence

313

the trajectory is also contingent on the variety of stimuli presented—the opponent’s moves, in the example of a board game. These are not very predictable if there is a wide choice of options. General concepts that appear as Holland builds his arguments include building blocks, or repeatable features that appear in systems with emergent properties. They are equivalent to holons or modules. They and the systems that they combine to form can be modeled by particularizing their salient features and the rules that they obey. Agents are holons that interact in the system, like billiard balls on a table at the physical level. Cells in a multicellular organism, or organisms in a social system are agents. The final neologism necessary for interpreting Holland is the “constrained generating procedure”—a system that combines the elements, rules and interactions that are capable of producing emergent novelty.49 Building upon these generalizations, Holland produces the following aphorisms. After each, I paraphrase his explanation and examples, together with my own comments and examples, in brackets. 1. Emergence occurs in systems that are generated. Emergences arise in systems composed of a finite array of multiple similar holons whose interaction changes their state. In the model system of a board game like chess the pieces have been manufactured as several sets of identical chessmen. [Similarly, in a cell there are various kinds of proteins, often existing as multiple copies. What is important here is change in state over time due to the operation of transition functions.] 2. The whole is more than the sum of the parts in these generated systems. [I have illustrated how this can be so by the simple example of symbionts that can turn detrimental properties into usable ones, or negative into positive.] Holland says that the overall behavior of the generated system cannot be predicted by knowing how the parts can behave: “The definition of the generated system, though it determines all the rest, is no more than a simply described starting point; subsequent activities can be determined only by extended examination and experiment. In this sense, more comes out than was put in.”50 Thus, a game like chess, played and studied for many centuries, continues to produce surprises. 3. Emergent phenomena in generated systems are, typically, persistent patterns with changing components. A physical illustration is the standing wave that persists in front of a rock in a swift stream, although the water molecules are constantly being replaced by new ones all

314

Chapter 8

the time. [Similarly, methylation and binding patterns that are initially induced by the environment can persist from one generation to the next. The most general biological example of a persistent pattern is the organism, whose molecular makeup turns over continuously. The pattern of the organism persists to a large extent in its offspring.] 4. The context in which a persistent emergent pattern is embedded determines its function. [This relates to organismal multifunctionality as well as to its environment. The persistent pattern of gill arch bones, for example, is used in different ways in a chronological evolutionary sequence: gill function; jaw function; ear function. This is a very productive model that I have touched upon in the physiological arena and will come back to in the next chapter.] 5. Interactions between persistent patterns add constraints and checks that provide increasing “competence” as the number of such patterns increases. [The constraints and checks in developmental canalization, for example, do make the developing embryo more “competent” at maturing true to type. This is an obstacle to evolutionary emergence.] But there is a saving qualification: “the possible sophistication of response rises extremely rapidly (factorially) with the number of interactants.”51 6. Persistent patterns often satisfy macrolaws. For example, predicting how an animal with a stereotyped behavior pattern will respond to a stimulus does not require “recourse to microlaws” such as the physics and chemistry of vision and chemoreception. [Nor are the persistent patterns of homology determined by genetic microlaws.] 7. Differential persistence is a typical consequence of the laws that generate emergent phenomena. [Amen to “consequences” coming after generative causes. Such persistence could be due to adaptability, and it is not smuggled in from high order biological systems, but is found in a simple computerized checkers-playing program that can alter its strategy by re-evaluation in terms of previously successful actions.] Holland also writes: “Generalists are much better tested for persistence than specialists, because the generalists receive many trials in a wide range of situations, in the same time that the specialists receive just a few trials. As a result, the generalists supply a firm niche for

The Re-invention of Emergence

315

the specialist that can interact with them.”52 [He overlooks the fact that generalists possess the qualities for persistence before they are tested. However, his point does bring out the significance of emergences involving symbiosis and ecological associations as well as the self-sufficient emergent property of adaptability.] 8. Higher-level generating procedures can result from enhanced persistence. Here Holland specifies symbiosis as a generator of enhanced persistence, commenting that it “yields patterns extremely unlikely on an a priori inspection of the original generator[s].” [This statement can also be usefully reversed to infer that adaptability is a generating condition for higher-level emergences. This applies to major adaptational changes as well. The persistent striving of the hippopotamoid ancestor of whales to swim out of its depth preceded the genetic assimilation of more whale-like anatomical phenotypic changes. But this all depended on its adaptable placental physiology.] My major criticism of Holland’s reductive rules is that some of them intuitively smuggle contraband from higher-order systems. This dilutes the mathematical rigor that he demands of emergentism and tries to provide for his simplest models. In taking the high road I have not bound myself by such constraints. I can smuggle and dilute—i.e. write about downward causation and interaction—to my heart’s content. Life is not a board game with finite rules that are changed by mutual agreement of the players. Instead, stable dynamic systems are disequilibrated into partial chaos by entities that break the rules. Physicochemical and biological contingencies inexorably impose new conditions. But it is more important that as emergent evolution progresses, it allows greater freedom of action for the organism, which has a feedback effect on its physiological makeup and morphogenesis. Holland does not exclude extrinsic effects, but underestimates them. It is alluring to draw parallels between the emergence of ideas from game playing, or any cognitive emergent process, and the emergence of biological innovations. But rigor is lost in the gaps between metaphor and actuality. Nevertheless, if Holland’s low road does not quite arrive at evolutionary emergence, it reaches, as he says, a “way station” for the expedition, where I can compare notes with him. Such an exercise shows us how the generative conditions for emergence include systems with multiple components that can be differentiated, and thereby become more adaptable. Emergence produces new macrolaws—they reflect the degree to which the whole is greater than the sum of its parts. And the further emergent evolution progresses the more it is likely to occur. These inferences have been frequently drawn by emergentists, and others I have conscripted. And when “show me” comes to “prove it,” Holland has the edge.

316

Chapter 8

William Wimsatt The philosopher William Wimsatt, who has been writing about the meaning of complexity and organization for a quarter-century, has recently paid more attention to emergence, by using systems reduction (or “simple systems” analysis). The success of this approach is illustrated by his 1997 essay “Aggregativity: Reductive heuristics for finding emergence.” Intuitively we might think that the study of emergence should take us directly to a case-by-case examination of those rare wholes that are obviously greater than the sums of their parts. But Wimsatt instead turns the problem on its head—an heuristic method that we often forget to apply. And according to Wimsatt’s rules, the upside-down mode is in this case more parsimonious than the intuitive one. Instead of seeking wholes that are greater than the sum of their parts he seeks wholes that are “mere aggregates.” When these are subtracted, the remainder are assumed to have emergent properties. The properties of a mere aggregate are independent of variations in its components, since the latter do not interact. The nature of the whole can be discovered by testing the effects of intersubstitution or rearrangement of parts, the effects of addition or subtraction of parts, and the effects of decomposition or reaggregation of parts. If none of these have any effect on the properties of the whole, it is an aggregate. If the whole has non-linear emergent properties arising from interactions between the components, some or all of the three tests will cause identifiable change in the original system.53 (I have adjusted his categories slightly to isolate the test conditions from the processes of interaction.) This approach demonstrates that mere aggregates are rare, even outside biological systems. Artifacts don’t count since human emergent properties are involved in their construction. Accordingly, wholes with emergent properties that don’t meet the test for simple aggregates, must be universal. Furthermore, Wimsatt’s three test categories suggest direct mental and practical experimental approaches to emergence. Yet, he cautions, “we tend to start with simple models of complex systems—models according to which the parts are more homogeneous, have simpler interactions, and in which many differentiated parts and relations are ignored (Wimsatt, 1980)—models which are more aggregative. But then as our models grow in realism, we should both capture more properties, and see more of them as organization dependent—or emergent.”54 Jeffrey Schwartz Schwartz enters the fray with Sudden Origins: Fossils, Genes, and the Emergence of Species (1999). It is inadequate to label him an anthropologist, since his interests in biology and paleontology are profound. The aforementioned title is one that leaps off the bookstore shelf at someone such as myself, though I do not ascribe to the idea that the emergence of species is fundamentally important (q.v. chapter 11). Furthermore, although he writes about emergences as sudden events, he uses the word “emergence” casually rather than formally. Nevertheless, Schwartz recognizes that molecular

The Re-invention of Emergence

317

biology has provided us with the information that could rationalize saltatory emergences, or hopeful monsters, especially the discoveries of regulatory Hox genes that I have outlined in chapter 6. He also appreciates that regression or deletion of gene expression has been important in the diversification of clades. I would only complement that notion with the proposal that orthogenesis/allometry, has been responsible for the exaggeration of what remains after deletion. Peter Corning Peter Corning has also addressed emergentism as a subset of a larger “Synergism Hypothesis” (1998). This is a valiant attempt to encourage the expansion of interdisciplinary thinking through the recognition that most of life and much of non-life is founded upon synergistic interactions. It is interesting that a complexity theorist can remark that “in contrast with the bloodless mathematical caricatures that are blind to the functional properties of the phenomenal world, the synergy paradigm draws our attention to the function aspect of cooperative effects.”55 Corning has also escaped the domination of natural selection found behind the Looking Glass. He clearly states that natural selection is the effect of emergent change: . . . causation in evolution runs backwards from our conventional view of this in evolution, functional effects are causes. To use Ernst Mayr’s (1961) well-known distinction, it is the “proximate” functional effects which result from any change in the organismal-environment relationship that are the causes of the “ultimate” (transgenerational) selective changes in the genotype, and the gene pool of a species.56

The larger category of synergism makes sense in the general context of evolution. Although emergentism is a subset, and an important one, emergent change is followed by a new dynamic stability. Both are synergistic, and both are part of the persistence and progress of life. Corning is especially taken with symbioses and social interactions that have unique properties arising from their emergence. Nonetheless, he extends synergism to include non-living complexes as well. When we consider pioneering thoughts about the nature of whole, we have to remember that Aristotle observed in Metaphysics that the loss of a single part may destroy the whole. Corning calls this the “synergy minus-one methodology” for testing synergy. He notes that in many biological examples the loss of a DNA base, or a protein amino acid, or a transfer RNA, or an endosymbiont, will bring about the collapse of the whole. This echoes Wimsatt’s test for an aggregate—take away a bit and nothing happens. So a population is an aggregate? A school of herring is behaviorally synergistic. Nevertheless, as Corning says, if a predator eats a herring from a feed ball, the whole is not destroyed, only diminished. But gobble a lot, and the population may be reduced to less than an effective breeding group. And even a breeding group resists a clear-cut distinction between aggregate and whole. It only takes two to tango,

318

Chapter 8

although two, by themselves, are deficient in biodiversity, and have a high probability of extinction. This train of thought brings us to the perennial debate over V. C. Wynne-Edwards’s 1962 “Group Selection Hypothesis.” Corning proposes that the rehabilitation of the group as a selective unit is justified on the grounds that there are mutually beneficial synergistic interactions between members of a group that are not close kin.57 Although I will delay dipping a toe into these muddy waters until my final chapter, I should register my agreement, and add that Corning also suggests comparative biological studies that would demonstrate that large groups operate more efficiently than small groups. Corning develops these ideas further in Nature’s Magic: Synergy in Evolution and the Fate of Humankind (2003). He continues to regard emergence as an important subset of synergism, but dislikes the term because of its plethora of synonyms. Although he gives natural selection more prominence in evolution than I do, there is considerable overlap in what we both find interesting in biological emergences. He emphasizes the proposition that behavior is a primary, though proximate cause of evolution, but takes no account of the physiological foundation of behavior. Harold Morowitz Just when I thought I had completed this chapter, I encountered Harold Morowitz’s The Emergence of Everything: How the World Became Complex (2002). Although he does indeed address all of the important emergences in the broadest sense, from the Big Bang to mind, he is not guilty of constructing a theory-of-everything-that-turns-outto-be-a-theory-of-nothing. Indeed, he does not construct a theory of anything; but manages to reinforce the concept of progressive complexification of the Universe through the emergence of the elements and their compounds, and novel chemical properties in general; the emergence of geospheres; life; membranes; metabolism; eukaryotic cells; multicellularity; nervous systems and intelligence. However, Morowitz runs into the same problems as Robert Chambers, Henri Bergson, C. L. Morgan, Samuel Alexander, and Jan Smuts. Since God is in their metaphysical machines, He, She, or It must have been there from the start. Chambers (1844) had God create a Babbagian computer program that produced apparent novelties that were really part of the original divine algorithm. Bergson (1908) began with a “primordial consciousness” that leapt in “vital sparks” to greater complexities. Alexander (1920) made God both the creator and the fabric of the cosmos, but regarded Deity as emergent from life. For a more comprehensive treatment of these implications, see chapters 5 and 6 of my 1985 book Evolutionary Theory: The Unfinished Synthesis. Morowitz, though very well qualified to elaborate mechanistically on the ascending hierarchy of emergent levels in physics, chemistry and biology, has created the same difficulties as the transcendentalists. As a result he is forced into a contradiction in terms by making the Universe “unfold,” despite his opinion that emergences are

The Re-invention of Emergence

319

unpredictable events with novel properties. And eventually he retrenches into a hylozoic/panpsychic position that mind is not, after all, a sudden emergence, but a property of cosmic maturation. “Mind emerges over a long time, not just over the last 500 million years. It is more deeply embedded in an evolving universe, and may have prebiotic roots.”58 Overview of Emergence Types of Emergence Developmental changes dominate the scheme of emergent evolution that Müller and Wagner present. But before epigenetics or morphogenesis became significant, life emerged, biochemical networks complexified, prokaryotes symbiosed, sexual reproduction originated, and simple multicellularity arose. Physiogenesis and the physiological and behavioral responses of organisms that accompanied developmental evolution are insufficiently explained by epigenetics. And societal relationships are largely independent of them. Nevertheless, Müller and Wagner provide principles that apply beyond developmental evolution. Therefore I incorporate them into a system based on my own interpretations of what has been suggested in the past. Although it is not the universal cause of emergences, reproduction is what makes them persist. Reproduction also contains processes that provide for evolvability. The kind of molecular reproductive apparatus needed to faithfully pass on some of the survivability of the parent to the offspring has to be able to make a copy of itself. Along with the flexibility to come apart temporarily for copying, comes the opportunity for the production of copies that are not exactly like the original, hence differentiation. Thus, the unavoidable consequences of reproduction are duplication and variation of the parental units. These natural experiments are complemented by a basic emergent feature of sexual reproduction—every generation is required to begin from scratch at the single cell stage. This makes it possible to progressively reorganize dynamic structures during the entire course of an organism’s individual development. In contrast, if it reproduces asexually, the existing components cannot be radically reorganized. However, adding on new features to actively growing or established organisms is not out of the question— especially in plants. Growing roots are epigenetically altered to make nodules that accommodate symbiotic nitrogen-fixing bacteria. Growing shoots in contact with damp soil put out roots instead of leaves. And somatic mutation can change the structure of new shoots and can be passed on to the next generation. Are such changes to be characterized as continuous or discontinuous? Saltatory Emergences In the sense that they are discontinuous, changes within organisms are almost all saltatory. A point mutation may produce a protein with a new function without any

320

Chapter 8

half-measures. However, I reserve the epithet for emergences that appear suddenly from radical complexifications and re-organizations. Although the emergence of life and mind are likely in this category, it is more tangibly exemplified by the endosymbiotic association of prokaryotes to produce the first eukaryotes. It has all of the typical connotations of emergence: unpredictability, greater adaptability, multifunctionality— a whole that is greater than the sum of its parts. This implies an adequate integrity at the point of emergence, causally prior to the involvement of the syndrome of natural selection in adjusting the self-organization and coordination of the new whole. Contingent associations of organisms, or molecular modules that had evolved independently, are clearly important constituents of saltatory evolution. So too are the emergences of hopeful monsters due to epigenetic changes. Backing the success of hopeful monsters is physiological, and hence behavioral adaptability. Environmental factors may catalyze saltatory emergences, for example, the availability of significant others for association, or physicochemical stimuli that may trigger chain reactions. Saltatory emergences are the most radical evolutionary changes. Critical-Point Emergences These may arise from thresholds in otherwise continuous epigenetic processes. Since critical-point emergences lend themselves to selectionistic interpretations I allow them a longer discussion than saltatory emergences. They may involve allometric or orthogenetic continuities of change in functional anatomy, as exemplified by the acquisition of true winged flight. The story of wings in the higher vertebrates brings in several interesting corollary aspects of emergent evolution. Knowing how the primitive vertebrate forelimb began as a swimming organ, and then became a creeping or walking organ, it is difficult to conceive of a hopeful flying monster appearing with fully developed wings, due to an all-or-nothing epigenetic event. Nevertheless, the flying reptile Coelurosauravus jaekeli supported its wings with rod-like bones that were not derived from pre-existing structures like forelimbs or ribs. Instead they arose de novo from dermal mineralizations, like battens in a sail, possibly because of physical lines of stress along which chondroblasts migrated and differentiated.59 As was mentioned in chapter 5, the universal pteroid bone that supports the proximal leading edge of pterosaur wings probably arose in the same way. The repression of Hox gene expression has been attributed to the simplification of the proto-dinosaur forelimb that led in one direction to bird wings—and could have been an all-or-none event.60 An allometrically growing, incipient, feathered wing would soon have gained an intermediate plateau of adaptability. It had multifunctional potential as a grasping hand, a balancing organ associated with rapid bipedal running or arborealism, and as a net for pulling insects out of the air.61 Even when the wing has been fully “adapted for flight” it can do other things. A bat can use its wing like a catcher’s mitt if the insect is not immediately seized by the mouth. And what

The Re-invention of Emergence

321

better way to go after flying prey than to take to the air? The proto-bat could figure that out for itself without the imaginary urging of selection pressure. Like the pterosaurs of the Jurassic, the New Zealand short-tailed bat Mystacina tuberculata continues to use its wings for terrestrial locomotion, having in this case gone to ground to forage for insects in leaf litter. I once saw a crow brought down in a pond by vengeful young ducks who had suffered the attacks of crows since fledglings. It painfully but effectively sculled its way to shore with its wings. As Darwin observed, the water ouzel, or dipper, can effectively “fly” underwater to forage. And the penguin’s wing was exapted as a flipper. Oddly, some flightless birds retain vestigial wings that have no apparent function. One might think that selection pressure would have turned them into something useful! Gliding and soaring could have arisen as the first critical-point emergence made possible by continued expansion of the wing area. Once gliding became a habit, lift would exceed the pull of gravity at the next critical point in allometric modification, through a combination of the expansion of wing area and flappability. Beyond that critical point true flight was an instantaneous reality. The new habit freed insects from predators and gave easy access to tall plants and their energy-rich reproductive organs. This aerial “plankton” was an extra attraction for birds and bats. These examples show that the generative conditions for emergences are combinations of epigenetic molecular changes, developmental shifts, and alterations of habitat, physiology and behavior. As organisms evolve progressively they explore new environments where restraints are different, behavioral choices wider, resources greater and competition in abeyance. From the association of organisms there arise social organizations that are greater than the sum of their individual members. Their emergent properties feed back to affect the individuals that constituted their generative level. Critical-point emergences in evolving organisms were conceptualized first by C. L. Morgan, but complexity emergentists frequently cite examples in non-living systems. Nicolis and Prigogine (1977) found that Bénard cells—columns of rising warm water surrounded by sheaths of descending cold water—form at a critical temperature point in the heating gradient. Per Bak’s theory of self-organized criticality arises from the same kind of inanimate phenomena.62 But applying it to evolution in How Nature Works (1996), he assumes that the build up of self-organization is due to the cumulative selection of adaptations. Moreover, at the critical point the build up is disintegrative, and his biological examples deal with extinction rather than progressive self-organizing evolution. He also concludes that exogenous catastrophic events like impacting asteroids are irrelevant to major extinctions—they would happen anyway as the result of minor disequilibrating events when critical points had been reached as a result of continuous selection! We think of continental drift as a gradual process on a geological time scale. But it too was probably punctuated by critical points where

322

Chapter 8

traditional movements of animals, and the dissemination of plant spores and seeds, stopped or began relatively suddenly. Indeed the impact of such geomorphological effects has been held responsible for major diversifications in marine invertebrates, birds, and mammals.63 The Causes of Emergence The causes of emergence are the causes of evolution. My use of the term “natural experiment” is intended to move the focus of attention toward the origins of biological novelty, and away from natural selection or “sorting.” It embraces all the causes of emergence as well as the origins of simple, adaptational modifications. If the question of how emergence occurs does not immediately elicit blinding insight, it has to be approached the hard way with analysis and categorization of evolutionary changes. This is the epistemological route chosen by Aristotle when he constructed the scala naturae, and developed by Francis Bacon in Novum Organum (1620) on the grounds that tabulation organizes phenomena, and allows the detection of hidden relationships. The most important natural experiments can be separated into three arenas of operation: the associative, involving intimate symbioses and social interactions, the physiological and behavioral, and the epigenetic, or developmental, which includes anatomy. All of these causal matrices are mutually influential. Moreover, they operate in, and are causally connected with the environments of cells, the organism’s internal milieu, and the world at large. Environmental influence is accommodated under extrinsic emergences. Intrinsic and Extrinsic Emergences Saltatory and critical-point emergences can be re-sorted according to whether they arise autonomously within a single organism (albeit through changes in part passed on from the previous generation), or through the interaction of the organism with external physicochemical, organismal, or ecological systems. C. L. Morgan used “immanent” and “transeunt,” but “intrinsic/endogenous/autonomous” and “extrinsic/exogenous” are less portentous to contemporary ears. Saltatory emergences come from both intrinsic and extrinsic generative conditions. Critical-point emergences tend to arise internally but they progress in relation to behavior and the conditions of the external environment. These two basic categories are therefore never mutually exclusive. Intrinsic Factors Some mechanisms may act spontaneously within the organism to produce emergent novelties affecting symbiosis, epigenesis and physiology. These include the following:

The Re-invention of Emergence

323

point mutations of structural DNA exon shuffling and consequent alterations in primary protein structure contingent recombination of structural genes and, consequently, protein domains mutations of modifier or regulator genes modifications of transcription and translation mechanisms transpositions, and resultant alterations in gene expression changes in methylation and histone-binding patterns, duplications of codons, exons, genes, chromosomes, genomes repetitive differentiation alterations of repair mechanisms self-amplifying holons that produce allometry/orthogenesis structural reorganization of the cytoplasm of gametes, especially ova. Although these are intrinsic occurrences, an extrinsic mutagen or trigger is sometimes involved. Once they occur, they will persist if they are adequately integrated into the organism. But even enhancement of overall adaptability may be of no particular use during the lifetime of the novel emergent. Either way, immediate improvement in fitness compared to the parental type is not assured; life might be precarious until the new system has been fine-tuned, and here natural selection qua dynamic stabilization has a role to play. Intrinsic experiments in emergent evolution that occur without reference to any extrinsic factors are the focus of complexity or antichaos theory. Stuart Kauffman asserts that most organized complexity arises from “tendencies for innate order to emerge in complex interactive networks.”64 Anticipated by Lamarck, Mivart, and Bergson, this also follows logically from an understanding of the nature of the molecular reproductive apparatus, even for biologists who are not up on history or chaos theory. If formal laws of innate complexification were to be framed, gradual trend to critical-point emergence would be a strong component, but some major saltatory emergences catalyzed by contingent effects from outside the organism would

324

Chapter 8

be excluded. A prokaryote cell could not have become a eukaryote unless there were other independently evolved, potential partners “out there.” Extrinsic Factors Environmental influences fall into two groups. The first have a broad impact, such as physiogeneses that alter the internal milieu, or affect functional anatomy. The second category consists of specific external environmental stimuli that trigger change in DNA expression, which may affect epigenesis. Development and physiology can be affected by the environment above the DNA level, and such changes might be converted to internal adaptations under genetic control. These are not just physical factors like light, temperature, and humidity but also more complex biological factors. Availability of food and consequent egg size affects development. Overcrowding makes salamanders into monstrous cannibals, and causes locusts to change color, enlarge their wings, and take to flying in swarms.64 Interspecific factors include freeliving bacteria that affect thallus form in seaweeds, nodule formation in legume roots, and gut development in mammals. Genetic assimilation can fix environmentally imposed change, in both developing and mature organisms. As Schmalhausen and Waddington realized, this leads to internalization of environmental stimuli of development, and ultimately tighter canalization. It presupposes an alchemical cauldron in the basement, seething with molecular experiments. Its products might be shelved, or recycled, or might just be immediately usable as a generative condition for a higher level in the organismal and environmental hierarchy. One consequence of the interaction of intrinsic and extrinsic mechanisms is increased adaptability and hence greater organismic integrity, but fixed adaptations or specializations can also be produced. In animals, the freedom conferred by adaptability permits behavioral experiments, and wider explorations of more extreme environments. Thus, there is a strong positive feedback loop with the environment that may bring about even more internal change. Adaptability in plants is partly physiological, but their strongest feature is ontogenic plasticity. A terrestrialized reproduction is important, and is linked with the exploratory behavior of pollinating animals. Prokaryotes have the adaptability to acquire additional genes by various routes, but endosymbiotic gene acquisition by eukaryotes was one of the most important evolutionary improvements in the adaptability of individual organisms. In symbioses the qualities of the participating partners complement each other, and in social interactions within groups of higher animals there is a parallel complementarity of both genetically and traditionally inherited talents. In the face of environmental changes the syndrome of natural selection may tolerate adaptable generalists over many generations. David Rollo suggests that “strategies that ensure being approximately right most of the time are favored over

The Re-invention of Emergence

325

those that may be locally superior, but that also run the risk of being precisely wrong.”65 Nevertheless, under stable conditions, specialists may continue to be precisely right for very long periods. Adaptable organisms must then hang on at environmental fringes and interfaces that fluctuate both rhythmically and irregularly, and where there is little competition. But to thrive they must have an emergent quality of unusual utility, or risk new environments, or find old environments that have been cleared of competition. Thus, the sense of physical movement from one environment to another is central to the concept of emergence. The separation of emergent phenomena into the intrinsic and the extrinsic is a simplifying exercise, but forgetting to reunite them in the final analysis is counterproductive to the quest. How well have complexity theorists succeeded in integrating intrinsic and extrinsic events that effect emergences? Kauffman’s generalization about evolution through cell differentiation is succinct but exclusive of environmental causes: . . . the existence of distinct cell types, the homeostatic stability of cell types, the number of cell types in an organism, the similarity in gene expression patterns in different cell types, the fact that development from the fertilized egg is organized around branching pathways of cell differentiation, and many other aspects of differentiation are all consequences of properties of self-organization so profoundly immanent in complex regulatory networks that selection cannot avoid that order. All aspects of differentiation appear to be properties of complex parallel-processing systems lying in the ordered regime. These properties may therefore reflect quasi-universal features of organisms due not to selection alone, but also to the spontaneous order of the systems on which selection has been privileged to act.66

This is an illustration of how constant genuflection to natural selection, even if ritualized to a momentary dip, gets in the way of a good ground-breaking stride. Holland also largely omits extrinsic factors. Yet, autonomous (intrinsic) self-ordering is a multiplex process that is difficult to isolate from the whole causal matrix. Matsuda’s work, and more recent books by Rollo, Jablonka and Lamb, Hall, Pearson, and Müller, and West-Eberhard, amply illustrate how the emergence of new orderings in the epigenetic arena are so often dependent upon external events. The simple availability of food, which might provide for larger eggs, is enough to trigger epigenetic novelties. The lack of an element in the food, for example iodine needed for the synthesis of thyroxine, may be enough to induce neoteny in an amphibian. The presence of a hormone in a juvenile termite’s food at a particular time of development determines its caste. In some cases environmental effects impose epigenetic regressions that are then followed by natural experiments with novel autonomous products. In the present book, I have gone beyond Schmalhausen and Waddington to show that the evolutionary path of animal homeostasis at the organism level—a paradigmatic self-organized complexity—has been dictated in part by physiogenic changes induced by the environment. Some emergences are contingent upon the co-existence

326

Chapter 8

of other independently evolved systems—proto-mitochondria and proto-chloroplasts had to evolve independently and then come together in the same place and at the same time to give rise to endosymbiotic eukaryotes. Some are profoundly influenced by catastrophic events. Natural selection is not involved in any of these except as a post hoc stabilizing influence. Some physical emergences into different environments, as well as evolutionary emergences, may be simply intolerable to the continued integrity of the emergent. They may also be rebuffed by the agents of natural selection, so that if the novel organisms are not strongly benefited by an immediately beneficial feature of the emergence, they will be unsuccessful in the face of competition. On the other hand, their progressive adaptability can operate in fluctuating environments, or in new environments where the established order of competition and predation do not exist, or where there has been a catastrophic environmental change that leaves the field open for new players. Catastrophe and emergence go hand in hand, and David Raup has gone so far as to suggest that progressive evolution would have been impossible on a planet that had not been subjected to major bolide impacts, however devastating they were in the short term.67 Emergences in Relation to Progressive and Adaptational Evolution In categorizing emergent evolution into saltatory and critical-point, and intrinsic/extrinsic events, it must not be forgotten that evolution is both progressive and adaptational. Progressive evolution involves the emergence of new levels of complexity/self-organization/adaptability. It is followed by a phase of diversification, involving orthogenesis/allometry, and specialization of habit in relation to habitat. I only need to confirm that all categories of emergence are involved in progressive and adaptational evolution, with a larger component of intrinsic/saltatory emergence in progress, and a larger component of extrinsic/critical-point emergence in adaptation. Stases As I use it here, the word “stasis” does not mean inertia, low entropy, or the arrest of time. In each causal arena where emergences persist, new dynamic equilibria develop around them. If variant properties that are adaptational to the newly emergent condition are selected, it is on the basis of integrative harmony and energetic efficiency—what Brian Goodwin simply calls “quality.” Natural selection, along with its agents, is the hypostasis of a stronger dynamic stability. By analogy, consider Genghis Khan’s conquest of China. The existing bureaucracy quickly re-arranged itself out of disarray to support the new dynasty, and thereby helped to ensure its sway for several centuries. Formerly resistant to change, the mandarins served to consolidate the revolution and to protect the new order from further encroachment. As time passes, stases become stronger and more resistant to the further modification. The

The Re-invention of Emergence

327

most familiar are homeorhesis in the epigenetic arena, homeostasis in the physiological arena, and ecostasis in the environment at large. Symbiostasis is the parallel stabilization of associative relationships and it demonstrates that evolution can progress despite stasis. Stasis is actually in a constant state of flux—the Red Queen is always running, and minor natural experiments lead to minor adaptations, without achieving any progress, although the Ultras call it “evolution.” The worldviews of uniformitarianism and gradualism were persuasive because stasis is the prevailing condition of life. Subjectively, as an organism, I would prefer to continue in stasis, rather than to “live in interesting times”—as the Chinese curse has it—when I might not survive long enough to enjoy the stay. For traditional evolutionists, what happens now is the model for evolution as a whole, and since stasis is the prevailing condition they must somehow explain it as evolutionary. Anything tainted with sudden change, catastrophe or saltation, is anathema. Those who have canonized natural selection might see its re-invention as hypostasis to be an act of destruction. However, they should be grateful to the originators of punctuated equilibrium for tacitly ceding to natural selection a significant role in punctuation as well as the major role in equilibrium phases. Moreover, there is a significant role for its hypostasis—generative entrenchment of dynamic structure provides a platform that can be built upon, although its destabilization has often been part of progressive emergence. Evolution would never have occurred if the motor of adaptation were the driving force, for its square wheels stop it as soon as it turns over. But the investigation of stasis is not superfluous to evolutionism. Knowing what blocks evolution helps us to understand what causes it, and it is useful to know that a period of relative stasis gives a better guarantee that new emergents will be internally well adjusted. Here are the tableaux vivants of stasis found in the rings of the evolutionary circus: 1. Symbiostasis. It may be extremely robust, or very vulnerable to change. Once entrenched, eukaryotic symbioses were never abandoned, remaining intact in the evolutionary lineages that possessed them and built upon them. Looser associations are more fragile, and may easily be destroyed, because harm to one is harm to both. Social emergences with innate behavior patterns tend to be inflexible but their members can be sacrificed without serious detriment to the whole society. At the human level a reservoir of adaptability resides in an array of independent individuals, which offsets the inflexibility of traditional behavior patterns. 2. Homeorhesis. The build up of developmental stasis results in tight canalization, with epigenetic locks that prevent disequilibration and keep the organism true to type. The archetypes of animal and plant morphologies may all have arisen at a time in biological history when canalization was loose, and natural selection in abeyance.

328

Chapter 8

Once established, homeorhesis could not easily be overridden. There remained potential mechanisms of epigenetic evolution, such as paedomorphic regrouping, hypermorphic additions, heterochronic alterations and radical deviations. Gene drive could overcome homeorhesis sufficiently to cause allometric shifts. 3. Homeostasis. Disequilibration of the internal milieu by physiogenesis overcame the weak resistance of homeostasis in early aquatic animals. Changes of this kind, though random, demonstrate the positive feedback loop between physiological adaptability and behavioral versatility. Commitment to terrestrial living locked in the chemical composition of the internal milieu as it was when the primitive tetrapods emerged onto land. It also potentiated more sophisticated self-organization, because the greater availability of oxygen levels could sustain the high metabolic rate ultimately provided for by homeothermy. Progressive improvement of homeostatic organization helped emergent organisms to diversify through a wider environmental range, since they had become more and more independent of the external milieu. Natural experimentation with behavior has been subject to stasis, through genetic fixing of patterns and loss of flexibility. In mammalian lineages freedom of choice has expanded, although traditive behavioral patterns in humans can be as rigid as genetic fixation. 4. Ecostasis. In the world at large, the resistance of natural selection to change was often overcome by ecological destabilization, which removed competitors and predators of emerging evolutionary novelties. Some hopeful monsters that emerged from natural experiments were only able to diversify after catastrophic, mass extinctions. Which of those succeeded might have been simply a matter of luck, of emerging in the right place at the right time. But the more adaptable the organism is, the more likely it is to arise like a phoenix from the ashes. Although this chapter illustrates how emergent phenomena have been given serious consideration by materialistic biologists, it also illustrates how adherence to selectionism limits emergentism’s analytical and creative scope. Most progress has been made by those who merely pay lip service to the role of natural selection, who ignore it entirely, or identify themselves as Contras. In the following chapter I will round up the causal particulars of emergence to establish additional generalities, and discover how feasible an emergence synthesis might be.

9 From the Particular to the General

We can only augur well for the sciences, when the ascent shall proceed by successive steps, without interruption or breach, from particulars to the lesser axioms, thence to the intermediate, and lastly, to the most general. —Francis Bacon, 16201

Francis Bacon was more a gradualist than an emergentist—no revolutionary interruptions or breaches for that staunch supporter of a stable monarchy. Yet the many relevant evolutionary particulars that have accumulated over the last century need some kind of ordering. And, as Arthur Lovejoy demonstrated, a Baconian epistemology can be just as easily applied to discontinuous evolutionary emergences as to continuous adaptational variation. “Lesser axioms” pertaining to the various “common and exceptional causes,” found in the causal rings of the emergence circus, have already been sorted out. Now we can move on to intermediate inductions. Identifying the necessary and sufficient generative conditions leads to the final goal— the most general inferences that apply to everything that is going on in the biosphere. The Symbiosis/Association Arena Associative emergences were the first as well as the latest stages of evolution— chemical associations gave rise to primitive cells, and humans associate in families and societies and global networks. They can be divided into four major subsets: intraorganismal (some intracellular) mutualistic symbioses, sexual associations, differentiated multicellularities, families/societies. There are also many loose commensal and cleansing relationships, and a variety of co-evolutionary interactions. Endosymbioses present us with the clearest illustration of saltatory emergence. Although the hosts and pro-symbionts might have been in each other’s physical presence for a long time, they came into a state of symbiosis suddenly, furnishing mutually beneficial qualities that the formerly independent organisms lacked. An important emergent feature of foundational symbioses is the acquisition of heritable characteristics. These make the new

330

Chapter 9

whole greater than the sum of the parts, and provide for transmission of that wholeness directly to the next generation. When the first eukaryotic cellular communes came into being, some features that were negative in the earlier state of independence became positive in the context of the new whole. This emergent property made up for the loss of independence, although, now, if something detrimental happened to one of the partners, the other would suffer as well, a condition that the multicellular condition was eventually to mitigate. Eukaryotes also lost the genetic flexibility that the constituent prokaryotes had, and still retain, in the form of transformation and conjugation. However, at the new level, transposon flexibility still persisted, as did viral and bacterial transduction. The latter provides a natural parallel for the genetic engineering process of transgenesis although the magnitude of its contribution of foreign genes is not well known. Genetic flexibility was restored in part by the emergent properties of sexual association. The natural experiment of meiosis not only provided for increased shuffling of the genetic deck, it also allowed the organism to identify and repair point mutations more thoroughly. Emergences involving multicellular associations, and their differentiation, complexification and organization, belong as much to the developmental and physiological arenas. That leaves the emergence and evolution of societies as the final performances in the association arena. These have had a particularly strong feedback effect on the evolution of the endocrine and nervous systems and on behavior. Here are the aphorisms of association that will contribute to a theory of association, and complement the developmental and physiological principles: 1. Associative emergences constitute saltations whose whole is greater than the sum of the parts. 2. They are physiologically more adaptable than their independent antecedents since they allow greater differentiation and coordination of functions. (Holland has cited symbioses as unpredictable generators of enhanced persistence.) 3. Photosynthesis of prokaryotes, enhanced by endosymbioses, effected a major geophysiological change: oxygenation of the biosphere. 4. Sexual associations provide for more genetic mixing and matching, and hence more evolvability, without totally sacrificing the unique qualities of the parents. 5. Sexual reproduction (as opposed to vegetative reproduction) guarantees that the entire developmental process from single cell to mature multicell is repeated in every individual, exposing every stage to epigenetic change and subsequent evolution.

From the Particular to the General

331

6. Although there are disadvantages that increase the vulnerability of partnerships, some associations like eukaryotic endosymbioses have been so vital, and ultimately so supportive of organismic integrity, that they have remained the foundation of most living organisms and ecosystems. 7. Symbioses sometimes have mutualistic, interspecific epigenetic effects between symbiont and host. 8. The emergence of societies whether by insects or vertebrates has tended at first to fall into genetically assimilated, “instinctive” dynamic stabilities. But such fixed action responses can be modulated and escaped. Nevertheless, the most flexible kinds of societies are those that have arisen in placental mammals that developed sophisticated nervous systems before they developed social behaviors. Thus, organisms that have brains and plastic behavior are always adaptable, although they sometimes run the risk of becoming creatures of habit. 9. The adaptability of human social associations generates an emergent non-genetic freedom of choice and intelligence that cannot be fixed by natural selection, although human societies often replace it with other rules and regulations. 10. Symbioses have founded and nurtured major aquatic and terrestrial ecosystems. Major emergences in themselves, they brought into being the “entangled bank[s],” the Darwinian symbols of evolutionary and ecosystemic complexity.2 Evolution by association evokes pithy principles: “mix and match,” “the whole is greater than the sum of its parts,” “the acquisition of heritable characteristics.” And, regarding generative conditions, “being in the same place at the same time.” Some of these also apply to some of the processes of physiological, behavioral, and developmental evolution. “Being in the right place at the right time” may be a matter of luck, but in the course of evolution it has often been essential. You may observe that the following arenas get lengthier treatment. This is not a measure of their relative evolutionary importance. In a general sense, the symbiosis/association arena might include multicellularity, but its differentiations and their functions have been abstracted into the physiological and developmental arenas. The Physiological Arena First, what happened in this most ignored ring of the evolutionary circus? Some general principles will emerge from a recapitulation of the main events. How did they happen? What were the generative conditions and mechanisms of physiological evolution?

332

Chapter 9

The Main Events The coordinated properties of reproduction and homeostasis are inherent in the emergence of life. The first cells needed enough physiological adaptability to preserve their integrity until they reproduced, and reproducibility had to be faithful enough to pass a genetic base of adaptability on to subsequent generations. While persistence of an organism and its lineage can be summed up by the words “survival” and “reproduction,” these are emergent properties of the first living cell—not the product of natural selection; quite the reverse. The most primitive cells could both persist in a stable environment and respond to changing conditions. In prokaryotes, this adaptability combined cellular homeostasis and the genetic ability to pass on adaptational improvements located in plasmids, through conjugation. When multicellularity emerged, the homeostatic power of regeneration was enhanced. The wholeness of the organism could be maintained by cell and tissue replacement. Then multicellularity took a three-dimensional form, with an internal milieu that reduced stress on the interior cells. But the next stage of physiological evolution had to accommodate to environmental change through conformity. Initially the physicochemical nature of the internal milieu was imposed by the environment. Primitive vertebrates were vulnerable to osmotic physiogenesis during their early aquatic phase. When they became terrestrial, some physical and chemical features were locked in, but further physical changes did occur, such as an increase in oxygen capacity, cheaper access to oxygen, and the addition of thermogenesis and homeothermy. During all these phases, existing coordinative conditions were refined, and innovative progressive improvements to the endocrine and nervous systems were major emergences. The culmination of homeostatic evolution in birds and mammals prepared them to survive major catastrophes, as well as to venture into zones that would have been deadly for their forebears. Independence from the environment allowed the higher vertebrates to experiment with new habits and habitats. Novel behaviors were correlated with progressive experiments in morphogenesis. Semi-independent features could be altered, such as the use of the limbs for flying, burrowing, swimming,, and running. Internal organ systems were adjusted, and the brain hypermorphosed. The quintessential feature of progressive physiological evolution is the adaptability to keep on doing the same things when external conditions change, and to do different things when external conditions stay the same—and anything in between. This underpins anatomical and behavioral specialization for specific habits and habitats. Thus, placental mammals can get themselves into strange situations, and persist until emergent morphological features and new behaviors are integrated. The stability of their internal milieu, in particular the constancy of body temperature, is especially significant for the developing embryo. The placenta allows maternal homeostasis to be shared intimately

From the Particular to the General

333

with the developing fetus until birth at a mature stage. This new level of homeostasis was also the foundation for the emergence of cerebral complexity and intelligence. Generative Conditions and Mechanisms of Physiological Evolution Biochemical novelties can emerge at any level of organization. One mechanism is the construction of novel proteins through shuffling or differential expression of exons. Gene sharing (i.e., combining all or bits of different genes to generate novel proteins) is another. These rearrangements of DNA sequences could hypothetically account for a great deal of novelty, but “new” proteins are relatively rare. More universally, repetitive differentiation of genes allow the extension and branching of biochemical pathways. These branchings parallel epigenetic bifurcations and variability. These metabolic amplifications contributed to the homeostasis of primitive cells and of multicellular organisms operating in fluctuating environments. When they first emerged, membrane proteins or proto-integrins of prokaryotes may have had multiple skeletal, catalytic, and absorptive functions. Their repetitive differentiations established physical adhesion among cells and between cells and environmental surfaces. This, in coordination with contractile microfilaments, allowed for a degree of cellular motility in multicellular organisms. Membrane proteins also gave rise to several membrane transport mechanisms, thus setting up the generative conditions for the evolutionary emergence of nerve cells. The (possibly orthogenetic) accumulation of ion channels in cell membranes finally reached the critical point where an adequate electropotential for the effective transmission of an impulse could cause the emergence of intercellular communication. Organization of the nervous system was made more complex by the repetitive differentiation of neuropeptides and their receptor molecules. Differentiation of sensory organs and their receptor molecules let animals find their place in nature and act appropriately. Regression of key enzymes in the metabolic maze, whether through repression or the lability of the responsible genes, has sometimes reinforced commitment to particular physiological modes, such as uric acid or urea excretion. Regression of hormones affects the options available to organisms inhabiting stable and longestablished environments. As a result of the “use it or lose it” effect, most fish have regressed to the point of being unable to travel back and forth between the sea and fresh water to visit their distant relatives. Nor is the sophisticated homeostasis of birds and mammals immune to regressive specialization. Some or all of the time, hummingbirds, camels, bats, hibernators, and mole rats abandon homeothermy, thereby saving water and metabolic energy. Some inflexible biochemical novelties that are immediately useful, and normally classed as adaptations, may also strongly enhance general adaptability. For example, the waxy waterproofing layer of the insect integument, an “adaptation to a dry environment,” improves the containment of body water. It is not physiologically

334

Chapter 9

adjustable, but it enhances the adaptability of the respiratory system, and supports flight, which would be worthless if the insect were to dehydrate in mid-air. Water conservation by the mammalian keratinous skin has similar roles. But water balance in insects requires the additional backing of physiologically adaptable Malpighian tubules, and mammals need adjustable kidneys. Multifunctional organs may become specialized for particular functions, or they might go through a series of exaptations in which only one function is to the fore. Proto-chordate feeding organs doubled as gills. From them, specialized gills and endocrine glands emerged. Then lungs evolved from gill pouches, and diversified into swim bladders and amphibian, reptilian, bird, and mammalian respiratory systems. And their bony anatomy became jaws and ear bones. Organ repetition and differentiation have provided evolvability and functional diversity. Physiogenic change imposed by the environment on the whole organism has been particularly important in some lineages. The salinity and oxygen content of the internal milieu have been so altered, and, with the shift from water to land, gravity affected functional anatomy. Catastrophic environmental change may be directly mutagenic, or contingently affective. The evolution of birds and mammals was thus influenced by the K-T and Eocene catastrophes, and the isolation of the continents in the later phases of continental drift also must be taken into account for both diversification and protection from competition. Behavioral evolution is demonstrably correlated with every aspect of physiological evolution. For much of the course of progressive change in animals, behavior may have begun as individual, phenotypic, and therefore variable responses to environmental conditions. It was then largely genetically accommodated as stereotyped patterns. For behavior to become more adaptable, genetically fixed actions, or engrained habits, have to have some plasticity. Disequilibration of the external environment comes into this too, by altering threshold conditions for behavioral changes. Finally, freedom of choice depends on the complexification of the nervous system, in the context of a dependable internal milieu. Since progressive physiological evolution permits more flexible behavior and consequent feedback to the organismal condition, the action of the organism guides the genetic assimilation of appropriate qualities of its DNA. Yet this is a principle that affects development as well. Summary of the Principles of Physiological Evolution Generalities with a Zoological Emphasis 1. There are two kinds of physiological evolution, the progressive advance of adaptability, which can be seen in toto as increasing homeostasis, and the accumulation of genetically fixed changes that are adaptational to particular conditions of the external and internal environments.

From the Particular to the General

335

2. Internal adaptations consolidate advances in adaptability. But adaptational changes will not usually create adaptability, since they substitute one condition for another, and are static rather than flexible. Evolution goes backwards when adaptations make an adaptable system regress to an inflexible specialized system. 3. Physiogenic changes in the internal milieu that became genetically fixed were involved in the evolution of adaptability. Nevertheless, the physicochemical changes that resulted do not constitute exclusive or necessary generative conditions for the emergence of higher levels of self-organization. Octopuses progressed quite nicely, while keeping their original internal milieux. 4. Adaptability that makes the organism independent of change in the environment is strongly associated with the greatest potential for behavioral and functional morphological diversification. Physiological emergences have been multifunctional. 5. As adaptability has evolved, the internal milieu has become more and more stable and resistant to further physicochemical change, without, however, resisting functional morphological progress. 6. With the independence from environmental fluctuation granted by homeostasis, freedom of behavioral choice has led to further progressive improvement in animals, enabling entry into novel environments, and experimentation with habits that involved them in subsequent functional morphological evolution and ongoing challenges to their adaptability. 7. In turn, behavioral experiments determine what is physiologically advantageous. The organism sets those rules. As Holland says, “The context in which a persistent emergent pattern is embedded determines its function.” (See previous chapter.) Rules for plant physiological evolution are quite different, largely because of the essential nature of the photosynthesizing chloroplast. The very small number of animals that have acquired chloroplasts as symbiotic organelles illustrate the limitations placed on photosynthesis by the need for freedom of motility. 8. Above the cellular level the homeostatic mechanisms of plants are limited, and they tolerate a wide range of internal changes. Waterproofing and transpirational adaptability minimize desiccation. To acquire water and mineral nutrients plants must root where they germinate, and conform to the conditions of the environment. 9. A unique ontogenetic plasticity has emerged—growth responses to the environment persist for the life of the plant. The system is sufficiently flexible that one plant,

336

Chapter 9

having branches in different micro-environments, may have different functional morphologies appropriate to their locales. (Lamarck used this to support his contention that the organism responded to a need imposed by the environment.) 10. Part of the ontogenic plasticity of plants is the ability to reconstruct established cellulose walls in response to light and gravity, through the mediation of hormones. How a wall of cellulose emerged in the ancient marine unicells that gave rise to the plants is unknown. However, it involved an increase in the duration of expression of the gene for the cell surface synthetase, and was correlated with photosynthesis. The cellulose wall is multifunctional, resisting ingestion and digestion, and physically resisting osmotic lysis—especially on entry into fresh water—while admitting enough light, water and mineral nutrients. Partially desiccation-proof itself, it was the foundation for a translucent waterproofing layer over the epidermis, and had the potential to combine in a variety of structures. Some of these were to be upright stems, later reinforced as woody skeletons. These materials were unavailable to animals that could not digest them nor utilize them fully until appropriate cellulolytic symbioses emerged. 11. The functional morphology of vascular plants had the adaptability to support the tree form. This created forest ecosystems that would not only accommodate much of the diversity of other terrestrial plants and animals, but also had a climatic, geophysiological influence. 12. In one physiological feature, plants have progressed further than animals. Above the level of the ferns, fully terrestrial sexual reproductive mechanisms have emerged, with desiccation-resistant male gametes in pollen grains, and ova and accessory cells that develop into desiccation-resistant, dispersible, food-storing seeds. In contrast, terrestrial animals employ archaic fluid fertilization mechanisms that necessitate intimate sexual conjugation. 13. Because animals are mobile, some able to fly, they can make close contact with other organisms in awkward places, whether for sex or food. This facilitated the reproduction of flowering plants. Although pollen, fruits and seeds, and color and odor attractants for pollinators were emergent novelties, their subsequent improvements were subject to the preferences of insects. It follows from the above principles of plant physiological evolution that plasticity of growth, innovations in pollinator attraction and in spore, pollen, and seed dispersal, and consequent adventures in novel environments provided the varied contexts that determined their functions.

From the Particular to the General

337

The very failure of animals to emulate the dry, terrestrialized reproduction of plants necessitated the retention of a primitive animal intimacy for the intromission of swimming sperm. This sexual togetherness then contributed to the evolution of higher social symbioses. Behavior Since theories of behavior are beyond my personal mandate, the only constructive action I can take is to integrate those aspects of behavior that I have touched upon in relation to physiology, association, and epigenetics. Nevertheless, I have no hesitation to jump into the thick of the fray when simplistic Darwinism is applied in “evolutionary psychology” and its anthropological equivalent. That conflict comes from the perennial “nature versus nurture” debate about human biology, which places genedetermined behaviors at one end of the spectrum, and intelligent actions at the other. Much behavior is emergent at the phenotypic level. I have touched upon behavioral changes that occur in relation to population density in ants. Schooling fish and flocking birds show emergent behavior not found when the individuals are isolated. The changes are generated from a critical number in the flock or school, combined with the avoidance of bumping into each other. Emergent behavior can spontaneously occur in humans in much the same way, as in crowded sports stadiums. The most important human example is in behavior that results from intelligent problem solving and planning. These behaviors are non-linear and have no innate programming. Even if some behaviors are innate, or heritable, we have to accept that innateness and heritability do not reside entirely in the genes. There are no genes for behavior; genes are “for” proteins, and “for” regulating other genes. And to carry out those functions they need to be prodded and helped by the rest of the organism. It is slightly closer to the truth to say that there are proteins for behavior, since behaviors are clearly catalyzed or colored by proteins such as hormones and neurotransmitters. Yet, again, although proteins are “for” a wide variety of functions they are not “for” behavior. They too need to be stimulated by the rest of the organism and its responses to environmental influences. Their amino acid sequences do not contain the structure of behavior, only functions that contribute to it. The challenge is to link information about the functions of proteins, and the actions of the organism in its environment, through the intermediate neuronal and hormonal sensitivities and flexibilities. Unquestionably there are organisms that exhibit little more than innate, automatic activities. And in simple organisms the links between stimulus and response have been worked out in terms of receptors, neuroanatomy, chemical messengers, and locomotory coordination. In some more complex organisms too; insects and birds, for example; there are inflexible instinctive behaviors

338

Chapter 9

that serve them well under the normal conditions of their life. Despite the advantages of flight, it limits body size and the ability of the ganglionic centers to expand to the point where greater freedom of thought and actions are possible. Instinct is an efficient way of packaging fast behavioral responses into small nervous systems, so long as the conditions of life remain the same. Since they do change, it is not surprising that birds and even insects are not complete automata, but have a measure of behavioral adaptability. Flexible exploratory actions are essential for finding food, shelter and mates, regardless of phylogenetic affinities. Complex innate behavior patterns provoke the question of how they came to be that way. Ethologists, whether genocentric ultra-Darwinists or not, would hardly accept that a complex gene-determined behavior pattern could appear as a saltatory macromutation or random combination of genes for behavior. By the Darwinist thesis, the buildup has to be gradual: one gene influences a little bit of behavior which is selected, another gets added on, and so a complex behavior pattern is constructed. The other alternative, one proposed by many ethologists, is that the behavior is initially very plastic and exploratory. The individual organism experiments by trial and error, and repeats those behaviors that bring food, shelter, or mates. Then genetic assimilation occurs through to genetic, epigenetic, and physiological experiments. This is the kind of process Jean Piaget outlined in Behavior and Evolution (1979). (Piaget would be historically classified as a structuralist, which is almost an emergentist.3) He also inferred that the greatest variety of experimental behavioral initiatives would occur on entry into new environments where the agents of natural selection were weak.4 From the adaptationist’s point of view, the success of particular behaviors that had become genetically fixed would be complemented by the kind of eusociality found in insects. Not only would every member of the society have the advantageous behavior; there would also be the potential for repetitive differentiation into castes—separation of offices and concurrences of efforts. We should however, keep in mind that those castes are determined by ontogenic hormonal manipulation. Moreover, the birds do very well without being eusocial. The birds and insects, despite some flight-associated convergences are also distinctively different physiologically. Some ethologists, including Gilbert Gottlieb (1976, 1992) and Simona Franková (1987), have seen parallels between behavioral and epigenetic evolution in terms of constraints, plasticity and canalization. And the argument of Stuart Newman and Gerd Müller (2000) that in epigenetic evolution the genomic role is to finally consolidate the processes that began as individual organismal experiments echoes Piaget’s view of behavioral evolution. So the process has parallels with evolutionary changes in epigenesis. Another common feature is the possibility for behavioral, and physiological changes at developmental thresholds which Eugene Balon has emphasized in his ichthyological discourses. No matter how genetically fixed the behavior of our fishy ancestors was, they had to have kept their options sufficiently

From the Particular to the General

339

open to invade the freshwater environment and to allow for their descendants becoming terrestrial. Now, how evolvable is behavior that has become entrenched, or instinctive? Presumably our ancestors among the mammals, reptiles and amphibians had more stereotypical behavior patterns than we. And perhaps, up to a point, we retain more of that kind of behavior than we would care to concede to the “evolutionary psychologists.” In physiology, fixation of homeostatic functions began early, though the physicochemical milieu was subjected to a series of physiogenic alterations resulting from behavioral and consequent environmental changes. Behavior changed the internal milieu and its physiological capacities. But the final entrenchment of emergent placental homeostasis was what gave the placental mammals their major boost in epigenetic, anatomical, and behavioral diversification and freedom—more of a slingshot than a bootstrap. While there are innate, though variable, factors of neuroendocrinology in the evolution of human familial associations, individuality and freedom of behavior has been retained. The playfulness of the young has been extended into adulthood. These freedoms have allowed us to enter into broader social relationships that are both flexible and fragile. The Developmental Arena Although the separation of the physiological arena of evolutionary causality from developmental physiology and morphogenesis is somewhat arbitrary, there is one striking distinction. Homeostasis in the mature organism is a relatively steady state of physicochemical and biotic conditions in the internal milieu. It is the foundation of diversifying evolution and potentiates further progressive evolution as well. Homeorhesis governs a chronological sequence of changes in gene expression, and in cellular differentiation, as well as in the internal milieu, resulting in a mature organism that is larger and more complex than the egg from which it came, while keeping it true to type. The tighter that control, the more difficult it is for epigenetic evolution to occur. Entrenchment of generative interactions, and strict canalization dominate epigenesis, though they can be supervened by novel additions or regressions in the late stages of development. However, innovative morphogeneses deep within development are known to produce cenogenic novelties of larval form and behavior. And some radical deviations that affect the whole life history can be detected. Also, many experiments involving the addition or removal of non-genetic materials and stimuli demonstrate that interference with morphogenesis sometimes changes the final product, without detriment to the organism. Natural evasive tactics such as paedomorphosis permit escape from generative entrenchment and tight canalization, allowing freedom to experiment with diversification.

340

Chapter 9

From a host of enlightening books and articles on epigenetics I will single out only one here, since it was the inspiration for the organization of this chapter. The novelties defined and detailed in Gerd Müller’s essay “Developmental mechanisms of the origin of morphological novelty: a side-effect hypothesis” (1990) are essentially epigenetic emergences. Müller’s aphorisms blaze a trail toward an evolutionary “theory of development”: 1. Novelties are defined as qualitative morphological changes with a discontinuous deviation from the ancestral state. 2. The majority of novelties arise as secondary by-products of epigenesis that appear when quantitative modifications of developmental processes reach a threshold of the affected system. 3. The causality for the origin of novel structures lies not within the genome but in epigenesis. 4. The specificity of morphological innovations depends on the reaction norms of developmental systems at their limits. 5. Intermediate structures and sequential switching of mechanisms provide opportunities for novelties to arise. 6. The plasticity prevailing at all levels of developmental systems facilitates the epigenetic integration of novelties. 7. The mechanisms named share a potential for rapid morphological transformation which may underlie several discontinuous phenomena of the evolutionary record.5 Müller is dealing specifically with morphogenesis, so a few additions are needed. At this stage in his quest, the role of natural selection in consolidating homeorhesis, or tightening canalization—hence reducing evolvability—is ignored, and environmental influences are not elaborated. The nature of epigenetic plasticity cited in aphorism 6 needs explanation, and specific molecular mechanisms—in contrast to morphological mechanisms of developmental evolution—are missing. Heterochrony, though implicit in item 7, could also be more explicit. However, he makes up for this in a subsequent essay co-authored with Günther Wagner in 1991. Cases that they cite include changes in the timing of cell lineage segregation in blastomeres of direct developing embryos, which lead to novel forms in nonfeeding larvae. Some primitive larval features are eliminated, and others appear earlier than usual.6 Also, skeletal evolution in vertebrate

From the Particular to the General

341

limbs has involved heterochronic simplification as well as innovative repatterning through “dissociation and recombination events.”7 According to aphorism 2, the majority of epigenetic novelties arise from a previous continuity of change, and so they fit the category of critical-point emergence. Discontinuous phenomena, or “rapid morphological transformations,” appear in aphorism 7. Paedomorphosis alone is a saltatory change, a regression, but without the loss of redundant DNA and associated molecular mechanisms. Those can be retooled for new functions. Other saltatory changes are progressive, such as topographic relocation of primordial cell lineages with epigenetic effects. These changes seem to be buffered by developmental plasticity, which Bateson called the “accommodatory mechanism . . . not generally recognized as existing, though when stated it seems obvious.”8 Associated with this is William Wimsatt’s “generative entrenchment,” which refers to a dynamic stability that results in canalization and resists evolutionary change. The degree of entrenchment relates to “the scope and structure of dependency relations among the parts of a system and the environment”9: The very existence and structure of dependencies in developmental programs is the metacondition that makes contingencies (and history) important in evolution—these contingencies, or those of other interacting lineages. Features (whether contingent or not) that accumulate many downstream dependencies become deep necessities, increasingly and ultimately irreplaceably important in the development of individual organisms. This causes them to be increasingly conservative in evolution, and this together with the inheritance of features down taxonomic lineages leads them to become taxonomic generalities of increasing scope, broadly represented across many organic types.10

I would prefer “phylogenetic commonalities” in place of “taxonomic generalities,” since the inference is that a generative entrenchment of a radially symmetrical body form, for example, remains a common feature of the organisms that arise from the archetype. Some entrenched qualities are universal, such as the relationship between mitochondria and their eukaryotic hosts. Similarly, chloroplasts are entrenched in plants. Symbioses are polyphyletic. However, although they have neither taxonomic nor phylogenetic general significance, the principle of generative entrenchment as stated by Wimsatt can still be followed. For example, bivalve sulfur-oxidizing symbioses led to an interdependency between cells with many mitochondria, and cells containing the new symbionts, and it also had epigenetic downstream effects on the form of the ctenidia that contain the bacteriocytes, as well as ventilation behavior (Reid 1990). Yet it also freed many of the symbiotic hosts from the trouble of developing digestive tracts, and so generative entrenchment does not imply permanent fixation of all the systems that were interdependent when the entrenchment occurred. Some could be free to regress, or to undergo exaptation. Wimsatt (1999) also makes a necessary distinction between canalization and generative entrenchment. Canalization is the degree of regulation of developmental

342

Chapter 9

pathways that are epigenetically downstream from the entrenched systems. It is therefore a sometimes advantageous concomitant of generative entrenchment, but not the same as generative entrenchment. Once essential interdependencies have been entrenched they are further stabilized by the syndrome of internal selection, as L. L. Whyte (1965) proposed. I alternatively use the term “co-adaptation” for the internal adjustments that are made in the wake of disequilibrating physiogeneses and symbiogeneses. Although we do not yet fully understand how they can be overcome, mechanisms that accommodate change clearly exist. Homeorhesis is nothing more than the whole package of accommodatory mechanisms that determines trueness to type in tightly canalized organisms. In loosely canalized epigenesis, interferences are accommodated so that the changed product of development is still functionally integrated. For example, the elongation of one toe in the horse lineage, at whatever stage in its evolution, could function in locomotion since it was always fully integrated with the other bones and neuromuscular features of the leg. This would still be true if the elongation involved a saltatory growth spurt instead of the imperceptible steps demanded by selectionism. As Goodwin says, this developmental robustness could not exist if the causes of epigenesis were nothing but the actions of structural genes and their genomic switches. And as Schmalhausen said, long before, physiological adaptability also accommodates developmental change. Many progressive epigenetic novelties are critical-point emergences, and can be explained in terms of constrained pathways, codon or gene amplification, and repetitive differentiation, the latter fitting Müller’s category of “secondary byproducts.” Internal selection might constantly tighten canalization to the point where every optional pathway is constrained. However, there is always room for experimentation with redundant DNA, especially if the pool of non-transcribing repetitive microsatellite sequences is routinely tapped to patch broken chromatid strands, without interfering with the prevailing dynamic stability.11 The range of such natural experimental methods is outlined in chapter 6. If the experiments are disintegrative they will be lost before they are found. If they are encumbrances, such as energy drains, their persistence will be diminished. When they become manifest in the phenotype they will be ignored if neutral, and preserved only in the narrow lineage in which they originate. If beneficial in some way, they will be incorporated into a new dynamic equilibrium and spread throughout the breeding population. Even if the innovations are advantageous, their subsequent enhancement may come from an autonomous, self-amplifying genetic drive, not the approval of selection. Allometric hopeful monsters are common, even in tightly canalized types. Their evolution is much easier to understand if they have been autonomously driven through the incipient experimental phase before they show any kind of advantage. DNA made redundant by the regression of ancestral pathways of development, along with

From the Particular to the General

343

repetitive gene and genome amplification and differentiation is a wellspring of experimental material and a key to understanding epigenetic plasticity. Epigenetically active hormones and enzymes that arise by repetitive differentiation may extend developmental pathways or cause them to diversify by branching. As Ryuichi Matsuda (1987) proposed, novel environmental influences are also important for epigenetic emergences. Also, their subsequent internalization into homeorhesis, as envisaged by Schmalhausen, could largely be explained as genetic assimilation—or epigenetic assimilation. This makes the canalization more independent of environmental vicissitudes, and establishes a new level of dynamic stability adjusted to the emergent novelty. Direct environmental effects may also be exaggerated temporarily by the effects of natural disasters. As well as allowing earlier emergences to escape from natural selection, catastrophes could conceivably cause direct changes through extreme physicochemical stress, such as heat shock. Although the significance of intermediate structures raised by Müller was anticipated by Riedl (1978) and Matsuda (1987), Eugene Balon’s development of the idea in 1986 is the most thorough: Life processes use bifurcations to create novelties and alternative answers, as and when required at any interval of ontogeny and evolution. . . . The saltatory mechanisms of epigenesis are responsible for a sequence of homeorhetic steps that are separated by relatively unstable thresholds, and . . . during those thresholds changes in life history as well as novelty are created.12

Balon agrees that there is a tendency toward tighter canalization (and, implicitly, generative entrenchment) as well as specialization during the course of evolution, but his research on fish indicates a number of possible escape routes at epigenetic thresholds, some of which are marked by larval and juvenile behavior changes during normal development. An external or internal environmental change at a threshold can result in a variant phase of development that may be effected by allometric shift or another epigenetic accommodation. An interesting case is that of Arctic char that are ice-locked as juveniles. At the stage of development when they are ready to forage for themselves they have no access to the atmosphere for filling their swim bladders. But they compensate by developing lighter skeletons than conspecifics who can more typically swallow air at the appropriate time.13 Large eggs buffer development through the early thresholds, and so offer greater homeorhetic stability as they head toward their established specialization. However, a sheltered childhood is not bound to have such an effect. In mammals, for example, it has a distinct advantage before adult physiological homeostasis stabilizes. But that has not prevented anatomical diversification. Developmental arrest and paedomorphosis are common regressions that sometimes follow through to bigger and better things, while leaving clues to their historical occurrence. For example, in hominid evolution, juvenile anatomy is prominent, while hypermorphosis increases brain size, and adjusts bipedalism, a vertical stance, and

344

Chapter 9

vocalization. In any case, Balon’s and Müller’s point is taken—discontinuous phases in development present opportunities for emergent evolutionary change. This is a nice fit with John Holland’s concept of a transition function whereby a change of state allows several different strategies to operate. It is appropriate to return here to the epigenetic influence of symbionts. The presence of nitrogen-fixing bacteria in host plants stimulates nodule formation, and I suspect that sulfur-oxidizing bacteria and dinoflagellate symbionts may also have had epigenetic effects on their hosts. The ability of retroviruses to be transmitted as oncogenes that locally disinhibit cell division, resulting in tumors, though detrimental in this case, is in the same category. Hypothetically, retroviral transgeneses could have beneficial, epigenetic effects. According to the human genome project, numerous transfers of genes from bacteria have occurred relatively recently, some of them with advantageous qualities. There is also growing evidence that gut microflora affect development of the alimentary tract. In humans, some relatively benign parasites such as pinworms trigger the development of defense mechanisms that prevent more serious ailments later in life. As I noted in “Evolution by Association,” gut bacteria also effect “normal” gut development and digestive and immunological functions in mice. Just as interesting are the free-living bacteria that stimulate the sea lettuce, Ulva, to make “lettuce-leaf” thalli instead of the thready form that they have in axenic (sterile) culture.14 Nitrogen-fixing bacteria stimulate the formation of root nodules in their host plants. Brian Hall (1992) classes these kinds of influences under “interspecific epigenetics.” We should also recall intraspecific effects such as alterations in feeding regimes that differentiate castes in ants and termites.15 There are also the consequences of overcrowding that change locust color and behavior, and that produce monstrous carnivores among salamanders. These changes overlap with the physiological factors. Body Plans The diversification of body plans among multicellular animals has been addressed by a number of authors, especially during the last decade. It is a subject that I have not elaborated. However, the notion of a single body plan common to all multicellular animals has an unavoidable theoretical allure. In The Plausibility of Life (2005), Marc Kirschner and John Gerhart propose that such a universal body plan has existed since the late Pre-Cambrian, though expressed in evolution in a number of diverse forms. It usually occurs in a bilaterally symmetrical organism that in its simplest state consists of a series of “compartments,” not necessarily visible, as in the case of segments in annelids and arthropods, or serially paired somites in vertebrates. The compartments are semi-autonomous regions whose forms and functions are regulated by differentially expressed Hox genes (not to mention many others,

From the Particular to the General

345

including Wnt). They constitute an invisible map that has no simple anatomical equivalent, since the compartments cooperate locally to produce visible morphological features, such as segments and limbs. The diversity of body forms, whose differences we tend to emphasize when we examine them, is thus the product of differential regulatory gene expression—to which we must, of course, add the effects of the external environment, the overall behavior and physiology or the organism, and related genomic changes. The Big Top The circus is an adequate metaphor for the complexity of the structures and functions of the biosphere. We can concentrate on the performances in a particular ring, or sit far enough back to view all three rings and take in the aerialists under the big top. It is complex, and dynamic, and the ringmaster is trying to keep it under control. He can’t keep his eye all the time on what the clowns—the generalists—are up to as they fool around with new acts in obscure corners, nor can we. Because it is mostly in a state of dynamic equilibrium, the circus metaphor has its limitations. To view progressive movements in evolution we need to look at the bootstrapping relationships between development, physiology, behavior, association, and environment, as well as their hierarchical structures. Curiously enough, in view of the claim of the Modern Synthesis to synthesize ecology and genetics, there is no existent “theory of environment” of much value to emergentism. Environmental biology comes in bits and pieces, and are assumed to be integrated under selection theory. But theories of demographics take environment too much for granted. In the context of adaptation by natural selection, environment is a mere backdrop to the real action, or a bag for the gene beans. Richard Lewontin provides a thoughtful epigraph for this section: Just as there is no organism without an environment, there can be no environment without an organism. . . . An environment is something that surrounds or encloses, but for there to be a surrounding there has to be something at the center to be surrounded.16

Lewontin writes in The Triple Helix (2000) that Darwin’s role in separating the organism and its environment was a necessary step for the analysis of natural selection and population studies and ultimately molecular biology. And he rightly argues that it is now time to put them back together again. We shouldn’t be too hard on Darwin; he knew that the bank was tangled, that populations of cats and mice and bumblebees were related, and that the earthworm, perfect in its adaptation, had a massive ecological impact. His greater sin was in separating the organism from its development and physiology, turning it into an abstracted pawn, to be sacrificed or crowned according to time, chance, and natural selection.

346

Chapter 9

One of the subjects of Lewontin’s critique is the Darwinist idea that an environmental niche can exist in anticipation of an organism’s occupation of it. As its proponents might say, the environment constrains and pressures the organism into a particular adaptational direction. Yet, as Lewontin points out, we can identify many such potential niches that have never been occupied. And I have already provided examples of potential adaptations that would require only the slightest nudge to pre-existent qualities. But the slightest nudge is still not a real force. Lewontin notes that birds that nest in trees do not eat the foliage of the trees. Maybe the hoatzin is an exception, since it can eat and digest foliage due to its cellulolytic symbionts. But it prefers juicy, low-growing marsh plants. In any case, most birds that dwell in trees do not eat the most abundant source of potential food. My daughter Clio has worked with the kakapo, an endangered species of flightless parrot, endemic to New Zealand, whose population numbers less than 100.17 A ground-nesting bird, it climbs trees to find fruit and seeds, and to roost. Its large size and relative inability to escape introduced predators, has brought it close to the fate of the similar dodo. Like the dodo, and other extinct flightless birds, it has done nothing constructive with its wings. It hasn’t turned them into hands. It doesn’t even use them to hold on when it climbs. At best it seems to use them as stabilizers when they jump or fall out of their trees. Handreared young wave their wings in apparent approval of human helpers bearing gifts of food—or to slap them in the face when they err in their ways. So there is a possibility that they have some secondary sexual signaling function during mating—which has not at the time of writing been directly observed. There are all kinds of potential niches beckoning to flightless birds with hands, but they have ignored the invitations to exapt. Why-not questions are interesting, but philosophically illegitimate. Therefore, we will not explore the significance of entrenched generative systems and epigenetically constrained pathways in this regard. As J. Scott Turner (2000) remarks, the early ecological theorists Whitehead, Clements, Tansley, Hutchinson, and Odum had roots in holism, and they did not draw a distinct boundary between the organism and its environment: Presently, biology that is not strictly materialist or reductionist is commonly regarded as somehow suspect or deficient in intellectual rigor. Even among ecologists, I think it fair to say, “good” ecology is defined by the distance one can put between it and the holistic philosophical leanings of the early ecologists.18

As a physiologist, Turner can understand that point of view without sympathizing with it. Physiologists only temporarily reduce systems to their parts before putting them back together again, and Turner goes further than most physiologists in restoring the organism to its environment (q.v. below). In my own case I can hardly listen to an ecologist say “I’m a reductionist and proud of it” without registering my disbelief.

From the Particular to the General

347

An environmental theory necessary for an emergence synthesis would look similar to a blend of Lamarckism and neo-Lamarckism, without the inheritance of acquired characteristics. Although Lamarck believed that the natural course of evolution was a gradual, intrinsic increase in complexity and organization, he knew that it was changed by ecological encounters: . . . by the influence of circumstances upon habits and then by habits on the state of the parts and even on organization, each animal can receive in its parts and its organization modifications susceptible of becoming quite considerable and of having given rise to the state in which we now see all the animals.19

Environment pervades the developmental, physiological and associative theories that I have just discussed, yet its importance demands an independent section of this chapter. In this instance I will follow a hierarchical approach from molecule to biosphere. The Molecular Environment The realization of genetic information depends on multiple factors, among which are the absence of repressors, presence of inducers, association with binding proteins and form-making proteins (prions and stress proteins such as chaperonins). Local ionic, electric, hydrophilic, or hydrofuge conditions, and organelle compartments, are also affective. The principle of homeostasis can be applied at this basement level, to explain how things stay the way they are: evolutionary experiments are blocked, and molecular glitches repaired. Change in these conditions can effect evolutionary change, and lift the lid of the molecular alchemical cauldron enough to let the occasional experiment escape. Weng et al (1999) have shown how the contingent association of biochemical pathways, and the use of molecular scaffolding to channel reactions and improve their efficiency, can all generate a tremendous range of novel properties. There is such great potential that the number of emergences is not an issue, but constraining them epigenetically and inhibiting them into ordered activity is. Homeotic gene mutations can produce saltatory epigenetic change when they get away with it. Experimentation is enhanced by the availability of spare parts through codon, exon, gene, and genome duplication, as well as the shuffling of exons, introns, and protein domains. The Cellular Environment Here we go outside the cell to the internal milieu, so that we can consider how cells are affected by that environment and intercommunicate through it. In varying degrees, cells influence each other’s mitotic cycles and differentiation. The fluid in which they are bathed is under homeostatic regulation, but change in local conditions allows escape from sameness. Differential adhesion of cells can affect organismal form

348

Chapter 9

through changes in self-assembly, even before genetic assimilation brings genes into play. The orientation of sheets of cultured cells has been shown to respond to light signals, and there is a faint possibility that photon emissions from metabolic processes can be detected by adjacent tissues.20 Light stimulation is known to be important in the development of cells and organs that use light as a source of energy or information. Migratory mesenchyme cells can experiment with the redistribution of tissues. Neural crest cells can alter major patterns of development of vertebrates. Cells such as neurons that grow probing and communicating axons can be powerful epigenetic inducers. Even more striking is the way in which they intercommunicate to generate evolutionary and ontogenic emergences that we call memory, intelligence, logic, language, consciousness, aesthetics—and, taking them together, mind. Koch and Laurent (1999) demonstrate that on top of the intracellular signaling potential of neurons, the emergent combinatorial potential of multiple dendrites, and intercellular signaling molecules is enormous. This emphasizes the conclusion reached by Wimsatt (1997) to the effect that once mere aggregates have been removed from consideration, the possibilities for emergent changes to wholes that are greater than the sums of their parts, both physical and biological, are almost infinite. Endosymbioses and other interspecific cellular relationships allow a nutritional mutualism between host and symbiont cells that is a significant emergent property. Furthermore, endosymbioses influence the epigenesis of the host. The Organism and Its External Environment Compared to the way in which the biotic, intraorganismal environment can be complexified by the endogenous activities of cells, the physicochemical effects of the external environment are simple, but nevertheless significant. They impose themselves on the internal milieu of primitive organisms. They may impose stresses that alter heat-stress protein functions and methylation patterns. Therefore, wherever their behavior takes organisms, both molecules and cells are affected, in their development and physiological function. Major evolutionary changes have been initiated by non-heritable influences of the external environment. Physical factors include light, temperature, high-energy radiation, diffusion, osmosis, gravity, electricity, electromagnetic fields, surface tension, friction, and mechanical stress. Chemical factors include the concentrations and proportions of ions, nutrients and the properties of colloids. Biotic factors include symbionts, conspecifics, other organisms, and airborne or water-borne biochemical messenger molecules. These may have direct influences on development, physiology and behavior, through sexual reproduction, the establishment of new symbioses and other associations, competition, disease, predation, or browsing. The more adaptable the organism, the more it can experiment with behavior and explore new environments. By consistently and persistently responding to a particular environment in a particular way an

From the Particular to the General

349

organism can direct the future of its lineage toward particular functional and anatomical adaptations. I have already touched on the larger aspects of geophysiology that have wrought diversions in the course of evolution. They have also contributed to Gaian homeostasis, to which I will return to it in the section below on “the biosphere.” Organisms can also effect small changes in the local environment that might be beneficial to them directly, or through consequent changes in local community structure. These local changes make up much of the substance of J. Scott Turner’s The Extended Organism (2000). He justifiably makes much of nest building and burrowing as habits that make environments less stressful, and often more productive of potential food organisms. I have a personal stake in such matters, having studied burrowing benthic marine animals for all of my career, and know how oxygenation and the consequent changes in bacterial communities can enrich the sediments. Such processes can be intermediaries in the invasion of hostile environments. Soil building by primitive plants is also a local effect with a global consequence. In rising through the hierarchical biological levels we come to populations and species. As populations expand in numbers and space they will encounter and be limited by stressful conditions. The most adaptable members of populations will be most likely to endure at the edge of chaos, trading off physiological stress for limited competition. They are not so much selected by the environment as self-selecting for the environment. Those that possess the appropriate emergent qualities can penetrate environmental interfaces to found new lineages that are genetically and epigenetically different from the parental population. They do not need to pass immigration inspection to do so. Since this is close enough to the conventional allopatric model I will not further expound it, except to say that large numbers of individuals might be capable of simultaneously penetrating the new environment. Thus, the emergent qualities of the new population are features that already existed at the interface, they are common to all of the new immigrants, and not the consequence of “new strong selection pressures.” The catalysis of stress is not limited to environmental boundaries. Overpopulation can stimulate some profound epigenetic changes. The example of the emergence of carnivorous monster salamanders that reduce the numbers of the normal vegetarian type is a case in point. Overcrowded locust hoppers are epigenetically stimulated to become flyers. During the evolution of these phenomena there might have been an initial “selection for” the flyers. But who is to say that their departure did not restore the parental population to its carrying capacity and so continued survival. One thing is sure, the novel deme that flew to pastures new was entirely composed of individuals who had the epigenetic plasticity to fly when the environment told them to. Thus, stable phases and turbulent thresholds exist not only at the developmental and physiological and behavioral levels, they can also be detected at the population

350

Chapter 9

level. An awareness of this is to be found in Mivart’s “intermittent stable equilibria” of organismal evolution, and in Eldredge and Gould’s initial exposition of punctuated equilibria of speciation. The Biosphere The word “biosphere” was coined by Austrian geologist Eduard Suess in 1875, and the concept was developed by Vladimir Vernadsky in Biosfera (1926). Vernadsky saw it as a heterogeneous sphere of living matter, which he sometimes called “animated water” since it was largely in and of the hydrosphere. The animated water of microscopic unicells, plants, and fungi continued from the seas and lakes onto the land. The biosphere was also in dynamic equilibrium with the lower geological substratum, and with the atmosphere. Beyond it lay an outer sphere of cosmic energy that came mainly from the sun. Before the 1998 publication of Vernadsky’s book in English as The Biosphere, James Lovelock (1979, 1988) had proposed his Gaia theory, which included the evolution (sensu lato) of the biosphere. He also recounted the development of Gaia, with well-argued details regarding atmospheric oxygenation, albedo effect, the climatic and nutrient roles of sulfur compounds, and other ways in which organisms interact with their environment. Lovelock’s mechanistically objective explanations more than compensate for what some of his readers find too metaphysical or even mystical. Like Vernadsky, Lovelock appreciates the feedback between organism and environment on a large scale, and like James Hutton (1785) sees that there is a geophysiology analogous to organismal homeostasis.21 Hutton, a physician and an agrarian by training and experience, was also a theoretical geologist. His correspondent Erasmus Darwin also saw animal physiology as a model for both organization and evolution of the Earth.22 Lovelock’s sometime collaborator Lynn Margulis has properly added symbiosis as a crucial emergent property of the biosphere. In a variety of ways, organisms can alter the biosphere locally and temporarily. Effects such as epidemics, population explosions of pests, parasites, and predators can seem devastating from the human perspective. Less dramatic alterations of soil building on land, and eutrophication of lakes and seas are, however, more important. Sudden rises of redox layers in aquatic environments can cause die-offs that increase biological oxygen demand and allow sulfide buildup. Instead of being decomposed back to carbon dioxide and water, organic debris might be partly sequestered, and fossilized, without consuming oxygen. Ecosystems, and patterns of climate can be altered by the evolution and redistribution of plants. Several symbioses have played major roles in such changes, though none can match the oxygenation of the atmosphere by symbiotic photosynthesizers. Large-scale changes have led to the biosphere with which we are familiar. But the dynamic stabilities of Gaia have not followed the same path as biological homeostasis. Although they involve some physicochemical stability through negative feedback,

From the Particular to the General

351

they have not advanced to a state of sophisticated resistance to change. That is just as well for biological evolution, which has needed the capricious clean sweeps of electromagnetic reversals, plate tectonics, volcanic activity, glaciation, deluges, sea-level changes, and the impact of comets and asteroids. These environmental alterations were part of the substrate of progressive emergent evolution. In progressing from molecules to the biosphere, we have taken one aspect of the environment so much for granted as to ignore it. Therefore, let’s go back to Vernadsky’s idea of animated water. We inhabit a planet where water is usually in a liquid phase that possesses anomalous physicochemical qualities. Life in all its manifestations would be impossible without the emergent wetness of H2O. Water is not only the universal solvent of the internal and external milieux. It buffers temperature change in all of its standing bodies, from puddles to oceans. My colleague Richard Ring found that a beetle, Pytho deplanatus, survives supercooling to –54°C because it accumulates the antifreezes glycerol and trehalose in its body fluids.23 At the other end of the temperature range, water, under sufficient pressure, remains liquid to its critical point of +374°C and 3,212 pounds per square inch. This makes it more plausible that life might have originated at high temperatures deep in the Earth’s crust. Water is a reactant in a wide range of biochemical processes, and its kinetic properties are particularly important in calcium-mediated functions. It is also a structural component of proteins, cartilage, and bone. Water is the medium of blood circulation; it cools by surface evaporation; its hydraulic behavior affects growth and moulting in crustaceans, locomotion in small insects, spiders, octopuses, slugs, snails, earthworms, and echinoderms. Under pressure, it also affects excretion in kidneys, feeding behavior in carnivorous clams, metamorphosis in insects, the epigenesis of eyes and brains, and sexual intercourse. At the surfaces of bodies of water, where the temperature falls below 0°C, its supramolecular structure and cohesive properties are capable of change, yet it does not immediately crystallize to ice without physical stimulation in the form of nucleating agents or mechanical disturbance. Its cohesiveness increases at biological surfaces, and it presents different fluid-dynamic challenges to microscopic and macroscopic creatures. There is no gene for water, any more than there are specific genes for particular behaviors. Yet, as much as behavior, water molded the early evolution of function and form. Then, when organisms emerged from water, they carried it in their internal milieu. And its very absence from the terrestrial environment molded them even more. Processes of Natural Experimentation and Their Emergent Results Now that the various causal arenas have been surveyed, it can be asked what processes and mechanisms of natural experiment and emergence are common to them all? The

352

Chapter 9

answers can be cross-referred to hierarchical levels. This may bring us closer to a unifying synthesis of emergence, or demonstrate that such a unification is impossible. Darwin’s “entangled bank” microcosm of biotic complexity was a product of “Growth with Reproduction; Inheritance which is almost implied by Reproduction; variability from the indirect and direct action of the external conditions of life, and from use and disuse; a ratio of increase so high as to lead to a Struggle for Life, and as a consequence to Natural Selection, entailing Divergence of character and the Extinction of less-improved forms.”24 Darwin was comprehensive enough for his time, especially after he co-opted neoLamarckism. And he had his priorities in the right order too, except that divergence of character is also prior to the consequence of natural selection. But what mechanisms and processes raise emergences from their generative foundations? Some answers have already been provided, especially in the “Where Are We?” section at the end of chapter 7 and in the introduction to the present chapter, but they bear elaboration. I begin with the simple process of reproduction that characterized the emergence of life, and then those mechanisms that emerged from simple reproduction to generate greater complexity and persistence in being. Organismal Reproduction Duplication of the whole organism requires duplication of the genome. In most sexually reproducing organisms a haploid copy of the genome is contained in an appropriately structured gamete. Of the gametes, the egg cell is most important to the developing embryo, since it provides a suitable milieu, molecular resources, and organelles. It also initiates a chronological series of epigenetic triggers not all of which are genetic. The size, structure, and constituents of the egg are crucial, as are environmental stimuli. Since the integrity of individual organisms eventually fails by time and by chance, reproduction ensures the continuity of life. It therefore generates persistent complex patterns on a scale quantitatively and qualitatively different from non-living systems. Starting from scratch in every generation from a single cell also exposes every organism to epigenetic experiment. Because of the inherently experimental nature of the copying process, biological reproduction and duplication of the constituent holons of the organism have the emergent potential for differentiation. But to reproduce, individuals must first survive, and that depends upon their integrity, the improvement of which is the mark of progressive evolution. Sexual reproduction was a novel experiment consisting of an emergent combination of genome duplication together with an intimate association of partners or their reproductive products. Even excluding gene duplication and mutation, it made possible repetitive differentiation at the genome level, and that allowed much greater freedom of experiment with phenotypic variability, while retaining sufficient reproductive faithfulness that the novelty would not be lost to subsequent generations of the lineage in which it occurred.

From the Particular to the General

353

Progressive Homeostasis The ability to persist in being was one of the first organisms’ fundamental characteristics that made evolution possible as soon as they emerged. Along with reproduction, self-maintenance was a sine qua non of evolvability, and the complexification of homeostasis was a major part of progressive evolution in animals, and in plants too, though the concepts differ somewhat. Many of the following mechanisms or processes have contributed to one of the most fundamental features of animal progressive evolution; i.e., the possession of an internal milieu that supports increased behavioral flexibility and freedom of choice. The early stages of this physiological evolution depended upon a bootstrapping process of feedback between behavior, environment, the physiological state, and development. Once the homeothermic condition was locked in as the final step in physicochemical homeostasis, it was possible to generate a metabolic rate necessary for sustaining a sophisticated nervous system, and consequently intelligent behavior. The placental mammal is the only type whose homeostasis is consistent throughout the entire life cycle. At this stage physiological homeostasis is the major generative condition of further evolutionary advances in behavior, anatomy, and association. Repetitive Differentiation This process of improvement of the organism’s integrity operates in every causal arena of evolution as a mechanism of overall enlargement, allometric shift, and the provision of differentiated units that can ultimately be coordinated in a complex hierarchy. While the mechanisms vary in the organismal hierarchy from the molecular to the social it amounts to a common process that constitutes a general principle of emergence that operates in space and time. At the biochemical level, repetitive differentiation of DNA produced gene families and hence multiple homologues of structural proteins and enzymes, permitting the buildup of biochemical pathways and cycles. As a result, different enzymes and hormones based on the same ur-DNA could provide for greater physiological adaptability, making even primitive organisms less vulnerable to the vicissitudes of environmental change. The mixing and matching of protein domains led to even more dramatic biochemical emergences. However, the number of protein domains is fewer than the total number of genes, since the polypeptide backbone and amino acid side chains have a constrained geometry. The same limitation applies to domain mixing and matching.25 At the organ level repetitive differentiation of anatomical units allowed the separation and concurrence of physiological offices. This was enhanced by progressive packaging, which also improved the organization of rapidly expanding tissues such as the neocortex of the brain. Even after the brain functions have become relatively fixed, differentiation of integration, in the form of rewiring, have produced

354

Chapter 9

evolutionary change. And novel dendritic connections have both phylogenetic and ontogenetic qualities. Multicellularity potentiated all of the epigenetic differentiations that built upon one another to construct the complexity of plant and animal life. This exposes the arbitrary nature of separating evolutionary causes into three arenas. Though convenient for simplicity of analysis, they cannot be left so disjointed. Physiological progress and anatomical diversification would not have happened in the absence of evolving multicellularity. Social associations among insects are analogous to repetitive differentiation of multiple cells: specialized castes come from repeated organismal units. Finally, family and social groups in higher animals emerged from generative conditions provided by epigenetic and physiological advances in neural self-organization. And, within large social groups, behavioral specialization could arise without commitment to genetic fixation. The evolutionary significance of repetitive differentiation as “varied repeats” has finally been appreciated by molecular biologists. The concept illustrates how natural experiments can take place out of sight of natural selection, and how the simplest kind of self-replicating system can spontaneously become complex and potentially selforganizing through multiple feedback controls. And it applies to the duplication of discrete codons, exons, introns, genes, chromosomes, karyotypes, cells, tissues, organs, segments, organisms, populations, and societies. The principle also embraces wholes that are greater than the sums of their parts, but also implies that they can progress to higher levels of complexity and wholeness. Repetitive differentiation also provides redundancy, which also fits the metaphor of evolution by natural experiment. Spare parts—genes, organelles, organs, clones, asexual organisms (in societies)—might be duplicates of those with essential functions. Or they may simply have become redundant because the original role is no longer necessary and thus freed to do something different. Therefore, Dohrn’s functionswechsel and Gould and Vrba’s exaptation are relevant to this kind of natural experiment.26 Mixing and Matching As soon as molecular, cellular, organ, and organismic holons began to be repeated and differentiated there was a wealth of material for experimental mixing and matching. At the genetic level Von Uexküll’s piano metaphor applies: the same set of keys can be played in a simple linear series as a scale, or in a huge variety of chords and tempos to become a concerto. If we were to take protein domains as the keys the keyboard would be very simple. If modifier genes as well as structural genes are taken as equivalent to the piano keys, many varieties of epigenesis are achievable. If the differentiating process involves the mixing and matching of exons, to produce biochemical novelty,

From the Particular to the General

355

the possibilities increase greatly, though the final products in the form of novel proteins are constrained as noted above. Intragenic shuffling of the existing pack of genes can also introduce some novel combinations, but the list of known examples remains quite brief. During epigenesis, mixing and matching of cells that influence development have been involved both in homeorhesis and in experimentation with epigenetic novelty. Implicit in genetic and epigenetic combinations is a canalizing algorithm that usually makes the organism true to type, but occasionally changes to produce experimental forms. At the organism level, mixing and matching of symbiotic proto-bionts, prokaryotes and eukaryotes have produced new types, sometimes with sufficient impact to found new ecosystems. By the same principle, at the ecosystem level, communities, and more intimate co-evolutionary relationships have emerged. The Acquisition of Heritable Characteristics Gene acquisition from extrinsic sources is a special case of mixing and matching. Natural transgenesis is so common in prokaryotes it makes phyletic relationships between their ancient ancestors highly complicated, and guarantees their descendants resistance to antibiotics. The acquisition of heritable characteristics that was part of endosymbiosis. was a major advantage to eukaryotes, but at that point their ability to get genes from foreign sources was greatly diminished. While retroviral gene acquisition has occurred in eukaryotes, it has not, as far as we yet know, been a major influence. If the acquisition of transgenes was a primitive source of evolutionary novelty, it has become redundant in the higher social animals, which depend on intelligence and learning. Self-Organization My readers would, I hope, agree that reproduction and self-maintenance were the fundamental characteristics of life, and that complexifications of those emergent properties the most central processes of evolution. Conceptually, self-organization could be equated with self-maintenance in the most primitive organisms. But as soon as something like a cell cycle emerged, self-organization became a distinctive feature that was also subject to emergent evolution. With the emergence of multicellular organisms, self-organization from a single cell, or small number of cells, to a differentiated condition became one of the most visible modes of evolution, most strongly linked with epigenetics. The following categories of evolutionary process are subsets of this one. Heterochrony Once the foundations of the structural genome were in place, and the switching mechanisms of the epigenome and its associated regulatory proteins wired in, there

356

Chapter 9

still remained a considerable degree of adjustability for the final form of the organism through the chronological rearrangement of switching, and consequent alteration of expression of the structural genes. Any organism that changes during its life cycle, including a prokaryote, is subject to the effects of the passage of time, and to associated intrinsic and extrinsic changes. Therefore, heterochronous changes are virtually universal natural experiments in evolution. Like Bateson, Goldschmidt, and Riedl, we might find it difficult to conceive of a complex organism changing at its lower hierarchical levels without destroying the upper ones. But we can evade Fisher’s total denial of the possibility by perceiving that complexification occurs in the dimension of time, and enjoys the pre-existent adaptability of the organism. In other words, the individual has the opportunity to respond appropriately throughout its development. Thus, in addition to recombinant variability, sexual reproduction has imposed development from a fertilized egg on every generation—providing for evolvability. Some timing alterations in development are made possible by repetitive differentiation. More proteins, for both construction and function, can increase rate of growth and more ribosomal RNA can get developmental processes off to a faster start. Heterochrony is also effected by environmental change, and the trigger can become internalized in epigenesis to result in greater homeorhesis. Allometry and Orthogenesis This could be regarded as a category of heterochrony, with strong causal links with repetitive differentiation as set out above. Some molecular events increase the likelihood of recurrence in future generations, thus constituting a kind of mechanistic drive. These include codon amplification through DNA replication slippage, gene duplication, concerted evolution and dosage amplification, resulting from unequal recombination, and corollary mechanisms provided by recombination enzymes and jumping genes. Along with redistribution of cells that affect epigenesis, these contribute to allometric growth changes, phenomena that used to be collectively called “orthogenesis.” Many instances of evolution by so-called directional selection are probably manifestations of allometric shifts caused by epigenetic drives, or by epigenetic constraints that allow change only along a single pathway. Orthogenesis is not a random process subjected to natural selection, so it can be carried to the point of ultramorphosis and extinction, since all the members of the phyletic lineage that possess it are driven in the same direction. But most known cases of allometric shift in the lineages of surviving organisms stopped before they went that far. Orthogenesis fits the category of an autonomous critical-point emergence, although it may proceed rapidly enough to appear to be saltatory. If it is adaptively neutral, the lineage in which it occurs will continue to possess it, and the gene and protein family will continue to be driven to further multiplication. This is the most interesting case,

From the Particular to the General

357

since it solves the problem exposed by Mivart, and worried over by Darwin, about the absence of adaptiveness in the incipient stages of evolution of complex functional anatomies. There is no question about the reality of this phenomenon or about its significance for anatomical diversification. Orthogenesis to ultramorphosis and extinction is a possible outcome, but most of the instances of allometric growth with which we are familiar stopped before the terminal stage of exaggeration was reached. Penetration of New Environments Another very general rule is the origin of novelty at the edge of chaos, or, more explicitly, under unstable conditions. The mechanisms vary, but the principle is the same. Innovatively adaptable organisms emerge at stressful environmental interfaces to penetrate into new conditions of life. Epigenetic innovations emerge at developmental thresholds after periods of stable growth. Some organisms that congregate at environmental fringes or interfaces may interact epigenetically or symbiotically. Populations expand in a stable manner until internal overcrowding or physicochemical limitations at the boundaries disequilibrate them. As a consequence, physiologically and behaviorally adaptable organisms may individually or en masse remove themselves to new environments. Speciation is a subsequent event. A variety of names have been applied to the rule that has just been outlined. Mivart’s “intermitting conditions of stable equilibrium” would have logically covered all of the implications, including saltation. “Punctuated equilibrium” is more succinct as well as more familiar. At a time when a speciation event was considered a pivotal evolutionary occurrence, Eldredge and Gould applied punctuated equilibrium to the relatively sudden origin of species. But the term has a more general theoretical potential that I will apply unless the proprietors object. Once new environments have been penetrated by adaptable organisms, they are more free to experiment with new developmental patterns, and exploratory behaviors that might not simply lead to physiological modifications, but also diversifying, adaptational evolution. This recourses back to heterochrony, repetitive differentiation, allometry/orthogenesis, and also links with the following. Genetic Assimilation Being simply the contingent imposition of physicochemical change in the organism by the environment, physiogenesis affects generative conditions, but is not a “tool” or mechanism of evolution. The genetic assimilation of such change is. As a general principle, internal adaptation of enzymes and organs to prevailing physicochemical conditions limits the survivability of an organism if those conditions are susceptible to change. But general adaptability to respond to environmental change finally produces an effective homeostasis that subsequently remains stable and persistent. Genetic assimilation begins with the internalization of contingent changes. Instead of

358

Chapter 9

being susceptible to an environmental cue for an epigenetic event, the organism produces its own trigger, which may be gene-based and thus may persist as an epigenetic mechanism. The ostrich, for example, develops appropriate calluses prior to frictional contact with the ground. Or if a dormant gene can be advantageously turned on by a drastic temperature change, variants of the gene that turn on spontaneously might have greater survival potential. The appearance of new genes through exon shuffling and tandem repeats, together with new proteins through domain mixing might significantly increase such adaptations and adaptabilities. But emphasis on gene contribution to genetic accommodation oversimplifies something that must involve the modification of existent mechanisms of epigenetics, physiology and behavior. The principle of genetic accommodation addresses how flexible behavior can become stereotyped. How then does an organism escape its instincts and progress to freedom of behavioral choice? The answer would seem to be an allometric expansion and reintegration of whatever parts of the nervous system can override the gene-determined behavior. Fundamentally, the organism does not, through natural selection, evolve solutions to problems set by the environment. The organism determines what is adaptive through its behavior, which is in turn substantiated by its physiology. Phase Transitions At the beginning of this chapter, following the examples of Kauffman (1993) and Goodwin (1994), I cryptically stated: “Where the generative conditions are sufficient, saltatory emergences may occur spontaneously through self-organizing phase transitions.” The concept is based on a physicochemical model. Under decreasing temperature an amorphous liquid phase of water will suddenly turn into an ordered crystalline phase of ice, sometimes speeded up by the presence of nucleating agents. Under increasing pressure, a colloid gel will go into the sol phase. The idea of such critical-point phase shifts might seem to apply to emergent evolution. The difference is that water and gels do not suddenly acquire the ability to change phase. They are fundamental physical qualities that are immanent in water and colloids. The same is true of an apparent phase shift from an aggregation of cells to a multicellular organism. It has long been known that if the cells of a sponge or a simple cnidarian are sifted completely apart they will reassemble themselves into something like the original animal. Therefore a disordered suspension of cells undergoes a “phase shift” into a self-organized organismic entity. But the cells already possessed the ability and the experience to form themselves into multicellular units. The same is true of the shift in the slime mold Dictyostelium from the unicellular ameboid phase to the multicellular fruiting body phase. Intercellular contact is the trigger for an epigenetic algorithm that already existed. The interesting question is: was the potential for multicellularity and alterations in cell products initially a non-heritable consequence of

From the Particular to the General

359

cell contact, which was then genetically assimilated. The simplest multicellular associations were sheets and spheres that might have formed with only the simplest of cell adhesion mechanisms. Once they existed, they must have been sufficiently exposed to varied environmental factors to autonomously begin differentiating. A biological model that complexity theorists like is the change in behavior of ants into a more recognizable rhythmic order when their population density increases to a critical point that can be defined as x ants per square centimeter, a system mathematically modeled by Cole (1991). Now, this built-in feature is not evolved anew in every generation of ants. Was the acquisition of ordered behavior a critical-point emergence at some time in the ants’ evolutionary history? The question is answerable through the experimental manipulation of population densities in the solitary relatives of social insects, and so is not totally speculative. The expression “phase transition” can be metaphorically applied to the threshold transitions in ontogenic and emergent evolutionary changes of state discussed by Balon, by Müller and Wagner, and by Holland. But it forces the metaphor too hard to apply it to cases of critical-point evolution such as the transition from non-flight to flight—where no flight had existed before. It is an impressive sight to see a flock of starlings feeding randomly on the ground, and then take to the air like a superorganism: “phase transition” would seem superficially appropriate. But it is inherent in the nature of starlings, not a novel emergent feature. When flight first evolved it too was relatively sudden. When allometric growth increased the size of wings there was a point in the life of individual proto-birds when true flight occurred. But this goes beyond the point where the metaphor of phase transition has any explanatory value. When I said that saltatory emergences may occur spontaneously through selforganizing phase transitions the emphasis is on the autonomous emergence of novel features, not on the shift to a phase that was already inherent in the physicochemical generative conditions. I might have been wiser to have avoided the metaphor altogether, but leave this section intact, since it is such a popular notion, even among the cognoscenti. Contingency-Dependent Emergences If you scan the headings in the previous list of tools and techniques of natural experimentation, you will detect a bias in favor of endogenous or intrinsic mechanisms. Also important are cases where the action of the organism has a feedback effect from the environment, and where accidents of nature are also causal. On many occasions the origin and success of intrinsic emergent novelties, including some of the threshold conditions frequently referred to above, has depended on such contingencies. Being in the right place at the right time is particularly apt when random climatic, tectonic, and catastrophic changes occur. However, physiological and behavioral adaptability is prime catastrophe insurance, even among the lucky ones. Some external conditions

360

Chapter 9

can be controlled, in the sense that the organism can choose to expose itself to them or avoid them, such as staying at home in the sea or venturing into fresh water. In addition, some chemical geophysiological environmental changes, such as the oxygenation of the biosphere, resulting from the action of emergent organisms, created new conditions that would affect subsequent evolution. Contingencies are part of the history of emergent evolution. For example, potential symbionts that had followed their own independent courses until the moment of symbiotic truth had to come together for long enough in sufficient numbers to reach that threshold. They had to be in the same place at the same time. Moreover, they had to have retained sufficient genetic compatibility since their original divergence from common ancestry. Modularity of molecular, and cellular structure was also significant in successful physiological emergences. Disequilibration The preceding note on contingencies emphasizes the importance of environmental change for the direction of emergent evolution. Environmental destabilization may also reduce the agents of natural selection to a point where evolutionary experiments have the freedom to diversify, prior to the inevitable re-establishment of ecostasis. Disequilibration of homeorhesis makes epigenetic experimentation possible. Destabilization of the internal milieu caused changes in primitive physiologies and presented challenges that only improvements in adaptability could meet. To what extent, then, is disequilibration important as a general evolutionary principle? It is certainly of great interest to complexity theorists. In physicochemical systems such as gases and fluids, temporary points of stability such as vortices, standing waves, and local concentrations, are highly vulnerable to destabilization, though order may be imposed by a phase shift such as crystallization that spreads through the system. However, these shifts to more stable systems are reversible to more chaotic states. In contrast, each emergent stage in progressive evolution “collapses chaos” as Cohen and Stewart (1994) phrase it. In other words these are changes in order that resist further interference. Greater degrees of physiological homeostasis also make the organism more adaptable, more able to modify itself in ways appropriate to new environmental conditions. This is not Lamarckian progressionism: many lineages have shown little progressive change since they first emerged. I have also shown that advances in adaptability can also regress into states that while less adaptable, conform better to the limitations that ecostases impose on the availability of resources. There may seem to be a paradox in the concept that homeorhetic and homeostatic stability are products of emergent evolution. It is logically balanced by the realization that they are accompanied by greater adaptability, and therefore freedom of choice. And in some animal lineages that means freedom of behavioral choices that can overcome tradition, both biological and social. At the latter level, a degree of chaos is bound to return.

From the Particular to the General

361

Closing Note on Baconian and Darwinian Epistemology Francis Bacon had an intuition that the chaos of human society required Baconian ordering in the form of a hierarchical division of labor—under the rule of philosopher kings and their intellectually adventurous ministers. For him, there was only one way to advance science—his way. Even his “interpreters of nature” were not permitted to fly up to “remote and most general axioms” and then apply their “supposed unshaken truth” to particular observations and experiments.27 Despite Darwin’s claim to espouse Baconian principles, he took the other way, and his descendants have been doing it ever since. If Bacon had always been followed dogmatically we would scarcely have emerged from the seventeenth century. But doctrinaire adherence to Darwin has stuck us in the nineteenth. Biology in general follows erratic paths between the two. Consequently, sufficient particulars and minor axioms of evolution have accumulated to allow us to ponder once more a synthesis that does not depend on the truism of natural selection. I trust that this makes the case that I am no more a Baconist than a dialectical materialist. If the former, I would have stuck more strictly to Bacon’s rules of tabulation, which require recording instances of the presence of a phenomenon; instances of the absence of a phenomenon; and instances of change in the phenomenon. Although all of these instances are in my brief, to stick to them too formally would result in the “long-windedness” that he affected to despise. Similarly, in the next chapter I am going to refer to a dialectical synthesis that does not conform strictly to Hegelian or Marxist doctrine. For example, what I will call “the thesis” is already a complex of complementary, not contradictory ideas. In the past they remained dissociated because most intellectual nets were not cast widely enough, or the few exceptions were treated with silence and contempt. Also, biology has become a collection of splintered subdisciplines within which are plenty of pet concepts or simply just metaphors that masquerade as explanations. However my thesis of complementary ideas must finally be united, if not in total harmony, with an antithetical neo-Darwinism. A harmonious unification of opposites is a contradiction; for if opposites can be unified they were only reified as opposites in pre-synthetic minds. The punch line is that both Baconian induction and dialectical materialism are epistemological procedures that offer inductive guidance, and as such they deserve the attention of students of biology. Any methodology can become a Procrustean bed that requires the amputation or racking of thought until it fits.

10 An Emergence Theory

The state in which we now see all the animals is on the one hand the product of the increasing composition of organization, which tends to form a regular gradation, and on the other hand that of the influences of a multitude of very different circumstances that continually tend to destroy the regularity in the gradation of the increasing composition of organization. —Jean-Baptiste Lamarck, 18091 [My concept of a] limited and contingent progress is very different from the deus ex machina of nineteenth-century thought, and our optimism may well be tempered by reflection on the difficulties to be overcome. None the less, the demonstration of the existence of a general trend which can legitimately be called progress, and the definition of its limitations, will remain as a fundamental contribution of evolutionary biology to human thought. —Julian Huxley, 19422 Transformation from the extremely unlikely to the likely is a major characteristic of systems exhibiting emergent phenomena. Even when the simplest persistent patterns are infrequent in a generating procedure, they will eventually occur if the system runs for any length of time. Once they occur, they will by definition persist, making them candidates for combination with other, persistent patterns (other copies or variants). At this point larger patterns with enhanced persistence and competence can occur. . . . The usual argument that evolution requires long sequences of improbable discoveries, and so is “slow,” misses this point. The unlikely will become likely if one allows for a layered series of generating procedures. —John Holland, 19983

Evolutionary progress, as I have presented it, results from episodic emergences of increasing, self-organizing, organismal complexity. Progress is induced and guided by various intrinsic and extrinsic causes, among which particular actions of organisms are important. Evolutionary emergences may be unpredictable saltations, or novelties that appear at thresholds in continuous series of change. For Lamarck, however, progress was gradual and continuous. Where gaps occurred, they were caused not by saltatory emergences, but by environmental obstacles and adaptational distractions. Since Lamarck lived in interesting intellectual and political times—in the milieu of the

364

Chapter 10

Enlightenment, the Industrial and French Revolutions, and not least the Fiber Theory—one might have expected of him a more revolutionary concept of evolutionary progress. Perhaps evolution alone was revolution enough for Lamarck. In place of divine law he called upon an inherent and steady progressive trend, the deus ex machina to which Julian Huxley subsequently referred in Evolution: The Modern Synthesis. Like Lamarck, Huxley wished to explain how organisms had evolved from the simple to the complex. He thought that natural selection, the ghost that Darwin had put in the machine, as both cause and explanation of evolutionary progress, was inadequate. Some of Darwin’s quarrelsome descendants have even greater difficulties with evolutionary progress, and they often reject it as a delusion. But here we are, discussing it! And there are our bacterial ancestors, who aren’t! It’s not vainglorious to say that we are different and more complex; to pretend otherwise is anti-intellectual. An emergence theory must therefore accept the reality of progressive evolution, and explain its mechanisms and processes. Huxley’s approach was to amalgamate all the pertinent aspects of biology, and he did a much better job of the Modern Synthesis than the American version, which had a hard enough time simply pulling population biology and genetics together with paleontology, and then losing the latter in short order. Huxley also asked how some organisms could improve their general adaptability, in contrast to organisms “armoured against progress”—locked into stasis by their specializations. Huxley thought the adaptable, persistent types might advance further if liberated into a more varied environment.4 He did not go so far as to say that they were being liberated from natural selection. Nevertheless, his impatience with adaptationism, and enthusiasm for progress as a product of developmental evolution led finally to his exclusion from the hagiography of the Modern Synthesis—he is barely credited with giving it its name. John Holland (quoted in the third epigraph) has the goal of finding the first principles of emergence in simple systems that can be mathematically formulated. He is concerned with persistent patterns that complexify by new combinations of variant building blocks as the system increases in size. Therefore his thesis emphasizes autonomous emergence by repetitive differentiation. Yet it does not exclude extrinsic causes that might effect epigenetic events and genetic mutations. All of this is relevant to progressive evolution. And although Holland does not refer to progress per se, it can be inferred from his remarks about complexity increasing through reorganization. It is also implicit in his statement that requiring evolution to be a long, slow sequence of “improbable discoveries” misses that point. At this juncture, I share Julian Huxley’s goal of clarifying progressive evolution, while I reject natural selection as its cause. In this chapter we will reach for a theory of emergent evolution by integrating the subtheories surveyed in the previous chapter. A dialectical synthesis that would accommodate both emergent evolution and the

An Emergence Theory

365

hypostasis of dynamic stability is desirable, and possible—but only at the higher epistemological level of the history of life. That will be attempted in the final chapter. In the meantime, we must not reject the selection syndrome outright. For one thing the consequences of progressive emergent evolution penetrate the early stages of re-equilibration as adaptational evolution leading to specialization and ecostasis. Moreover, we must take the selection syndrome into account as a prevalent constraint in the history of life, and consider how dynamic stability provides a pause during which sudden innovations can shake down into smoothly interacting modules and wholes. Then we can further explore how its obstructiveness has been periodically overcome. We have attended performances in the causal rings of the evolutionary circus, exploring association, physiology, behavior and development, and we have seen how they leap and swing and interweave in patterns under the big top of the environment. Emergent evolution has been sorted into categories of intrinsic and extrinsic causation as well as saltatory and critical-point processes. And I have summarized their mechanisms, along with the contingencies that affected their course and enhanced their operation. Now for a recap of universal features, before a deeper plunge into the dim depths of formal theory. Important points that have been discussed earlier are interpolated in brackets, to remind us that there are subalterns under the generals. Then the “theories” of symbiosis/association, physiology/behavior, development and environment can be integrated into a more natural system. Evolutionary emergentism implies discontinuity in the generation of novelty. Even a mutation that gives rise to a minor phenotypic variation is discontinuous. Emergences may occur at critical points in a continuously evolving allometric series, suddenly bringing functional novelties into play (gliding to flight). Saltations may emerge from a spontaneous coming together of appropriate generative conditions (origin of life; endosymbiosis; sex; multicellularity). Or they may be due to repetitions and rearrangements of existing conditions (codon, exon and gene duplications and differentiations; exon and protein domain shuffling; gene transposition and conversion; chromosome mutations; homeotic shifts). Divergent saltations may leap in new directions from generative thresholds, at intermediate stages in established continuities of embryonic development (radical deviations, larval digressions that return to the adult norm, innovations arising from paedomorphosis). Emergences may be largely intrinsic (self-generating, or autonomous) or extrinsic (imposed by or responding to the environment), but hardly ever exclusively one or the other. Their persistence is often characterized by adaptability, multifunctionality, and ontogenic flexibility.5 And these amount to an integrated novelty—a whole that is greater than the sum of its parts (quaternary protein structures, contingent interactions of molecules or physiological systems, symbioses and societies). Emergences are generated by natural experiments involving mechanisms of repetitive differentiation, mixing and matching, heterochrony, exploratory behavior

366

Chapter 10

(both developmental and organismal), physiological and behavioral responsiveness to the environment, and unpredictable external contingencies (physicochemical, climatic, catastrophic impacts). When molecular biologists talk about “episodic evolution” there is usually duplication of a gene or a genome at the heart of it (growthhormone genes, genome duplication and reduplication in early vertebrates). These are crucial evolutionary processes, but the emergence principle applies at all the higher levels, from organogenesis through organismal physiology and behavior to the interactions of societies and communities. Emergences are episodic, lacking periodicity and hence predictability. But are they altogether random? Darwin assumed total randomness in most biological change, because omnipresent, non-random, natural selection would ensure that adaptive random changes were saved and reproduced. (Moreover, selection theory would also be saved and reproduced!) Emergentists, unlike Darwin, are not trying to impose randomness of evolutionary change in order the magnify the importance of natural selection. We are more directly concerned with the causes of evolutionary change, and therefore able to recognize that many molecular changes are non-random (hypermutability, transposable elements, self-amplification), and with how the extremely unlikely becomes the likely. Moreover, the significance of random change depends on where and when it happens. If it occurs in the context of appropriate generative conditions it may catalyze an emergence with a constellation of novel qualities—a whole that is greater than the sum of its parts. For example, a purine synthesized in an interstellar dust cloud might persist, yet would be unlikely to participate in a biological emergence. In contrast, a purine synthesized in a wet place containing other abiogenic organic molecules is a lot more likely to be involved in some kind of emergent system. I will discuss this further in the context of Holland’s “layered series of generating procedures.” Once biological emergences appear, they fall into dynamically stable organismal and ecological states that resist change. Paradoxically, the higher the level of emergence, the easier it becomes to escape stasis. This is because of greater freedom of choice for the organism, and more alternatives of habit and habitat from which to choose. My zoological bias is again disclosed by this statement, but the same applies to plants to a less obvious degree. Though neglected by reductionist evolutionists, the bootstrapping interaction of development, association, physiology, behavior and environment has been to the fore in animal evolution. As a result of such feedback, internal stasis can be disequilibrated from without. Also, ecostasis can be disrupted by large-scale emergences, avoided if the emergents move to unexploited environments, or simply eradicated by catastrophe. Saltations can leap the barriers of stasis. The clearest examples of truly saltatory emergence are to be found in the associative arena, with epigenetic deviations following close behind. Coalescence of self-reproducing proto-bionts generated the

An Emergence Theory

367

first self-repairing and self-organizing cells. Endosymbiosis and sexual association, which were the next major emergences, brought an array of physiological and behavioral changes with them, even at the unicellular stage. Principles of association and epigenetics apply to the differentiation of multicellularity, another foundation of emergent evolution. The disadvantages of independence are turned to the advantage of the whole. Thus, improved wholeness, because it is more artful at getting out from under the agents of dynamic stability, is a formal characteristic of saltatory emergence. Although, for the moment, epigenetic innovations are subsumed as part of the potential provided by multicellular associations, emergences of the distinct body plans of the animal phyla deserve particular emphasis, and these also illustrate saltatory and critical-point processes. Associative emergences have had the most significant biotic impact on geophysiology, initially through the oxygenation of the biosphere. Once they invaded the land, symbiotic fungal-algal plants constructed soil, extended their range, and affected global albedo and climatic patterns. For emergent plants and animals, novel physicochemical conditions had immediate direct impact on functional morphology. Symbiosis continued to have major consequences. Photosynthesizing symbiosis, in conjunction with rigid plant cell walls, effected progressive ecosystematic changes and raised oxygen levels in the air. Cellulose became the foundation of a new trophic pyramid once animals acquired symbiotic micro-organisms to digest it. Plant evolution extended diversity of habitat for animals, and correlated changes in animal habits affected their morphology. However, the greater their physiological and behavioral adaptability, the more they could take advantage of the full range of available environments. The last great emergences, intelligence and mind, were products of developmental, physiological, behavioral, and social evolution. Some elements may have simply appeared at critical points in the expansion of the cerebrum, the increased capacity of its neurons to form new dendritic connections, and repetitive variation of chemical messengers that allowed the brain to distinguish between new functions. Nevertheless, epigenetic algorithms, orthogenesis and constraints, directional selection, and hypermorphosis are insufficient explanations. For example, the allometrically expanding hemispheres of the cerebrum benefited from a new link, the corpus callosum. That integration then bootstrapped further hypermorphosis, since the two hemispheres could become more differentiated without being of different minds. Innovative integrations of brain areas for logic, language, memory, vocalization, and aesthetic sensibility were part of the emergent constellation. To be meaningful they had to be connected with hand-eye coordination. To be fully effective they had to be able to override older, hereditary, behavioral mechanisms. This required further reorganization and bypassing of existent neural connections. The emergent result was greater freedom from gene determination.

368

Chapter 10

The evolution of this web of dynamic structures occurred in a broader environmental context. Big-headed babies could develop fully and survive only if their mothers were well nourished, so they needed extended families. Putting it another way, extended families were part of the generative conditions that allowed big-headed babies to become successful monsters. Once they survived, they needed social interaction to continue their neural complexification to the point where its adaptability could be fully exploited. Unfortunately, one of the larger consequences of socialization is a renewed assault on geophysiology, through deforestation, greater carbon dioxide production with its greenhouse consequences, and emissions that destroy the ozone layer. Geophysiology will always be in some kind of homeostasis, but not necessarily one that is kind to interfering humans. An Emergence Theory holds that evolution is emergent, and explains its generation, and its qualities. It might emphasize intrinsic or extrinsic causes, but does not stray too far from a holistic, or interactionistic evaluation of evolution. Nor does it simplistically characterize evolution as either saltatory or gradual. In resolving these apparent contradictions an emergence theory shows dialectical competence. The generative conditions of emergences combine interacting associative, physiological, behavioral and developmental causes. Post-Lamarckism (with its emphasis on the causal nature of external factors, but without the inheritance of acquired characters) is integrated. The syndrome of secondary causes and effects known loosely as natural selection is interpreted as barely adequate to simple adaptational evolution, but more importantly as an ultimate obstacle to the evolutionary diversification of subsequent emergents. Phenotype sorting, differential survival and reproduction, and “agents” such as competition, predation, and sexual selection are recognized as the consequences of emergent evolution. Nevertheless, phases of intrinsic and extrinsic dynamic stability provide time for the shakedown of emergent novelty. Selection theory is not only secondary to evolutionary causation, it also lacks historical priority. That falls to Lamarck, who for all his inadequacies and errors saw progressive evolution as the thing to explain. And priority over neo-Darwinism falls to the neo- (or post-) Lamarckists, who not only addressed the right issues, namely generative causes, but gave them a far more holistic environmental context. Does Emergentism Provide an Adequate Theory of Evolution? That evolution is a historical reality was proposed by Lamarck and established by Darwin. Whatever doubts remain about the chronological details and phyletic affinities, the historical theory has become axiomatic—not “just a theory” (which many non-scientists or non-philosophers equate with “just a speculation”). It is the theory of causation, or the evolutionary mechanism, that is at stake. Emergentism might be colloquially referred to as a “theory,” but to really qualify it would not only

An Emergence Theory

369

need to increase the understanding of the relevant phenomena. It would also be expected to make an inspiringly simple, formal statement about its nature; preferably mathematical or at least translatable into something quantifiable. It should give some indication of where to look for mechanisms and how to experimentally test them. A formal theory of emergent complexification would have to address not only generative causes, but distinguish between the two different processes of saltatory and critical-point emergence. Then there is the category of haphazard contingency, which includes a multitude of feedbacks from environment to organism and back again, and includes the evolutionary significance of catastrophe. I have been ignoring the applicability of emergentism to non-biological phenomena, partly because of my bias as a biologist and partly to keep matters simple. Most current theoretical studies of emergence do not involve biological systems but presume to infer biological significance. The emergence of biological reproduction was a major discontinuity. It also, with the corollary qualities of self-maintenance and selforganization, established the novelty of biological evolution. Therefore, I have always balked at making the emergences of life an extension of cosmological evolution, in the original sense of “unfolding” or “unrolling.” I suspected that a try for an Emergence Theory of Everything could effectively produce an Emergence Theory of Nothing, especially if physical principles are expected to be its generators. A catchy name that purports to save all the appearances does not make a theory, whether it be “emergentism,” “synergism,” “natural selectionism,” or Lamarckism qua “the inheritance of acquired characteristics.” Yet, during the time I have spent on this book I have come to the opinion that a complete synthesis should accommodate all natural phenomena, from the cosmological to the cognitive. Several complexity theorists who understand the importance of emergence have tried for mathematical abstractions that would apply from physical phenomena to biological evolution. It can’t be done, because new rules appear with every new emergence. In Signs of Life (2000), Ricard Solé and Brian Goodwin effectively demonstrate how mathematical descriptors of emergent systems can be achieved. But a mathematical descriptor no more explains how an emergence originated than a landscape painting explains how the landscape came to be. Both may provide clues to the unknown, and both have large audiences that are comfortable with the techniques, and enthusiastic about the results—consequently more minds are ready to invest in the subject. However, there is one proven mathematical approach to the generation of change, with which many modern biologists are familiar: the transformational theory of D’Arcy Thompson. The coordination grid diagrams from the last chapter of his book On Growth and Form (1917) are still frequently reproduced to demonstrate that change of overall form, in skull structure, for example, fit mathematical formulae. Therefore their generation is either driven in part, or at least severely constrained by mechanical forces. That seems to indicate how a mathematico-

370

Chapter 10

reductive approach can successfully be taken to critical-point emergence. D’Arcy Thompson was himself wary of oversimplification—he admitted his omission of the essential biological mechanisms that actually did the building in obedience to the mechanical constraints and forces. The more you try to dissect emergent evolution the more it holds together—no theoretically dominant reduction, mathematical or otherwise, pops out. Furthermore, for a comprehensive theory, due attention needs to be paid, not only to all the parts, but also to the exceptions and contradictions. Those are what mathematical formulations are likely to miss. The best theory must model a web of related causes and effects and must show how emergent features originate, how they affect all the other causal strands, and how they all connect to the biospheric whole. Emergentism does not have to qualify technically as a theory to demand that more attention be given to the generation of novelties, and less to their subsequent demographic fate. That would certainly clarify evolution—if shifting to the relevant from the irrelevant matters. Theoretical Potential A simple way of testing the theoretical potential of emergentism is to ask how it would change traditional interpretations of evolutionary history. The conviction that evolution is a real process that followed a particular historical course was instilled by evidence originally set out in The Origin of Species. Much of that evidence has proved incontrovertible. Early molecular biological evidence that partly explained the origin of variations was added to what had already been concluded from functional morphology and embryology, but awkward questions about gene expression were ignored. Now, if the completion of the human genome project has demonstrated anything of value to evolutionary theory, it is that knowing the genes is not nearly enough. Some of us were already aware of that before the project was undertaken. We do not need a new theory to tell us that that evolutionary history was discontinuous. The evidence has been staring us in the face ever since marine fossils were discovered by the ancient Greeks in strata high above sea level. Later, demarcations between the strata could be seen to have been imposed by sudden catastrophes or rapid climatic changes, as Cuvier said nearly 200 years ago. We can also surmise that discontinuity was generated by autonomous evolutionary emergence as well. One of the consequences, in vertebrate evolution at least, was a growing potential for behavioral adaptability that had a positive feedback effect on further emergent events. All genetic molecular changes are saltatory, many are non-random, and some are selfamplifying. The molecular biology of embryology and homology also supports saltatory epigenetic interpretations. And the more discontinuity of evolutionary history is illuminated, the dimmer becomes the light of Darwinian gradualism. The fossil record, and the diversification and numerical distribution of modern organisms, suggest which emergences have actually been most significant—most

An Emergence Theory

371

analysts settle for life, sex, the eukaryotic condition, epigenetic complexification, and mind. Progress to new emergent levels, and the components of advances in adaptability, can be tentatively identified. Artificial selection, a bulwark of the Darwinian historical theory, is a better model for the performance of emergent novelty in the absence of natural selection than it is for the creativity of natural selection. Thus, some of Darwin’s original evidence can be removed, some reinterpreted, and some reversed. The superstructure of the Modern Synthesis could be renovated out of nostalgia, but a new theory is needed to integrate new historical and causal interpretations of emergent evolution without the stultification of selectionism. How much can we get out of an emergence theory at its present inchoate stage of construction? It proposes that the first living cell, with the emergent properties of selfmaintenance and reproduction, contained the potential for progressive evolution. Emergent novelty could suddenly appear at critical points or thresholds in allometric evolutionary continuities, or through qualitative saltations. Preservation of integrity at every stage is a given, but that quality alone gives no guarantee of survival or success in the face of strong competition and predation. Therefore emergent evolution was unlikely to succeed unless its products had immediately advantageous properties that made the old competition irrelevant. (Recall that such properties are inherent in the emergence, not created by prevailing extrinsic circumstances—though perhaps, by chance, appropriate to them). Alternatively, emergent properties allowed the discovery of environments where agents of natural selection were absent, or had been removed. Hence the congregation of adaptable organisms at environmental interfaces was important for evolutionary progress. These circumstances were pivotal, but so rare that successful progressive changes were brief evolutionary squirts, interspersed with periods of diversification, and longer periods of dynamic stasis. When major emergences occur they bring with them new rules of operation that can only be predicted in very general ways from a knowledge of lower levels of emergence. That simple life could become more complex is predictable enough. However, to abstract principles of complexification to a new emergent level, theory requires the canny approach taken in Holland’s Emergence: Order from Chaos (1998) and by Solé and Goodwin in Signs of Life (2000). Following that route it might ultimately be possible to write an epigenetic program for the developmental complexification of a specific type of zygote. Remember, however, that a complete description of the process in an animal with a completely charted genome, such as the nematode worm Caenorhabditis elegans, does not explain it. It takes us part of the way, since it might suggest how new algorithms might have emerged from the epigenetic nodes. Yet such abstractions are complicated by environmental, associative, physiological, and biochemical variables. The algorithms in question operate through different factors at every hierarchical level—environmental or organismal—as well as through the genes and their protein products; however, the algorithms are not determined by

372

Chapter 10

genes. Components that do reside in genes have been historically instructed by the organism and its environment. Whatever information the genes might consequently have, they need to be told what to do by higher organismal levels. The basic instruction embedded in an algorithm is “If this is the current situation, then that is the necessary response.” And for an organism to do “that,” it needs more than the genes; it needs modifiers, signals, functional morphologies, and experiences that constitute “this.” Furthermore, the constructive impact of “this” is accompanied by negative effects. For example, progressive DNA repression, methylation and histone binding constrain how much of “that” might be possible—a necessary means of separating organismal offices prior to the refinement of their concurrent efforts. This ensures that the ontogenic whole is not only greater than the sum of its parts, but quite different from the whole that constitutes the early embryo. The course of cell-line differentiation and the dynamic form of the mature organism are results of non-linear processes. Their evolutionary pathways cannot be predicted from a simple generalizing principle. Even if they could, the routes of their progress are influenced by contingencies that are themselves partly to be expected and partly a matter of pure chance. Thus, the parsimony of “It’s all in the genes” is as spurious as that of “God did it.” Sidney Brenner, who originated the C. elegans studies, has come to understand that the genetic information is not enough, and that theoretical biology must deal with the flow of that information.6 To do so requires a holistic/interactionistic ability to integrate all of the factors operating at every hierarchical level—and a practical as well as theoretical facility with biology. Apart from the materialist’s truism that life has no programmer, there are interesting parallels between progressive improvement of the design of mechanical artifacts, and the evolution of organisms. Electronics, for example, followed a progressive sequence of experiments punctuated with key discoveries like an alternating current electricity supply, servomechanisms, diodes, transistors, and microchips. Between inventions, societies explored their uses and exploited their benefits, although some novelties that defied application were shelved for the time being, and then dusted off when their potentials were recognized. These engineered emergences provided both economic and theoretical boosts, and their uses often diversified widely. The electronically engineered future is somewhat predictable—high-temperature superconductors, biochips, and fuzzy-logic computation are being pursued. To stay with the engineering parallels, biological evolution progressed to becoming greater wholes by acquiring add-on parts, through symbiosis, and through multicellularity. Once there, surplus capacity made redundant components available for retooling at the molecular and organismal levels. Although the analogy takes us part of the way it is insufficient to clarify biological evolution. There is still a ghost in the organismal machine. For an organismal whole to become greater than the sum of its parts it must be progressively self-engineering. A creature with a multiplicity of parts needs a hierarchy of

An Emergence Theory

373

regulators to become an organized complex—first, promoters of protein synthesis that respond algorithmically by binding peptides and steroids; next some kind of hormonal system that produces those activating and inhibiting molecules, and so directs epigenesis, and signals between cells in the mature organism. Then (in the case of animals), a central nervous system to provide rapid and dedicated communication. These kinds of complexifications yielded greater functional efficacy and energetic efficiency. There were key innovations along the way, such as the myelin that insulated the wiring and speeded up signals without being anatomically unwieldy, and the emergence of migratory neural crest cells that complemented these and other functions. And then there came rerouting and retrofitting to bypass primitive command centers without interfering with their essential functions. The emergence of the corpus callosum connected the increasingly isolated cerebral hemispheres. But the basic physical and biological mechanisms for generating electrical impulses were already present. Neurons are omnipresent in the simplest and most primitive multicellular animals. Nervous systems co-evolved with hormonal systems that worked by diffusion before the emergence of body-fluid circulation. Such systems are always subject to phenotypic modification by the actions of the organism and the environment, as well as genotypic modification. If the parallel is made with an automobile, not only must it be re-engineered while still in motion with “all systems go”; in addition, its modification is directed in part by the route it is taking. There is a pit stop—sexual reproduction—during which the whole vehicle is rebuilt. But although the genetic engine is re-organized and repaired at that stage, it is also maintained and modified by maternal mechanics. They not only provide fuel and spare parts for the early part of the next lap, their quantity and redistribution can cause drastic changes in the phenotypic track performance of the vehicle. The Predictability of Emergences, and the Predictiveness of an Emergence Theory Conventionally, scientific epistemology requires that any formal theory should predict the consequences of a given set of interacting conditions—in particular, the likely results of scientific experiments. Such predictiveness should focus ideas, research and experimental design, whether through practical or thought experiments. Paradoxically, emergences are generally believed to be unpredictable, the reverse of the coin being that wholes that are greater than the sums of their parts are irreducible. One of the stumbling blocks to comprehension is that redundancy and circularity tend to creep into definitions of predictability. For example, it is not helpful to be told that non-linear dynamics are characterized by unpredictable results—then to be told that because emergent effects are unpredictable they must be the products of nonlinear dynamics—thus unpredictability is a sine qua non of emergence. There do exist emergent wholes that can be entirely reduced, but such an exercise does not remove the whole nor its emergent properties from further consideration. In any case, the

374

Chapter 10

putative unpredictability of emergences, and the predictiveness of an emergence theory are two epistemologically distinct issues. It is quite consistent for a theory that deals with unpredictable phenomena to be predictive in the sense that it can guide empirical and theoretical research. Predictability of events, and predictiveness of theory have a tenuous logical connection. Therefore, I deal with both in this section. C. L. Morgan surmised that we could not predict emergences because of our imperfect knowledge of their antecedents. Thus, if we dug deeper and thought harder, predictability might be achieved. However, in the natural biological world, the generative conditions for emergences contain too many variables. And there are other problems. All known biological emergences are historical—we can only make reductive sense of them through hindsight. Second, the nature of future emergences is in the realm of science fiction. There has been no shortage of predictions of that kind, and time and again those of a linear technical nature have come to pass—space ships, artificial satellites, moon landings, space stations, and so forth. The unpredictability of emergent properties is an aspect of their novelty relative to the older, simpler levels of the hierarchy. Physical phenomena such as superconductivity, superfluidity, ferromagnetism, and crystal structure have emergent properties that would not have been predicted from physical principles and elemental chemistry.7 By John Stuart Mill’s example, a complete knowledge of the properties of oxygen and hydrogen would predict the empirical formula H2O, and might suggest some qualities of its gaseous and crystalline phases, but would not predict the “every snowflake is different” aspect of ice formation. A knowledge of the behavior of water’s hydrogen bonds would predict anomalous properties of its liquid state. But it helps a lot if an a priori knowledge of the emergent—water—is smuggled down to its generative level. Ricard Solé and Brian Goodwin (2000) carry the discussion beyond this point by demonstrating that the Navier-Stokes equations that circumscribe the behavior of water are based on its properties of incompressibility, cohesion, and fluidity, none of which are predictable from a knowledge of hydrogen, oxygen, or the water molecule in isolation.8 In turn, the Navier-Stokes equations are incompetent to predict Bénard cells—complex, orderly patterns of fluid flow that arise from changes in thermal gradients. These too can be captured mathematically once they have emerged and have been quantified. However, reducible complexes are not necessarily predictable from physicochemical principles. Some biological emergences are completely reducible. The tetrameric mammalian hemoglobin molecule has the emergent property of cooperativity, which makes oxygen loading and unloading much more efficient than is possible for the four constituent monomers acting alone All of that can be explained reductively, provided the necessity of having an organism to synthesize the molecules is left out. However, there are radical differences between physicochemical and biological emergences. For example, before they ever met in appropriate circumstances, atoms of oxygen and

An Emergence Theory

375

hydrogen already had the propensity to combine as a peculiar liquid at the appropriate temperature and pressure, with all of the properties allowed by the hydrogen bond. Simple molecules of H2O can become more complex structures, such as Bénard cells, icebergs, waves, and whirlpools. But those are temporary states of complexity that are easily undone back to a simple compound or its elemental components. Therefore, although the complexities of water are unpredictable, once they are known they are reducible to physical properties. Therefore, water is distinct from evolutionary emergents that depend not only on changes in the simple constituents but also on biological reproduction, a hierarchical foundation of earlier emergences, and multiple contingencies. Copy error is a major component of emergent evolution that does not apply to non-living systems—unless there is a deus ex machina, in the form of a human experimenter who makes it so. Moreover, if there are any Special Creationists or Intelligent Designers still with us, what, pray, would be the point of a capacity for making mistakes, as opposed to the creation or design of exactly what was wanted? All known evolutionary emergences have already occurred, and Solé and Goodwin have done a creditable job of showing that their qualities can be discovered, captured by mathematical descriptors, and demonstrated not to contradict rules operating at their generative level. The question of predictability of higher levels from the qualities of lower levels needs to be tackled with honest post-predictions that ask whether knowledge of the generative level could have allowed prediction of the new focal level—no smuggling down of its emergent properties allowed! The pragmatic problem for biologists is: how were known emergences generated? Some critical-point emergences seem to be linear and predictable. I refer to selfamplifying allometric shifts, such as the increase in the size of a forelimb that lets it function as a wing with true flight, at the point where lift exceeds drag. It is true that the aerodynamics of flight was not captured mathematically until it had been observed as a real function, but all kinds of predictions could be applied from simple models to untried biological aviation experiments. Here the unpredictability problem would seem to be spurious, but it has to be compounded with the metabolic and behavioral and environmental requirements. Whatever the answer, I am not going to exclude these kinds of innovative critical-point biological emergences just because they can’t be stuffed into the pigeonhole of unpredictability. Some philosophers save them by calling them “weak emergences,” in contrast with strong saltatory emergences. The problem of predictability is compounded by contingencies that significantly affected the course of evolution. However, symbiogeneses, earthquakes, volcanic eruptions, and bolide strikes are probable, if not predictable as to timing. Holland suggests that a long-running experiment in the combination of persistent patterns will inevitably result in their complexification at higher emergent levels. That the unlikely becomes the likely under these circumstance is a predictive generalization, but Holland adds that such effects are unpredictable. As a human organism he

376

Chapter 10

intuitively knew that before he became a theorist. It may be predicted that emergences will keep on happening; it’s the timing and the exact nature of a particular emergence that is unpredictable. Furthermore, their occurrence can be narrowed down from global to local generative sites. A likely location for natural experimental freedom is where one ecosystem meets another, where their constituent organisms can interact. Others are interfaces between physicochemical or biological systems—where clay and detrital surfaces meet the water column—where the organismal epidermis meets the external milieu—where the non-adaptable is torture-tested. At the microscopic level, typical venues are where cells or molecules come into contact at cell membranes, especially synapses, junctions, and intracellular membranes and microtubules. At an even lower level, the surfaces of proteins, the way they fold, and their interactions with water molecules, ions, organic compounds, and each other, have emergent potential. David Rollo (1994) argues that the problem of predicting the emergent stems from the astronomical combinatorial possibilities for multiple holonic variants at the generative level. But the same difficulty would apply to the epistemological process of inducing any effective hypothesis from among an infinity of options. The problem is routinely overcome by lateral thinking, or intuitive inference, and alertness to the unexpected.9 One commentator on emergence tells me that predictability is a red herring, meaning that predictability and non-predictability are unnatural categories imposed by the limitations of logic. If that is so, we should simply ask the obvious questions—Can emergence occur, has it occurred, and will it occur?—and we should anticipate the obvious answer: Yes. The hypothetical possibilities for biological emergence are limited by boundaries of integrity, beyond which disintegration is their likely fate. Yet the following general predictions can easily be made: Predictable Changes 1. Self-maintaining integrity and the ability to reproduce and differentiate are emergent properties of the first living systems. With those qualities they will inevitably become more complex, especially if they are made up of reproducible modules, such as DNA, cells, and organs. This is a post-prediction as far as living systems are concerned, but is predictable from non-biological complexity models. 2. Out of an array of redundant genes, some of their variants will serve novel enzymatic, hormonal and other regulatory functions that increase complexity. 3. There are corollaries to 1 and 2, at and above the gene level: in a holonic (or modular) system that is organized into a hierarchy, change in redundant holons can increase the complexity of order. Some of the older systems may be bypassed or reorganized.

An Emergence Theory

377

4. Simple organisms—proto-bionts, prokaryotes, eukaryotes—will form symbiotic relationships with each other and with more complex multicellular organisms. 5. Unicells will form multicellular complexes, which will ultimately differentiate. 6. Complexity will also be generated by topographic changes that affect the expression of homeotic gene clusters and by movements of embryonic cells that effect developmental organizational changes. 7. Hierarchical layering is both spatial and temporal in organisms. During development the branching and interaction of cell lineages follow algorithms that are susceptible to heterochronous variation. In the mature organism the spatial layers (molecules, cells, organs, whole organism) interact constantly. They will be further influenced by temporal changes that are either random or cyclical. 8. In practical terms hierarchically ordered complexity may be governed by physiological communication such as hormonal and nervous systems, i.e., both “wireless broadcast” and “hard-wired.” These are subject to emergent changes. 9. Differentiation at the organismal level can lead to phyletic divergence. It is predictable that diverse types may become concentrated together in the same place at the same time. Therefore it is predictable that symbioses and looser associations will emerge in such environments. 10. Symbioses and other emergences can also reshape the environment to make more diverse ecosystems that are initially free of agents of natural selection such as competition and predation. 11. Environmental interfaces will be inhabited by adaptable organisms, some of which will penetrate to the adjacent zone, exposing them to physiogenic change. 12. External, internal and intracellular environments will impose physicochemical change. Organisms will respond behaviorally, physiologically, and at the molecular level. 13. Terrestrial plants will experiment with the a variety of forms that will establish major ecosystems. (Tree forms produce forests, grasses produce savannas, and so on.) 14. In some animals, experiments in behavior will tend at first to be genetically assimilated. Then they will tend to escape to greater degrees of individual freedom.

378

Chapter 10

Progress in behavioral evolution, if unconstrained, will increasingly influence the nature of functional-morphological emergences. (E.g., insect size is physically constrained by the exoskeleton, which limits the size of the nervous system, which restricts them to stereotyped behavior patterns.) 15. Environments with stable high temperatures will physiogenically impose a quasihomeothermy on their inhabitants. 16. Thermogenesis, insulation, and the construction of microenvironments will allow some animals to behaviorally experiment with cooler environments. 17. With dependable homeothermy will come the expansion and complexification of the central nervous system, producing more exploratory or curious behavior until intelligence emerges. 18. Social interactions will produce effective wholes beyond the organismal entity. It would be simple enough to select an appropriate progressive series of emergences, and conclude that it has to be that way. Julian Huxley remarked: Evolution is thus seen as a series of blind alleys. Some are extremely short—those leading to new genera and species that either remain stable or become extinct. Others are longer—the lines of adaptive radiation within a group such as a class or sub-class, which run for tens of millions of years before coming up against their terminal blank wall. Others are still longer—the lines that have in the past led to the development of the major phyla. . . . But all in the long run have terminated blindly. . . . Only along one single line is progress and its future possibility being continued—the line of man.10

The octopus, for example, achieved some kind of intelligence without going through all of the stages required of homeotherms. Once there, it could not take advantage of its flexible behavior to progress further. It was in a blind alley because its general physiological adaptability was inadequate for further change to a freshwater or terrestrial existence. Yet this shows that a variety of routes to the emergence of complex organisms might be taken, and that despite the advancement of humans we are not the only model worth investigating. There are some inductive guides to focus prediction of progress. Evidence of an allometric or proportional growth shift in the past predicts the possibility of its exaggeration in the future, and novel functions may emerge at critical points. Allometry and molecular homology also give parallel and convergent shifts some predictability. Natural selection will be manifested in consolidation of the gains of emergent experiments. The consequent dynamic stability will present barriers that only saltatory experiments will leap over, and that only autonomous drives will be able to ignore. Predictability of emergence as it has just

An Emergence Theory

379

been outlined is qualitatively different from empirical predictiveness. So now we have to ask if emergentism has enough predictiveness to suggest experiments to refute its own implications, as distinct from assessing the post-predictability of particular emergences that have already occurred. Can it point to obvious avenues of research to which the older theories did not point? Has it a unifying formal structure that applies to a broad spectrum of biological phenomena? Emergentism predicts that conditions where the barriers of natural selection are low or absent will encourage the numerical increase of existing emergent types. Natural experiments may often have failed because such opportunities were too rare. Nevertheless, the more adaptable an organism becomes, the more operational freedom it has, and the greater scope to diversify. This is a dimension that enlarges the “unoccupied niche space” of neo-Darwinists, and the “morphospace” of developmental evolutionists. However, as diversifying emergents multiply in ecospace, dynamic stability will be re-asserted; new wholes will become more finely tuned, and future emergences will have more hurdles to leap. Therefore, progressive evolution, whether structural or functional, must, on the rare occasions when it happens, advance by saltation: not just genetic, sometimes behavioral, and sometimes amidst cataclysmic clearing of laboratory benches. Problems of predictability and predictiveness can be addressed by thought experiments, and may not be beyond observations and experiments in the field or in the laboratory, but for there to be a way there has to be a will. In the meantime, scientific analysis has largely employed computer models that are unavoidably simplistic. Empirical Tests Whereas selectionism addresses the maintenance of the status quo, emergentism is in a position to institute new interpretations of evidence already available. The historical records of artificial breeding programs, for example, provide a wide range of data pertaining to the fate of phenomena that appear and persist under what Darwin referred to as the changed conditions of life. Dmitry Belyaev (1970) called it “destabilizing selection”—a misnomer, since differential reproduction did not come into the case, and he meant that it was an array of effects caused by a variety of living conditions. The agents of natural dynamic stabilization were absent through human manipulation, but hormonal/behavioral bias, genetic isolation and founder effect were imposed. With some attention to emergentism it should be possible to manipulate variables in breeding experiments more appropriately. And the effects of eliminating the differentiality of survival and reproduction are testable. Emergentism would also encourage greater investigation of naturally destabilizing conditions, whose potential was suspected by Belyaev and his predecessor Ivan Schmalhausen. The original question that brought us to this pass was “Does emergentism give rise to a new causal theory of evolution?”z Its causes, effects, and categories of emergence

380

Chapter 10

differ from those proffered by the Modern Synthesis not only in re-inventing natural selection and emergence. It also has some power to direct future analysis and synthesis, and to predict future events as well as post-predicting the generalities of past evolution. It is non-Laplacean—axiomatically unable to tell the future history of life in any detail from a total knowledge of life as it is. But it does say that life will progress, and that the end of natural selection is not the end of evolution. The reverse is true. We already have the means to modify our future evolution, without anything like natural selection having a role. We can make specific changes and proceed from there. Going back one step in history, plant and animal breeders worked with natural experiments that might have arisen coincidentally from conditions that they have imposed. Again, natural selection was excluded. Nature does much the same with its experiments. Its successes arise from their qualities. Natural selection is no more relevant there than it is with deliberately designed modifications. Karl Popper alarmed theorists when he inferred that it was impossible to test a null hypothesis for natural selection, as defined by neo-Darwinists, and that selection theory was therefore unscientific.11 But null-hypothesis experiments on natural selection have been going on for about 10,000 years, without the experimenters’ being aware of it. Plant and animal breeders have always done their best to remove the influence of natural selection, so that any sports or hopeful monsters that seemed interesting or useful to them would be preserved from competition, predation, disease, the vagaries of climate, and genetic dilution. Embryologists and epigeneticists ensure that such agents of natural selection are minimized. If they were to recognize this, and act upon it, they might better understand how natural experiments occur and succeed, and also grasp the role of natural selection in consolidating saltatory emergences. J. W. Beament’s observations on how competition limits the full expression of the potential of an animal’s adaptability could be an inductive guide for such investigations.12 And the heuristic computer models developed by complexity theorists are improved by factoring in progressive adaptability as well as a re-invented natural selection. Beyond predictions about the course of evolution, a well-knit emergence theory could have the heuristic potential to suggest avenues of investigation that would refute it or support it. Finding a null hypothesis with sufficient scope to refute this nascent synthesis is difficult since there is already general agreement that major transitions to innovative states have occurred in the past, and it comes down to refuting their emergent nature, as opposed to their gradual accumulation through “insensible” adaptational changes. Major genetic changes can take place outside the scrutiny of natural selection, and saltatory and critical-point emergences have occurred during evolutionary history. The ultimate null hypothesis for an emergence theory would be tested by the discovery of an environment where evolution by the selectionist definition, changes in the distribution of alleles, is quite lively, with natural selection firmly in charge, but

An Emergence Theory

381

with no progress made. When we look at the biological past through narrow windows of time, this is the kind of stasis we almost invariably observe. Paradoxically, progress must have occurred, in brief emergent spurts, when we were not looking, otherwise we would not be here to consider the problem. One obvious truth embraces all of the Baconian lists and tables. In any physicochemical environment from which a system that reproduces, duplicates, and differentiates can emerge, all of the above can happen, has happened, and will happen again, given similar opportunities. However, before we leave the non-progressive null hypothesis, I should admit that the first 2 billion years of the history of life on Earth would almost seem to confirm it. Although I have been writing about the possibility of molecular, epigenetic, and life cycle emergences on scales of milliseconds, weeks, and months, the emergence of successful eukaryotic unicells did not occur until 2 billion years after the emergence of life, and it was almost another billion before successful multicellular organisms emerged. Little wonder that the benign interference of an Intelligent Designer is so alluring. What then would an emergence theory usefully have to say about it? Going back to where we started, there is always the possibility that the global environment was so homogeneous, and the dynamic stability imposed by the agents of natural selection so imperturbable, that the first tentative experiments in progressive evolution were doomed to failure. Maybe that is part of the explanation. Though experiments might be tried anywhere, they often survive best in stressful places where competition is low. Those often exist at environmental interfaces. And that is where symbioses and other associations often emerge. What, then, were the environmental interfaces when life was first getting to know itself? Life was traditionally thought to have originated in oxygen-free environments where there was a concentration of abiogenic organic molecules. A popular current contender is the high-pressure, sulfurous, marine volcanic vent environment. Whichever it was (perhaps both), the two kinds of environment probably co-existed, and between them was a barrier where the pickings were slim for any life forms, regardless of origin. As to the interface between aquatic and terrestrial conditions, one was wet, the other dry. And the less challenging interface between marine and freshwater environments was between the salty and the dilute. The latter was still stressful enough for unicells, and a long wait for the appropriate generative conditions to arise is not surprising. All these environments were anoxic and would stay so until the emergence of photosynthetic hydrolysis systems. Once those had appeared, the buildup of oxygen to the point where it would make a difference was certainly on a geological time scale. Once it did reach that point, two well-defined interfaces came into being. The less obvious of these was at the fluctuating redox point between oxygenated and anoxic layers in benthic sediments and water columns. Here there was likely to be congregation of biochemically different unicells, some with the capacity to detoxify oxygen,

382

Chapter 10

and even to put it to good use. These had the potential to engage in symbioses. As to the interface between water and the atmosphere: long before the increase in atmospheric oxygen reached present-day levels, it reached a partial pressure much higher than is possible for oxygen dissolved in water. Here was a chance for cells and multicells that could protect themselves from ultraviolet light, and repair the damage it caused, to get on with the business of making the terrestrial environment more attractive to a wider assortment of organisms. Thus did the unlikely become the likely, very slowly at first, but with an accelerating pace. Through progressive changes in adaptability, higher animals were finally able to explore new environments and add to their physiological qualities. Those that behaved consistently and persistently, in relation to each other and to their environment, accelerated anatomical adaptation. The lack of such adaptability in primitive unicells contributed to the slow pace of their advancement. The Magic Formula We cannot leave the question of how much of a scientific theory emergentism can provide without asking if there really is a holy grail to be found. To some extent I have satisfied Stephen Jay Gould’s agenda of seeking causes, strengths of causes, levels of causes, and contingencies, and have added a few more relevant items. I have also tried to meet the challenge of saying more about the emergent property than that it is simply the emergent property. (Please don’t laugh out loud.) Predictiveness and predictability, which belong to formal theory, are to some extent manageable, even if we lack explicit forecasts. Now, what about the greater challenge of constructing a universal formula that would apply all the way up the ladder of nature, ascending from emergences in physico-chemical systems, through the emergence of life, to the emergence of mind? Doyne Farmer (as quoted by Mitchell Waldrop in 1992) usefully returns us to that question: I’m of the school of thought that life and organization are inexorable . . . just as inexorable as the increase in entropy. They just seem more fluky because they proceed in fits and starts, and they build on themselves. Life is a reflection of a much more general phenomenon that I’d like to believe is described by some counterpart of the second law of thermodynamics—some law that would describe the tendency of matter to organize itself, and that would predict the general properties of organization we’d expect to see in the universe.13

Matter organizing itself? Crystallization meets that requirement. Spontaneous protein folding does as well. But like the structure and behavior of wet water, they are the expression of an inherent property of atoms or compounds realized by appropriate physicochemical conditions. The Second Law of Thermodynamics was once much discussed by theoretical biologists who attempted to reduce life to thermodynamic principles. This gave physicists and chemists an opportunity to add their theories

An Emergence Theory

383

about life, unsullied by its leaky, slimy, stinky messiness. (Scientists, including some biologists, seem never to have come to terms with living in a human body!) But all that could be said initially was that living organisms do not defy the Second Law. They are open (or dissipative) systems that become more complex ontogenically and phylogenetically, and in so doing they generate massive amounts of entropy by the expenditure of useful energy. That life obeys the Second Law says nothing useful that we did not already know about organisms and their emergent properties. Fits and starts and an ability to build on what already exists could certainly apply to cosmological emergences, and obey the principles of quantum and particle physics. However, a “law that would describe the tendency of matter to organize itself” would need corollary laws of emergent reproduction, and the repetitive differentiation of biological modules. These are emergences that do not happen in non-biological systems. Further ad hoc laws would then be needed to contain the origin and development of mind. How these biological properties emerged will not be answered by physical laws, or by a universal mathematical formula. Although biological emergences do not defy the laws governing the lower levels, their causes are usually identifiable only at the transition point. They have new properties that cannot be assessed before they have arrived. Interesting though a physical law of the complexification of non-living matter might be, its universality would fail at the barrier of biological reproduction. That does not write off the possibility that emergent increase in complexity could be inexorable, and that when patterns are biological their establishment may be fast and their effects persistent. Note, however, that th plurals in the Holland quotation at the chapter’s head (“copies,” “variants”) may indicate some embarrassment with regard to the multiplicity of biological possibilities. Because of the difficulties I have just outlined, I have been trying to sneak up on a universal formal theory, without expecting to catch it. Nevertheless the game is afoot. I am at a way station on the quest, coming down the mountain a few stops back from the summit of biological complexity, while Holland, Solé, and Goodwin are a few stops on from rigorous simplicity. In addition to the question of inexorability of complexification, including the biological kind, there is the question of how many fundamental emergences there have been during the history of evolution. And can the multiplicity of living examples be simplified to something approximating a general biological rule? Maybe we need separate rules for the emergence of life and the emergence of mind, with sexual reproduction, endosymbiosis, multicellularity, and other major innovations between the two. That they are all saltatory complexifications is enough to suggest a universal patterning, but not a common formula. Some biologists still hope that there can be a universal theory. Antonio Lima-de-Faria proposed that autonomous complexification is the result of the organism’s physical nature, without much reference to the prevailing conditions of its life. Thus, he titled his 1988 book Evolution without

384

Chapter 10

Selection: Form and Function by Autoevolution. In the same mode, Stuart Kauffman asked in The Origins of Order (1993) if there is some principle in physical nature whereby physicochemical systems must produce an integrated, reproducing biological entity. There is no question that since the Big Bang the cosmos has become more heterogeneous and complex, and has evolved in the old-fashioned sense of unfolding in time, revealing the potential of its original emergent properties. Even so, it would be impossible to predict where and when galaxies and their component star systems would emerge. Farmer’s organizational equivalent to the Second Law of Thermodynamics, if formalized, might apply across the non-living board. But biological evolution, regardless of ideological prejudice, requires biological reproduction, self-maintenance, and a kind of self-organization unique to living systems. The Big Bang was the Enormous Emergence, but it was qualitatively different from the emergence of life. Or am I engaging in a semantic quibble? Has the cosmos “evolved” the familiar chemical elements in such a way that certain random combinations such as water and carbon compounds and the rest must arise, stable enough to hold together, but unstable enough to be pulled apart and brought into more complex configurations? And have other natural physicochemical experiments brought them together to reach a critical mass of chemical complexity out of which life must eventually, if unpredictably, emerge—thus making us “at home in the universe”? This would assuage Kauffman’s nostalgia, and if he is right then life is a consequence of the emergent properties of the Big Bang. Cosmologists may now start jumping up and down, and shouting “That’s what we’ve been trying to tell you.” But neither metaphysical absolutes nor ontological assertions about the nature of being have explanatory power. Life has new rules intrinsic to the emergent properties of reproduction and self-maintenance. And mind, though a consequence of biological reproduction and differentiation, has a novel set of rules arising from the unique emergent properties of an expanded and reorganized brain. The semantic reduction of biological evolution to cosmic evolution is either delusion or dishonesty. If cosmologists find their equation, it will say nothing about the higher rules of biological emergence. To say that life will emerge in a universe that has carbon may be true, but that it has happened in this instance is not evidence enough to generalize. To argue that the concept of autoevolution applies generally all the way from the Big Bang to mind is a variant of the hylozoic (or panpsychic) fallacy. It would imply that life and mind must somehow be contained in subatomic particles and energies, and progressive biological evolution must be an unfolding and multiplication of their initial features. Emergentism saves us from hylozoism. Life and mind are not properties of the Enormous Emergence, any more than the mammalian neocortex is emboîté in the primitive chordate nervous system. As Holland puts it, there are layered series of emergent patterns, and each has its own novel emergent properties. Each

An Emergence Theory

385

level may have the propensity to generate the next one up when the appropriate conditions arise. But the lower layers do not possess the properties of the higher ones. Kauffman and his associates have devoted a great deal of time to the emergence of complexity within non-living systems, and find it difficult enough to pin down at that level, far less in the evolution of organisms. Thus, when John Horgan proclaims “the end of chaoplexity” he concludes that complexity theories tend to describe what is going on rather than to explain them: Self-organized criticality is not really a theory at all. Like punctuated equilibrium, self-organized criticality is merely a description, one of many, of the random fluctuations, the noise permeating nature. By Bak’s own admission, his model can generate neither specific predictions about nature nor meaningful insights. What good is it, then?14

The simple answer is that these ideas begin to present alternatives to a prevailing theory that for seven decades has focused the attention of self-styled evolutionists on minor, non-evolutionary fluctuations of dynamically stable systems. That theory has generated no useful predictions about or insights into progressive evolution. It is “armoured against it.” The central role that Julian Huxley gave evolutionary progress has been written out of the neo-Darwinist drama. Instead of predicting “the end of evolutionary biology,” Horgan might better ask “What good is an evolutionary biology that ignores the phenomenon it purports to address? And when are we going to see something that doesn’t ignore it?” In the meantime, Holland has taken positive steps toward formal theory, if not toward a single law of organizational evolution. Even if his material consists only of “fragments,” they are indeed “lawful.”15 As he well understands, their explanation of emergent evolution is limited because they are modeled on “learning” novelties in complex systems that were designed by humans. There is a major leap from here to systems that self-complexify in a manner that is novel and heritable. Nevertheless, Holland’s generalizations, cited at the end of chapter 8, interdigitate with those I have drawn from the various arenas of evolutionary performance. My aphorisms may not be “lawful,” but they amalgamate the relevant information, and compare and contrast different points of view, to bring out relationships or generalities, and reveal exceptions that prove the rules. The generative levels of emergence, biological or intellectual, need a congruence of all the necessary parts. So if all of these are put together, how much closer do we get to an integrating explanation than does the mixed bag of tricks found in the Modern Synthesis? A reexamination of the construction of the whole is in order. Layered Series of Generative Procedures Holland remarks that new wholes emerge through layered series of generative procedures. This refers to hierarchical dynamic structures whose complexity in real life and evolution defies a formal theory of evolutionary emergence. To simplify the

386

Chapter 10

problem of identifying the layers of progressive evolution, we can look at them from two perspectives: historical layering, and the layering of structural grades. Earlier chapters and sections dealing with the chronology of emergences have given some idea of the complexity of historical layering, and the fact that it is not entirely separable from structural gradation. In reviewing them here, geophysiology and the random impact of contingencies are left out for the sake of simplicity. The epigenesis of a complex, mature organism from a fertilized egg provides a crude recapitulatory model of historical layering in evolution. In both development and evolution, the most visible chronological series of progressive emergences is anatomical. This, however, is not the most important from the emergentist point of view, since it can be molded by behavioral changes made possible by physiological advances. Form follows function, and function follows form. Progressively they bootstrap each other into the future. Take vertebrate evolution as an example. The bootstrapping relationship among behavior, physiology, and anatomy was accompanied by complexification of the central nervous system and further tuning of the hormonal system—all dependent on a reliably constant internal milieu. Developmental changes underlying anatomical evolution involved allometric shifts, hypermorphic additions, and interpolations at earlier stages in epigenesis, as well as paedomorphic processes. While there were no radical changes in functional anatomy once a humanoid form was assumed, cranial expansion, an effective voice box, and dexterous hands were some of the necessary generative factors accompanying neopallial allometry during the last lap of the course to humanization. Furthermore, the final structural grade of the cerebral cortex has added a new layer: acquired mental characteristics that can be inherited through education. Thus, the historical layers, or hierarchical levels, are internested, and have effective feedback between them. Novelties are added to early layers as well as to the top of the pile (not the easiest kind of system to model mathematically), but progressive evolution requires separation of offices and concurrence of efforts. The other way of looking at emergent levels is simpler, though more static. It can be called “the hierarchy of structural grades,” or “a compositional hierarchy” (i.e., molecules < organelles < cells < tissues < organs < whole organisms < societies < demes < communities < ecosystems, and so on). The development of a zygote into the hierarchical modular structure of the early embryo would seem to recapitulate evolution. Yet that is an oversimplification, since the embryo is a whole at any stage, and its structural grades mingle, interact, expand, and contract throughout development. As to the mature whole organism, its hierarchical levels and their regulation are conventionally studied as if they stood alone. And, historically, they have attracted fickle attention. Once, cellular biology was all the rage; now it is molecular biology. We can reach an understanding of progressive evolution only by studying them multidimensionally, as wholes moving through time. Though we might tend to concentrate on

An Emergence Theory

387

the regulation of gene expression as “the explanation” of evolution of the higher grades we must remember that gene expression can be altered by changes in the investment of maternal effects, as the result of behavioral bifurcations, alteration of external stimuli, internal catalysts, mediator cascade effects, and so on, with genetic or epigenetic accommodation last if not least. One implication of Holland’s analysis goes beyond the layering I have just described, being a feature of emergences from the non-living to the mental. Holland says that once persistent patterns occur they become candidates for “combination with other, persistent patterns (other copies or variants),” and that “at this point larger patterns with enhanced persistence and competence can occur.” These need not be vertically layered compositions. Persistent patters may be abiotic coacervates or abiotic organic chemical films on surfaces. In organisms they may be adjacent holons at the same focal level. The patterns may be different organisms with a propensity to form symbioses. Or they may be members of the same species whose interactions change with population density, reproductive cycle, and season. Interactions also change with such simple physical rhythms as day and night. Male grizzly bears, for example, aggressively keep other males at a distance during the day, but in the dark they have been observed in close, non-aggressive contact.16 Combination of independently evolved persistent patterns occurs only where generative conditions are appropriate. For example, they have to be in the same place at the same time if the unlikely is to become the likely—randomness takes on a different meaning where the odds are significantly changed. The toss of a coin gives a 50 percent chance of a win, whereas the purchase of a lottery ticket is a long shot. In the selectionist rulebook, astronomical odds don’t matter if there is enough time and persistent trial-and-error variation. In contrast, the emergentist rulebook considers how organisms might shorten the evolutionary odds in their own favor. Instead of treating natural experimentation as random point mutations of DNA that though heritable might not persist, emergentism provides a holistic synthesis of multidimensional factors. Thus, it meets the demand for a truly “interactionist” interpretation of evolution that Susan Oyama calls for in The Ontogeny of Information (1985). The Holism of Emergentism Examination of the construction of layered series has brought us back to the dynamic structure of the whole, which requires examination of the organism’s internal relationships as well as its relationship with its environment, rather than reductionism. In the genocentric universe, the genes of cats that reproduce most successfully come to dominate catdom. However, as Jack Cohen and Ian Stewart say in The Collapse of Chaos (1994), cats eat mice but cat molecules are indifferent to mouse molecules. A total knowledge of cat and mouse molecules would never allow a prediction about

388

Chapter 10

how mice affect cats’ behavior, since that has progressed through several emergent levels. I might add that cats, to successfully pass on “their” genes, would do just as well to cosset mice, or simply ensure that the bacteria in their feces were allowed to reproduce. Cat, mouse, and bacterial genes are much of a muchness; it’s the organismal packages that differ distinctively. Nevertheless, reduction is a necessary tool for identifying causes and mechanisms that participate in emergent evolution, and systems reduction is valuable for understanding the rules of self-organization. With the basic emergent properties of life (reproduction, self-maintenance, selforganization) come repetitive differentiation, and combination. It can be predicted that nature will experiment, organismal complexity will emerge, and stasis or equilibrium among existent organisms will be established as numbers rise and resources become limited. Escape from stasis is made possible by improvement in selforganizing complexity, expressed as adaptability. This extends the freedom of the organism from conformity to the environment, and allows some freedom of choice of behavior, and hence a possible escape route via environmental fringes and interfaces. The above tenets approach a general biological principle of emergence based upon the fundamental emergent properties of the first living organisms. These are not qualities of non-living systems, so they are unashamedly vitalistic, no longer invoking supernatural entelechies or mysterious black-box powers. This much has been intuitively obvious to many evolutionary “essentialists” for at least a century. Although classical genetics separated the organism from the environment, and the double helix separated the gene from the organism, they, and the works of modern molecular biologists and epigeneticists, have contributed to bringing out vitality as a rational and materialistic emergent evolutionary principle. Since disequilibration of stasis permits escape from it, and enhances the scope for further emergences, it is predictable that stress, small accidents, and large catastrophes will be significant. The nature of some of the particular contingencies likely to be encountered is unpredictable, and the randomness of natural accidents makes it impossible to predict what emergents, in what phylogenetic lineages, will survive and thrive, although adaptability is more reliable than luck. By asserting the repetitive differentiation of DNA as a fundamental emergent mechanism, I risk decaying into gene orbit. Of course, the principle applies to epigenesis and associations too. I must also re-emphasize the axiom that emergences are the expression of all of the generative conditions of organism and environment as a whole, and that they result in a constellation of novel properties. Stuart Newman and Gerd Müller (2000) make the case even more strongly when they argue that “the correlation of an organism’s form with its genotype, rather than being a defining condition of morphological evolution, is a highly derived property.” Then they conclude that “genetic change . . . mainly plays a consolidating role, rather than an innovating one.”17

An Emergence Theory

389

The need for holism (or interactionism) in an emergence synthesis brings us back to the question of how the whole can be greater than the sum of its parts. I have already given the example of how the functional properties of quaternary protein structure constitute a greater whole than the sum of its constituent monomers. At a higher structural grade, the symbiosis of two different organisms can turn surplus energy and toxic waste products to mutual benefit. Interactions within societies can alter the constituent individuals in a variety of ways. Wholeness also resides in the hierarchical structure of the organism. Each emergent level in the hierarchy has properties to which the parts contribute, but which do not reside in the parts. Having such an organization ensures increased efficiency and effectiveness. One important element of emergence, intrinsic complexification of differentiated cell types, is overall an exponential function of reproduction and time—quite a simple equation in the absence of natural selection. As a function of the number of differentiated cell types, the curve for all of animal life flattens after the emergence of the vertebrates.18 But the historical curve of some lineages, especially that of hominids, fits the simple exponential equation, its logarithmic slope theoretically determined by the fact that the acceleration of complexification is virtually equivalent to increasing adaptability and freedom to explore unexploited environments. It could continue to rise more steeply if we did not voluntarily limit experiments in human genetic engineering. Looked at more closely, the curve would be seen to be a staircase, consisting of a number of steep (critical-point) or vertical (saltatory) risers representing emergent events separated by steps or landings whose breadth is proportional to the duration of periods of dynamic stability. If the randomness of natural experiments and accidental contingencies is included, the whole graph becomes irregular, although the overall shape is not altered. To express complexification as the increase in the number of differentiated cell types does not address the final stages of primate evolution. Once a particular cell type, the neocortical neuron, was acquired, it did not go on to produce variant repeats. The cell type could of course differentiate ontogenically in the number and topography of dendrites, in its cell membrane receptors, and in its DNA modifiers. With numerical increase of neurons, progressive packaging became more important, since each of those cells had the potential to exchange information with thousands of others. Its interaction with surrounding glial cells complemented its versatility. Consequently the whole brain could further differentiate ontogenically to an astronomical degree, prior to its genetic consolidation (or accommodation). A near infinity of connections at the level of the individual underlies the emergent quality of the human mind, one of the other big problems of evolution that emergentism might address. Yet here I have already fallen into the trap of oversimplification. There is more to mind than neurons and glial cells with multiple connections and inborn powers of intelligence and memory. We must not ignore the roles of nutrition, experience, introspection,

390

Chapter 10

education, and social interaction. Ontogenic emergences—bright ideas—come when unusual connections or bisociations are made between cells and brain. Gene-determined hard wiring would prevent them. Thus, one generalization about the emergent properties of mind covers the whole constellation: freedom from fixed action patterns and stereotyped behavior, and freedom to embrace conventional ways; freedom from competition and predation, and freedom to challenge them; freedom from conformity, and freedom to conform through cooperation. “Persistent patterns” become free to plan and create future patterns. Emergence as a Dialectical Synthesis A dialectical synthesis involves bringing together apparently contradictory ideas as thesis and antithesis so that a new synthesis will emerge. There is a serendipitous parallel between this interpretation of the history of ideas and the emergence of novelty from its generative conditions. But beyond this dialectic is the bisociation of existing facts and ideas that no one had thought of putting together. That is the proper goal of interdisciplinary research, and from it new concepts sometimes emerge. This is what Ernst Mayr, without invoking Hegelian or Marxist language, claims for his version of the Modern Synthesis. But there the contradictions are subsumed under the thesis, put in black boxes, rejected, or vilified. The reductionism does not resolve complexity; it destroys it. Emergence does not generate a simple Hegelian synthesis. If proposed as a theory or a thesis, it already integrates saltation with gradual change, and the intrinsic with the extrinsic. It is a theory of evolution, which selection theory is not. Therefore, emergentism rejects natural selection as an evolutionary cause, but establishes its theoretical importance as a meaningful consequence of emergent evolution and as a dominating condition of existence. I close this chapter with a demonstration of how the contradictions of emergentism and selectionism can, to some extent, be synthesized, and come to the conclusion that a comprehensive dialectical synthesis cannot be achieved. Therefore, to come to a dialectical synthesis requires transcending an emergence theory of evolution, to integrate it with a selection theory of non-evolutionary dynamic stability, and that task will be undertaken at the beginning of the next chapter. Summary The following tables are intended to “bring home the Bacon,” to outline other general principles, and expand on the contrasts and comparisons of emergentism provided in the “field trip guide” late in chapter 3. In this context, bringing home the Bacon does not infer the wholesale consumption of the old ham, but only his use of tabulation to compare and contrast different aspects of nature. Some of the vernacular slogans in

An Emergence Theory

391

current use are included in brackets. If any have slipped your mind, look down to the end of this chapter, where they are redefined. Thesis: evolution by emergence

Antithesis: evolution by natural selection

1 Emergence ranges from variation of the existent to large-scale innovation.

Natural selection (defined as differential survival and reproduction) is assumed to act on small-scale heritable variations.

Commentary Regardless of the extent of change, large or small, whether at the molecular, organ, organismal, or population levels, the quality of that change will out. In other words its outcome will be realized or constrained by how it interacts with the internal and external environments. If its emergent quality is related to integrity/persistence in being, it may survive until different conditions arise, when it might then diversify. 2 Emergentism synthesizes intrinsic and extrinsic change. Adaptation to environment is a consequence of organismal change—sometimes behavioral—often set off by the environment itself, but also possible as autonomous emergence.

Natural selection is most evident in relation to extrinsic change, i.e., adaptation to environment. Intrinsic self-adaptations, or coadaptations are tacitly acknowledged, but ignored.

Commentary The failure of selection theory to incorporate intrinsic change is not entirely a matter of neglect. But it is unable to logically exclude such change. If molecular and organismal adaptation is seen to result from the interaction of genes, organisms, and their behavior in relation to their habitats, some of the apparent contradictions disappear. 3 Emergentism synthesizes saltatory and gradual change.

By convention, selectionism depends on gradually continuous variation and rejects saltations.

Commentary There is no longer any reason to reject the reality of saltatory change. In the case of critical-point emergence, an appearance of saltation results from the sudden realization of a potential inherent in a gradually increasing continuity. That might be assumed to be through adaptational accumulation. Yet the question that remains is “To what extent might continuity of variation be due to an autonomous drive?” Here

392

Chapter 10

there might be no convenient accommodation of the emergent and selectionist positions. Nevertheless, see the next item. 4 Autonomous directional change, or orthogenesis, can be equated with selfamplifying genetic and epigenetic systems. It may be responsible for much of anatomical divergence.

The appearance of orthogenesis is in reality orthoselection of mutations whose phenotypic expression is more exaggerated and therefore more strongly selected than the earlier variation. This apparent trend is limited by natural selection when it reaches a point of maladaptation or disintegrity.

Commentary Self-amplifying autonomous processes occur at the level of codons, larger base sequences, and possibly epigenetic interrelationships. At present these are in or close to mainstream evolutionary theory as molecular drive and concerted evolution. For these, mechanisms have been expounded that are acceptable to all parties. Their effects can be continuous and rapid, being amplified from generation to generation. Some are known to be disintegrative, and thereby finally irrelevant to evolution, needing no supernumerary force to remove them. Self-amplifying processes might also be stopped before the point of instability—every lineage can be shown to have undergone allometric growth shifts that produced their characteristic anatomies. Conceptually, orthogenesis gets over the selectionist’s difficulty in understanding how the incipient stages of allometric shifts might have sufficient selective value. They do not require such a value, since the processes are autonomous. 5 There are several generative causal arenas: symbiosis/association; epigenetic/developmental processes; physiology and behavior.

Major causal categories are normalizing and directional selection acting on DNA variants.

Commentary The need to accept saltatory evolution is reinforced when the performances of generative causes are observed. It is unfortunate that selection theory ignores them and doubly unfortunate that it does so because they do not fit the theory. Yet its emphasis on the normalized condition should not be surprising, since dynamic stability dominates biological history. It thus provides the time for new types to integrate and fine-tune their innovative qualities.

An Emergence Theory

6 All the previous categories involve real organisms, and real congregations of them interacting within their demes as well as interacting with similar groups of other species in communities.

393

The previous categories are operational at the level of pools of genes in populations. This requires the working hypothesis of an external causal agent—natural selection.

Commentary Here I must admit that it is difficult to integrate reductionist population thinking. However, populations are theoretically and practically significant. For me to abandon their reality at this juncture would be as anti-intellectual as the populationists’ evisceration of real populations to turn them into gene pools. Once the population has been re-invented as an interbreeding assemblage of organisms in a real environment, the problems disappear. The integrating qualities and behaviors of such populations are offset by geographical limitations. Under limiting, stressful conditions, they are vulnerable to disequilibration. Thus, periods of equilibrium that lead to expansion in space and time reach thresholds where population bifurcation is possible. This parallels in space the temporal epigenetic conditions of alternating stable growth and expansion, punctuated by unstable interphenes where developmental bifurcations are possible. 7 Genetic assimilation results from the consistent and persistent behavior of individual organisms. Appropriate molecular experiments are then incorporated. Phenotypic form and function precede genotypic changes.

A metaphorical external agent, selection pressure, is adduced to explain the generation of genetic changes appropriate to the prevailing conditions of life.

Commentary There was never much of a logical difference between Lamarck’s organismsresponding-to-needs and the Darwinists’ evolution-responding-to-selection-pressures. But Lamarck came closer to the truth by putting the organism, rather than the gene pool, on the line. The behavior of the organism determines what is adaptive. Whether we say that it constrains the viability of natural experiments or that it changes the selection pressure, it comes down to the same thing: What the individual organism does is important in the evolutionary scheme. 8 In all of the causal arenas, saltatory emergences may occur. Both saltatory and critical-point emergences may be sufficient to escape existing dynamic stability

Selection theory considers saltatory change to be detrimental, and to diminish the causal role of natural selection acting on gradual change or imperceptible variation.

394

Chapter 10

Commentary This revisits items 3 and 5 above. But what may be added is the realization that differential survival and reproduction are normally present, whether or not change is saltatory or gradual. Whichever it might be, the experimental quality is worked out in relation to the life and environment of its organismal possessor. This does not confirm the inference that the working out of those qualities in environments is evolutionarily causal, nor that natural selection, despite its ubiquity, is creative. 9 All of the above items are correlated with environmental change and the response of organisms to such change, both ontogenic, and by genetic, epigenetic and physiological accommodation.

Environmental change alters selection pressures.

Commentary In this case the selectionist antithesis is specious—metaphorical forces need no logical synthesis. The real and positive qualities outlined in item 8 are self-sufficient. Much as it might so proclaim, and much as its reluctant brides might close their eyes, selection theory has no droit de seigneur over genetics and molecular biology. 10 The potential for emergence resides in the primary qualities of life: simple persistence through reproduction, selfmaintenance, and self-organization.

Natural selection, qua differential survival and reproduction, created the primary qualities of life through the preservation of persistent systems.

Commentary Competition is an effect of emergent qualities once their possessors have sufficiently increased in number to make resources limiting. Natural selection is the outcome of competitive interactions. The result is that certain emergent qualities, whether simple adaptational gene mutations, or physiological/behavioral adaptabilities may come to dominate certain populations. Although competition and natural selection are realities, they are not creative, but are the consequences of creative natural experiments. 11 Intrinsic complexity is increased by multiplying and integrating the fundamental genetic, epigenetic, physiological and morphological units of organisms.

Complexity is a coincidental consequence of the accumulation of adaptations.

An Emergence Theory

395

Commentary Theoretically, biological self-organizing systems (i.e., those that involve reproduction) will increase in complexity and integrity without reference to external conditions. Integrity infers a degree of autonomous regulation that could be loosely described as “internal adaptation.” The modification of these integrated qualities may be randomly generated from existing modules, but their persistence is proportional to their integrative qualities. There is nothing of random aggregation in this process. All organisms inhabit environments, and modifications that are perceived as adaptationsto-environment are consequences of the adaptability of well-integrated wholes that allows their entry into and survival in new habitats. 12 Complexity is further increased by differentiation and reorganization of the multiple units—molecules, cells, organs, organisms. (Mixing and matching. Separation of offices and concurrence of efforts. Repetitive differentiation.)

Natural selection favors complexity and adaptability when they have high fitness. It favors regression and simplification when those have high fitness.

Commentary Complexification through multiplication of subunits as presented under thesis 11 is an oversimplification, since biological versatility depends upon differentiation as well. Differentiation, in turn, is ineffective, unless its products are regulated. Therefore a multiplicity of epigenetic and physiological processes need to interact simultaneously. This is one of the relatively unknown factors of emergent evolution. The selectionist antithesis is correct as far as it goes. Yet it reinforces the realization that much of the time complexity and adaptability do not have high fitness, although their emergence will out, regardless of their ultimate fate. The reverse of the coin is simplification and regression, which happens under prolonged conditions of dynamic stability. (Marine fish lose the ability to return to fresh water and vice versa.) I will not press the point, but this kind of evolution-by-natural-selection strikes me as more akin to devolution. 13 Generative conditions for emergences are strong at environmental interfaces where there is opportunity for novel organismal associations. (More mixing and matching. Being in the same place at the same time.)

Key innovations are caused by selection pressure, which is strong at environmental interfaces.

Commentary Here there is no fundamental contradiction between thesis and antithesis. Both accept that evolutionary conditions change at environmental interfaces. “Selection pressure”

396

Chapter 10

under these circumstances is actually a release from the pressure of normalizing selection. 14 At environmental interfaces there are opportunities and benefits for escape from old environments and their stases.

Environments act as filters that remove incompetent variants.

Commentary Here again there is no serious conflict. It is, however, more to the point to appreciate the positive emergent qualities of natural experiments than to emphasize the negative qualities of variants that fail. 15 Critical-point emergences find an analogy with ontogenic changes at epigenetic thresholds. Saltatory bifurcations are also possible at these nodes.

Epigenetic metamorphoses are no more than selectively advantageous accelerations of changes that previously followed a path of gradual continuity.

Commentary Without argument, epigenetic alternatives must sink or swim according to prevailing internal and external conditions. But they are implicitly subject to loose canalization and may have no initial gene determination. That too will pass, so that ultimately the alternatives will often be found to have become tightly genetically assimilated or canalized. The concept of selection pressure as a constraining or creative influence is superfluous. 16 Populations reach density thresholds. There they may be limited to fluctuation around the point of carrying capacity. But physiological and behavioral alternatives that affect populations en masse may crop up.

Population-density effects are gene determined and selectable.

Commentary In response to population-density stress, entire demes may remove themselves (starving locusts becoming flying swarms). Only epigenetic exceptions—“flyers”— would be involved in the first instance, but at that point they would then constitute an innovative lineage, removed from the parent population. And all members of the new population would have that quality. The stress response might be an expression of atavistic traits or epigenetic novelty that has no fixed DNA determinant. Either way, the standard explanations of allele concentrations at the peaks of genetic landscapes are inadequate.

An Emergence Theory

17 Adaptable organisms may concentrate activities in a particular behavioral and environmental direction. This determines the adaptiveness of subsequent genetic experiments (organism driven).

397

The organism changes the selection pressures.

Commentary This thesis restores the participation of the organism in its evolutionary future. NeoDarwinism reduced the organism to an ephemeral bag of alleles. Yet the antithesis that the organism can change selection pressures is already close enough to the thesis that physiology and behavior of the individual can lead to genetic assimilation and diversification. 18 The study of evolution is the study of the generation of emergent properties.

The study of evolution is the study of the numerical distribution of traits in populations.

Commentary This is not too difficult. The study of evolution is indeed the study of emergent properties. The study of the numerical distribution of traits in populations is also the study of emergent properties. However, the latter case is limited to the prevailing ecological conditions. What we need to integrate are changes in the prevailing ecological, developmental, physiological, and associative conditions. And we also need to rethink the meaning of adaptiveness in the context of emergence. 19 Adaptiveness is synonymous with “emergent property” and the integrity of the whole organism. It does not require metaphorical assistance.

The natural selection of adaptiveness is the essence of evolution.

Commentary Along with several earlier articles, this final one returns us to matters of “biological advantage,” “usefulness,” and “adaptiveness” and how they can be synthesized with self-organization (both in development and in the operation of regulatory systems in the matured organism), self-maintenance (molecular repair, tissue regeneration, and homeostasis at cellular and whole organisms levels), emergent properties that make wholes greater than the sums of their parts, and those that have lesser roles in survival and reproduction. “So!” says the selectionist. “You mean new traits that enhance survival and reproduction, and so are spread throughout the population by natural selection.” I am indeed talking about certain new traits that enhance survival and

398

Chapter 10

reproduction and do indeed spread throughout the population—under certain ecological circumstances. But I am not forgetting that the qualities of those new traits came into being with them, not as value added by natural selection. And I am remembering emergent adaptabilities that may not be called into immediate use, but have a potential to be advantageous under contingencies that often involve the diminution of the agents of natural selection, such as competition and predation. Structures and functions that were already integrated gave rise to those adaptabilities, and they, in turn, made the whole organism even greater than the sum of its parts. In conclusion, it should be clear that some of the selectionist antitheses presented here are irreconcilable with an emergence thesis of evolution. The potential for a synthesis is improved by the realization that consequent to major emergences, adaptational diversifications, are genuine evolutionary changes that are part of the re-equilibration process. Adaptive radiation is variations on a theme, rather than progressive complexifications. Yet it too relies largely on minor emergences. The remaining question about this adaptive diversification is: to what extent does it depend on orthogenesis by self-amplifying mechanisms, or the more conventional process of random change “accumulated by natural selection”? Since the final dynamic equilibrium usually involves an inescapable degree of specialization, adaptational evolution peters out. Therefore emergent, progressive evolution can only be dialectically synthesized with the equilibrium state and the selection syndrome above the level of evolution, as a synthesis of biology, which is the subject of the final chapter. Definitions of Terms Used in Tables Being in the same place at the same time Points to generative conditions where the probability of emergent novelty is high—where the unlikely has become the likely. See also mixing and matching. Constellation of emergent qualities Emergences often have multifunctional consequences. Only one may be immediately advantageous, but the others represent a potential for diversification or evolvability. Critical-point emergence In a continuous series of changes a new or enhanced function may appear in the manner of a saltation (e.g., wing enlarges to point where lift exceeds drag and gravity—gliding to true flight). Generative conditions (in chemistry, physics and cosmology: initial conditions) combination of circumstances that contribute to an emergence.

The

Genetic assimilation Phenotypic changes induced by the behavior of the organism in relation to its environment become genetically fixed, wherever genes exist that can

An Emergence Theory

399

mimic the effect, or wherever the genome can proffer experimental innovations that might match. Thus the organism determines what is adaptive, and phenotypic change adds to organismal quality. Some of these effects involve complexification through self-organization, i.e., through non-DNA epigenetic assimilations. Geophysiology Organisms affect the environment directly (e.g., produce atmospheric oxygen by photosynthesis), and organisms contribute to biospheric equilibria (affecting climatic cycles and changes etc.) La vie libre Bernard’s expression for the independence of life afforded by homeostasis. Mixing and matching Organisms that congregate with others of different species in particular environments may interact epigenetically or symbiotically. Successful symbiotic and sexual associations are those that match well, by complementing each other’s qualities. Mixing and matching also applies to natural experiments with DNA exons and protein domains. Persistence in being The Darwinian term for survival through “innate flexibility of constitution” (i.e., through adaptability/homeostasis). Physiogenesis The environment directly imposes physicochemical change on the organism; e.g., the environment cools, the organism cools. Repetitive differentiation Holons (modules) at the molecular, organ, and organismal levels are duplicated, reduplicated, and differentiated—natural experiments that increase adaptability (Aristotle’s separation of offices and concurrence of efforts). Saltatory emergence Sudden large-scale change that is rapidly manifested on an organismal time scale (e.g., all-or-nothing establishment of proto-mitochondrial prokaryotes as endosymbionts in eukaryotic cells; deviations in early epigenesis). The unlikely becomes the likely Holland’s shorthand for heterogeneous distributions of persistent patterns that have the potential to interact and to become more complex and therefore more persistent. Use it or lose it Regression is a common process that need not be detrimental, even as overall organization increases. Sometimes it redefines a particular way of life. Some labile biochemical pathways may be redirected by the loss of a particular enzyme. But the genetic machinery may remain dormant, to be revived later.

11 A Biological Synthesis

Taking the route you would be likely to take From the place you would be likely to come from . . . What you thought you came for Is only a shell, a husk of meaning From which the purpose breaks only when it is fulfilled If at all. Either you had no purpose Or the purpose is beyond the end you figured And is altered in fulfillment —T. S. Eliot, 19431

These lines from Eliot’s poem “Little Gidding” are particularly apt to my pilgrimage to this point. The figured end was not a synthesis of biology. First, my route pointed to the replacement of selection theory with an emergence theory. Then I tried for a synthesis of the two. In the previous chapter, I set out a table of contrasts between emergence theses and selectionist antitheses, intending to integrate the contradictions, to produce an “Emergence Synthesis of Evolution.” It is a near-impossible task, since the emergence thesis deals with the major mechanisms and processes of evolutionary change, while the selection antithesis deals largely with non-evolutionary fluctuations in the demographics and ecology of populations. The Modern Synthesis “solves” this problem by fiat: stasis is evolution. However, selectionism cannot be dialectically synthesized with emergentism within evolutionary theory any more than apples can be synthesized with oranges. They are both legitimate yet separate categories of biology, dealing with different phenomena. A synthesis of an emergence thesis and a selectionist antithesis can only be achieved at a higher dialectical level, namely life itself, in the form of a biological synthesis. To be more specific, it is a synthesis of life’s history in both its evolutionary and static phases. If readers were to start this book at the final chapter—not unusual among bookstore browsers—they might wonder what such a synthesis might mean, and why such a

402

Chapter 11

thing should be necessary. It is necessary because understanding the history of life needs both the component of evolution, which involves discontinuous, complexification on a biological time scale, and the component of dynamic stability (i.e., the selection syndrome), which has dominated the history of life on a geological time scale. It is necessary to see the two components as different from one another, and that emergent evolution results in new phases of diversification and dynamic stability. It is necessary to end selectionism’s claim to all evolutionary change, and to disestablish the Modern Synthesis. This then is my “purpose breaking only as it is fulfilled.” My particular synthesis of biology is not original. It has historical roots in Lamarck, who saw the importance of progressive evolution, emphasized the individual organism, and suspected that adaptation to environment was a disruption of evolutionary gradation. Its historical roots in Darwin are weaker, since he did not know how to deal with the evolution of the “lower” to the “higher,” denied the saltatory nature of emergent evolution, addressed mainly the processes of dynamic stability, and elevated natural selection to the status of a secular creator. His critic St. George Jackson Mivart was much closer to an evolutionary synthesis of life. His sense of the history of life as a series of intermittent equilibria and disequilibria was combined with a protostructuralistic mechanism of autonomous progress. Much of the theorizing of Mivart’s contemporaries, the neo-Lamarckists, was also relevant to a biological synthesis. Curiously, the form given the Modern Synthesis by its British founder, Julian Huxley, was originally close to a theory of biology. Its goal was to understand evolutionary progress, by amalgamating as many biological subdisciplines relevant to evolution as possible. However, his version succumbed to the mathematical neoDarwinists and evolutionary ecologists. The American version, which opposed the very idea of evolutionary progress, was never a comprehensive synthesis in the first place, and was little more than genetics and population biology in an ecological context. Thirty years ago I was assaulted by a book title during an innocent visit to the library stacks of Glasgow University. It was Space, Time, Form: The Biological Synthesis, published in 1962 by the gadfly evolutionist Léon Croizat. And it came close to a synthesis of biological history. Its major theme is the diversification of evolutionary lineages in relation to biogeographical changes, but Croizat understood the importance of the progressive evolution of adaptability, and also proposed mechanisms of orthogenesis. The evolutionist Gareth Nelson, also much intrigued by Croizat, proposed a “theory of comparative biology” that also came at the central problem from the periphery of diversification (Nelson, 1970). My present perception of a synthesis of biology has a thematic commonality with Niles Eldredge and Stephen Jay Gould’s punctuated equilibrium. Theirs was, however, closer to the mainstream of evolutionism. They proposed that speciation was relatively rapid, and was followed by a long phase of dynamic equilibrium until the

A Biological Synthesis

403

species perhaps became extinct, or again rapidly bifurcated. The concept was flawed in that it erroneously emphasized speciation as a pivotal process of evolution, and did not adequately explain the processes of change, which, moreover, operate at higher speeds than they suggested. Furthermore, instead of following through, both authors took a defensive stance in the face of strong selectionist opposition, and apparently decided to opt for respectability over revolution. In an essay co-authored with Elisabeth Vrba, Eldredge also wrote: “A general theory of biology is a theory of hierarchical levels—how they arise and interact.”2 This is equivalent to my argument that new hierarchical levels are emergences; their interaction within the organism is the ground of adaptability, though influencing and influenced by the external environment. Later in this chapter, I will return to the vexing issue of speciation as an evolutionary event, since it is relevant both to emergentism and selectionism. A comprehensive synthesis of biology would be required to explain the origin of life, as well as its hierarchical composition and subsequent history. That origin was certainly a grand emergence, if not a set of multiple emergences. But the analysis of emergent evolution provided in the previous chapters deals with organisms that are capable of reproduction. However much it might hint at the possibilities, it is insufficient to explain the origin of biological reproduction from non-living generative conditions. Therefore it falls short of the requirements of a complete synthesis. Nevertheless, it has a strong dialectical component in its integration of emergent evolution with stasis. It asserts that life has followed an irregular pattern of rapid evolutionary advances in complexity and adaptability. Where the agents of natural selection are diminished or absent, major emergences have undergone diversification, which has a component of non-progressive minor emergences combined with adaptational change. The latter becomes canalized into specialization, and the conditions are now set for extensive phases of dynamic equilibrium. For convenience, the two components of change and equilibrium are temporarily isolated and their processes set out; evolution first: Progress means an increase in complexity, which equates with improved self-organization and greater adaptability. Emergent evolution may be saltatory, or it may spontaneously occur at critical points, or thresholds, in continuous processes. There are both extrinsic and intrinsic causes of emergent evolution. Natural experiments in evolutionary change can occur randomly and nonrandomly at any time where the generative conditions exist, but they are resisted by existent equilibria.

404

Chapter 11

Natural experiments in evolutionary change are rapid in a time scale of microseconds for biological molecules, seconds for physiological changes, hours for behavioral changes, days to months for epigenetic changes. If these changes are not resisted by existing dynamic equilibria, emergent organisms can undergo adaptive diversification. Major emergences increase adaptability by having multifunctional features. Resistance to change can be overcome by emergent innovations that are immediately more successful than the qualities of organisms that are part of the existent equilibria. Resistance to change is avoidable by innovative, adaptable emergences that can range more widely and find environments where resources are favorable, and where predation/browsing is low. Resistance to change can be removed by the catastrophic disequilibration of existent stabilities. Here, adaptability is at a premium. The removal of existent stabilities allows diversification of innovative organisms that were previously obstructed. Progressive emergent evolution in animals means greater freedom to choose how and when to act. That animals should have such greater freedom increases their individual roles in generating further evolutionary change. Next, the nature of ecostases (dynamic ecological equilibria) that have dominated the history of life: Equilibria may be organismal (i.e., physiological) or ecological (involving behavior, the physicochemical environment, and intraspecific and interspecific relationships). Organismal equilibria, such as cellular and organismal homeostasis and homeorhesis, resist change but can be disequlibrated, usually by external conditions. Ecostasis is reached when the interactions of organisms in their environment result in only minor fluctuations of allele and species distribution. Adaptation to and specialization for environment, within species and genera, increase the diversity of species, and strengthens the final state of dynamic equilibrium. Adaptation and specialization are therefore components of diversifying

A Biological Synthesis

405

evolution that straddles the emergence thesis and the selection antithesis. Diversifying evolution is emergent evolution but has the consequence of ecostasis. A change-resistant ecostasis can be disequilibrated by a large-scale emergent evolutionary change in one of its interacting species, or by the emergence of a new type. Catastrophic disequilibration of ecostasis is more likely and more effective than the previous cause. It allows diversification and regrouping of the more innovative and adaptable types, which finally reach a novel equilibrium that will persist until a sufficiently disequilibrating cause occurs once again. As was noted in the introduction to this chapter, the biological synthesis contains both the component of progressive evolution, which involves discontinuous, complexification on a biological time scale, and the component of dynamic stability (i.e., the selection syndrome), which has dominated the history of life on a geological time scale. It proposes that the two components are different from one another, and that emergent evolution results in a slower phase of diversification with both emergentist and adaptationist implications, and finally long phases of dynamic stability. To give the synthesis a more distinct character, I propose that selectionism’s claim to all evolutionary change be dropped. To help disestablish the Modern Synthesis, what follows emphasizes the distinction between emergent causes and selection effects. To be candid, rather than leave neo-Darwinist population biologists entirely in peace, I would advise them to get rid of the expressions “natural selection” and “selection pressure” altogether, and find ways to express themselves without archaic, jejune metaphors. What Difference Does It Make? “What of it? What difference does it make?” cried Herbert Spencer Jennings in 1927, when biological emergentism was a current Idol of the Theatre.3 He hoped that one purpose would be the escape of biology from the domination of reductionism, by demonstrating that new emergent levels have irreducible novel properties. Biologists who thought evolution was progress from lower to higher might be rescued from accusations of anthropocentrism. Faint hopes; such criticisms are still heard from conventional evolutionists, and reductionism rides tall in its journey to the empty shell of ultra-Darwinism. A higher synthesis will integrate an emergent evolutionary thesis with the classical neo-Darwinist antithesis, recognizing that their complementarity clarifies the history of life. The big difference is that evolution has been removed from the neo-Darwinists’ Modern Synthesis and natural selection is no longer considered to have any causal

406

Chapter 11

validity for evolution. We have barely scraped the surface of the study of emergent evolution, so there is plenty of scope within the theory to proceed in the major causal arenas of symbiosis/association; physiology/behavior and development. If these are not taken to be trivial “proximate” causes, with natural selection as the governing “ultimate” cause, some intellectual progress is inevitable. The major goal of biological emergentism is to understand progressive evolution. Maybe I should take this final opportunity to emphasize that by progressive evolution I mean increase in complexity. And complexity is not simply a multiplicity of parts, but an effective ordering of the parts through self-organization. By self-organization I infer the interactivity of the parts that arranges them in a hierarchical dynamic structure and also maintains it—the way a single-cell human zygote does, on its way to becoming and being a person. There doesn’t have to be an increase in the total number of parts nor the total number of differentiated cell types to achieve greater self-organization, though some such correlations do exist. Progressive evolution is characterized by improved physiological and behavioral attributes. The consequent freedom of choice and action has a feedback effect on both developing and mature anatomy. Symbioses were major emergences en route to the origin of eukaryotes. Sexual associations and cellular differentiations were major emergences en route to the evolution of complex multicellular organisms. Social interactions also had feedback effects on behavior and physiology. And the organismal whole was influenced by environmental changes. Some resulted from biological activity that oxygenated the biosphere. Some came through the invasion of new environments. Some came out of unpredictable environmental contingencies: continental drift, volcanic catastrophes, bolide impacts, and wobbles in the tilt of Earth’s axis, along with severe climate changes. Contingencies or not, progressive evolution was inevitable in some evolutionary lines, because of the generative potential of the first living organisms. Therefore, Lamarck was not altogether wrong in his progressionism. These generalizations are not limited to animals, but affect the other kingdoms, such as the plants, to a lesser degree. Now that I have recapped the previous ten chapters in a single paragraph, back to my original purpose of linking progressive evolution and emergentism. To establish that link, I had first to shift your attention from demographics to the generation of change—from natural selection to variation. If you had felt that the random identification of adaptive traits through the quantification of fitness was enough, I had to persuade you to consider the more direct approach: the comprehensive examination of emergent novelty and an assessment of its place in the evolution of organismal wholes. When that was achieved it became possible to return to the syndrome of natural selection, as the establishment of prolonged periods of dynamic stability. Then it could be reintegrated with emergence in the biological synthesis. Without such a regrouping, the current quest for genes for everything, in order to validate a nineteenth-century superstition, would continue to make a mockery of biology.

A Biological Synthesis

407

Furthermore, it had to be appreciated that no emergent hierarchical level, from gene to society, is more important than any other. Those who figure that the end of evolutionary biology will be nigh when we’ve sorted out the proteins as well as the genes, must transcend that delusion too, and realize we will still be scratching the surface. The pundits who have pronounced the end of evolutionary biology are themselves finished. The rest of us have an exciting future of new ideas and new discoveries. That is how emergentism can gain fulfillment. The new purpose that transcends that figured end is a synthesis of biology. In Agnes Arber’s seminal work on the philosophy of biology, The Mind and the Eye (1954), she notes how the mind can apply different ways of seeing the same evidence, a theme adopted by the structuralists A. J. Hughes and David Lambert (1984). It is a different way of seeing to observe the genesis of evolutionary novelty, as opposed to its fate. The clay of life periodically, autonomously, and spontaneously emerges to more complex levels. When generative conditions are right, a pot is produced—a thing in itself; there is no magic potter for it to turn against, as the Darwinist E. B. Poulton objected. The market may find it wanting, of low value, and reject it; but that affects neither the pot, nor its intrinsic nature, nor its origin, nor even its continued production. A different perception of natural selection is necessary for a new outlook. Another purpose of emergentism is to boost out of the genocentric universe to one populated by real organisms. Ultra-Darwinism, in combination with reductionism and media hyperbole, has polarized modern biology, the social sciences, and public opinion. May we now ask what difference it would make if natural selection were seen as a consequence of, and subsequently an obstacle to, evolution, not as its cause? How then would diversification be explained, and where would speciation fit into the new scheme? What difference would it make to biology in general if an emergence theory of evolution were accepted? What difference is there between futures predicted on the basis of neoDarwinism and emergentism? What difference would it make to chemistry, physics, and cosmology; or, going the other way, to anthropology, psychology, sociology, and economics? Does it suggest why humans do what they do? What difference does it make to ontology, and to the gulf between quality and quantity and that between subjectivity and objectivity? Natural Selection as a Barrier to Evolution? What if, without prejudging alternatives, natural selection were better understood as the hypostasis (or imposer) of dynamic stability? That is how selectionists actually see it, when they open their eyes. The conventional journey has already taken us close to that vista since normalizing selection and “stable evolutionary” strategies are the most common conditions. The exception, directional selection, may only be the

408

Chapter 11

consequence of autonomous amplifications and epigenetic constraints. The reverse of the question is more important: what might be biologically possible in the absence of the agents of natural selection? There is no longer any heresy in the notion that the Eden of life was a natural laboratory where a warm dilute incubating medium could have been seasoned with organic spices from outer space, or that it could have been a pressure-cooked Hadean brew of sulfur and seawater. But, before all, it was vacant. Emerging founders were free to experiment with all possible variations that met the primary qualifications of integrity and reproduction. They increased in number and experimented further without impediment, until they saturated the available space and ran short of resources. The obstacle of natural selection was encountered for the first time when they began to compete with each other and prey on each other. Freewheeling experimentation was not terminated. But at that point the only experiments that enjoyed immediate success were those that adjusted organisms to existing dynamic equilibria of the environment and their own inner coordinative conditions. Thereafter, only the rare experiment could transcend the hypostasis of natural selection to re-enter Eden. Sometimes the laboratory itself was reconstructed, as when the biosphere was oxygenated. Sometimes a random catastrophe cleared the bench to allow the survivors to try out their own new experiments. Sometimes adaptable emergents moved to a brand-new environment, from sea to fresh water and then to land. However, where environmental change was gradual, natural selection continued to resist evolutionary experiments. Although dramatic demographic shifts might have occurred, they involved existent forms, except on rare occasions when novel emergents had such radically improved or novel features that the old competition ceased to matter. Natural selectionism, reductionism, population thinking, selfish-gene-ism, ultraDarwinism, “scientific” journalism, and fear of the unknown demand that the universe be comfortably simple, and therefore genocentric, if not merely molecular. Thus, it would take the full perception of natural selection as a barrier to evolution, and appreciation of the generative evolutionary processes of emergence, to shift to a holistic universe, where organisms are recognized as real entities that internally order the action of their genes and externally respond to the outer environment, as well as changing it on occasion. Nevertheless, a synthesis of biology must retain an understanding of stable equilibria in populations and species, and so identify the value of conventional population genetics and ecology, devoid of the logical and polemical excesses of the old selectionism and ultra-Darwinism. Here is a brief example of what is right and wrong about current interpretations of neo-Darwinism. The following came in 2001 from an editor of Science, commenting on an original article in the American Naturalist (once the organ of American neo-Lamarckism); its authors are not responsible for what the reviewer writes:

A Biological Synthesis

409

Natural selection is the pervasive force shaping the evolution of living organisms. Selection can take several forms—directional, stabilizing, disruptive, indirect—and can act in different ways on different organismal traits. In recent decades much research has been devoted to measuring the strength of the various types of selection on phenotypes and quantitative traits both in the wild and in the laboratory. Kingsolver et al. [2001] analyze this literature and uncover some unexpected patterns. In both vertebrates and plants, the strength of selection on morphological traits was twice as great as on life-history traits. Strength of selection on some components of fitness, such as fecundity or mating success was greater than on others such as survival; the strength and frequency of stabilizing selection, which keeps a trait constant, was no greater than that of disruptive selections. This synthesis provides a fresh view of the complexities of the evolutionary landscape and of the statistical hurdles that need to be cleared.4

Pervasive, shape-shifting, and versatile, the ghost is still in the machine—the magic potter is still throwing the pots. And revelation will come out of a bigger computer with a better stats package. Apples are compared with oranges, though some fruits, such as physiology and behavior, are still forbidden. Well, maybe survival as opposed to fecundity has something to do with physiological adaptability, and it’s usually the specialists that are numerically superior to the generalists. To be fair, Kingsolver and his colleagues usefully tested common assumptions of neo-Darwinism—for example, that normalizing selection should be more common than disruptive or catastrophic selection. It isn’t. But it’s an exception that proves my rule. The dynamic stability that usually predominates in an ecosystem does not have to be homogeneous. In the classical case of disruptive selection, the various color morphs of the escargot Cepaea nemoralis are adapted to microenvironments where there might be dark woodland, open heath, short grass, clumped rushes, and hedgerows. So the population shakes down into a mosaic of different morphs living in appropriate conditions—assuming that there is strict genetic determination of the morphs, and not a large ecophenotypic component, which is always possible. Despite the disruptive epithet, it is merely a lumpy dynamic stability. As to stronger selection for anatomy than life-history traits, there is a lot of environmentally sensitive flexibility (or broad reaction norms), even in strongly canalized morphogenesis, so the two cannot be arbitrarily separated. The rarity of directional selection reinforces my argument that natural selection—qua ecostasis—powerfully resists innovation. I am oversimplifying, but should not need to remind you tenacious readers that natural selection does not prevent evolution. Natural experiments in gene mutation, and in more complex arrangements, occur all the time, where the generative conditions are suitable. Natural selection blocks these innovations from demographic significance. If they show progressive improvements in adaptability, they may persist, but their numbers may not increase sufficiently to flag them with superior selection coefficients. This re-emphasizes how selectionists look at the adaptational distribution of alleles in abstract populations. These gene pools don’t really

410

Chapter 11

exist in real environments that are home to real organisms in an entangled bank of physicochemical and biotic conditions. Such analyses and comparisons give an indirect measure of the emergent qualities of organisms under the prevailing conditions of competition or cooperation—i.e., interaction. When those change, a different feature may be found to be the “best adaptation.” In the meantime, one of the common results of competition is regression, not progressive evolution. Another epistemological caution should be raised in this discussion. Kingsolver and his colleagues have tested unwarranted assumptions and taken a broader comparative view of what they take to be the mechanisms of evolution. If they were to meet the real challenge—testing the fundamental assumption that natural selection is an evolutionary cause—they could really shake the scales from their readers’ eyes. Here is another exercise in translating from selectionspeak to a more objective language. In Making Sense of Life (2002), Evelyn Fox Keller attributes the following aphorism to Garrett Odell: “Robust gene networks are the only networks natural selection can evolve.”5 I have already (in chapter 5) remarked positively on the value of Odell’s models for demonstrating how whole network systems have emergent qualities that are not found in the parts. That is not in dispute. But what does the aphorism mean to you, or to Keller and Odell? The sense is that robust networks are the only kind that can persist in fluctuating environments; in other words they are adaptable. But that property comes at the point of origin of the system, not as the subsequent effect of a metaphorical force that somehow metamorphoses into a mechanistic cause. Odell and his colleagues deal with system flexibility in a way that makes ad hoc figurative language unnecessary. So putting it back again is counterproductive. The Barrier of Natural Selection and the Rate of Evolution I have made it clear in earlier chapters that emergent evolution is rapid. Looking through a narrow window in time shows us molecular events occurring in microseconds; physiological changes taking milliseconds, hours, or perhaps days; and epigenetic experiments being field tested in days to months. Speciation might take years, but since that is not a pivotal aspect of evolution, maybe simply a slow consequence of emergent evolution, I do not count it seriously (q.v. below). In the absence of the agents of natural selection the continued diversification of natural experiments that preserve the integrity of the organism would potentially occur on a limited, biological time scale. In contrast to the emergentist view, the selectionist looks at the length of time taken from the origin of life to the present time, and that involves a geological time scale of several billion years. This concept of evolutionary time has led to the averaging of the natural selection of random point mutations to create potentially misleading molecular clocks. Where the paleontological evidence doesn’t fit, molecular biologists

A Biological Synthesis

411

complain about “the tyranny of the fossil record.” They need to accommodate the knowledge that molecular change is not only instantaneous but clustered, especially in the wake of large-scale genomic changes in the form of sequence repetitions, gene duplications, and chromosome and whole genome duplications. Then they need to subtract the long periods of dynamic stasis, when, even if molecular and organismal experiments are being tried, they do not register in the DNA of the type. Then they might begin to find ways of accurately estimating when emergences occurred and how long bifurcating lineages took to acquire their visible features. This is the difference that emergentism makes to the whole conception of evolutionary rates. What Difference Does Emergence Make to the Concepts of Adaptation and Adaptive Radiation? For selectionists, real evolution is adaptation under competition—useful, genetically determined variations are selected and become general features of populations. Therefore, the question “What is it for?” drives their primary methodology, closely followed by “What was the selection pressure that popped it?” I can demonstrate the inadequacy of this approach from personal experience. Tom Fenchel and Rupert Riedl conceived the thiobios—a trophic structure based on the energy of sulfur—in 1970. Shortly after they published, Fenchel explained to me that the benthic thiobios coexisted with the primary productivity of photosynthesis in the water column above. Since either kind of primary product was useable as the main food source of the clams that I researched, I was mildly baffled at the significance of the thiobios. However, it was illustrated dramatically when thiobiotic marine thermal vent communities were discovered about a decade later. The point was driven home in my later studies of lucinoid bivalves, whose symbiotic nature and place in the thiobios had just been discovered. During that research, I came across several earlier authoritative answers to the anatomical “What is it for?” question.6 Those ingenious adaptationist interpretations had nothing to do with the sulfur oxidation that is the primary source of energy for these animals. Unfamiliarity with the symbiosis led to meaningless explanations of anatomical variations in terms of their selective advantage. Yet because any selectionistic conclusion was enough to terminate the search before the real answer had been found, no one had questioned its deficiencies. It was valid within the paradigm, but it did not represent any kind of intermediate step on the way to the truth; rather it was an impediment to further progress.7 An emergentistic interpretation is radically different. The virtually all-or-none establishment of a sulfur-oxidizing symbiosis had multiple effects, including behavioral change, digestive-tract regression, paedomorphosis, and structural alterations in gill filaments that housed the symbionts. There was no selection pressure for any of these—there never actually is a selection pressure. The generative conditions were the clam’s epigenetic and physiological adaptabilities, in the context of a high sulfide

412

Chapter 11

environment that was inclement for many other animals, but happened to be inhabited at its fringes by sulfur-oxidizing bacteria. These then became symbionts when larvae fed on them. Even at that stage, the symbiotic lucinoid clams persisted at interfacial environments with access to reduced sulfur for nearly 200 million years, without major diversification. That only occurred in the wake of K-T, the Cretaceous termination event, when decomposing competitors spiked the soup with sulfide. For most neo-Darwinists, adaptive radiation, or any kind of large-scale evolution, is no more than slow, cumulative adaptational divergence. From the emergentist’s point of view, the typological, essentialistic organism, long despised by Ernst Mayr and his disciples, must make a comeback. The multifunctional emergent properties of an archetype are what make specialization of divergent adaptive lines possible. Although the directional exaggeration of simple adaptational features through selection are also possible, a number of the component steps are allometric critical-point emergences, and some may be saltations that occur despite selection. Once the archetypes of the animal phyla were in place in the early Cambrian, further saltatory epigenetic emergences produced the founding members of the classes and lower taxa. These were not progressive in all cases. In mollusks, experiments in shell form, combined with allometric shifts, exemplified by the huge expansion of the posterior region of giant clams, could have been all that was necessary for diversification. Paedomorphosis, which is common, is a regressive process. Regression of segments and limbs was important in the diversification of Arthropoda, following their initial repetitive differentiation. Class Crustacea has a great diversity of forms. In comparison, a casual observer might think of insects as being all much of a muchness in body plan, although diverse in habit. But primitive insects had to go through various emergent stages. Reduction of the number of legs as a function of Hox gene mutation has been demonstrated by experiments on the multi-limbed brine shrimp Artemia.8 Somehow the six-legged condition became an archetypal feature of all future insects. That sets us another problem. The number of legs in spiders is eight, and there are ten in decapod crustaceans. An analogous case is that of the unsegmented Cephalopoda: octopuses have eight tentacles; squid have ten. Are those numbers so crucially adaptational that they occurred and were selected numerous times in each group? Or did each group diversify from a few archetypal organisms that coincidentally happened to have the characteristic number of limbs along with a constellation of more meaningful qualities? In the case of the insects, a pair of antennae were also characteristic of their multiple properties, but wings were not. Numerous when they were first acquired, they were reduced to four and then to two.9 Later on, some acquired holometabolic metamorphosis—the striking caterpillar to butterfly and maggot to blowfly transformations. To achieve this, epigenetic innovations, especially those that caused anatomical differentiations and heterochronous delays in the larvae, were required.

A Biological Synthesis

413

It is easy to get lost in the evolution of form when discussing diversification. But some of the most significant emergences among insects came with the acquisition of symbionts, and the biochemistry and physiology that accompanied them. Wings are a sine qua non for taking to the air. But certain physiological functions and behavior are also crucial for flight, as well as for reverting to the water. As for the vertebrates, it cannot be emphasized enough how a sophisticated homeostasis is the foundation of their diversification, and how physiology, behavior, and development have bootstrapped each other’s evolution. Nor is the environment simply a backdrop to be adapted to. It has directly interacted with the internal milieu, imposed epigenetic change, limited some behaviors, and afforded open access to others. And while all that was going on, the environment was being geophysiologically changed by organisms. Saltatory emergence and orthogenesis are too unruly for the Modern Synthesis, which stands or falls by the premises that natural experiments are all small and random, and that those that are adaptational are gradually accumulated by natural selection. However, orthogenetic trends—in officialese, allometric shifts—are major components of so-called adaptive radiation. They are also quite “ruly” by epigenetic standards. Although the initiating experiment in replication slippage, gene duplication, cell duplication, or cellular topography is probably random, it can continue to amplify itself in its own lineage without involving “selective advantage” in its incipient stages. If it finally becomes disintegrative, it does not need a metaphorical force to supervise its extinction. When orthogenetic trends reach a critical point of improved utility the emergent potential can be magnified—organisms can change their behavior to let function follow form. As easily as Slijper’s little goat took to bipedalism, a reptile or a placental mammal whose forelimbs had been epigenetically reduced could do the same. Then it could choose to use forelimbs as hands, butterfly nets, stabilizers, gliding organs, and ultimately wings. But it would stop using them to dig burrows or to break into termite nests. Orthogenesis, as a motor of diversification, would displace natural selection to the end of the allometric assembly line. Simple adaptations, genetic variants that dominate in particular external or internal environments, are real phenomena. Their accumulation cannot result in progressive evolution, instead they draw our attention away from it. Eventually, most experiments in emergent evolution were to fall into dynamic stability, with the most efficient use of resources, specialization into new niches partitions, and diminution of the potential for further change. This is what the Darwinist image of the adaptive radiation wheel conveys. There is a hub, represented by the ancestral type, and its divergent, specializing descendants radiate out gradually, like spokes. It is a metaphor required by a system of thought based on the directional and gradual accumulation of adaptations. Ad hoc hypotheses explained the absence of intermediate forms: they were outcompeted by their descendants, and so rare as to

414

Chapter 11

escape fossilization. This is oversimplification, if not fabrication; most of the links are missing because they never existed. Because of these complications, the purest textbook explanations of adaptive radiation are given at the genus level. Darwin’s finches are perennial exemplars of evolution in action through purely adaptational diversification. But how pure is it; how might the revision of natural selection and application of emergentism affect its interpretation? Are the Galapagos finches the products of gradual selection of insensible adaptational steps, or of saltatory emergence? Their functional anatomy and behavior diversified relatively recently from that of a South American seed-eating ancestor. The beak and associated head structure of these finches are discontinuous, differential expressions of neural crest organizer cell action, and bone morphological protein genes (q.v. chapter 5). However, convention ignores the epigenetic origin of beak “monstrosities” to look instead at “the one true cause”— natural selection—especially when the conditions of life change. This essentially comes down to competition for food. If only small seeds are available, small-beaked birds have the highest fitness. Also, the exclusion of large-beaked finches may additionally change the genetic environment so that the beak size may be affected, say by increased homozygosity. Just how much speciation in the conventional sense has occurred in the Galapagos genus Geospiza is debatable. Peter and Rosemary Grant, those sterling observers of the Galapagos finch species, have found that their hybrids are viable.10 John Gerhart and Marc Kirschner argue that developmental studies can help us “to understand the specific factors and interactions altered in the course of selectiondriven evolution.”11 From what they immediately add, I can only understand such a statement as another genuflection to selection by authors wary of assault by ultraDarwinists: The rapid diversification of beaks suggests that any of a variety of mutational changes may be enough to generate new morphologies and change size, perhaps reflecting the manifold regulatory contingency of beak development. The emergence of successful or functional morphologies in variants may not be a complete accident, but may reflect the kinds of hierarchies and compartmentalization of developmental processes, ones that tend to produce functional or at least nonlethal anatomies, rather than ones with serious mechanical defects. This shows up on the robustness and flexibility of the processes. . . . In light of the network’s characteristics, change is probably not random.12

And yet it is “selection-driven”? Besides the fecundity and brief generation time of Darwin’s finches—birds born in the spring can be grandparents by the fall—there are two possible reasons for rapid demographic responses to environmental change. Beak epigenesis might still be more loosely canalized than in the ancestral continental populations. But it may just be a pre-existing polymorphism, some of the beak morphs normally occurring only in small numbers at the fringes, or recessively lurking out of sight.

A Biological Synthesis

415

From the emergentist’s point of view, the finches that were rafted or blown to the Galapagos from South America entered a pristine environment almost devoid of avian competition, but which immediately called into action remarkable physiological and behavioral adaptabilities that the birds already possessed, prior to adaptational adjustment. Clio’s African grey parrot Mungo—a seed eater, and very fond of sunflower seeds and pistachios—is intrigued by the taste of snails, and the crunchiness of their shells, but does not care to have his beak gummed up with mucus. However, he will ingest and digest mangoes, ice cream, earthworms, wireworms, chicken wings, prawns, crab legs (shell on), barbecued steak, and parts of visiting primates that get within his reach. His behavioral repertoire goes beyond mimicking sounds, to associating words with objects, reading body language, and cajoling us to communicate, entertain, and preen him.13 But his wild conspecifics are constrained by competition and predation to be less versatile. This might underrate parrot cognition. Clio has also done some field research on the New Zealand alpine kea parrot under the direction of Ludwig Huber. She remarks that wild kea are not neophobic, i.e., scared away by strange things. They show great interest in human activity and can be induced to explore human artifacts. But they also exercise their option of becoming bored and flying away, rather than sticking around and responding to the more complex experiments that are possible with captive birds. In some regards birds are more adaptable than humans. A television documentary was made about the “death zone” above 6,000 meters on Everest, where so many climbers have died, even when equipped with supplementary oxygen. It failed to remark on the Himalayan crows that flew overhead, wondering what all the fuss was about. G. Ledyard Stebbins’s comment that the ability of the bar-headed goose to fly high in the Himalayas was due to the selection of a mutated gene for hemoglobin was unmindful that it was a small part of the emergent constellation of birdness, because that, of course, would be essentialism.14 And there are the Arctic cormorants whose adaptability allows them to regress to an apparently non-adaptive condition—to diminish their thermoregulatory ability under conditions of extreme cold stress—and yet thrive. Darwin’s finches are birds; not disembodied beaks, far less genes for beaks. When they first arrived in the Galapagos, their avian adaptabilities enabled them to deal with dry or wet, hot or cold conditions, to eat and digest seeds, fruit, insects, or snails. Founder effect and hormonal destabilization may have strongly influenced epigenetic homeorhesis in the small population of invaders, restoring a primitive plasticity to beak formation until populations filled out and selection took charge, putting a premium on birds with beaks that could keep up with the competition. They could experiment further with functional morphology, complemented by behavioral and physiological adaptability, given more freedom from natural selection. But selectionism is forced into ad hoc adaptational speculation. In contrast to conventional

416

Chapter 11

adaptive radiation, Darwin’s finches illustrate the constraining effects of competition, and the speed at which adaptability and diversifying experimentation can act in its absence. I once looked after a baby robin whose exploratory bent had taken far enough for its nest to be unfindable. After a day, and much gulping of worms from a pair of forceps, it imprinted on me, flopping along as fast as it could to stay located between my feet. Every time I glanced down to avoid stepping on it, it looked up at a 6-foot, 200-pound colossus and gaped for more food. This hopeful little monster would not have lasted another hour in the wild, under the scrutiny of ravens and bald eagles. But given protection from the slings and arrows of natural selection it might have started a new lineage of robin pioneers, ready to take full advantage of their essential birdness. Another example of adaptive radiation is an even better trial of the Darwinist rule. The cichlid fish can “evolve at a dizzying pace—hundreds of species live within just three African lakes, and many of them seem to have emerged overnight.” This is a 1999 Scientific American editorial comment regarding an article by Melanie Stiassny and Axel Meyer, “Cichlids of the Rift Lakes.”15 “Overnight” is not an exaggeration in a Darwinist time frame, since much of the speciation in Lake Victoria occurred after it dried out and then begin to refill 14,000 years ago.16 The windows of opportunity got narrower in Lake Nabugabo, which was isolated 4,000 years ago, and has five endemic species, and at the South end of Lake Malawi that was “dry two centuries ago, but now has numerous species and color morphs that are found nowhere else.”17 By this time you should not be surprised, since you know that by our interpretation of emergent evolution any organism invading a pristine environment can diversify as fast as its isolation, the resources, its behavior, its physiology, and epigenetic and genetic accommodation will allow. And we have already been talking in terms of months and years, not geological epochs. This argument is complemented by Stiassny’s and Meyer’s own explanation of the generative conditions of the cichlids’ evolutionary diversification. First they point to the provision of isolated environments because of climatic changes. Next comes the adaptability of the double jaw structure, which can be put to use for eating plants, grazing on encrusting algae, picking off insects, mollusks, ingesting sand and sorting out edible infauna. The cichlids may eat zooplankton and the eggs, young, and adults of other fish, and they may even nip off the scales of other cichlids.18 The inner jaws evolved in the same way as, but later than, the outer mandibles of all gnathostome fish—by exaptation of gill arches. Although the feeding behavior is usually species specific, the double jaws remain epigenetically malleable: “They can change form even within the lifetime of a single animal. (Even the teeth might transform, so that sharp, pointed piercers become flat, molarlike crushers.) Cichlids that are fed one kind of diet rather than another can turn out to look very different.”19 Loose canalization of development has allowed the parallel evolution of five behavioral and morphological specialists in Lake Tanganyika and

A Biological Synthesis

417

Lake Malawi. Another condition for emergent diversification is an unusual degree of parental care. Many cichlids protect their eggs and young in their mouths, where they can also share scraps of food. This parental care allows persistence of the species despite relatively small numbers of offspring and small populations. Where Does Speciation Fit In? The diversification of Darwin’s finches and African cichlids brings us right to the speciation problem. We can deal immediately with one evolutionary aspect of speciation, without serious objection from any faction. It pins the ratchet of evolutionary change by diminishing the loss of phenotypic, organismal uniqueness through sexual reproduction. Imagine a world without biological species, where all types could interbreed—where pollen in the air could produce miscegenies as well as allergies. The speciation curb on natural experimentation reduces the generation of ragweed babies—gimcrack hybrids that would fail the integrity test. There is a tradeoff between evolvability and speciation, just as there is between evolvability and the phenotypic integrity of the organism. To refer, as I have, to species as a “mental construct” is not entirely dismissive—all our theories and concepts are mental constructs whose survivability is independent of nature. To me the biological species concept, based on the reproductive viability of the members of populations, comes close to a satisfactory pragmatic definition, if such a definition is required to justify the continued use of “species” as a universal taxonomic unit. Even so, it has fuzzy edges wherein lie “recognition,” and accessibility factors.20 My real objection was to the Darwinian notion of speciation by accumulated adaptations, which is a different can of (perfectly adapted) worms. Linnaean systematics had few evolutionary connotations, and so engendered little dispute, while making the immense diversity of living organisms seem more manageable. Classification into species brings order into confusing diversity, which was attractive to Enlightenment philosophers. Furthermore, the giving of names and the naming of names has always had mythic proportions for humans. Even in science there are metaphysical implications of possession and control by namers and definers. So the psychological impact of the species concept was already strong before Darwin came along. Therefore, Darwin’s explanation of the origin of such real categories must have been highly alluring or provocative, according to the mindsets of his individual readers. One way or another, they were all excited, which helped to mask the fact that Darwin actually had little to say about how species originated, and a lot to say about what happened afterwards. The conventional wisdom of The Origin of Species proposed that once new species had arisen through the adaptational diversification of a founding group then the major evolutionary rite of passage had been satisfied. “Speciation events” are still

418

Chapter 11

portentous, even for biologists who doubt the primacy of adaptation. Could it be that they are putting the cart before the horse; that they are instead referring to emergence events or simply environmental events that are then followed by speciation? That being the case, speciation could be epiphenomenal to emergent evolution and drastic environmental changes, rather than a causal process. This is not to say that epiphenomena are insignificant. For example, mind is an emergence, and is also an epiphenomenon of brain structure and function. And the causal impact of mind takes it far from being a mere babble of a brook. Nor should we rule out adaptational diversification as a process that hastens speciation. Yet speciation could also occur in the absence of emergent or adaptational evolution, as a consequence of isolation and genetic drift. To evaluate these alternatives, it helps to know how and when speciation has occurred. Most of the species with which we are presently familiar, through observation of generations of living representatives, or through the fossil record, or both, as in the example of Homo sapiens, are remarkably fixed. For selectionists, however, it is evolution when cryptic characteristics appear phenotypically under unusual environmental change, or simply when the numerical distribution of alleles is altered. To the contrary, the species as a whole may still actually be in stasis. Regardless of which stage of cladogenesis an organism has reached—the final bifurcation of a genus, or the primary bifurcation of a class—it belongs to what we would pragmatically call a species. But the former might be doomed to stay where it is, while the latter could be the archetype of the whole class, perhaps because it was in the right place at the right time, or because an apomorphic or key emergent feature gave it a specialized function, or an unusual adaptability, or because it was in an unusually loose state of primitive epigenetic canalization. In summary, speciation may occur in the wake of various events that provide some kind of isolation. These include any emergence that makes members of the old population unavailable for breeding and results in sympatric speciation. Second, it may come after catastrophic change in the environment that disequilibrates the old dynamic stability. Third, it may follow from the invasion of a new environment, with organismal relocation (usually referred to as “gene flow” in the literature of population genetics) obstructed somehow; i.e., standard allopatric speciation. These ideas are to be found in any textbook. But their conventional interpretation leaves the impression that speciation is an evolutionary cause rather than an effect, and would reject the notion that speciation is epiphenomenal. Random changes in isolated populations that culminate in biological speciation need not amount to significant evolutionary change—in contrast with evolution defined as changes in the distribution of alleles in populations. The event needs to be logically separated from the speciation. Another aspect of speciation that might be affected by an emergence synthesis takes us back through the Looking Glass to species as units of selection. These are the hierarchical levels at which natural selection is purported to act. In ascending order they

A Biological Synthesis

419

include genes, epigenetic cell lines (or creodes), organs, organisms, groups, species and ecosystems. Darwin sensibly settled for selection at the organism level, though his position was made uncomfortable by sexual selection, which seemed to go in the opposite direction to natural selection. Still, sexual choices happened in nature, so they must be natural too, though eliminated if they went too far astray. I have argued that selection is a consequence of emergence, but does emergentism see any real biological phenomena that are reflected in the mirror as selection of this or that unit? Where, as the selectionist might say, is adaptiveness manifested? Below the organism level, it is conceivable that any gene mutation, cellular differentiation, or organ exaptation might result in an improvement in survival—other things being equal at all the other levels. The only arguments that I would give selectionists on this point are, first, that all other things are never equal; there are epigenetic and homeostatic adjustments, and other differentially advantageous events that have to be considered in the organism as a whole. Second, such improvements might be dramatically sudden, rather than through slow adaptational accumulation. In any case, it is above the organism level that the Looking Glass war of selection units is waged. To keep the smoke in the mirror to a minimum, we might start by extending Darwin’s view of sexual selection. Now, although Darwin usually concentrated on the fitness of the individual organism, here the interaction of a mating couple came into the picture. And the selective choices were real. The sexual synergy of the pair makes it a whole that is greater than the sum of its parts—especially one that successfully produces offspring. The concept of a whole can be extended beyond reproductive interactions to groups that cooperate in defense, shelter, and nutrition as well. Conceptually we might be dealing with anything from biofilms made up of various unicellular organisms to insect societies or human social groups. Removal of an individual from such a whole might have a negative effect, as might rearranging the dispersal or behavior of individuals within the group. Thus, William Wimsatt’s definition and test of an emergent whole, in contrast to an aggregate, applies to such groups; q.v. chapter 7 of the present volume and Wimsatt (1997) on aggregativity. Remember that we are looking at the whole from the point of view of its “proximate” emergent properties, rather than its “ultimate” differential reproduction. To view the group in isolation in this way is still too restrictive, because there is interaction between it and the groups of other species in the community as well, not just as competitors, or predators or prey, but sometimes cooperatively or symbiotically. The emergent properties of the group, or the ecosystem for that matter, make it not just a “unit,” but a whole that is greater than the sum of its organisms. Moreover, it might expand, or, more likely, extend satellites of itself into new environments, where it might take over from the groups already in occupation of them. This is more a matter of ecological succession than evolution, yet it crudely recapitulates evolutionary history.

420

Chapter 11

To say that groups have holistic qualities is as far as it need be taken. It is unnecessary to venture through the Looking Glass, to insist that such groups are selected, or that they were created by natural selection in the first place. It is even less profitable to gaze through the glass darkly to where species are selected, because species, and for the most part populations, are not emergent wholes but aggregates. In a recent personal communication, Michael Ghiselin, whose idea of “the species as an individual” inspired the popularity of species selection, says that “species do all sorts of things that organisms do not. For instance, they speciate.” That puts any species in the ranks of emergent wholes. However, it is one thing to invent an inspirational analogy, and another to give it material reality by fiat. If we could escape the population-thinking bind of the mid twentieth century, we might see that species cannot speciate. Only individual organisms and some of their close relatives speciate. Those can be selected in the conventional sense because they are integrated wholes. But the rest of the species members get left behind. Unless a species is so few in number that all its members interact, it is not a whole but an aggregate. Therefore, species have nothing “selectable” that is not found in their parts, i.e. the individual organisms. Indeed, to put it more provocatively, species may simply be reifications (i.e., “There’s this big important word ‘species,’ so it must represent something real”). To take away, or rearrange the individual members of a species might have no holistic impact whatsoever, until those changes reduced the totality of the species to a group or groups whose emergent properties would then be affected by further attrition. For these reasons, I am inclined to accept V. C. Wynne-Edwards’s discredited 1962 concept of group selection and to dismiss the currently popular species selection. What Difference Doesn’t It Make? In the absence of appropriate theories or syntheses, life goes on, and so does biology. Most of the work that I have cited in this book was done by people who would automatically call themselves Darwinists or neo-Darwinists. Good biology is done by people who are fascinated by living organisms, their internal interactions, and their environment, regardless of the kind of theory that they superimpose upon them. If you look at the outstanding general textbooks on introductory biology that are presently on the market, you will read chapter after chapter, dishing out good biology without reference to natural selection, although the Find-The-Selection-Pressure game is becoming more intrusive. Good biology is even done by people who start from ultra-Darwinist theory and attempt to confirm, verify, or nullify it. Everything that is seen within the Looking Glass obeys the physical rules of action and reaction—and optics. It “makes sense,” even in mirror image. Studies of adaptations and selective advantage reflect the qualities of emergences under particular circumstances; so they are an indirect path to reality. But would life not be simpler, and logically more sufficient, if all those good biologists went directly to the crux of evolutionary change? That gives us part of a response to the next awkward question.

A Biological Synthesis

421

Does It Even Matter? In a 1988 essay, Marjorie Grene asks “Is evolution at a crossroads?” and concludes that the New Biologists and complexity theorists present a significant challenge to orthodoxy. However, “considering the history of the Darwinian tradition, the challenges it has overcome and assimilated in the past, I would predict that in this case again it is probably a Kauffmanlike expansion rather than a Goodwinlike replacement that will occur.”21 In my opinion, recent expansions (or shifts of evidence) have pushed the envelope of the Modern Synthesis to its limits, demonstrating that replacement is the only real alternative (West-Eberhard 2003; Kirschner and Gerhart 2005). We are stuck with a synthesis that has perennially resisted any kind of revolution, or assault, or even sober Baconian diplomacy. It has instead absorbed and expanded and forgotten where it has been, until the next lot of sea lawyers rise in the mess to tell it where it is now, and where it ought to be going, which is nowhere. And it is not easy to say so without the “irritated tone of voice” that Grene and West-Eberhard dislike, because the success of the Darwinian tradition has not depended on logic or evidence, but on sophistry, polemic, authoritarianism, me-tooism, and, worst of all, indifference. For many evolutionists, the statement “if it exists it must have been selected [because it must therefore be adaptive]” is a truism that makes anything to the contrary seem irrelevant. In attempting a theory of emergent evolution, and its synthesis with a re-invented natural selection, I have no claim to originality or priority. Nor does dissidence of itself generate good alternatives. But ever since The Origin of Species was published, those of us suspicious of selectionist explanations of evolution have looked to the more fundamental causes of biological novelty. Independently we have drifted into the three-ring circus, finding ourselves surrounded by empty seats and the ghosts of Mivart, Cope, Bateson, Schmalhausen, Goldschmidt, Croizat, and company. We know it’s the right place to be once we have seen what performances can be mounted, after the conservative ringmaster, natural selection, has been sent off. The idea that natural experimentation is what you put in its place was initially a glib response to divert attention away from the causal role of natural selection, en route to finding the intrinsic origins of evolutionary novelty. It left natural selection a complementary role of consolidating successful emergences by contributing to the re-establishment of physiological and ecological equilibria. What is “selected” is emergent quality, and it is the evolution of that quality that poses the major question. Natural experimentation generates other metaphors, some pertaining to laboratory experiments in molecular biology. It operates randomly within the constraints of its generative level, like scientists who have just developed a new technique saying “Let’s try it this way now, and see what happens.” But out of random experiments there emerge self-generating processes that continue along particular “lines of research.”

422

Chapter 11

Already in general use, the metaphor of natural experimentation is particularly relevant to epigenetics. Hopeful monsters are natural experiments. What kinds of conditions make them successful, and how complicated can such experiments be? They involve whole organisms in complex environments, not simply mutant genes, whether they arise randomly or under direction from the environment. We have plenty of evidence from nature and from the laboratory to show how heritable characteristics can be acquired by transgenesis, and how radical changes can be induced in epigenesis. How many more of our laboratory experiments might have been tried naturally during evolutionary history? A final response to “What do you put in its place?” must deal with the metaphysical gap that would be left by the removal of natural selection from popular consciousness. This brings us back full circle to adaptiveness, a usage I have tried to avoid because it conflates adaptability and adaptation, and is in lockstep with natural selection. Adaptiveness, if it must be retained as a biological term, should have the primary sense of contribution to the persistence of organismal wholeness, which is often self-sufficiently independent of external conditions. Postmodernism wants to leave the wreck of the Modern Synthesis as an Ozymandian memorial to paradigmatic obtuseness. I can’t be too critical of the movement since I myself am skeptical of institutionalized ideas. But I have a mission to launch a theory of emergent evolution, and try for a biological synthesis, rather than to bob around with the motley flotilla that presents the postmodern smorgasbord. I don’t want to go beyond reductionism with Koestler and Smythies (1969), beyond neo-Darwinism with Ho and Saunders (1984), or beyond natural selection with Wesson (1991). I want reductionism and neo-Darwinism and natural selection to be seen for what they actually are, and then find a different synthesis. Such a synthesis would be interdisciplinary as well as dialectical, since it must involve more than the conciliation of a set of apparent contradictions. Post-Lamarckism, structuralism, complexity theory, the lucky-strike paradigm of neo-catastrophism, evo-devo, and symbiosis studies all focus on important elements of evolutionary causation. But their individual adherents, whether modern mutineers or postmodern privateers, lack the resolve to escape the vortex of Darwinism. If they do not all hang together in a new synthesis they will all hang separately, to be scavenged by the Modern Synthesis, stuck in the hold, and forgotten. That seems to have been the fate of “the New Biology” of the 1970s. Although it tried to change the paradigm, it needed more than groupthink enthusiasm to hold it together. From the outset I have shown that the most important emergences involve progressive improvements in the general features of biological self-organization, and physiological and behavioral adaptability. In the case of plants—lest my zoocentrism makes me forget—the emphasis has been on reproductive adaptability, bootstrapped by the evolution of animal pollinators. These progressive evolutionary changes hang

A Biological Synthesis

423

on through self-sufficiency, without high fitness, during the prevailing times of dynamic stability. In other words they persist regardless of natural selection qua differential reproduction. Emergentism does not contradict the perception that most dynamic structures in their development and actions are correlated with the persistence of the organism in its own being and the persistence of its descendants in future generations. In that sense it does not contradict the role attributed to natural selection in popular consciousness, and it accepts that the metaphor of selection pressure has been an inductive guide on occasion. However, it rejects outright the reification of metaphor as a creative force, whether it be selection pressure or Lamarck’s organism-respondingto-a-need. The re-invention of natural selection should be enough to shiver the timbers of the Modern Synthesis, and confront ultra-Darwinistic extrapolations that have already infected anthropology, psychology, economics, and philosophy. Since the universe never unfolds the way it ought, selectionists, faced with such a prospect, are more likely to go into a state of denial, or to shoot the messenger. The spectators hate to be told that the Emperor is naked, because they knew it all along, and it is less humbling to continue the mass pretence than to come out and admit self-delusion. But beyond whistle blowing, emergentism proffers a holistic emphasis on the importance and connectedness of every aspect of life. For this a metaphysical yearning existed long before Darwin, and far beyond biology. Deeming it irrelevant, analytical selection theory has failed to satisfy that need. As a minor effect the constant adduction of natural selection to any evolutionary insight would no longer be necessary, and this drain on freedom of thought would cease. Casting around for the explanation of any biological phenomenon in terms of gradual accumulations, subtle correlations of insensible advantages, and arcane selection pressures would be unnecessary. The weak knees of dissenters would enjoy relief from genuflection to the hypostasis of natural selection. The significance of adaptiveness would not be diminished, but would not be a Looking Glass reflection of emergent quality. As adaptability, it would be seen in the context of complexification of the essentials of life, and as adaptation it would remain a relative feature that contributes to stable equilibria, whether physiological, developmental, or environmental. The larger problem for biology is simplistic reductionism, to which ultra-Darwinism is naturally allied. The solution lies with realistic holism, also known as interactionism, the natural ally of emergentism. Nevertheless reduction would remain paramount in some thought experiments, and in the laboratory, where molecular biology would continue to demonstrate mechanisms that participate in emergence. The reductive and synthetic consilience that E. O. Wilson calls for is necessary, but not in a form where reductionism leads as “the way and the light.”

424

Chapter 11

Realistic holism need not lead to an unfocused research program that tries to do everything at once. Particularized research would continue as usual. A comprehensive comparative biology with an emergentistic outlook would be a check on the hubris of specialists who believe that their topic is the most important in the business. No single discipline has a lock on the hidden treasures of evolution. Particular research results need to be cross-referred to other programs, so that nobody is out of the loop. Holistic evolutionary studies should discover the lateral relationships between traditional disciplines—the three ring circus of epigenetics/form, physiology/ behavior/function, symbiosis/society under the big top of an ecology that does not forever strain to abstract the environment to numbers. Within them there are hierarchical relationships—functional morphology relates up to behavior and down to biological molecules. Intolerant reductionism has no place in an emergence program; nor does a hylozoism that forces the characteristics of higher emergent levels onto lower ones. New rules emerge at each new level of progressive evolution, and their identification requires knowledge of the organism and its relationships. These common sense guidelines are already followed by a good many biologists, but at a time when organismal biology is nearing extinction they need to be loudly proclaimed. The significance of founder effect would be elevated, since innovation finds its best expression among a small number of organisms with closely related genetic and epigenetic combinations that are not overwhelmed by a disapproving crowd of old family relatives. Speciation is already being played down as a central issue of evolutionary theory within the Modern Synthesis, so emergentism would only re-emphasize that the origin of species is not the pivotal process of evolution. Cladistic systematics, which concentrate on emergent characteristics of bifurcations, would be supported and extended, since emergentism emphasizes a comprehensive understanding of generative conditions and the full constellation of emergent properties. However, some distinctive apomorphic features of a clade-founding type may have no adaptive meaning. Gout is not characteristic of Dalmatians because it has some cryptic adaptiveness for spotted firehouse dogs, but because the few original sports were careless with the genes for nitrogen metabolism; and the same may be true for gout in the hominid lineage. At a given focal generative level there is insufficient time and discrimination for internal relationships to be fine-tuned to perfection before it takes off again to a new emergent level. Emergentism would give different direction to research. The greatest unknowns were in epigenetics and developmental evolution. The Modern Synthesis has traditionally been hostile to these because of their saltatory implications. “Evo-devos” are increasing in number. Old hypotheses based on physicochemical morphogenetic fields are being revived. Once an almost perennial theme of evolutionary embryology, they had been discarded by the genocentric phase of the Modern Synthesis. The

A Biological Synthesis

425

widespread phenotypic consequences of early changes in cell lineages are being analyzed. And the significance of the deep homology of homeotic gene groupings that universally organize limb and eye structures has become so widely recognized that one neo-Darwinist calls them “cliché stories”—in other words, “tiresomely awkward for the conventional explanation, so please stop going on about them.”22 We would be in a better position to make predictions if the accommodatory mechanisms for radical epigenetic evolution were fully understood. The human genome project is spinning off information relevant to the role of viral transduction, especially of bacterial genes, and its other discoveries might be taken further with an emergentistic rather than a selectionistic impetus. And the same goes for the evaluation of epigenetic drives in allometry or orthogenesis, if they are allowed to come in from the cold. There is no shortage of causes to be investigated in the light of emergentism. One outstanding but relatively unexamined database is in agriculture. And where better to start than Darwin’s 1868 workbook The Variation of Plants and Animals under Domestication? Though all of them might be recognized as legitimate by conventional evolutionism, they still need to be integrated into a single synthesis for which emergentism offers a framework. Who Cares? After about 70 years of neo-Darwinism and ultra-Darwinism, there are more than a few biologists and philosophers who see the limitations of selection theory, and some who, like me, reject it altogether as a significant evolutionary cause. Throughout this book I have alluded to the others I find significant. Eugene Balon (2004) provides further insight into the paradigmatic crisis that I mention in my introduction. “Irritated voices” abound. A final broadside at the ultras would be out of place here. However, there is one case history that I find striking. The biological historian William Provine is the author of The Origins of Theoretical Population Biology (1971), a book that I found most helpful while writing my 1985 book Evolutionary Theory: The Unfinished Synthesis. Provine attributes the strength of neo-Darwinism to the popular appeal of P. M. Sheppard and Theodosius Dobzhansky. The influence and success of Richard Dawkins and Stephen Jay Gould would seem to bear Provine out. In his 1988 essay “Progress in evolution and meaning in life,” Provine eventually came round to the opinion that the Modern Synthesis in America had been a constriction of evolutionary theory, more characterized by what it rejected than what it included. Interestingly, his afterword to the 2001 reissue of The Origins of Theoretical Population Genetics indicates that Provine has become more impatient with the wait. He now believes that natural selection has no causal role in evolution. Furthermore, he manages to avoid the “irritated tone of voice”:

426

Chapter 11

In 1970 I could see the origins of theoretical population genetics as being an unalloyed good for evolutionary biology, and thus obviously a great subject for a historian. Now I see these same theoretical models of the early 1930s, still widely used today, as an impediment to understanding evolutionary biology, and their amazing persistence in textbooks and classrooms as a great topic for other historians.23

Welcome to the circus. I suspect that many conventional biologists have the subconscious feeling that the current paradigm is bolstered by too many ad hoc hypotheses, and that the trend to ultra-Darwinism is puritanical rather than rational. Darwinist cant about fitness has already extended as excellence in everything from sport to education: the winner excels over all the rest, who therefore do not count, except as an admiring audience. Emergentism looks out for the rest, especially those who avoid the competition. Liberation from these burdens should be inspirational in itself. The Cosmological Foundation What difference does it really make to big stuff such as cosmology and the origin of life? This question plunges deeply into the metaphysics of emergentism. Asteroid and comet strikes are small potatoes in this context, although the one that created the moon about 4 billion years ago is hard to ignore since it probably extinguished any life that might have existed by that time. It gave our planet tides that were initially devastating to microorganisms with terrestrial ambitions, and leached the rocks at a faster pace than at present. And tides continue to make the water/land interface interesting. That vast primeval collision also affected planetary rotation and tilt, with climatic consequences that make Earth more hospitable to adaptable emergents. But these factors pale by comparison with the origin of the Universe. The appearance of non-biological cosmological novelties fits my general definition of emergence as a spontaneous innovation arising from the interaction of generative conditions. Calling the cosmological Big Bang an “enormous emergence” points to its instantaneous nature, and the subsequent complexification of matter with literal constellations of emergent properties. Cosmologists have a major problem regarding the initial, generative conditions, since they lay outside this universe. They were nonetheless critical if the resulting universe was to have the inherent property of producing and supporting organisms. It is easy to talk loosely about the “evolution” of the chemical elements, stars, planets, and pre-biotic chemical compounds. Given its unique initial (generative) conditions, such complexifications over the course of time seem to have been inherent in the Big Bang—not just probable, but somewhere and sometime inevitable under the right physical forces. Timing and location remain, however, unpredictable. In such a universe, with the passage of time, chaos must fall into an ordered complexity of attractors consisting of galaxies, stars, planets, elements, chemical

A Biological Synthesis

427

compounds, and organic molecules that constitute the generative conditions of the emergence of life. But this “evolution,” though it has emergent properties, has the oldfashioned sense of an unfolding of pre-existent characteristics, unlike biological evolution, which is reproductive and progressive. This was the position taken by the psychologist William McDougall in his critique of emergentism, Modern Materialism and Emergent Evolution (1929). He admitted that the formation of a water molecule and a biological emergence were analogous. There was a synthesis; the whole was greater than the sum of its parts, its properties could not be predicted from an intimate knowledge of oxygen and hydrogen. But he saw that the ability to form water was inherent in its elements; the synthesis was bound to occur under particular physical conditions, and only in a narrow temperature range would the fluid phase show its anomalies. Although the forms of crystalline water varied considerably, they were obedient to local physical conditions. Once such a phase shift had occurred it could regress to a more chaotic state, but a new emergent level could not build upon the structure of ice. Biological emergent evolution needs reproduction, which transcends chemistry, solid-state physics and quantum mechanics. Only living organisms can reproduce their qualities, and project their histories into the future. Only they evince the self-maintenance and adaptability that can differentiate, specialize, and become more complex in ways that further enhanced those fundamental properties. Even so, it is not quite that clean cut. There are intermediate areas of physiology and epigenetics where external physicochemical influences can persist for millennia without effecting reproducible gene changes. Although I have given an accurate account of McDougall’s distinction between physicochemical and biological emergences, Modern Materialism and Emergent Evolution was a critique of the “curate’s egg” style of emergence concepts found in the writings of Alexander, Morgan, Strong, Noble, Broad, Sellars, Wheeler, and Jennings. (A visiting curate was served a rotten egg for breakfast. When the Bishop asked him “How is your egg?” he diplomatically replied “Parts of it are excellent.”) The word McDougall used for heredity was “memory.” By equating the two he was attempting to bolster his argument that biological evolution was purposeful, under the influence of memory and mind. He was thus imposing a hylozoic point of view on all of life. Hylozoism makes for some strange bedfellows. McDougall also believed in the inheritance of acquired characteristics, on the basis of his own observations. It is tempting to go along with McDougall part of the way and leave it there. But is the qualitative barrier between non-life and life real, or is it illusory? Was the evolution of our cosmos bound to generate conditions for the emergence of life, with all of those novel properties? If so, the next phase of progressive complexification of its offspring would be inevitable, and mind must finally emerge, despite obstacles placed in its way by natural selection, and because of catastrophic events so disastrous as to almost wipe out any progress that had already been made. Therefore, Stuart

428

Chapter 11

Kauffman would be right in saying that we are “at home in the Universe” instead of leading a lonely, unique existence arising from almost impossible coincidences. Emergent evolution is about the erratic journey from the origin of life to mind. We can post-predict the conditions through which evolutionary progress was made, and say that terrestrial bipedal homeotherms are probably the best candidates for the final passage, but there are no inevitabilities about what lineage would have made it, or where, or when, or how. Paul Davies makes a useful metaphysical argument in The Fifth Miracle (1999). He says we have to choose between two alternatives. One is a universe where the origin of life is so improbable as to be almost miraculous, where, as Stephen Jay Gould said, the tape of biological history could never be replayed the same way twice. The other alternative is a panpsychic, deterministic universe, where mind is inherent in the Big Bang, and must inevitably be manifested. However, in an emergent universe there is a third, intermediate course of events. In the wake of the Big Bang, mass and energy are likely to become heterogeneously distributed. Some locations predictably produce carbon, oxygen, hydrogen, nitrogen, etc. Then there are rarer but still predictable gravity wells where they will all accumulate, tend to stay together, and interact. There, the potential for the origin of life might be high, though astronomically improbable in the rest of the universe. The evidence keeps pointing to an early emergence of life on Earth, almost as soon as it cooled sufficiently for liquid water to form. When life emerged, it had no pre-determined role to act out. It did, however, have the ability to act, and the ability to maintain itself, to reproduce, and to become more complex. We might see prokaryotes as helpless victims of powerful physicochemical forces. But in their microenvironments they do not behave entirely at random. This increases the likelihood that they will be raised, through intrinsic and extrinsic influences, at some place and time, to a new level of self-organization, to a new relatedness, which has new rules of performance that do not contradict, but may supersede the old rules. Thus, mind, as a manifestation of those novel internal relationships, becomes a likely outcome at the higher levels. But it is not predetermined by the early generative conditions. And it certainly does not reside hylozoically in the Big Bang nor in the simplicities of solid-state physics. Eric Chaisson might take a different stance. He is a natural successor of Alfred Lotka, the thermodynamicist who noted that evolution is related to the increase in the flux of energy through the biosphere.24 Chaisson’s survey of the expansion of “energy rate density” as a function of increasing complexity in the Universe is set out in Cosmic Evolution: The Rise of Complexity in Nature (2001). According to Chaisson, evolution, sensu lato, is driven by energy flow. It then follows that “the ultimate source” of order and complexity is cosmic expansion. In other words, biological evolution is not a cause, but an effect. This conclusion, he writes, is “sure to dismay most biologists.”25 Well, he sure got that right; when I first encountered the argument, I felt hoist with

A Biological Synthesis

429

my own petard—but not for long. Somebody else’s parallel analysis puts it this way: “For want of a nail, a shoe is lost. For want of a shoe, a horse is lost. For want of a horse, a rider is lost. For want of a rider, a battle is lost. For want of a battle a kingdom is lost.”26 So horseshoe nails, or the lack thereof, are the causes of political revolution? Give me a break! At consequent emergent levels there are spare nails, farriers, fresh horses, reinforcements, smart generals, and the hearts and minds of the people. An expanding universe has the potential to develop carbon etc. Carbon etc. have the potential to originate life. Life has the potential to complexify through reproduction— now there’s something the expanding universe didn’t have in its hylozoic mind—and it is as a result of biological evolution that energy flow increases in the biosphere. Cosmological emergences take on a different significance in the context of ideas proposed by Andrei Linde, and extended by Lee Smolin in The Life of The Cosmos (1996).27 Their fundamental principle is that in a meta-universal dimension, universes generate other universes that all may have slightly different characteristics, depending on their initial conditions. This excites one commentator to say the following: Then even if most universes originally were tiny and very short-lived, natural selection, the key Darwinian process, is bound to occur. Selection for what? For universes that are bigger and longer-lasting, since those will produce the most black holes and effectively “out-breed” the others.28

By this thesis, the type of universe most likely selected is one that generates stars that generate black holes that generate new universes of a kind likely to contain carbon and the potential for the origin and evolution of organisms. Now, not-so-gentle but rather case-hardened reader, if you have actually read the book in your hand instead of just picking it up to skim the last chapter, you should know what to make of all this, cosmological mysteries notwithstanding. But let us take it one last time, assuming that it is reasonable to apply the term “emergence” to cosmology. Emergences are sudden and spontaneous, whether saltatory or of the critical-point type. Their properties depend upon the combinatorial complexities of the generative level—matter/energy, gravity and whatever other ineffable forces might be involved in cosmic emergences. If the emergent organism/universe has the wrong total mass, or rapidly implodes, it dies without issue, and no additional agency is relevant. If it is able to reproduce, it will do so without depending on any other causal agency, and its offspring will reproduce according to their fecundity and the availability of matter/energy. Even by conventional standards of argument, natural selection only comes in where there is competition that effects differential reproduction. In the cosmic meta-universe, as well as under biotic conditions, space and resources have to be finite for competition to make a difference. A novel organism in the absence of competition is free to survive and reproduce according to its emergent properties. There is room for

430

Chapter 11

natural selection only where there is no room for all offspring to survive and reproduce. That universes with carbon-based life forms are most likely to produce similar offspring comes from their original emergent nature, not from natural selection. As a closed system (except that it might be spewing mass into the meta-universe through its black holes), a universe is subject to no competition. If it has offspring, their energy and matter are generated by the nature of the black holes from which they emerge, so there should be no competition in the meta-universe either. All that counts, according to Smolin, is that the type of universe that has the highest fecundity is the winner: It must also be stressed that at this formal level, concepts like “survival of the fittest” or “competition for resources” play no role. What matters is only that rate of reproduction varies strongly.29

A general emergence theory might help to focus the question of cosmic origins and evolution, but selection theory only applies to biological systems where differential reproduction is possible. And in fact a closer reading of Smolin reveals something more akin to emergence than selection; the birth of a cosmos, for example, is a “bounce.” Also, in complex systems, “new collective effects emerge” (including cosmic symbioses), and, in extreme cases of disequilibration, evolution is rapid until new stability is achieved.30 Though he acknowledges that conventional treatments of selection theory emphasize competition, “an important theme of evolution might be the ability to invent new ways of living, in order to minimize the actual competition among the species.”31 But then he lets Richard Dawkins justify his adoption of a selectionist stance: The theory of evolution by natural selection is the only theory we know of that is, in principle, capable of explaining the existence of organized complexity. Even if the evidence did not favor it, it would still be the best theory available.32

Thus Smolin commits to natural selection as a universal law, going as far as to equate evolution with “the history of selection.” But since competition and survival of the fittest do not enter his thesis at any level, what he is really proposing is a cosmological theory of fecundity. Moreover, what he finds interesting about evolution is progressive self-organization, which despite Dawkins is an issue distinct from selection theory. The Life of the Cosmos, and its review that I cited earlier, illustrates what a tangled web ultra-Darwinism can weave. As ever, Monkey is free to roam at will, provided he stays in the hand of Darwin, or perhaps Dawkins. When James Lovelock portrayed Earth, or Gaia, as an evolving organismic whole, Richard Dawkins objected that since Gaia had never been in competition with others of its type, natural selection had no part in its development, and so, by definition, it could not have evolved.33 The emergentist interpretation is that the absence of inter-

A Biological Synthesis

431

planetary competition meant that the only obstacles to Gaia becoming more complex and self-organized were its physical limitations, and the stases in its biological systems. It has gone through periods of emergent biospheric change, such as the conversion to an oxygenated atmosphere, and the emergence of complex ecosystems that interact in a planetary homeostasis, disrupted and reorganized by occasional contingencies such as asteroid hits. Gaia also has the potential to reproduce, by sending out space habitats and terraforming the other planets. But without that ever happening, it can be seen to have auto-evolved, in the sense of having generated life, whose biomass has increased and become more ordered, in the absence of competing Gaias. The Gaia concept also reinforces the idea that dynamic stability can be changed, both progressively and regressively. Such changes can be saltatory, as when a major bolide makes an impact, or they can be critical-point emergences, as when the gradual accumulation of atmospheric oxygen reaches the threshold where animals can leave the water and occupy the land. Since humans can effect such major changes, we have to take care in deciding what kind we want. Metaphysical Aspects of Emergence The following aphorism is J. B. S. Haldane’s: The fact about science is that everyone who has made a serious contribution to it is aware, or very strongly suspects, that the world is not only queerer than anyone has imagined, but queerer than anyone can imagine. This is a most disturbing thought, and one flees from it by stating the exact opposite.34

Is it not time that we confronted that queerness and tried to understand it instead of fleeing from it? The aphorism suggests why it has taken us so long to accept some version or other of a comprehensive theory of biology, instead of one that equates evolution with stasis. There are two poles that attract different kinds of extremists: one lot flees from mystery, by operating as if the world is neither queer nor complex. Like Democritus, they prefix their pronouncements with the phrase “nothing but atoms and the void (or molecules and genes).” The “nothing-but” reductionists may have fled from fear of the unknown, but they have also fled toward a strategic metaphysical position where they dominate not only biological philosophy but also command most of the practical resources of biology. Their evangelical zeal nearly comes full circle, resembling the romantic extreme of transcendental holism, which celebrates complexity for its mystery. Emergentists, or realistic holists, or interactionists, if I guess their responses correctly, celebrate the unknown for the challenge of understanding it. Any scientists who feel a warm glow at the realization that they do not have the slightest inkling of how to explain the problem nature has just set them, will appreciate this. In contrast, fleeing from the queerness of the universe, or sheltering it from investigation by pretending that it is simple or irrelevant, will certainly prolong the journey to the next level of understanding.

432

Chapter 11

Another potential philosophical value of emergentism was appreciated by C. L. Prosser, one of the most influential comparative physiologists of the twentieth century. While he unfortunately did not elaborate on emergent systems in physiology and ecology, he thought they had a deep metaphysical power equivalent to that of categories like beauty and value—not a notion to inspire reductionists.35 In the same vein, Brian Goodwin proposes an emergentistic science of qualities. Among human emergent qualities he counts harmonious, self-regulating (as opposed to authoritative) societies, holistic medicine, and play, all of which have strong subjective or uniquely individual features. Wisdom, as an emergent quality of human individuals, and, one would hope, their libraries and teachers, does not exist at the lower levels. Yet ultraDarwinism suggests that wisdom, along with human acts of systematic murder, rape with malice aforethought, militarism, and other aspects of the will to power were already programmed into the genes, and hence into behavioral algorithms, by the end of the Pleistocene, through the action of natural selection. By their argument, our best option is to be nobly objective, and not to hide from the conclusion that our evolutionary success has depended on these genetically predetermined features, which must be acknowledged and transcended. Any kind of nastiness can be rationalized in this way, although some who buy the argument are the first to demand that we take the miscreants and “lock ’em up, then hang ’em high!” The more insidious philosophical sin is to diminish our values in order to validate our prejudices. Emergentism does not promise a gentler and kinder universe; only a more thorough realistic, holistic analysis. There are many benign emergent properties of the human organism, such as aesthetic enjoyment and creativity, empathy and mystical awareness, together with more objective abilities of logical and empirical analysis, judgment, experience, memory and communicative skill. But, as Arthur Koestler pointed out, they can be overridden by mass adherence to an emergent ideology. People who are individually pacific can collectively go marching off to war (global or guerilla), to perform acts that to call “subhuman” demeans the rest of the animal kingdom.36 The structure of our behavior is not genetically determined. Nevertheless, benign human qualities emerged out of archaic neural structures and ancient hormonal ferments that can still rise to swamp them, often through a positive feedback amplification triggered by social stimuli. This emphasizes the urgent need for a true evolutionary psychology that does not beg the question that all behavior is the genetically determined product of natural selection. One small gain for emergentism could be the realization of Jennings’s hope for a release from accusations of anthropocentrism, when we talk about unique emergent advances in human evolution, and emergence to higher levels of organization in general. These evolutionary steps can be characterized in a variety of ways, including number of differentiated cell types, and the complexity of supervening levels of coordination. To say that the human is higher does not disparage the worm, but implies

A Biological Synthesis

433

that perfection-of-adaptation-to-environment is a totally inadequate assessment. Jennings’s hope for release from reductionism is also realized. Hylozoism claims that all features of complex organisms, having gradually accumulated, must be found to a degree in the simplest organisms. To the contrary, innovations that occur at each emergence have new properties that did not exist before. Quantity, Quality, and Subjectivity Goodwin’s science of qualities is an admirable goal. Unfortunately, historic fear of unknown qualities that might harbor vitalism, transcendentalism or intelligent design, has driven the popularity of simplistic and intolerant reductionism. Design has also been sanitized as a popular selectionist metaphor; “but of course we all know what such language really means.” Quantitative rigor has become an end in itself, to be substituted for qualitative observation, scrutiny of basic premises, and speculative induction. It is easily applied to the big questions, but it is anti-intellectual if cases where it is not as easily applied are dismissed as “tiny and uninteresting.” Here we can reply Yes to the question “Even if it makes a difference, does it matter?” If we require our students to prioritize quantifiability for their research proposals, we block all of the interesting and original observations that might be made in the field. Precise quantification must be secondary to identifying quality, and figuring out how it comes into being. Is there a metaphysician in the house? An ontological diagnosis, please!37 Can we make anything of C. L. Prosser’s suggestion that the schism between subjective and objective worldviews might be bridged by the insight of emergentism? Consider the difference between mind and intelligence, difficult enough to define objectively, though we try. “Intelligence” is a general term for an emergent property of a brain whose protein structure is genetically determined and anatomy genetically based. However, its development is modified by experience, and differs significantly at the level of neuronal connections, even in genetically identical twins. Mind is the larger subjective package of intelligence, together with education, memory, sensuousness, consciousness, and subconscious activity. During the entire life of the individual, feedback between mind and dynamic neuroendocrinal structure results in the modification of both. Why do humans do what they do? The constellation of mental properties represented by the word “mind” are not gene-determined. Nor can they be, if mind is to be adaptable, conferring a freedom of thought from traditional conditioning that makes intellectual progress difficult. This we can contrast with the stereotyped behavior patterns of insects and birds that have been strongly reinforced through natural selection. In humans, an analogous mechanism of repression and regression reinforces obedience to tradition. In recognizing that, emergentism has one hopeful ontological quality. By demonstrating that biological evolutionary progress can be made, despite the self-correction of stasis and its obstacles to change, it holds out some

434

Chapter 11

hope for theoretical progress as well. Mind is naturally highly subjective, even solipsistic. As Michael Polanyi argued, the closest that we can ever come to objectivity is a personal knowledge that has been subject to all of the structure and modification of our heredity, experience, education and social conditioning.38 Doctrinaire reductionism is a flight from such subjectivity. There is plenty to be afraid of, doctrinaire reductionism for example. But that is the price of imagination and creativity. Emergentists and their predecessors often had a nose for ineffable qualities like vital force and animal magnetism, many of which were finally identified and measured because suspicion of their reality persisted until experimental validation became technically possible. Cognition and freedom of thought are other such vital forces. The invention of some kind of artificial intelligence that will explain mind mechanistically, without resort to semantic reductionism is still an open question. One of the emergent properties that it would need to display is the subjective flash of insight that catalyzes change. That is a quality that gives a new meaning to “social butterflies” like Confucius, Buddha, Moses, Jesus, Mohammed, Darwin, Marx, and Einstein, who fluttered by with new ideas and rearranged the flowers on a global scale. As Koestler pointed out in The Act of Creation (1964), intellectual emergent properties have multiple manifestations in the humanities, arts and sciences, and not the least of these properties is humor. Jokes mimic creative insight, making us laugh at the bisociation of the punch line, and at ourselves for letting the narrative deliberately lead us astray. Nature does not mislead us; we are quite capable of doing it to ourselves, until parts that seemed dissociated converge and emerge as mental wholes. And we laugh at our own inability to see it before. Out of all that individual subjectivity comes productive thought, both qualitative and quantitative. And, if we know that emergences (both mental and evolutionary) are part of reality, we know that the end of biotic evolution is not nigh, and that we should continue to expect—and respect— surprises.39 I began this journey by identifying a need to understand the progressive evolution of biological persistence out of the emergent properties of life: self-maintenance, and reproduction—simple enough to say! Now I am back where I started, knowing that we still have not explained how life itself emerged with those properties, nor how mind emerged from them. We cannot depend on the discovery of a universal formulation for emergence in all its manifestations since new rules are generated at every level and are influenced by improbable contingencies. John Holland (1998) consoles us with the thought that in any system that runs long enough, unlikely persistent patterns will emerge. They are then candidates for combination with other persistent patterns, to emerge as larger patterns with enhanced persistence. His closing question for future study is, essentially, “How do they do it?” He cautions that “the Baconian approach of gathering data until significant relations emerge is unlikely to work because systems exhibiting emergence are so complex.”40

A Biological Synthesis

435

The latter warning came too late for me to heed. In any case, a Baconian study of the interactive arenas of biology gave me a modus operandi that provides multiple examples of the generative processes of association and repetitive differentiation (or modularity). It also clarifies the particulars of the intrinsic and extrinsic causes of the two fundamental types of emergent evolution: critical-point and saltatory. The comprehensive Baconian approach includes contingencies within the category of extrinsic causes and so makes it possible to synthesize all the aspects of emergent evolutionary causation. The need to explain the generative processes of evolution, to grant a central role for repetitive differentiation, to accommodate saltatory events, and to include extrinsic and intrinsic causes has been intuitively obvious to some evolutionists for more than a century, so I have actually been engaged in analyzing hypotheses to which others had already leapt. Nevertheless, I hope that my case histories have contributed to an understanding of how wholes not only have properties that do not exist in the parts, but also how such wholes come into existence. They can’t but fail to do so. Millennial ideas about the end of progress and “the end of science” are popular these days.41 Ever since Aristotle, we have suffered in every generation from the hubris of scientists who think that they have most of it sorted out. Some biologists believe that Darwin already had evolution all ship-shape and Bristol-fashion in The Origin of Species. Darwin knew otherwise, but he had high hopes for the differences that his theory would make to biology: When we no longer look at an organic being as a savage looks at a ship, as at something wholly beyond his comprehension; when we regard every production of nature as one which has had a history; when we contemplate every complex structure and instinct as the summing up of many contrivances, each useful to the possessor, nearly in the same way as when we look at any great mechanical invention as the summing up of the labour, the experience, the reason, and even the blunders of numerous workmen; when we thus view each organic being, how far more interesting, I speak from experience, will the study of natural history become! A grand and almost untrodden field of inquiry will be opened, on the causes and laws of variation, on correlation of growth, on the effects of use and disuse, on the direct action of external conditions, and so forth.42

Darwin went on to suggest future revelations from systematics, breeding experiments, and the study of archetypal body forms through paleontology and embryology. All have been realized in varying degrees. And there is no question that he did make the study of natural history far more interesting. On the other hand, the fields mentioned in his second paragraph remain “almost untrodden,” perhaps because “summing up” is an inadequate way of synthesizing evolution. During the passage of Darwinian Theory through the Modern Synthesis to ultra-Darwinism, Darwin’s unknowns of evolution have been cast away for being too neo-Lamarckist or too saltatory. What remains is little more than a hulk of contrivances, tingles, and dubious new ladings.

436

Chapter 11

The organism is still regarded as savages would have seen a ship. Analyzing its bits of wood and nails and sails does not lead to the realization that the ship is greater than the sum of its parts. A theory of emergence, and its synthesis with selection theory, are hopeful epistemological monsters. Whether as ideas or organisms, rough beasts have always slouched along the track of evolution.43 It would not have happened without them. Now that humans are able to manipulate future evolution there will be rougher still. But if we continue to canonize natural selection we will tread too much the dark side of the path. I would hope that after the stasis of the twentieth century, a nascent, generative theory of emergent evolution, and a general biological synthesis, would help to carry Darwin’s program forward. T. S. Eliot reassures us that some ideas have a greater destiny than their termination as meaningless husks: What we call the beginning is often the end And to make an end is to make a beginning. The end is where we start from. . . . We shall not cease from exploration And the end of all our exploring Will be to arrive where we started And know the place for the first time.44

Notes

Introduction 1. Williams 1966, p. 139. 2. Endler 1986, p. 51. 3. Goodwin 1994, p. 531. 4. Hawkes 1997, p. 14. 5. The phrase “Big Bang of Biology” was popularized by the newsweekly Time in its issue of December 4, 1995. It had earlier been used in Scientific American by J. S. Levinton (1992). 6. Grassé 1977, p. 5. This is the English edition of the original 1973 publication in French. 7. Wilson 1970, p. 369, citing Lyell’s journal entries of March 1860. 8. Aristotle referred to the rungs on the ladder of life as hypostases. See Lovejoy 1936. 9. Gould 1989, p. 154, citing Conway Morris 1977. Hallucigenia was presented as a creature that walked on spines and had dorsal tentacles. At first sight Conway Morris took it to be a bit of something else. Now he has turned it upside-down, making the tentacles legs, and identifying it as a kind of arthropod (Conway Morris 1998). Complexity theorists call bizarre experiments “kludges.” 10. Von Bertalanffy 1967, p. 82. 11. Darwin 1872, p. 60. 12. Szalay 1998, p. 342 (emphasis added here). 13. Muir and Howard (1999) warn that transgenes used in genetically modified food organisms might turn out to be such “Trojan genes.” 14. Bertalanffy 1952, p. 92. 15. Dobzhansky et al. 1977, p. 153. 16. Darwin 1859, p. 490. This was unchanged in the sixth edition of Origin, but Darwin had hedged on p. 95: “It is, however, an error to suppose that there would be no struggle for existence,

438

Notes to Introduction

and, consequently, no natural selection, until many forms had been produced: variations in a single species inhabiting an isolated station might be beneficial, and thus the whole mass of individuals might be modified, or two distinct forms might arise.” He should have left well enough alone. 17. Gould 1991. 18. The use of “high table” as a metaphor for the self-appointed elite of evolutionism is attributed by Eldredge (1995) to Maynard Smith. 19. According to Adams (1979), the term “gene pool” was introduced by Dobzhansky (1950) and popularized by Mayr (1963). 20. Gerhardt and Kirschner 1997, p. 205. 21. Mackie 1999. My colleague prefers to be called a Darwinist rather than a neo-Darwinist. The insertion of “God willing” into his discourse is a tongue-in-cheek response to a reviewer of Gerhart and Kirschner 1997 who wondered if the cubomedusan eye was evidence against the existence of God. The debate emphasizes my point that natural selection has been a secular replacement for the creator and that both require a certain degree of faith. 22. Bryan Sykes is leader of the group that analyzed the mitochondrial DNA of Cheddar Man. In Sykes 1999, he puts Cheddar Man in a larger context of European human ancestry. 23. For a useful elaboration of this, see Moss 2003. 24. The phrase “blanket utility” is from Baldwin 1896. 25. Wimsatt 1998, p. 271. 26. Gould 1983c, p. 152. 27. Therefore, the way of life and the exploratory behavior of organisms may initiate adaptational changes. 28. For a comparison of cosmological and biological evolution, see chapter 11. 29. This is paraphrased from an account of a Muller lecture by his graduate student (later my professor of genetics) Guido Pontecorvo (Cohen 1997). 30. Muller 1949. 31. One of the first people I find to have used “experiment” in this sense was Max Delbrück (1949). For fuller treatments of the uses and abuses of figurative language in biology, see Arber 1954, Oyama 1985, and Keller 1995. 32. Desmond and Moore 1991, p. 420f. 33. When I was writing Evolutionary Theory: The Unfinished Synthesis (1985), J. A. Shapiro had already used this molecular biological metaphor. He later extended the idea in his essay “Natural genetic engineering in evolution” (1992).

Notes to Chapter 1

439

34. Zambryski 1988, 1989. 35. Maynard Smith and Szathmáry 1995, p. 8. 36. Ibid., p. 145. 37. One reader of my manuscript thought I might be criticized for my assumed superiority in taking the “high road.” The metaphor is drawn from the song “Loch Lomond,” in which the low road is by far the preferable route; the high road is death by hanging. 38. The expression “art of the possible” was applied by Herophilos to medicine in the fourth century B.C. 39. Williams 1966, p. 139. Chapter 1 1. Eldredge 1995, p. 36f. I ask his indulgence for reversing his order of march. 2. Berry 2000, p. 117. 3. In Darwin 1872, see p. 60 for natural selection as a “false term” and p. 47 for “a power incessantly ready for action, and as immeasurably superior to man’s feeble efforts as the works of Nature are to those of Art.” 4. Darwin 1872, p. 392. 5. In Evolutionary Theory: The Unfinished Synthesis (1985a, p. 144ff.), I discuss the objections that were raised against the reductionism of cell theory by the “organismalists.” These objections were ignored by the next generation of reductionists who were to out-reduce the cell theorists by making DNA the lowest common denominator of life. 6. For discussions, see Greene 1971; Shapere 1971; Mayr 1974. 7. My historian colleague Judith Friedman volunteered this opinion about evolutionary paradigms prior to discourse about my own thoughts. For an earlier discussion, see Reid 1985a, p. 107. In afterthoughts about scientific revolutions, Kuhn (2000) thought that the more gradualistic model suggested by the Modern Synthesis could be closer to the truth than the mutationist/revolutionist model. “Alas, poor Yorick!” I thought I knew him. 8. Poulton 1908, p. 97. 9. Goudge 1961. 10. Cope 1887. 11. Wallace 1858, p. 60. 12. Schmalhausen 1949, p. 53. 13. The example of grassfinch behavior is from pp. 593–594 of Futuyama 1998. It was brought to my attention by Camilla Berry, whose comment is cited from p. 60 of her 2000 thesis. Robert

440

Notes to Chapter 1

Wesson devotes a section of his 1991 book Beyond Natural Selection to the caprices of creative selection pressure. 14. Williams 1966, p. 34. 15. Ibid., p. 255f. 16. Cohen and Stewart 1997, p. 136. 17. Horgan (1996, p. 255) uses “theological Darwinians.” Horgan vacillates on his own adherence to this faith. On natural selection in the guise of creator, see Skolimowski 1974 and Paterson 1982. 18. Schwartz 1999, p. 304f. 19. Ruse 1996, p. 493. 20. Ibid., p. 331. 21. Smocovitis (1996), citing Provine 1992. 22. Mayr 1980, p. 40, also citing Laudun 1977. 23. Ibid., p. 40. 24. Many a paradigm buff must have thought of this line. I thought John Casti (1989) had priority for using it as a book title, but a search of the Internet reveals that it is polyphyletic. 25. Kuhn 1970, p. 6. 26. Ibid., p. 67f. 27. I borrow this image from Eldredge (1995, p. 4). Eldredge calls it “High Table,” an expression attributed to J. Maynard Smith. To avoid mixing my thematic nautical metaphor, I prefer “captain’s table,” which still conveys the sense of privilege, elitism, etc. Eldredge perceives that paleontologists have received reluctant invitations to join the company at the table, but that does not make it any less elitist since epigeneticists, physiologists and those who research symbiosis and evolution are conspicuous by their absence. 28. Kuhn 1970, p. 88. 29. See Goodwin in Brockman 1995, p. 105. 30. Campbell 1993, p. 494. 31. Endler 1989, p. 51. 32. Ibid., p. 97. 33. Dawkins 1976. 34. One of the best current discourses in favor of the importance of organismal and the inadequacy of selfish gene concepts is developed by David Rollo (1995). The word “phenotype” is unfortunately corrupted by its genocentric definition as an expression of the genotype. See also Dover 2000.

Notes to Chapter 1

441

35. Darwin 1872, p. 107. 36. Clark 1975, p. 99, translating Spinoza’s Ethics III. 7. 37. Shapere 1980, p. 393. 38. Hamburger 1980, p. 99f. 39. Arthur 1997, p. 218, citing Horder 1994. 40. Nagel 1978, chapter 13. 41. Wallace, as cited by Mivart (1871) on evolution of mind. 42. Mivart 1871, p. 313. Mivart borrowed Galton’s metaphor of a multifaceted stone that would roll or stay put according to the energy applied. 43. Herbert Spencer (1862), cited by Oldroyd (1980, p. 205). 44. Mivart 1871, p. 113, citing Galton. Punctuated equilibrium was described by a waggish gradualist as “evolution by jerks,” but Eldredge (1995) exacts a fine revenge. Eldredge (1985) and Gould (2003) reassess punctuated equilibrium. 45. Keller 1983. 46. Mayr 1954. 47. Kimura 1960. 48. Ohno 1970, preface. 49. Butler 1886. 50. Lincoln et al. 1998. This idea comes from Van Valen (1965). 51. Darwin 1859, p. 41. 52. Wallace 1858, p. 61. 53. Darwin 1872, p. 59. 54. Kolesnikova 1987. 55. Belyaev died in 1985. For a summary of his Russian publications, see Sumnyj and Ruvinskij 1987. See also Belyaev and Trut 1987. For a list of Belyaev’s English publications, see Jablonka and Lamb 1995. 56. E. Darwin 1794, p. 320f. See also McNeil 1987, p. 104. 57. For a succinct discussion of Mayr’s treatment of proximate and ultimate causes, see pp. 77–79 of Cor van der Weele 1999. 58. Mayr 1991, p. 145. 59. Van der Weele 1999, citing Mayr 1994.

442

Notes to Chapter 2

60. Williams 1966, p. 4. 61. Ibid., p. 5. 62. Ibid., p. 270f. 63. Ibid., p. 55. 64. Dawkins, in Brockman 1995, p. 48. Re Biblical misquotation: In John 8:12, Jesus refers to himself as “the light of the world.” This is often mixed up with John 14:6 “I am the way, the truth and the life”—odd how “the truth” was left out by Wilson (1998b, p. 60) and Williams (1966, p. 273). Jesus himself may have misremembered his quotes; see Job 38:19 “Where is the way where light dwelleth?” 65. Steve Jones, in Brockman 1995, p. 95. 66. On Luther and reason, see Honderich 1995, p. 514. 67. Eldredge 1995, p. 226. 68. Bacon 1620, p. 5. 69. Løvtrup 1975, p. 511. 70. This quotation from the bacteriologist Melvin Cohn is borrowed from p. ix of Bibel 1992, which attempts a synthesis in some ways more ambitious than the present work. A bacteriologist and an immunologist, Bibel is the kind of rare bird mentioned by her source. I find it reassuring that there are other biologists who are not afraid to go naked in the market place. She does not detail the source of the quotation. 71. Bateson 1894, p. 13. 72. Ibid., p. 13. Chapter 2 1. Cohen and Stewart 1994, p. 436. 2. Holland 1998, p. 1. 3. On natural selection as “book-keeping,” see Wimsatt 1980. 4. See, for example, Brooks and Wiley 1986 and Prigogene and Stengers 1984. 5. Conrad 1990, p. 79. 6. Ibid., p. 61. 7. Kauffman 1992, p. xiv. 8. Wilson 1998b, p. 60. 9. See note 67 to chapter 1.

Notes to Chapter 2

443

10. Woodger 1929, p. xv. 11. Goodwin 1994, p. xi. 12. This seems to be the equivalent of “hierarchical reductionism” (Dawkins 1989). 13. Cohen and Stewart 1994, p. 220. 14. Ibid., p. 228. 15. Ibid., p. 376. 16. Ibid., p. 380. 17. Kim (1999, p. 33), citing Galen (~200 A.D.) On The Elements According to Hippocrates 1. 3. 70: 15–74. p. 23. 18. Drummond 1883, p. 405. 19. Alexander 1920, p. 46f. 20. Holland 1998, p. 190. 21. Morgan 1931, p. 5. 22. Bak and Kan Chen 1991; Bak 1996. 23. Caudwell 1986, p. 176. Foster (2000) explains that the British Communist Party, which had published Caudwell’s other works, held back Development and Heredity because its emergentism seemed too idealistic. Engels, on the other hand, had too many political commitments to see The Dialectics of Nature published promptly. His Anti-Düring (1878) was the only hint of his emergentism in the meantime. 24. Sperry 1983, p. 79. 25. Oyama 2000, p. 181f. Note also that Cohen is the biologist and Stewart is the mathematician in their collaborative authorship of The Collapse of Chaos (1994). 26. Kauffman 1995, p. 23. 27. Holland 1998, p. 239. 28. Corning 1998, p. 11. 29. El-Hani and Pihlström 2002, p. 30. 30. Dawkins (1995, p. 83) lays claim to the question of the “evolution of evolvability.” 31. Benford (1995, p. 136) uses the phrase “The Enormous Emergence.” The quotation is from p. 274 of Martin and Benford 1996. 32. Otte and Endler 1989. 33. Lovelock (1988) is a proponent of geophysiology.

444

Notes to Chapter 3

34. Biologists are forever arguing about teleology and might be suspicious of “goal-directed” activities. What I mean is any series of specific related actions that terminates at a predetermined point—sometimes called “teleonomic,” as distinct from teleological. It may be the final product of a biochemical pathway, the healing of a wound by clotting and tissue regeneration, or a sequence of behaviors. For a discussion, see pp. 278–282 of Reid 1985a. 35. Dobzhansky 1967, p. 58. Dobzhansky attributes the idea of a unique, emergent humanum to Brunner (1952). 36. Ibid., p. 57. 37. Ibid., p. 61. 38. Gehring 1998, p. 99. Chapter 3 1. Margulis 1981, p. 201. 2. Margulis 1998, p. 8. 3. Jennings 1927, p. 22. 4. Margulis 1981, pp. 222–224. 5. Margulis 1998, p. 11. 6. Sapp and I probably share a quirky appreciation of the word-play parallel between our title and the idiom “guilt by association.” While acknowledging his outstanding contribution to the subject, I will not plead guilty to plagiarism, since I have been using this title for seminars since 1986, and I used it in an abstract published in 1987. 7. Sapp 1994, p. 131f. 8. Sapp (1994, p. 159) refers to the Wollmans’ unpublished inferences on paraheredity. 9. I use Meyer-Abich’s 1964 coinage to draw attention to his little appreciated contribution to symbiosis studies. 10. Fitt 1985. 11. On aspects of lichen symbioses, see Douglas 1994, pp. 15, 17, 51–52. 12. Aristotle, De Motu Animalium. 13. Aristotle, Politics; De Motu, 703a29f. 14. Heddi et al. 2001, citing Perru 1997. 15. Haines 2002. 16. Aksoy 2000

Notes to Chapter 3

445

17. Buchner 1965; Clark et al. 2000. 18. See the review by Boursaux-Eude and Gross (2000). 19. Yoshida et al. 2001. 20. Fukatsu and Ishakawa 1993. 21. Imms 1957, p. 800. 22. Dohlen et al. 2001. 23. Curtis and Walter 1995. 24. Caetano and Da Cruz-Landim 1985, cited by Boursaux-Eude and Gross (2000). 25. On leafcutter ants, see Van Borm et al. 2002. They also survey the status of ant symbioses in general. 26. Currie et al. 1999, cited in Boursaux-Eude and Gross 2000. 27. Reviewed by Haines 2002, citing Morin et al. 2000. 28. Marcotrigiano 1999. 29. Cohen (2001), interviewing Luis Villarreal of the University of California at Irvine; Muir et al. 2004. 30. Greider 1998. 31. Maynard Smith and Szathmáry 1995, p. 131. 32. Cavalier-Smith 1985. 33. Edelman 1987. 34. Reid 1985a, p. 144ff. 35. Jennings 1927, p. 22. 36. Wilson 1975, p. 7. 37. Shostak and Kolluri 1995. 38. Rudman 1981. 39. Pirozynski and Malloch 1975. 40. Douglas 1994, p. 19. 41. Maser et al. 1978. 42. Margulis 1981, citing Cleveland and Grimstone 1964. 43. Margulis 1981, p. 12. 44. Klenk et al. 1997.

446

Notes to Chapter 3

45. Martin and Müller 1998. 46. Doolittle 1998. 47. Ibid. 48. Margulis and Sagan (2002) base these assumptions on the work of R. Gupta. 49. Margulis and Sagan 2002, p. 158. 50. Martin and Müller 1998. 51. Niklas 1997, p. 133f. 52. Ibid., p. 143. 53. Sapp 1994, p. 40. 54. Ibid., p. 185, citing Uzell and Spolsky 1974. 55. Schwartz and Dayhoff 1978. 56. Margulis 1981, p. 18f. 57. Bryant et al. 1998. 58. Fankboner and Reid 1981. 59. Reid 1990. 60. Douglas 1994, p. 54. 61. Tatewaki et al. 1983 62. Margulis and Sagan, 2002, p. 60f. 63. Bry et al. 1996; Hooper et al. 2001 64. Fankboner 1971. 65. Douglas 1994, pp. 101–103. 66. The definitions set out here are to be found in chapter 2 of Wilson 1975. 67. Costerton et al. 1995 68. England et al. 1999; Dunny and Wians 1999. 69. Shimkets 1990; Dworkin 1996. 70. Burkholder 1999. 71. Bever and Simms 2000. 72. Shapiro 1998. 73. Engelberg-Kulka and Glaser 1999.

Notes to Chapter 4

447

74. Crespi 2001, p. 182. 75. Virchow 1858. This refers to his major work. For a summary of his preliminary opinions on the cell theory, see Hall 1969, volume 2, p. 280ff. 76. Solé and Goodwin (2000, p. 149) recommend the more extensive comparisons of societies and neural organization made by Gordon et al. (1992) and Gordon (1999). 77. I may be putting words into the mouths of Solé and Goodwin here, but this is what their comparison means to me. 78. Whewell 1840, cited in Wilson 1998b, p. 8. 79. Maynard Smith and Szathmáry 1995, p. 145. Chapter 4 1. Murphy 1869, p. 68. Nicely said, but the French physiologist Milne Edwards promoted the idea of a division of physiological labor in 1834. 2. Child 1924, p. 43. 3. J. Huxley 1942, p. 564f. As I argued in chapter 1, it was Huxley’s sensitivity to the importance of such progressive evolution that put him beyond the pale of the Modern Synthesis that he helped to create. 4. Woodger 1929, p. 41. 5. Darwin 1859, p. 141. 6. Darwin 1872, p. 349. 7. Arthur 1997, pp. 287–289. 8. Bernard, strictly speaking, was not post-Darwinian, but his most important work was done after the publication of The Origin of Species. The physiologist Viscount J. S. Haldane (father of J. B. S. Haldane) was fixated on the holistic aspects of physiology, but did not focus strongly on how it evolved. See Reid 1985a, p. 109f. 9. Schmalhausen 1949, p. 232. 10. Whyte’s early ideas on the subject were formulated before he encountered Schmalhausen’s book, and although the parallels are remarkable he gave it short shrift. As well as having priority, Schmalhausen’s biological treatment was the more substantial. On Whyte’s contribution and the broader historical context of physiological organization, see chapter 14 of Reid 1985a. 11. Rollo 1994, p. 256. 12. Conrad 1990, p. 79. 13. Conrad 1983, p. 10.

448

Notes to Chapter 4

14. Harbo 1997. My interest in this species began with honours thesis research by my student Mia Parker in 1995. I then made additional informal studies of ecological distribution and feeding behavior, some of which are published here for the first time. 15. Rollo 1994, citing and paraphrasing Alexander 1991, p. 446. 16. I believe Bernard first made his frequently quoted comment “La fixité du milieu intérieur; c’est la condition de la vie libre” in a lecture given in 1872. 17. For a summary of the theories of Severtsov and Schmalhausen, see Adams 1980. 18. Waddington 1957. 19. Gould and Vrba 1982. This was a re-invention of the Funktionswechsels wheel (Dohrn 1875). 20. Schmalhausen 1949, p. 233. I have inserted my own translations of the words “adaptation” and “adaptive,” since he used them less definitively than he might have. I suspect his original Russian was more precise. 21. Darwin 1872, p. 135. 22. The full citation list for these studies includes Wynne-Edwards 1998; McMillan and WynneEdwards 1998, 1999; and Wynne-Edwards et al. 1999. 23. Le Maho 1977. 24. Sherman et al. 1992. 25. Grémillet and Wanless 2000; Grémillet et al. 2005. 26. See De Beer 1940, p. 95. 27. For reviews of genetic assimilation, see Reid 1985a, Hall 2001, and West-Eberhard 2003b. 28. Woodger 1929, p. xv. The “semantic reductions” that I mentioned in the introduction are epistemological intolerant abstractions. 29. For a useful review of the origin-of-life hypotheses, see Maynard Smith and Szathmáry 1996. 30. Schueller 1998. 31. Ibid. 32. Orgel 1994. 33. Kauffman 1995, p. 50. 34. Ibid., p. 62. 35. Rebeck 1994. 36. Schwartz’s ideas were obtained by means of a telephone interview. At the time of writing, his intended book had not been published. 37. For the fundamental arguments, see Zuckerkandl 1975.

Notes to Chapter 4

449

38. Kauffman 1995, chapter 4, p. 71ff. 39. See Ganti 1974 for chemotons and Woese and Fox 1977 for progenotes. See also the 1995 review by Maynard Smith and Szathmáry. 40. Doolittle and Brown 1994. 41. Weng et al. 1999, p. 93. 42. Ibid., p. 94. 43. The literature is awash with neologisms that are redundant or misleading, or invented for self-serving reasons. I am reluctant to compound these errors, preferring to resurrect the historical name unless it is already redundant or misleading. Although the process is presently known to molecular biologists as the production of variant repeats (see chapter 6), to call it “variant repetition” puts the cart before the horse. 44. Hochachka (1973) reviews his salmonid kinase research. 45. Ohno 1970, p. i. 46. Britten and Davidson 1971. 47. Pray 2004. 48. Wallis 1996. 49. Jiminez et al. 1993. 50. Li and Graur 1991, p. 80, citing Goodman 1981 and Czelusniak et al. 1982. 51. Segre and Goldring 1993. 52. Gerhart and Kirschner 1997, p. 223. 53. Fleming and Allison 1922. 54. Gerhart and Kirschner 1997, p. 204, citing Wistow 1993. 55. Newcombe et al. 1997, cited by Lewontin 2000, p. 117. 56. Mangum and Towle 1977. 57. Hochachka 1973. 58. On photosynthesis evolution, see Niklas 1997. 59. Pearson 1986, p. 12. 60. Pearson 2004, p. 77. 61. Schmalhausen 1949, p. 11. 62. Uvarov 1977. See also chapter 5 of the present volume. 63. Collins and Cheek 1983.

450

Notes to Chapter 5

Chapter 5 1. Waddington 1957, p. 13. 2. Ho and Saunders 1982, p. 93. 3. Müller and Wagner 1991, p. 230. 4. Woodger 1929, p. 372. 5. Bacon et al. 2001. It’s nice to see a historical and philosophical perspective on the subject. Nevertheless, as in the original model, “Socrates” is the inevitable winner of the debate. 6. See Hall 1992 and Arthur 1997. 7. See chapters 1 and 4 re Belyaev (1970) on sports and on breeding programs that destabilize epigenetic patterns. 8. Darwin 1859, p. 12; p. 21 for pigeons; p. 31 for the Sebright quotation. 9. Geoffroy 1826. 10. T. H. Huxley 1864, p. 34. 11. On Murphy and Mivart, see chapter 3 of Reid 1985a. 12. For appreciations of Cope, see Gould 1977a and Reid 1985a. 13. Brooks 1883. 14. Bateson 1928, p. 42f. 15. Reid 2003. 16. De Vries 1901–1903. 17. Spemann 1914. 18. Arthur 1998, p. 99, citing Spemann 1938, p. 367f. 19. E. B.Wilson (1928) had made the distinction between pragmatic experimenters and the more “romantic” generalists. 20. Kukalova-Peck 1983. 21. Murray et al. 1981; Palka et al. 1983. 22. Rollo 1994, p. 105. 23. The role of neural crest cells in avian facial prominences was confirmed for chicken embryos by Le Lièvre (1978), Noden (1983), and Couly et al. (1998). Lee et al. (2001) show the subsequent involvement of bone morphogenetic protein and retinoic acid. 24. For example, see Lloyd 1984.

Notes to Chapter 5

451

25. West-Eberhard 1989, p. 256f. For a more extensive treatment of this subject, see WestEberhard 2003b. 26. West-Eberhard 2003a; Larsen 2004. 27. Slijper 1942, summarized in West-Eberhard 1989. West-Eberhard provides more examples of these kinds of change on p. 255 of the aforementioned work; she gives more details of the goat case in West-Eberhard 2003b. 28. Waddington 1969, p. 221. 29. Saunders 1984, p. 255. 30. Garstang, one of several biologists interested in radical changes in invertebrate development, also distinguished himself by writing about them in verse (1951). See Hall 2000 on Garstang and De Beer, and also on the earlier, nowadays overlooked work of the developmental evolutionist Frank Balfour, a student of Ray Lankester and a teacher of Bateson and Weldon. 31. Hardy 1965, p. 217. 32. Gerhart and Kirschner 1997, pp. 331ff. On chronological expression of Hox, see Gilbert 1997, chapter 16. 33. Schmalhausen 1949, p. 233. 34. Müller 1990, p. 121. 35. Ibid., p. 109. 36. Ibid., p. 104. 37. Ibid. Müller cites Long 1976 and Brylski and Hall 1988 on cheek pouches. 38. Carroll (1988) and Burke (1989) on the subject of the sudden evolution of the chelonid carapace, cited by Müller (1990, p. 107). 39. Raff and Wray 1989, Raff et al. 1990, and Wray and Raff 1990, cited in Muller and Wagner 1991, p. 234. 40. Müller and Wagner 1991, p. 234. 41. On the developmental evolution of bird skeletons, see Müller 1989, 1991. 42. Futuyama 1998. 43. Müller 1991. 44. Newman and Müller 2000. 45. All quotations in this paragraph from Keller 2002, p. 252, citing Dassow et al. 2000. See also Meir et al. 2002. 46. Hall 2001, p. 219.

452

Notes to Chapter 5

47. Osborn 1896, p. 312. 48. Kane 1995. 49. Darwin 1871, p. 119f. 50. Waddington 1942. 51. West-Eberhard 2003b, p. 147. 52. Wake 1989, 2003. 53. This is an unpublished observation regarding an unidentified species of the Orchestia type in Haida Gwaii (Queen Charlotte Islands) of British Columbia. Since both beach and sea levels went up and down like a pair of unsynchronized yo-yos in the wake of the last glaciation (Hetherington et al. 2002), Matsuda may have been right on both counts. 54. Referenced on pp. 107–109 of van der Weele 1999. 55. Newman and Müller 2000, p. 305, citing Magee 1997. 56. Ibid., p. 305, citing Deeming and Ferguson 1988. 57. Ibid., p. 305, citing McLaren and Michie 1958. 58. Balon 1983 and 1986; see also Bruton 1994. 59. Oyama 2000, p. 39. 60. Garstang 1922, p. 81, cited in Hall 2000. 61. Newman and Müller 2000, p. 303. 62. Ibid., p. 305. 63. Brenner 1972; Sulston et al. 1980; Wood et al. 1988. 64. Hill et al. 1992, p. 60, cited in Gehring 1998. 65. Gehring 1998, p. 204, quoting calculations made by Rubin. 66. Newman and Müller 2000, p. 307, citing Kazmierczak and Degens 1986. 67. Newman (1994) includes epiboly, involution and delamination. See Steinberg 2003 for a more comprehensive account of differential adhesive reactions. 68. Goodwin 1989, p. 92. 69. Newman and Müller 2000, p. 307, citing the formation of lumens in tumors (Tsarfaty et al. 1992). 70. Nüsslein-Volhard 1996, p. 61. This author points out that the importance of gradients was emphasized by Boveri a century ago. See also Haraway 1976. 71. Newman and Müller 2000, p. 316.

Notes to Chapter 6

453

72. Newman and Müller 2000, p. 307, citing Palmeirim et al. 1997 and Newman 1993. See also Nüsslein-Volhard 1996. 73. This paraphrases Newman and Müller 2000, p. 310, citing various authors. 74. Frey et al. 1997. 75. I discuss the controversial usage of “goals” in Reid 1985a. 76. Newman and Müller 2000, p. 313. Chapter 6 1. Campbell 1985, cited by Jablonka and Lamb 1994, p. 72. 2. Matsuda 1987, p. 53. 3. Newman and Müller 2000, p. 304. 4. Woodger 1929, p. 32. Woodger probably borrowed this construction from G. K. Chesterton. 5. The cuisine metaphor is borrowed, with thanks, from an unpublished essay on epigenetics by my former student Kevin Little. It pops up in the literature in such a way as to suggest polyphyletic origins. 6. Novick and Weiner 1957, cited in Jablonka and Lamb 1994, p. 82. 7. Johnston 2004. 8. Wolffe and Matzke 1999. The verbal communication metaphor is Kevin Little’s (personal communication). 9. Pelling 1959; Clever and Karlson 1960. 10. Baker 1997. 11. Li and Graur 1991, p. 159. 12. Tittiger et al. 1993. 13. Li and Graur 1991, p. 15. 14. Li and Graur 2000, p. 294f. 15. Nowak 1994. 16. Li et al. 2001. 17. Lewin 1997, p. 525ff. 18. See Caporale 2003, p. 86ff. She notes also that the acronym SOS (“Save our ship”) was suggested by Miroslav Radman, a molecular biologist with a family background in fishing. 19. On HR and NHEJ, see Lin and Waldman 2001 and Willet-Brozick et al. 2001.

454

Notes to Chapter 6

20. Pirkkala et al. 2001 is a good general reference for heat-shock proteins. 21. Brenner et al. 1993; see also Elgar et al. 1999. 22. Jablonka and Lamb 1994, p. 69f. 23. Prody et al. 1989. 24. These examples are from Jablonka and Lamb 1994 and Prody et al. 1989. 25. Moxon and Wills 1999. 26. King 1994; King, Soller, and Kashi 1994. 27. Moxon and Wills 1999, p. 99. 28. Wurtele et al. 2003; Little and Chartrand 2004. 29. Lin and Waldman 2001; Willett-Brozick et al. 2001. 30. Li et al. 2001. 31. Lachmansingh and Rollo 1994. 32. Baker 1997, p. 101. 33. Laudet et al. 1992 Laudet 1999. 34. For a brief history of concerted evolution, see Graur and Li 2000, p. 304ff. See also Dover 2000. 35. Dover 1982, p. 111. 36. Li and Graur 1991, p. 164. 37. Ibid., p. 166. 38. Ibid., p. 140f. 39. Graur and Li 2000, p. 290f. 40. Li and Graur 1991, p. 188. 41. Gerhart and Kirschner 1997, p. 228. 42. Li and Graur 1991, p. 193f. 43. Hurst and Randerson 2002. 44. Starr and Cline 2002. 45. Ting et al. 1992. 46. Gerhart and Kirschner 1997, pp. 309–314. 47. Walter Gehring (1998) gives a remarkable account of the recent history of this research, in which he played a leading role.

Notes to Chapter 6

455

48. Raff (1996), discussing the work of Kukalova-Pek (1983). 49. Kukalova-Pek 1983. 50. Carroll 1995, p. 482. 51. Ibid., p. 481. 52. Ronshaugen et al. 2002 53. For additional references, see Carroll 1995, p. 483. 54. Gerhart and Kirschner 1997, pp. 532–535. 55. Burke et al. 1995, cited by Carroll 1995. 56. Tomarev et al. 1997. 57. Experiments with eyeless were conducted in 1994 in Gehring’s laboratory. Gehring credits Gerry Rubin with the calculation that 2,500 genes are involved in eye construction. 58. Gerhardt and Kirschner 1997, p. 205. George Mackie’s (1999) response is also cited in the introduction to the present work. 59. Goodwin 1994, p. 167f. He does not use the term “generative conditions of emergence,” and I have expanded these beyond those that he cites. 60. Behe 1996, p. 41. 61. Gould 1983n, p. 186, citing Darwin 1868. 62. Grimes and Aufderheide 1991, cited by Jablonka and Lamb 1994, p. 87. Also Jablonka and Lamb 1994, p. 89. 63. Jablonka and Lamb 1994, p. 90. 64. Ibid., p. 102. 65. Monk 1995. 66. Surani 2001. 67. Jablonka and Lamb 1994, p. 130. 68. Pray 2004, p. 1. 69. Jablonka and Lamb are, however, conspicuous by their absence from Pray’s references. 70. Jablonka and Lamb 1994, p. 232, citing Bottema 1989. 71. Jablonka and Lamb 1994, p. 205. 72. Arthur 1997, pp. 54–63; Wray 1998. 73. Jablonka and Lamb 1994, p. 171f.

456

Notes to Chapter 7

74. Jegalian and Lahn 2001, p. 58. See also Lahn and Page 1997, 1999; Jegalian and Page 1998. 75. Dover 2000, p. 108ff. 76. Wells 1813. 77. Shapiro 1992, p. 101f. 78. Yuh and Davidson 1998. 79. Driever and Nüsslein-Volhard 1988a,b. Chapter 7 1. Darwin 1872, p. 157. 2. Eimer 1898, p. 467. 3. Waagen 1867; Haake 1895. See Bowler 1979 and Reid 1985a. 4. Goldschmidt 1940, p. 22. 5. J. Huxley 1942, p. 304f., citing Watson 1926. 6. Eldredge 1995, p. 129ff. 7. The “pull of the present” metaphor is used by Raff (1996, p. 129). 8. Simpson 1944, p. 150. 9. Ibid., p. 155. 10. Thompson 1941, p. 807f. 11. Ibid., p. 1094f. 12. Reid 1985a, p. 199. At the time I had overlooked Bowler 1979 and Grehan 1984. Mea maxima culpa! Thanks to John Grehan for a gentle nudge. 13. Grehan and Ainsworth (1985) regarding Dover’s response to orthogenetic implications of molecular drive (1982). 14. Arthur 1997, p. 213. 15. Britten and Davidson 1971, p. 118. 16. Lima-De-Faria 1988, p. 211, citing Ritossa et al. 1971, Tartof 1974, Brown and Dawid 1968, and Perkowska et al. 1968. See also Jablonka and Lamb 1995, p. 70. 17. Vrba 2004 and personal communication. 18. Gould 1974. 19. On Triceratops see Gerhard and Kirschner 1998, p. 551. For a survey of the range of effects of neural crest organizer cells on head anatomies, see ibid., pp. 552–554.

Notes to Chapter 8

457

20. On the theoretical significance of horse allometry, see Eldredge 1995, p. 129ff. 21. Thomson 1992, p. 136. 22. Ibid., p. 136. 23. Ibid., p. 137. 24. Raff 1996, p. 25. 25. Ashitawa et al. 1994. 26. On Huntington’s disease see Rubinsztein et al. 1994. On fragile-X pathology see Parrish et al. 1994. I am most grateful to the biological historian Judith Friedmann for priming me on the subject of anticipation. 27. Ashitawa et al. 1994. 28. Dover 2000, p. 101. 29. Ibid., p. 209. 30. Ibid., p. 32. 31. For a review of the RNA experiments see Eigen and Schuster 1979. 32. Brown et al. 2004; Morwood et al. 2004. 33. Fondon and Gardner 2004, p. 18060. 34. Ibid., p. 18061. Chapter 8 1. Vrba 1989, p. 382. 2. Kauffman 1995, p. 23. 3. Gould 1996, in Horgan 1996, p. 124f. 4. Kauffman 1993, p. 180. 5. See also Reid 1985b. 6. Mayr 1960, p. 249. I have already mentioned Darwin’s musings on “generative variability.” 7. Mayr 1960, p. 351. Some developmental evolutionists have tried to distinguish between major “innovation” and trivial “novelty.” That I do find stultifying. 8. Ibid., p. 364. 9. Ibid., p. 364. 10. Relevant publications from the Brenowitz group include Smith et al. 1997 and Tramontin et al. 1998.

458

Notes to Chapter 8

11. Mayr 1960, p. 368. 12. Ibid., p. 369. 13. Ibid., p. 371. 14. Ibid., p. 377. 15. Ibid., p. 371. 16. Ibid., p. 374. 17. Darwin 1872, pp. 112, 365. 18. Schmalhausen 1949, p. 33. 19. I am grateful to my colleague John Taylor for some material used here. See also Taylor and Raes 2005. 20. Maynard Smith 1988, p. 222. 21. Ibid., p. 222f. 22. Rollo 1994, p. 8. 23. Riedl 1978, p. 110. 24. Ibid., p. 272. 25. Ibid., p. 117. 26. Ibid., p. 126. 27. Ibid., p. 127. 28. Ibid., p. 129. 29. Ibid., pp. 132–138. 30. Vrba and Eldredge 1984, p. 146. 31. Vrba 1989, p. 382. 32. Ibid., p. 390. 33. Salthe 1985, p. 155. 34. Ibid., p. 201. 35. Campbell 1993, p. 5. I refer to the third edition. Although I am aware its success has resulted in subsequent editions, these generalities about evolution have not changed, despite the fact that the 1999 fifth edition, co-authored with Reece and Mitchell, chose Dawkins as the key interviewee on evolution. 36. Ibid. 1993, p. 7.

Notes to Chapter 8

459

37. Polanyi 1958, p. 151f. 38. Müller and Wagner 1991, p. 232. 39. Ibid., p. 243. 40. Ibid., p. 242. 41. Ibid., p. 245. 42. Ibid., p. 251. 43. Cohen and Stewart 1994, p. 232. 44. Ibid., p. 441. 45. Kauffman 1993, p. xiv. 46. Ibid., p. xv. 47. Ibid., p. xvi. 48. Holland 1998, p. 222ff. 49. Ibid., p. 125ff. 50. Ibid., p. 225. 51. Ibid., p. 227. 52. Ibid., p. 229. 53. Wimsatt 1997, p. S376. 54. Ibid., p. S382. 55. Corning 1998 p. 23. 56. Ibid., p. 15f. 57. Wynne-Edwards 1962. Corning (1997) singles out D. S. Wilson (1975 and subsequent publications) as the leader of the attempt to bring back group selection. 58. Morowitz 2002, p. 104f. 59. Frey et al. 1997. 60. J. Schwartz 1999, p. 342. 61. Bakker hints at this function in his seminal 1975 paper in Scientific American. 62. Bak and Chen 1971. 63. Kauffman 1993, p. xv. 64. On locusts, see Uvarov 1977. On cannibalism in salamanders, see Collins and Cheek 1983 and Maret and Collins 1997; see also Hall 1999, p. 303.

460

Notes to Chapter 9

65. Rollo 1994, p. xii. 66. Kauffman 1993, p. xvii. 67. Raup 1986, p. 22. Chapter 9 1. I have slightly abbreviated this version of Bacon’s Novum Organum, without, I hope, losing any of his original intentions (Bacon 1620; p. 82 in my 1901 edition). 2. Darwin wrote “entangled” for his renowned final paragraph in the first edition of Origin. My sixth edition says “tangled,” but this may be a printer’s error. 3. Piaget 1968. 4. Piaget 1979, p. 23. A similar notion about behavioral innovation in the absence of natural selection had been expressed by R. Chauvin (1977). 5. Müller 1990, p. 123f. This author’s assertion in point 6, regarding plasticity at all developmental levels, anticipates Kirschner and Gerhart’s (2005) emphasis on epigenetic adaptability and “exploratory behavior” during development. 6. Müller and Wagner 1991, p. 234. I have paraphrased their citations of Raff and Wray 1989 and Raff et al. 1990. 7. Ibid., p. 235, citing Wake and Roth 1989. 8. Bateson (1888) in a letter to his sister Anna, cited on pp. 42–43 of Bateson 1928. 9. Wimsatt 1998, p. 270. 10. Wimsatt 1999, p. 142. 11. See, e.g., Little and Chartrand 2004. 12. Balon, cited by Bruton 1994, p. 271. See also Balon 1986. 13. Balon 1983, p. 2049. 14. See Hooper et al. 2001 on bacterial commensals of mammals and Tatewaki et al. 1983 on epigenetic effect of bacteria on thallus-formation in Ulva and related green macroalgae. 15. West-Eberhard 2003b. 16. Lewontin 2000, p. 48. 17. See www.kakaporecovery.org.nz/involved/reading.html. 18. Turner 2000, p. 213. 19. Lamarck 1809 I, p. 267. 20. Kobayashi and Inaba 2000.

Notes to Chapter 10

461

21. Lovelock 1988, p. 10. Hutton’s two papers on “The theory of the earth” were read to the Royal Society of Edinburgh in 1785, and published in its proceedings in 1788. 22. McNeil 1987. 23. Ring 1982. For a review of the cold adaptations and cold adaptabilities of Arctic insects, see Ring and Tesar 1981. 24. Darwin 1859, p. 489f. 25. See Vendruscolo and Dobson 2005; Trinh Xuan Hoang et al. 2004; Chotia et al. 2003. 26. Dohrn 1875; Gould and Vrba 1982. 27. Bacon 1620, p. 82. See also his New Atlantis (1628) on the division of scientific labor. Chapter 10 1. Lamarck 1809 I, p. 221. 2. J. Huxley 1942, p. 578. 3. Holland 1998, p. 231. 4. J. Huxley 1942, p. 568. 5. West-Eberhard (2003b) uses the expression “phenotypic plasticity” for the latter properties. However, although her treatment of the subject is comprehensive, phenotypic plasticity in earlier and narrower neo-Darwinist applications is a misnomer equivalent to the taxonomist’s nomen nudum: i.e., a name that cannot be used because as originally published it failed to meet taxonomic standards; in the case of phenotypic plasticity, failure to meet the explicit meaning of the word “plasticity”! 6. Brenner 1999, p. 1964. 7. Anderson 1994, cited by Solé and Goodwin 2000, p. 18. 8. Solé and Goodwin 2000, p. 13ff. 9. Rollo 1994, p. 14. 10. J. Huxley 1942, p. 571. 11. Popper 1972. 12. Beament 1961. 13. Waldrop 1992, p. 288. 14. Horgan 1996, p. 206. 15. Holland 1998, p. 231. 16. Personal communication from Tom Reimchen, University of Victoria.

462

Notes to Chapter 11

17. Newman and Müller 2000, pp. 304 and 315 respectively. 18. Valentine 1986. Chapter 11 1. T. S. Eliot, “Little Gidding,” in Four Quartets (1943). 2. Vrba and Eldredge 1984, p. 146. 3. Jennings 1927, p. 22. 4. Science 293 (2001), p. 1086. 5. Keller 2002, p. 254. 6. Reid and Brand 1986; Reid 1990, 1998. 7. I should temper this conclusion by noting that the studies in question did provide morphological details from which valid inferences could be made about the symbiosis, regardless of their original interpretation. 8. Ronshaugen et al. 2002. 9. Kukalova-Peck 1983. 10. Grant and Grant 1994. 11. Gerhart and Kirschner 1997, p. 553. 12. Ibid., p. 553. 13. Mungo is a bred-in-captivity parrot whose acquisition was prompted by the original work of Irene Pepperberg with Alex, her African grey. See Pepperberg 1998. 14. Stebbins, 1987 lecture at Cornell University, cited in Grene 1988, p. 55. 15. Re Stiassny and Meyer 1999, p. 3. 16. Ibid., p. 68. 17. Ibid., p. 69. 18. Fryer and Iles 1972; Greenwood, 1974; Stiassny 1991; Stiassny and Meyer 1999. 19. Stiassny and Meyer 1999, p. 66. 20. Paterson 1985. 21. Grene 1988, p. 68. 22. Leroi 1998, p. 82. 23. Provine 2001, p. 204.

Notes to Chapter 11

463

24. Lotka 1922. 25. Chaisson 2001, p. 216. 26. George Herbert (1593–1633), popularized by Benjamin Franklin (1758). See pp. 270 and 374 of Bartlett’s Familiar Quotations (1980). 27. Linde 1994. Horgan (1996, pp. 98–102) discusses his views. 28. Dyer 1997. 29. Smolin 1997, p. 104. 30. Ibid., p. 150f. 31. Ibid., p. 185. 32. Ibid., p. 105f, citing Dawkins The Blind Watchmaker; no page reference given. 33. Cohen and Stewart (1994, p. 375) attribute this criticism to Dawkins. Sheri Tepper (1998) gives an interesting fictional account of a Gaian superorganism that has complexified to the point of having intelligent and semi-independent subunits that lack sexual reproduction, depending instead on cooperation with the parental whole to persist. 34. Clark 1968, citing a letter from Haldane to Robert Graves. 35. Prosser 1965, p. 364. 36. Koestler 1967, p. 349. 37. My philosopher friend C. B. (Danny) Daniels, a movie buff, has waited in vain to see the question “Is there a metaphysician in the house?” flashed on the screen. I hereby indulge him and thank him for his reflections on Spinoza, single-malts, and oysters on the half-shell. 38. Polanyi 1958. 39. Peirce 1898. 40. Holland 1998, p. 242. 41. Gunther Stent (1969) started the trend in biology a bit early for the millenium, though it is obvious from his use of Robert Frost’s “It Is Almost the Year Two Thousand” as an introductory quotation that he had the millenium in mind. John Horgan (1996) took up the refrain. I am not sure that he actually believes that the end of science is nigh, but I got a lot of mileage out of his discourse in any case. 42. Darwin 1859, p. 485f. 43. The “rough beast” is the hopeful monster in “The Second Coming” (Yeats 1920). 44. Eliot 1943.

Bibliography

After I published Evolutionary Theory: The Unfinished Synthesis, I returned to my research on mollusks. After ten years, no saltatory emergences had occurred in evolutionary theory, and a chance encounter with Brian Goodwin and David Lambert in Auckland goaded me to return to the fray. I would like to acknowledge some works that I found particularly stimulating on my re-entry to the lists. The first group includes, in the order in which I encountered them, Stephen Jay Gould’s 1994 Scientific American essay “The evolution of life on Earth,” David Rollo’s Phenotypes: Their Epigenetics, Ecology and Evolution (1994), Eva Jablonka and Marion Lamb’s Epigenetic Inheritance and Evolution: The Lamarckian Dimension (1995), Antonio Lima-de-Faria’s Evolution without Selection: Form and Function by Autoevolution (1988), Niles Eldredge’s The Reinvention of Darwin (1995), Jack Cohen and Ian Stewart’s The Collapse of Chaos (1994), and Stuart Kauffman’s At Home in the Universe (1995). John Horgan’s The End of Science (1996) was also provocatively stimulating. I am not in complete agreement with these authors’ views of evolutionary theory, since they all allow themselves to be bogged down by natural selection. (It operates that way in theory as well as in nature.) Nevertheless, I have great respect for their talents and freely admit their positive influence upon me. Publications of conference and workshop proceedings leave far too much space between the stools that the important stuff can fall into. Nevertheless, I found Matthew Nitecki’s University of Chicago series, in particular Evolutionary Progress (1988) and Evolutionary Innovations (1990), to be useful. And I have to confess to personal contributions to the 1988 American Zoological Society’s symposium “Is the Organism Necessary” and to the 2001 KLI workshop on Ryuichi Matsuda and “ecoevo-devo,” where I met strangers who seemed like old friends immediately. Although Brian Goodwin had hinted darkly to me about the contents of his book How the Leopard Changed Its Spots before its publication, I did not catch up with it until I had finished my first draft. But his 1989 essay “Evolution and the generative order” had already had a catalytic effect, and How the Leopard Changed Its Spots later precipitated some of my thoughts into a more solid form.

466

Bibliography

Of all the major works on the subject of complexity by authors who start from the bottom, I found John Holland’s Emergence: From Chaos to Order (1998) the most valuable for comparative purposes, and more often on my side of the Looking Glass than not. Although I like to work in a state of “enlightened ignorance,” to avoid constructing a pastiche of other peoples’ ideas that would hinder the development of my own, I probably should have taken a fuller account of Susan Oyama from the outset. Furthermore, I have not specified here the numerous works by Lynn Margulis and her co-author Dorion Sagan that have emerged along the rocky road of symbiosis theory. But I was already intellectually charged by her ideas long before I conceived of the present work. In the time between my submission of the final manuscript and the copy-editing stage, several important books appeared. I have minimally mentioned some of them in the text. They include Callebaut and Raskin-Gutmann’s Modularity: Understanding the Development and Evolution of Natural Complex Systems, Jablonka and Lamb’s Epigenetic Inheritance and Evolution: The Lamarckian Dimension, and Kirschner and Gerhart’s The Plausibility of Life: Resolving Darwin’s Dilemma, all published in 2005. In a planned future volume that will deal with the history of life as envisioned by an emergentist, I expect to give the latter works fuller consideration, as well as Müller and Newman’s Origination of Organismal Form: Beyond the Gene in Developmental and Evolutionary Biology. I have received much reference material from so many colleagues, correspondents, and students that I am unable to name them all; some of their names appear in the endnotes. But a general thank you is in order. I have listed the first editions of books, followed by the editions whose pages I have cited. Adams MB (1979) From “gene fund” to “gene pool”: On the evolution of evolutionary language. Studies in the History of Biology 3: 241–285. Adams MB (1980) Severtsov and Schmalhausen: Russian morphology and the evolutionary synthesis. In The Evolutionary Synthesis, ed. Mayr E, Provine W. Cambridge: Harvard University Press. Adler MJ (2001) A novel mechanism for evolution? Science 294: 53. Aksoy S (2000) Tsetse—a haven for microorganisms. Parasitology Today 16: 114–118. Alexander RD (1991) Some unanswered questions about naked mole rats. In The Biology of the Naked Mole-Rat, ed. Sherman P, Jarvis J, Alexander R. Princeton: Princeton University Press. Alexander S (1920) Space, Time and Deity. London: Macmillan. Alt FW, Kellems RE, Bertino JR, Schinke RT (1978) Selective multiplications of dihydrofolate reductase genes in methotrexate-resistant variants of cultured murine cells. J. Biol. Chem. 253: 1357–1370.

Bibliography

467

Altmann K (1890, 1894) Die Elementerorganismen und ihre Beziehungen zu den Zellen. Leipzig: Von Veit. Anderson PW (1994) A Career in Theoretical Physics. Singapore: World Scientific Publishing. Appel TA (1987) The Cuvier-Geoffroy Debate: French Biology in the Decades before Darwin. Oxford: Oxford University Press. Arber A (1954) The Mind and the Eye. Cambridge: Cambridge University Press. Arthur W (1997) The Origin of Animal Body Plans: A Study in Evolutionary Developmental Biology. Cambridge: Cambridge University Press. Arthur W (1998) The emerging conceptual framework of evolutionary developmental biology. Nature 415: 757–764. Ashby WR (1952) Design for a Brain. New York: Wiley. Ashizawa T, Anvret M, Baiget M, Barcelo JM, Brunner H, Cobo AM, Dallapiccola B, Fenwick RGG, Grandell W, Harley H, Junien C, Koch MC, Korneluk RG, Lavedan C, Miki T, Mulley JC, de Munian AL, Novelli G, Roses AD, Seltzer WK, Shaw dJ, Smeets H, Sutherland R, Yamagata H, Harper PS (1994) Characteristics of intergenerational contractions of the CTG repeat in myotonic dystrophy. Am J. Human Genetics 54: 414–423. Babbage C (1838) The Ninth Bridgewater Treatise: A Fragment, second edition. London: Murray. Bacon F (1620) Novum Organum. London. 1902 edition, New York: Collier. Bacon M, Wu C-T, Morris J (2001) Genes, genetics, and epigenetics: A correspondence. Science 293: 1103–1105. Bak P (1996). How Nature Works: The Science of Self-Organized Criticality. New York: Springer. Bak P, Chen K (1991) Self-organized criticality. Sci. Am. 264, January: 46–53. Baker MW (1997) Steroid receptor phylogeny and vertebrate origins. Molecular and Cellular Endocrinology 135: 101–107. Bakker R (1975) Dinosaur renaissance. Sci. Am. 232, July: 58–78. Baldwin JM (1896) A new factor in evolution. Am. Nat. 30: 354–536. Baldwin JM (1897) Organic selection. Nature 55: 558. Balon EK (1983) Epigenetic mechanisms: Reflections on evolutionary processes. Canadian Journal of Fisheries and Aquatic Sciences 40: 2045–2058. Balon EK (1986) Saltatory ontogeny and evolution. Rivista di Biologia Biology Forum 79: 151–190. Balon EK (l989) The epigenetic mechanisms of bifurcation and alternative life-history styles. In Alternative Life-History Styles of Animals, ed. Bruton M. Dordrecht: Kluwer. Balon EK (2004) Alternative ontogenies and evolution: A farewell to gradualism. In Environment, Development and Evolution, ed. Hall B, Pearson R, and Müller G. Cambridge: MIT Press. Bardou F, Jaeger L (2004) Large phenotype jumps in biomolecular evolution. Physical Review E 69: 031908-1–031908-7.

468

Bibliography

Barrie VJ, Conway KW (2002) Rapid sea level change and coastal evolution on the Pacific margin of Canada. J. Sedimentary Geol. 150: 171–183. Bartlett J (1980) Bartlett’s Familiar Quotations, fifteenth edition, ed. Beck E. Boston: Little, Brown. Bateson B (1928) William Bateson F.R.S. Cambridge: Cambridge University Press. Bateson W (1894) Materials for the Study of Variation, with Especial Regard to Discontinuity in the Origin of the Species. Cambridge: Cambridge University Press. Reprinted by Johns Hopkins University Press, 1992. Beament JW (1961) The role of physiology in adaptation and competition between animals. Symp. Soc. Exptl. Biol. 15: 62–71. Beament JW (1964) The active transport and passive movement of water in insects. In Advances in Insect Physiology, volume 2, ed. Beament J, Treherne J, Wigglesworth V. Academic Press. Beard DB, Dotson EM, Pennington PM, Eichler S, Cordon-Rosales C, Durvasula RV (2001) Bacterial symbiosis and paratransgenic control of vector-borne Chagas disease. International Journal for Parasitology 31: 621–627. Beardsley T (1997) Evolution evolving. Sci. Am., September: 15–16. Behe MJ (1996) Darwin’s Black Box. New York: Simon and Schuster. Bell G (1988) Sex and Death in Protozoa: The History of an Obsession. Cambridge: Cambridge University Press. Belyaev DK (1969) Domestication of animals. Sci. J. 1: 47–52. Belyaev DK (1979) Destabilizing selection as a factor in domestication. Journal of Heredity 70: 301–308. Belyaev DK, Trut LN (1989) The role of behaviour in the evolutionary organization of wild animals into domestic forms. In Behaviour as One of the Main Factors of Evolution, ed. Leonovicová V, Novák V. Prague: Czechoslovak Academy of Sciences. Benford G (1995) Sailing Bright Eternity. New York: Bantam Spectra. Benveniste RE (1985) The contributions of retroviruses to the study of mammalian evolution. In Molecular Evolutionary Genetics, ed. MacIntyre R. New York: Plenum. Berg LS (1926) Nomogenesis or Evolution Determined by Law. London: Constable. First published in Russian in 1922. English edition reissued in 1969 by MIT Press. Berg R (1988) Acquired Traits. New York: Viking Penguin. Bergson H (1908) L’évolution créatrice. Paris: Alcan. Bernard C (1878) Leçons sur les phénomènes de la vie, communes aux animaux et aux végétaux. Paris: Baillière. Berry CV (2000) Bulldogs, Watchdogs and Underdogs: Is Evolutionary Theory a Paradigm or a Kennel? MSc thesis, University of Victoria, B.C. Bertalanffy L von (1952) Problems of Life. New York: Wiley.

Bibliography

469

Bertalanffy L von (1967) Robots, Men, and Minds. New York: Braziller. Bertalanffy L von (1968) General System Theory. New York: Braziller. Bestor TH, Chandler VL, Feinberg AP (1994) Epigenetic effects in eukaryotic gene expression. Developmental Genetics 15: 458–462. Beurlen K (1930) Vergleichende Stammesgeschichte, Grundlagen, Methoden, Probleme, unter besonderer Berucksiehtigung der hoheren Krebse. Fortschr. Geol. 7: 317–586. Bever JD and Simms EL (2000) Evolution of nitrogen fixation in spatially structured populations of Rhizobium. Heredity 85: 366–372. Bibel DJ (1992) Freeing the Goose in the Bottle: Discovering Zen through Science, Understanding Science through Zen. Oakland, CA: Elie Metchnikoff Memorial Library. Bolk L (1929) The origin of racial characteristics in man. Am. J. Phys. Anth. 13: 1–28. Bonner JT (1974) On Development. Cambridge: Harvard University Press. Bonner JT (1988) The Evolution of Complexity by Means of Natural Selection. Princeton: Princeton University Press. Bottema S (1989) Some observations on the modern domestication processes. In The Walking Larder, ed. Clutton-Beck J. London: Unwin Hyman. Boursaux-Eude C, Gross R (2000) New insights into symbiotic associations between ants and bacteria. Res. Microbiol. 151: 513–519. Bowler PJ (1979) Theodor Eimer. Evolution by “definitely directed evolution.” J. Hist. Med. Allied Sci. 34: 40–73. Brenner SE (1974) The genetics of Caenorhabditis elegans. Genetics 77: 71–94. Brenner SE (1999) Theoretical biology in the third millenium. Phil. Trans. Roy. Soc. Lond. 354: 1963–1965. Brenner SE, Elgar G, Sandford R, Macrea A, Venkatesh B, Aparicio S (1993) Characterization of the pufferfish (fugu) genome as a compact model vertebrate genome. Nature 366: 265–268. Breznac JA, Brune A (1994) Role of microorganisms in digestion of lignocellulose by termites. Ann. Rev. Entomology 39: 453–488. Britten RJ, Davidson EH (1969) Gene regulation for higher cells: A theory. Science 165: 349–357. Britten RJ, Davidson EH (1971) Repetitive and non-repetitive DNA sequences and a speculation on the origins of evolutionary novelty. Quart. Rev. Biol. 46: 111–138. Brockman J (1995) The Third Culture: Beyond the Scientific Revolution. New York: Simon and Schuster. Brooks DR, Wiley EO (1986) Evolution as Entropy: Toward a Unified Theory of Biology. Chicago: University of Chicago Press. Brooks WK (1883) The Law of Heredity: A Study of the Cause of Variation, and the Origin of Living Organisms. Baltimore: Murphy.

470

Bibliography

Brown DD, David IB (1968) Specific gene amplification in oocytes. Science 160: 272–280. Brown DD, Wensink PC, Jordan E (1972) A comparison of the ribosomal DNAs of Xenopus laevis and Xenopus mülleri: Evolution of tandem genes. J. Mol. Biol. 63: 57–73. Brown P, Sutikna T, Morwood MJ, Soejono RP, Jatmiko, Septomo EW, Rokus AD A new smallbodied hominin from the Late Pleistocene of Flores, Indonesia. Nature 431: 1055–1061. Brunner E (1952) The Christian Doctrine of Creation and Redemption. New York: Westminster. Bruton MN (1994) The epigenesis of an epigeneticist: An interview with Eugene Balon. South African Journal of Science 90: 270–275. Bry L, Falk PG, Midtvedt T, Gordon JI (1996) A model of host-microbial interactions in an open mammalian ecosystem. Science 273: 1380–1383. Bryant D, Burke L, McManus J, Spalding M (1998) Reefs at Risk. Washington: World Resources Institute. Brylski P, Hall BK (1988) Ontogeny of a macroevolutionary phenotype: The external cheek pouches of geomyoid rodents. Evolution 42: 391–395. Buchner P (1965) Endosymbiosis of Animals with Plant Micro-organisms. New York: Wiley Interscience. Burke AC (1989) The development of the turtle carapace: Implications for the evolution of a novel bauplan. J. Morphology 199: 363–378. Burke AC, Nelson CE, Morgan BA, Tabin CL (1995) Hox genes and the evolution of vertebrate axial morphology. Development 121: 333–346. Burkhardt RW (1977) The Spirit of System: Lamarck and Evolutionary Biology. Cambridge: Harvard University Press. Burkholder JM (1999) The lurking perils of Pfiesteria. Sci. Am. 281, August: 42–49. Buss LW (1987) The Evolution of Individuality. Princeton: Princeton University Press. Butler S (1879) Evolution Old and New. London: Hardwicke & Bogue. Butler S (1886) Luck or Cunning. London: Fifield. Cai H, Levine M (1995) Modulation of enhancer-promoter interactions by insulators in the Drosophila embryo. Nature 376: 533–536. Cain AJ, Sheppard PM (1954) Natural selection in Cepaea. Genetics 39: 89–116. Cairns-Smith AG (1971) The Life Puzzle. Edinburgh: Oliver and Boyd. Callebaut W, Rasskin-Gutman D, eds. (2005) Modularity: Understanding the Development and Evolution of Natural Complex Systems. Cambridge: MIT Press. Campbell JH (1985) An organizational interpretation of evolution. In Evolution at a Crossroads, ed. Depew D, Weber B. Cambridge: MIT Press. Campbell NA (1993) Biology, third edition. Redwood City, CA: Benjamin/Cummings. Campbell NA, Reece JB, Mitchell LG (1999) Biology, fifth edition. Menlo Park, CA: Benjamin/Cummings.

Bibliography

471

Cannon WB (1932) The Wisdom of the Body. New York: Norton. Caporale LH (2003) Darwin in the Genome: Molecular Strategies in Biological Evolution. McGraw-Hill. Caporale LH (2003) Natural selection and the emergence of a mutational phenotype: An update of the evolutionary synthesis considering mechanisms that affect genome variation. Ann. Rev. Microbiol. 57: 467–485. Carroll L (1867, 1887). Alice through the Looking Glass. Reissued by Folio Society, London, 1962. Carroll R (1988) Paleontology and Vertebrate Evolution. New York: Freeman. Carroll R (1997) Patterns and Processes of Vertebrate Evolution. Cambridge: Cambridge University Press. Carroll SB (1995) Homeotic genes and the evolution of arthropods and chordates. Nature 376: 479–485. Casti J (1989) Paradigms Lost: Images of Man in the Mirror of Science. New York: Morrow. Caudwell C (1986) Heredity and development. In Caudwell, Scenes and Actions. London: Routledge and Kegan Paul. Cavalier-Smith T (1985) The Evolution of Genome Size. Chichester: Wiley. Cavalier-Smith T (1987) The origin of cells: A symbiosis between genes, catalysts and membranes. Cold Spring Harbor Symp. Quant. Biol. 52: 805–842. Cavanaugh CM (1983) Symbiotic chemoautotrophic bacteria in marine invertebrates from sulphide-rich habitats. Nature 302: 58–61. Chaisson EJ (2001) Cosmic Evolution: The Rise of Complexity in Nature. Cambridge: Harvard University Press. Chambers R (1844) Vestiges of the Natural History of Creation. London: Churchill. Reissued 1994 in facsimile by University of Chicago Press. Chan WWS, Chong KY, Madinovich C, Simerly C, Schallen G (2001) Transgenic monkeys produced by retroviral gene transfer into mature oocytes. Science 291: 309–312. Chauvin R (1977) Ethology: The Biological Study of Animal Behavior. New York: International Universities Press. Cheng G, Aksoy S (1999) Tissue tropism, transmission and expression of foreign genes in vivo in midgut symbionts of tsetse flies. Insect Molecular Biology 8, 125–132. Child CM (1924) Physiological Foundations of Behaviour. New York: Holt. Chothia C, Gough J, Vogel C, Teichmann SA (2003) Evolution of the protein repertoire. Science 300: 1701–1703. Churchill FB (1980) The modern evolutionary synthesis and the biogenetic law. In The Evolutionary Synthesis, ed. Mayr E, Provine W. Cambridge: Harvard University Press. Clark MA, Moran NA, Baumann P, Wernegreen JJ (2000) Cospeciation between bacterial endosymbionts (Buchnera) and a recent radiation of aphids (Uroleucon) and pitfalls of testing for phylogenetic congruence. Evolution 54: 517–525.

472

Bibliography

Clark R (1968) J.B.S.: The Life and Work of J.B.S. Haldane. London: Hodder and Stoughton. Clark SRL (1975) Aristotle’s Man. Oxford: Clarendon. Cleveland LR, Grimstone RV (1964) The fine structure of the flagellate Mixotricha paradoxa and its associated micro-organisms. Proc. Roy. Soc. B159: 668–686. Clever U, Karlson P (1960) Induktion von Puff-Veränderungen in den Speicheldrüsenchromosomen von Chironomus tentans durch Ecdyson. Exp. Cell Res. 20: 623–626. Cohen B (1997) Genetics: A heritage of excellence. The Avenue (Glasgow University) 22: 20–22. Cohen J, Stewart I (1994) The Collapse of Chaos. New York: Penguin. Cohen P (2001) Mother’s little helper. New Scientist, February 24: 10. Cole BJ (1991) Is animal behaviour chaotic? Evidence from the activity of ants. Proc. Roy. Soc. B224: 253–259. Collins JH, Cheek JE (1983) Effect of food and density on development of typical and cannibalistic salamander larvae in Ambystoma tigrinum nebulosum. Amer. Zool. 23: 77–84. Conrad M (1983) Adaptability: The Significance of Variability from Molecule to Ecosystem. New York: Plenum. Conrad M (1990) The geometry of evolution. Biosystems 24: 61–81. Conway Morris S (1977) A new metazoan from the Burgess Shale. Paleontology 20: 623–640. Conway Morris S (1995) Ecology in deep time. Trends in Ecology and Evolution 10: 290–294. Conway Morris S (1998) The Crucible of Creation. Oxford: Oxford University Press. Cook LM (1961) The edge effect in population genetics. Am. Nat. 95: 295–307. Cope ED (1887) The Origin of the Fittest. Chicago: Open Court Cope ED (1896) The Primary Factors of Organic Evolution. Chicago: Open Court. Copleston F (1960) A History of Philosophy. Modern Philosophy, volume 4: Descartes to Leibniz. Birmingham: Newman. Garden City, NY: Doubleday Image Edition, 1963. Corning PA (1997) Holistic Darwinism: “Synergistic selection” and the evolutionary process. Journal of Social and Evolutionary Systems 20: 363–401. Corning PA (1998) The synergism hypothesis. On the concept of synergy and its role in the evolution of complex systems. Journal of Social and Evolutionary Systems 21: 1–35. Corning PA (2002) Nature’s Magic: Synergy in Evolution and the Fate of Humankind. Cambridge: Cambridge University Press. Costerton JM, Lewandowski Z (1995) Microbial biofilms. Annual Review of Microbiology 49: 711–745. Couly G, Grapin-Botton A, Coltey P, Ruhin B, Le Douarin NM (1998) Determination of the identity of the derivatives of the cephalic neural crest: Incompatability between Hox gene expression and lower jaw development. Development 125: 3445–3459.

Bibliography

473

Covello PS, Gray MW (1992) Silent mitochondrial and active nuclear genes for subunit 2 of cytochrome oxidase (cox 2) in soybean: Evidence for RNA-mediated gene transfer. EMBO J. 11: 3815–3820. Cremer T, Cremer C (2001) Chromosome territories, nuclear architecture and gene regulation in mammalian cells. Nature Reviews Genetics 2: 292–301. Crespi BJ (2001) The evolution of social behavior in microorganisms. Trends in Ecology and Evolution 16: 178–183. Croizat L (1962) Space, Time, Form: The Biological Synthesis. Caracas: Published by the author. Cruse D (1999) Science and intentionality: Exploring a serious contradiction in the logic of scientific materialism. Unpublished ms. received by R. G. B. Reid from the author. Curtis AD, Waller DA (1997) Variation in rates of nitrogen fixation in termites: Response to dietary nitrogen in the field and laboratory. Physiological Entomology 22: 303. Cuthill S (1994) Cellular epigenetics and the origin of cancer. BioEssays 16 (6): 393–394. Czelusniak J, Goodman M, Hewett-Emmett D, Weiss ML, Venta PJ, Tashian RE (1982) Phylogenetic origins and adaptive evolution of avian and mammalian haemoglobin genes. Nature 298: 297–300. Darwin C (1859) On the Origin of Species by Means of Natural Selection. London: Murray. Darwin C (1868) The Variation of Animals and Plants under Domestication. London: Murray. Darwin C (1871) The Descent of Man. London: Murray. Darwin C (1872) On the Origin of Species by Means of Natural Selection, sixth edition. London: Murray. Darwin E (1794) Zoonomia. Johnson. Third edition, 1801. Darwin F, ed. (1892) The Autobiography of Charles Darwin and Selected Letters. London: Appleton. Facsimile edition: New York: Dover, 1961. Dassow GE von, Meir E, Munro GM, Odell M (2000) The segment polarity network is a robust developmental module. Nature 406: 188–192. Davidson EH, Peterson KJ, Cameron RA (1995) Origin of bilaterian body plans: Evolution of developmental regulatory mechanisms. Science 270: 1319–1325. Davies P (1999) The Fifth Miracle. The Search for the Origin and Meaning of Life. New York: Touchstone. Daviss B (1998) Cast out of the Garden of Eden. New Scientist 158: 26–30. Davitaschvili L (1966) The Contemporary Status of Evolutionary Studies in the West (published in Russian). Moscow: Acad. Sci. USSR, Div. Gen. Biol., Science Publ. House. Dawkins R (1976) The Selfish Gene. Oxford: Oxford University Press. Dawkins R (1982) The Extended Phenotype. Oxford: Oxford University Press. Dawkins R (1985) The Blind Watchmaker. London: Longman.

474

Bibliography

Dawkins R (1989) The evolution of evolvability. In Artificial Life, volume 1, ed. Langton C. Redwood City CA: Addison-Wesley. Dawkins R (1995) A survival machine. In The Third Culture, ed. Brockman J. New York: Simon and Schuster. De Bary A (1879) Die Erscheinung der Symbiose. Vortrag auf der Versammlung der Naturforscher und Artze zu Cassel. Strassburg: Trubner. De Beer G (1940) Embryos and Ancestors. Oxford: Oxford University Press. Deeming DC, Ferguson MW (1988) Environmental regulation of sex determination in reptiles. Phil. Trans. Roy. Soc. B322: 19–39. De la Mettrie JO (1748) L’Homme machine. Leiden. Delbrück M (1949) A physicist looks at biology. Trans. Conn. Acad. Arts Sci. 38: 173–190. Desmond A, Moore J (1991) Darwin. The Life of a Tormented Evolutionist. New York: Norton. Detera-Wadleigh SD, Fanning TG (1994) Phylogeny of the steroid receptor superfamily. Molecular Phylogenetics and Evolution 3: 192–205. De Vries H (1901–03) Die Mutationstheorie. Leipzig: Von Veit. De Vries H (1909–10) Mutation Theory. Chicago: Open Court. Dobzhansky T (1937) Genetics and the Origin of Species. New York: Columbia University Press. Dobzhansky T (1967) The Biology of Ultimate Concern. New York: New American Library. Dobzhansky T, Ayala FJ, Stebbins GL, Valentine JW (1977) Evolution. San Francisco: Freeman. Dohlen CD von, Kohler S, Alsop ST, McManus WR (2001) Mealybug β-proteobacterial endosymbionts contain γ-proteobacterial symbionts. Nature 412: 433–436. Dohrn A (1875) Der Ursprung der Wirbelthiere und das Prinzip des Funktionswechsels: Genealogische Skizzen. Doolittle WF (1987) The origin and function of intervening sequences: A review. Am. Nat. 130: 915–928. Doolittle WF (1991) The nature of introns. Current Biol. UK 1: 145–146. Doolittle WF (2002) Microbial genomes multiply. Nature 416: 697–700. Doolittle WF, Brown JR (1994) Tempo, mode, the progenote and the universal root. Proc. Natl. Acad. Sci. 91: 6721–6728. Douglas AE (1994) Symbiotic Interactions. Oxford: Oxford University Press. Dover GA (1982) Molecular drive: A cohesive mode of species evolution. Nature 299: 111–117. Dover GA (1986) Molecular drive in multigene families. How biological novelties arise, spread, are assimilated. Trends in Genetics 2: 159–165. Dover GA (2000) Anti-Dawkins. In Alas, Poor Darwin: Arguments against Evolutionary Psychology, ed. Rose H, Rose S. London: Jonathan Cape.

Bibliography

475

Dover GA (2000) Dear Mr. Darwin. Berkeley: University of California Press. Driesch H (1908) The Science and Philosophy of the Organism. London: Black. Driever W, Nüsslein-Volhard C (1988a) The bicoid protein determines position in the Drosophila embryo in a concentration-dependent manner. Cell 54: 83–93. Driever W, Nüsslein-Volhard C (1988b) A gradient of bicoid protein in Drosophila embryos. Cell 54: 95–104. Drummond H (1883) Natural Law in the Spiritual World. London: Hodder and Stoughton. Drummond H (1894) The Ascent of Man. London: Hodder and Stoughton. Duchesne LC, Larson DW (1989) Cellulose and the evolution of plant life. BioScience 39: 238–241. Dworkin J, Allamonda LJ, Deamer DW, Sandford SA (2001) Self-assembling amphiphilic molecules: Synthesis in simulated interstellar/precometary ice. Proc. Natl. Acad. Sci. 98: 815–819. Dyer G (1997) Darwin theory proving to be survival of fittest. Victoria Times Colonist, August 24. Edelman GM (1987) Cell adhesion and the evolutionary origins of immunity. Immunological Reviews 100: 11–45. Eimer T (1898) On Orthogenesis and the Impotence of Natural Selection in Species Formation. Chicago: Open Court. Eldredge N (1985) Time Frames: The Rethinking of Darwinian Evolution and the Theory of Punctuated Equilibria. New York: Simon and Schuster. Eldredge N (1985) Unfinished Synthesis: Biological Hierarchies and Modern Evolutionary Thought. Oxford: Oxford University Press. Eldredge N (1995) Reinventing Darwin: The Great Debate at the High Table of Evolutionary Theory. New York: Wiley. Eldredge N, Gould SJ (1972) Punctuated equilibria: An alternative to phyletic Gradualism. In Models in Paleobiology, ed. San Francisco: Schopf T. Freeman, Cooper. El-Hani CN, Pihlström S (2002) Emergence theories and pragmatic realism. http://www .helsinki.fi/science/commens/papers/emergentism.pdf. Eliot TS (1943) Little Gidding. In Four Quartets. New York: Harcourt, Brace. Emerson AE (1939) Social coordination and the superorganism. Amer. Midl. Nat. 21: 182–209. Endler JA (1986) Natural Selection in the Wild. Princeton: Princeton University Press. Engelberg-Kulka H, Glaser G (1999) Addiction modules and programmed cell death and antideath in bacterial cultures. Ann. Rev. Microbiol. 53: 43–70. Engels F (1878) Anti-Düring. Leipzig. Engels (1954) The Dialectics of Nature. Moscow: Foreign Languages Publishing. First published in Russian and German in 1925. Escriva H, Manzon L, Youson J, Laudet V (2002) Analysis of lamprey and hagfish genes reveals a complexity of gene duplication during early vertebrate evolution. Mol. Biol. Evol. 19: 1440–1450.

476

Bibliography

Fankboner PV (1971) Intracellular digestion of symbiotic zooxanthellae by host amoebocytes in giant clams (Bivalvia: Tridacnidae), with a note on the role of the hypertrophied siphonal epidermis. Biol. Bull. Mar. Biol. Lab. Woods Hole 141: 222–234. Fankboner PV, Reid RGB (1981) Mass expulsion of zooxanthellae by heat-stressed corals: A source of food for giant clams? Experientia 37: 251–252. Farmer C (1997) Did lungs and the intercardiac shunt evolve to oxygenate the heart in vertebrates? Paleobiology 23: 358–372. Fenchel TL, Riedl RJ (1971) The sulfide system: A new biotic community underneath the oxidized layer of marine sand bottoms. Mar. Biol. 7: 255–268. Fisher RA (1930) The Genetical Theory of Natural Selection. Oxford: Clarendon. Fitch WM, Ayala FJ, eds. (1995) Tempo and Mode in Evolution: Genetics and Paleontology 50 Years after Simpson. Washington: National Academy Press. Fitt WK (1985) The effect of different strains of the zooxanthellae Symbiodinium microadriaticum on the growth and survival of their coelenterate and molluscan hosts. Proceedings of the International Congress on Coral Reefs 6: 131–136. Flamm WG, Walker MB, McCallum M (1969) Some properties of the single strands isolated from the DNA of the nuclear satellite of the mouse (Mus musculus). J. Mol. Biol. 40: 320–443. Fleming A, Allison VC (1922) Observations on a bacteriolytic substance (“lysozyme”) found in secretions and tissues. British J. Exp. Pathology 3: 252–260. Fondon JW III, Garner HR (2004) Molecular origins of rapid and continuous morphological evolution. Proc. Natl. Acad. Sci. 101: 18058–18063. Foster JB (2000) Marx’s Ecology. New York: Monthly Review Press. Frank AB (1877) Ueber der biologischen Verhältnisse des thallus eineger Krustenflechten. Beitrage zur Biologie der Pflanzen 195: 123–200. Franková S (1987) Contribution of the concepts of constraints and developmental plasticity to the study of behavioral ontogeny and phylogeny. In Behaviour as One of the Main Factors of Evolution, ed. Leonovicová V, Novák V. Prague: Czechoslovak Academy of Sciences. ffrench-Constant RH, Bowen DJ (2000) Novel insecticidal toxins from nematode-symbiotic bacteria. Cell. Mol. Life Sci. 57: 828–833. Frey E, Sues HD, Munk W (1997) Gliding mechanism in the Late Permian reptile Coelurosauravus. Science 275: 1450–1452. Fruton J (1999) Proteins, Enzymes and Genes: The Interplay of Chemistry and Biology. New Haven: Yale University Press. Fryer G, Iles TD (1972) The Cichlid Fishes of the Great Lakes of Africa: Their Biology and Evolution. Edinburgh: Oliver and Boyd. Fukatsu T, Ishikawa H (1992) Occurrence of chaperonin 60 and chaperonin 10 in primary and secondary bacterial symbionts of aphids: Implications for the evolution of an endosymbiotic system in aphids. J. Mol. Evol. 36: 568–577.

Bibliography

477

Futuyama DJ (1998) Evolutionary Biology, third edition. Sunderland MA: Sinauer. Galton F (1869) Hereditary Genius, and Inquiry into Its Laws. London: Macmillan. Gánti T (1975) Organization of chemical reactions into dividing and metabolizing units: The chemotons. BioSystems 7: 189–195. Garstang W (1928) The origin and evolution of larval forms. Proceedings of the British Association for the Advancement of Science D: 1–23. Garstang W (1951) Larval Forms and Other Zoological Verses. Oxford: Blackwell. Gehring WJ (1998) Master Control Genes in Development and Evolution: The Homeobox Story. New Haven: Yale University Press. Geoffroy é de St H (1818) Philosophie anatomique, volume 1. Paris. Geoffroy é de St H (1822) Philosophie anatomique, volume 2. Paris. Geoffroy é de St H (1833) Mémoire sur le degré d’influence du monde ambiant pour modifier les formes animals: Question intéressant l’origine des espèces télésauriennes et successivement celle des animaux de l’époque actuelle. Mémoires de l’Académie des Sciences 12: 63–92. Geoffroy I de St H (1832–1836) l’Histoire génerale et particulière des anomalies d’organisation chez l’homme et les animaux (Traité de tératologie). Gerard RW (1942) Higher levels of integration. In Biological Symposia VIII, ed. Redfield R. Lancaster. Gerhart J, Kirschner M (1997) Cells, Embryos, and Evolution. Oxford: Blackwell Science. Ghiselin MT (1969) The Triumph of the Darwinian Method. Berkeley: University of California Press. Ghiselin MT (1974) A radical solution to the species problem. Systematic Zoology 23: 536–44. Gilbert LI, Tata JR, eds. (1996) Metamorphosis: Postembryonic Reprogramming of Gene Expression in Amphibian and Insect Cells. Academic Press. Gilbert SF (1994) Developmental Biology, fourth edition. Sunderland MA: Sinauer. Gilbert W (1991). A vision of the grail. In The Code of Codes, ed. Kevles D, Hood L. Cambridge: Harvard University Press. Gillespie J (1991) The Causes of Molecular Evolution. Oxford: Oxford University Press. Gleick J (1987) Chaos: Making a New Science. New York: Penguin. Goebel W, Gross R (2001) Intracellular survival strategies of mutualistic and parasitic prokaryotes. Trends in Microbiology 9: 267–273. Goldschmidt RB (1940) The Material Basis of Evolution. New Haven: Yale University Press. Goldschmidt RB (1960) In and Out of the Ivory Tower. Seattle: University of Washington Press. Goodman M (1981) Decoding the pattern of protein evolution. Prog. Biophys. Mol. Biol. 38: 105–164.

478

Bibliography

Goodrich ES (1913) Metameric segmentation and homology. Quart. J. Micr. Sci. 59: 227–248. Goodwin BC (1989) Evolution and the generative order. In Theoretical Biology, ed. Goodwin B, Saunders P. Edinburgh: Edinburgh University Press. Goodwin BC (1991) Development. London: Hodder and Stoughton and Open University. Goodwin BC (1994) How the Leopard Changed its Spots: The Evolution of Complexity. New York: Simon and Schuster. Goodwin BC (1995) Biology is just a dance. In The Third Culture, ed. Brockman J. New York: Simon and Schuster. Goodwin BC, Saunders PT, eds. (1989) Theoretical Biology: Epigenetic and Evolutionary Order from Complex Systems. Edinburgh: Edinburgh University Press. Gorces VG (1995) Keeping enhancers under control. Nature 376: 462–463. Gordon DM (1999) Ants at Work. New York: Free Press. Gordon DM, Goodwin BC, Trainor LH (1992) A parallel distributed model of ant colony behaviour. J. Theor. Biol. 156: 293–307. Gottlieb G (1976) The roles of experience in the development of behavior and the nervous system. In Neural and Behavioral Specificity, ed. Gottlieb G. Academic Press. Gottlieb G (1987) The developmental basis of evolutionary change. J. Comp. Psych. 101: 262–271. Gottlieb G (1992) Individual Development and Evolution. Oxford: Oxford University Press. Goudge TA (1961) The Ascent of Life: A Philosophical Study of the Theory of Evolution. Toronto: University of Toronto Press. Gould SJ (1974) The evolutionary significance of “bizarre” structures: Antler size and skull size in the “Irish elk,” Megaloceros giganteus. Evolution 28: 191–220. Gould SJ (1977a) Ontogeny and Phylogeny. Cambridge: Belknap. Gould SJ (1977b) Ever Since Darwin: Reflections in Natural History. New York: Norton. Gould SJ (1980a) The panda’s thumb. In The Panda’s Thumb. New York: Norton. Gould SJ (1980b) Is a new and general theory of evolution emerging? Paleobiology 6: 119–130. Gould SJ (1983a) Helpful monsters. In Hen’s Teeth and Horse’s Toes. New York: Norton. Gould SJ (1983b) Quaggas, coiled oysters, and flimsy facts. In Hen’s Teeth and Horse’s Toes. New York: Norton. Gould SJ (1983c) Hen’s teeth and horses’ toes. In Hen’s Teeth and Horse’s Toes. New York: Norton. Gould SJ (1989). Wonderful Life: The Burgess Shale and the Nature of History. New York: Norton. Gould SJ (1991) The disparity of the Burgess Shale arthropod fauna and the limits to cladistic analysis: Why we must strive to quantify morphospace. Paleobiology 8: 4–15. Gould SJ (1994) The evolution of life on the earth. Sci. Am. 271 October: 85–91.

Bibliography

479

Gould SJ (1995) The Macroevolutionary reconstruction of Darwinism. In Tempo and Mode in Evolution, ed. Fitch W, Ayala F. Washington: National Academy Press. Gould SJ (1997) Darwinian fundamentalism. New York Review of Books, June 12. Gould SJ (2002) The Structure of Evolutionary Theory. Cambridge: Belknap. Gould SJ, Lewontin R (1979) The spandrels of San Marco and the Panglossian paradigm: A critique of the adaptationist program. Proc. Roy. Soc. B205: 581–598. Gould SJ, Vrba E (1982) Exaptation—a missing term in the science of form. Paleobiology 8: 4–15. Grady D (1996) Quick-change pathogens gain an evolutionary edge. Science 274: 1081. Grant PR, Grant BR (1994) Phenotypic and genetic effects of hybridization in Darwin’s finches. Evolution 48: 297–316. Grassé P-P, ed. (1948–1975) Traité de zoologie. Paris: Masson. Grassé P-P, ed. (1973) L’évolution du vivant. Paris: Albin Michel. Grassé P-P, ed. (1977) Evolution of Living Organisms. New York: Academic Press. Graur D, Li WH (2000) Fundamentals of Molecular Evolution, second edition. Sunderland MA: Sinauer. Greene JC (1971) Science, Ideology, and World View. Berkeley: University of California Press. Greene RW (1970) Symbiosis in sacoglossan opisthobranchs: Symbiosis with algal chloroplasts. Malacologia 10: 357–368. Greenwood PH (1974) The cichlid fishes of Lake Victoria, East Africa: The biology and evolution of a species flock. Bull. British Mus. Nat. Hist (Zool.) Supplement 6: 1–134. Gregorini O (1986) Evolution, Development and Cancero-genesis. Il Programma Vita Edizione Minarelli. Gregory WK (1934) Polyisomerism and anisomerism in cranial and dental evolution among vertebrates. Proc. Natl. Acad. Sci. 20: 1–9. Grehan JR (1984) Evolution by law. Croizat’s “orthogeny” and Darwin’s “Laws of Growth.” Tuatara 27: 14–19. Grehan JR, Ainsworth R (1985) Orthogenesis and evolution. Syst. Zool. 34: 174–192. Greider CW (1998) Telomeres and senescence: The history, the experiment, the future. Current Biology 8: R178–R181. Greksa LP (1996) Evidence of a genetic basis to the enhanced total lung capacities of Andean Highlanders. Human Biology 68: 119–129. Grémillet D, Chauvin C, Wilson RP, Le Maho Y, Wanless S (2005) Unusual feather structure allows partial plumage wettability in diving great cormorants Phalacrocorax carbo. Journal of Avian Biology 36: 57–63. Grémillet D, Wanless S (2000) Fast food dive-in. NREC news release, August 30. Grene M (1988) Is evolution at a crossroads? Evolutionary Biology 24: 51–81.

480

Bibliography

Grimes GW, Aufderheide KJ (1991) Cellular Aspects of Pattern Formation: The Problem of Assembly. Basel: Karger. Grunstein M (1992) Histones as regulators of genes. Sci. Am. 267: 68–74. Gustafson R, Reid RGB (1988a) Larval and post-larval morphogenesis in the gutless protobranch bivalve Solemya reidi (Cryptodonta: Solemyidae). Mar. Biol. 97: 373–387. Gustafson R, Reid RGB (1988b) Association of bacteria with larvae of the gutless protobranch bivalve Solemya reidi (Cryptodonta: Solemyidae). Mar. Biol. 97: 389–401. Haake W (1895) Gestalten und Vererbung. Leipzig: Veigel. Haines LR (2002) The Identification and Partial Characterization of Molecules Found within the Midgut of the Tsetse Fly. MSc Thesis, University of Victoria, B.C. Haldane JBS (1932) The Causes of Evolution. London: Longman. Haldane JS (1913) Mechanism, Life and Personality. London: Murray. Halder AM, Callaerts O, Gehring WJ (1995) Induction of ectopic eyes by targeted expression of the eyeless gene in Drosophila. Science 267: 1788–1792. Hall BK (1978) Developmental and Cellular Skeletal Biology. New York: Academic Press. Hall BK (1992) Evolutionary Developmental Biology. London: Chapman and Hall. Hall BK, ed. (1994) Homology. San Diego: Academic Press. Hall BK (1999a) Evolutionary Developmental Biology, second edition. Dordrecht: Kluwer. Hall BK (1999b) The Neural Crest in Development and Evolution. New York: Springer-Verlag. Hall BK (2000) Balfour, Garstang, de Beer: The first century of evolutionary embryology. Amer. Zool. 40: 718–728. Hall BK (2001) Organic selection: Proximate environmental effects on the evolution of morphology and behaviour. Biology and Philosophy 16: 215–237. Hall BK, Pearson R, Müller GB, eds. (2003) Environment, Development and Evolution: Toward a Synthesis. Cambridge: MIT Press. Hall TS (1969) Ideas of Life and Matter. Chicago: University of Chicago Press. Hamburger V (1980) Embryology and the modern synthesis in evolutionary theory. In The Evolutionary Synthesis, ed. Mayr E, Provine W. Cambridge: Harvard University Press. Haraway DJ (1976) Crystals, Fabrics and Fields: Metaphors of Organicism in Twentieth Century Developmental Biology. New Haven: Yale University Press. Hardy AC (1954) Escape from specialization. In Evolution as a Process, ed. Huxley J, Hardy A, Ford E. London: Allen and Unwin. Hardy AC (1965) The Living Stream. London: Collins. Hawkes N (1997) Is science nearing the final frontier? Review of John Horgan’s The End of Science. Times (London), May 5.

Bibliography

481

Heddi A, Charles H, Khatchadourian C (2001) Intracellular bacterial symbiosis in the genus Sitophilus: The “biological individual” concept revisited. Res. Microbiol. 152: 431–457. Hedges SB, Parker PH, Sibley CG, Kumar S (1996) Continental breakup and the ordinal diversification of birds and mammals. Nature 381: 226–229. Henderson LJ (1917) The Order of Nature. Cambridge: Harvard University Press. Herring SW (1993) Formation of the vertebrate face: Epigenetic and functional influences. American Zoologist 33: 472–483. Heslop-Harrison J (1983) Chromosomes, cladism and the new evolutionary debate. In Kew Chromosomal Conference II, ed. Brandham P, Bennett M. London: Allen and Unwin. Himmelfarb G (1959) Darwin and the Darwinian Revolution. Garden City NY: Doubleday. His W (1874) über unsere Körperform. Leipzig: Vogel. Ho M-W (1984) Environment and heredity in development and evolution. In Beyond NeoDarwinism, ed. Ho M-W, Saunders P. London: Academic Press. Ho M-W, Saunders PT (1982) Adaptation and natural selection: Mechanisms and teleology. In Towards a Liberatory Biology, ed. Rose S. London: Allison and Busby. Hochachka P (1973) Comparative intermediary metabolism. In Comparative Animal Physiology, third edition, ed. Prosser C. Saunders. Holland JH (1998) Emergence. From Chaos to Order. Reading MA: Addison-Wesley. Holliday R (1993) Epigenetic inheritance based on DNA methylation. In DNA Methylation, ed. Jost J, Saluz H. Birkhauser. Holliday R (1994). Epigenetics: An overview. Developmental Genetics 15: 453–457. Honderich T, ed. (1995) The Oxford Companion to Philosophy. Oxford: Oxford University Press. Hooper LV, Wong MH, Thalin A, Hasson L, Falk PG, Gordon JI (2001) Molecular analysis of commensal host-microbial relationships in the intestine. Science 291: 881–884. Horder TJ (1994) Partial truths: A review of the use of concepts in evolutionary science. In Models in Phylogeny Reconstruction, ed. Scotland R, Siebert D, Williams D. Oxford: Clarendon. Horgan J (1996) The End of Science. New York: Addison-Wesley (Broadway Books, 1997). Horowitz NH (1945) Evolution of biochemical syntheses. Proc. Nat. Acad. Sci. 31: 153–157. Hovasse R (1950) Adaptation et évolution. Paris: Hermann. Hughes AJ, Lambert DM (1984) Structuralism, functionalism, and “ways of seeing.” J. Theor. Biol. 111: 787–800. Hume D (1748) Philosophical Essays Concerning Human Understanding. London. Humphrey N. 1995. The thick moment. In The Third Culture, ed. Brockman J. New York: Simon and Schuster. Hurst LD (1992) Intragenomic conflict as an evolutionary force. Proc. Roy. Soc. B248: 135–140. Hurst LD, Randerson JP (2002) Parasitic sex puppeteers. Sci. Am. 287, April: 56–61.

482

Bibliography

Hutton FW (1899) Darwinism and Lamarckism. London: Duckworth. Huxley J (1942) Evolution: The Modern Synthesis. London: Allen & Unwin. Second edition, 1962; third edition, 1974. Huxley J (1948) Man in the Modern World. New York: Harper and Row, Mentor. Huxley TH (1864) Criticisms on “The Origin of Species.” Nat. Hist. Rev. New Series 4: 25–36. Huxley TH (1882) The Coming of Age of the “Origin of Species.” Proc. Roy. Inst. 9: 361–368. Huxley TH (1894) Owen’s position in the history of anatomical science: Appendix to volume 3. In R. Owen, The Life of Richard Owen. London: Murray. Imms AD (1957) A General Textbook of Entomology, ninth edition. London: Methuen. Jablonka E, Lamb MJ (1995) Epigenetic Inheritance and Evolution: The Lamarckian Dimension. Oxford: Oxford University Press. Jablonka E, Lamb MJ (2005) Evolution in Four Dimensions: Genetic, Epigenetic, Behavioral and Symbiolic Variation in the History of Life. Cambridge: MIT Press. Jablonski D (1986) Larval ecology and macroevolution in marine invertebrates. Bull. Marine Sci. 39: 565–587. Jacob F, Monod J (1961) Genetic regulatory mechanisms in the synthesis of proteins. J. Mol. Biol. 3: 318–56. Jegalian K, Page DC (1998) A proposed path by which genes common to mammalian X and Y chromosomes evolve to become X inactivated. Nature 394: 776–780. Jegalian K, Lahn BR (2001) Why the Y is so weird. Sci. Am. 284, February: 56–61. Jennings HS (1927) Diverse doctrines of evolution, their relation to the practice of science and of life. Science 65: 19–25. Jiminez G, Ford AM, Enver T, Boronat A (1993) Multiple changes in chromatin structure precede the transcriptional activation of the human growth hormone locus in placental cells. Molecular and Cellular Endocrinology 96: 53–60. Johnson PE (1993) Darwin on Trial. Downer’s Grove IN: InterVarsity. Johnston N (2004) Seeds of a micro revolution. The-scientist.com. Jones FW (1943) Habit and Heritage. London: Routledge. Jones S (1995) Why is there so much genetic diversity? In The Third Culture, ed. Brockman J. New York: Simon and Schuster. Kane J (1995) Savages. Vancouver: Douglas and Stewart. Kauffman SA (1989) Origins of order in evolution: Self-organization and selection. In Theoretical Biology, ed. Goodwin B, Saunders P. Edinburgh: Edinburgh University Press. Kauffman SA (1991) Antichaos and adaptation. Sci. Am. 265, August: 78–84. Kauffman SA (1993) Origins of Order: Self-Organization and Selection in Evolution. New York: Oxford University Press.

Bibliography

483

Kauffman SA (1995) At Home in the Universe. New York: Oxford University Press. Kayzer W (1996) A Glorious Accident. TV Matters. Keller EF (1983) A Feeling for the Organism. San Francisco: Freeman. Keller EF (1995) Refiguring Life: Metaphors of Twentieth-Century Biology. New York: Columbia University Press. Keller EF (2002) Making Sense of Life. Cambridge: Harvard University Press. Kellner RL (2002) Molecular identification of an endosymbiotic bacterium associated with pederin biosynthesis in Paeder sabaeus (Coleoptera: Staphylinidae). Insect Biochem. Mol. Biol. 32: 389–395. Kennison JA (1993) Transcriptional activation of Drosophila homeotic genes from distant regulatory sites. Trends Genet. 9: 75–79. Kim J (1999) Making sense of emergence. Phil. Studies 95: 3–36. Kimura M (1960) Optimum mutation rate and degree of dominance as determined by the principle of minimum genetic load. J. Genetics 57: 21–34. King DG (1994) Triple repeat DNA as a highly mutable regulatory mechanism. Science 263: 595–596. King DG, Soller M, Kashi Y (1994) Evolutionary tuning knobs. Endeavour 21: 36–40. King RC, Stansfield WD (1984) A Dictionary of Genetics, third edition. Oxford: Oxford University Press. Kingsolver JG, Hoekstra JM, Berrigan D, Vignieri SN, Hill CE, Hoang A, Gibert P, Beerli P (2001) The strength of phenotypic selection in natural populations. Am. Nat. 157: 245–261. Kirschner MW, Gerhart JC (2005) The Plausibility of Life: Resolving Darwin’s Dilemma. New Haven: Yale University Press. Klenk H-P et al. (1997) The complete genome sequence of the hyperthermophilic, sulphatereducing archaeon Archaeoglobus fulgidus. Nature 390: 364–370. Kobayashi M, Inaba H (2000) Photon statistics and correlation analysis of ultraweak light originating from living organisms. Applied Optics 39: 183. Koch AL (1993) Genetic response of microbes to extreme challenges. J. Theor. Biol. 160: 1–21. Koch C, Laurent G (1999) Complexity and the nervous system. Science 284: 96–101. Koestler A (1964) The Act of Creation. London: Hutchinson. Koestler A (1967) The Ghost in the Machine. London: Hutchinson. Koestler A, Smythies JR (1969) Beyond Reductionism: New Perspectives in the Life Sciences. London: Hutchinson. Kolesnikova LA (1987) Evolution of the pineal gland and its changes during selection of silver foxes for domestic behaviour. In Behaviour as One of the Main Factors of Evolution, ed. Leonovicová V, Novák V. Prague: Czechoslovak Academy of Sciences.

484

Bibliography

Kollar EJ, Fisher C (1980) Tooth induction in chick epithelium: Expression of quiescent genes for enamel synthesis. Science 207: 993–995. Kölliker RA von (1864) Ueber die Darwinsche Schopfungstheorie ein Vortrag. Zeit. Wiss. Zool. 14. Kolnicki R (1999) Karyotypic fission theory applied: Kinetochore reproduction and lemur evolution. Symbiosis 26: 123–141. Kolnicki R (2000) Kinetochore reproduction in animal evolution: Call biological explanation of karyotypic fission theory. Proc. Natl. Acad. Sci. 97: 9493–9497. Kornegay JR, Schilling JW, Wilson AC (1994) Molecular adaptation of a leaf eating bird: Stomach lysozyme of the hoatzin. Mol. Biol. Evol. 11: 921–928. Krebs AI, Houston JI (1989) Optimization in ecology. In Ecological Concepts, ed. Cherrett M. Blackwell Scientific. Kuhn TS (1962) The Structure of Scientific Revolutions. Chicago: University of Chicago Press (2nd Edition 1970). Kuhn TS (2000) The Road since Structure, ed. Conant J, Haugeland J. Chicago: University of Chicago Press. Kukalova-Peck J (1983) Origin of the insect wing and wing articulation from the arthropodan leg. Can. J. Zool. 61: 1618–165. Lachmansingh E, Rollo CD (1994) Evidence of a trade-off between growth and behavioural activity in giant “supermice” genetically engineered with extra growth hormone genes. Can. J. Zool. 72: 2158–2168. Lahn BR, Page DC (1997) Functional coherence of the human Y chromosome. Science 278: 675–680. Lahn BR, Page DC (1999) Four evolutionary strata on the human X chromosome. Science 286: 964–967. Lamarck J-B (1809) La philosophie zoologique. Paris. Reissued in Naturalis Classica series (Wheldon and Wesley). Lamarck J-B (1815–22) L’Histoire naturelle des animaux sans vertebres. Paris. Langton CB (1992) Adaptation to the edge of chaos. In Artificial Life II, ed. Langton C, Farmer J, Rasmussen A, Taylor C. Reading MA: Addison-Wesley. Larsen EW (2004) A view of phenotypic plasticity from molecules to morphogenesis. In Environment, Development and Evolution, ed. Hall B, Pearson R, Müller G. Cambridge: MIT Press. Laudet V (1997) Evolution of the nuclear receptor superfamily: Early diversification from an ancestral orphan receptor. J. Mol. Endocrinology 19: 207–226. Laudet V, Hanni C, Coll J, Catzeflis F, Stehelin D (1992) Evolution of the nuclear receptor gene superfamily. EMBO J. 11: 1003–1013. Law MS, Change KW, Fung TK, Chan YH, Yu KL, Chan KM (1996) Isolation and characterization of two distinct growth hormone cDNAs from the goldfish, Carassius uratus. Archives of Biochemistry and Biophysics 33: 19–23.

Bibliography

485

Lederberg J (1951) Genetic studies in bacteria. In Genetics in the Twentieth Century, ed. Dunn L. New York: Macmillan. Lee S-H, Fu KK, Hui JN, Richman JM (2001) Noggin and retinoic acid transform the identity of avian facial prominences. Nature 414: 909–912. Le Lièvre CS (1978) Participation of neural crest-derived cells in the genesis of the skull in birds. J. Embryol. Exp. Morph. 47: 16–37. Le Maho Y (1977) The emperor penguin: A strategy to live and breed in the cold. Am. Sci. 65: 680–696. Leroi AM (1998) The burden of the bauplan. Trends in Ecology and Evolution 13: 82–83. Levinton JS (1992) The big-bang of animal evolution. Sci. Am. 267, November: 84–91. Levit G, Hossfeld U (2006) The forgotten “old-fashioned” synthesis: The evolutionary theory of Ludwig H. Plate (1862–1937). Internationale Zeitschrift für Geschichte und Ethik der Naturwissenschaften, Technik und Medizin 14: 9–25. Lewes GH (1874, 1875). Problems of Life and Mind. London: Trubner. Lewin B (1997) Genes VI. New York: Oxford University Press. Lewin R (1988) A heresy in evolutionary biology. Research News, September 16: 1431. Lewis JA, Davide JP, Melera PW (1982) Selective amplification of polymorphic dihydrofolate reductase gene loci in Chinese hamster lung cells. Proc. Natl. Acad. Sci. 79: 6961–6965. Lewontin R (2000) The Triple Helix: Gene, Organism, and Environment. Cambridge: Harvard University Press. Li W-H, Graur D (1991). Fundamentals of Molecular Evolution. Sunderland MA: Sinauer. Li W-H, Gu Z, Wang H, Nekrutenko A (2001) Evolutionary analyses of the human genome. Nature 409: 847–849. Lima-de-Faria A (1988) Evolution without Selection: Form and Function by Autoevolution. Amsterdam: Elsevier. Lin Y, Waldman AS (2001) Capture of DNA sequences at double-strand breaks in mammalian chromosomes. Genetics 158: 1665–1674. Lincoln R, Boxshall G, Clark P (1998) A Dictionary of Ecology, Evolution and Systematics, second edition. Oxford: Oxford University Press. Linde A (1994) The self-reproducing inflationary universe. Sci. Am. 271, November: 48–55. Little KC, Chartrand P (2004) Genomic DNA is captured and amplified during double-strand break (DSB) repair in human cells. Oncogene 23: 4166–4172. Livant W (1988) The rise of reproduction: Ten propositions. In Evolution of Social Behavior and Integrative Levels, ed. Greenberg G, Tobach E. New Jersey: Lawrence Erlbaum Assoc. Livant W (1998) Bill Livant’s cure for baldness. Science and Society 62: 471–474. Lloyd RE (1914) What Is Adaptation? London: Longman.

486

Bibliography

Logsdon JF, Doolittle RF (1997) Origin of antifreeze protein genes: A cool tale in molecular evolution. Proc. Natl. Acad. Sci. 94: 3485–3487. Long CA (1976) Evolution of mammalian cheek pouches and a possibly discontinuous origin of a higher taxon (Geomyoidea). Am. Nat. 110: 1093–1111. Lotka AJ (1922) Contribution to the energetics of evolution. Proc. Natl. Acad. Sci. 8: 147–162. Lovejoy AO (1927) The meaning of emergence and its modes. J. Philos. Studies 2: 167–189. Lovejoy AO (1936) The Great Chain of Being. Cambridge: Harvard University Press. Lovelock JE (1979) Gaia: A New Look at Life on Earth. Oxford: Oxford University Press. Lovelock JE (1988) The Ages of Gaia. New York: Norton. Løvtrup S (1974) Epigenetics. New York: Wiley. Løvtrup S (1975) Letter. Syst. Zool. 24: 507–511. Løvtrup S (1983) Reduction and emergence. Rivista di Biologia—Biology Forum 76: 437–461. Løvtrup S (1992) Human evolution and epigenesis. Human Evolution 7: 25–31. Lwoff A (1944) L’évolution physiologique. Paris: Hennann. Lyell C (1830–33) Principles of Geology. London: John Murray. Macbeth N (1971) Darwin Retried. New York: Dell. MacFadden BJ (1986) Fossil horses from “Eohippus” to Equus: Scaling, Cope’s law, and the evolution of body size. Paleobiology 12: 355–369. Mackie GO (1999) Coelenterate organs. Mar. Fresh. Behav. Physiol. 32: 113–127. Magee PT (1997) Which came first, the hypha or the yeast? Science 277: 52–53. Magnum CP, Towle D (1977) Physiological adaptation to unstable environments. Am. Sci. 65: 67–75. Marcotrigniano M (1999) Chimeras and variegation: Patterns of deceit. Horticultural Science 32: 773–784. Maret TJ, Collins JP (1997) Ecological origin of morphological diversity: A study of alternative trophic phenotypes in larval salamanders. Evolution 51: 898–905. Margulis L (1970) Origin of Eukaryotic Cells. New Haven: Yale University Press. Margulis L (1981) Symbiosis in Cell Evolution: Life and Its Environment in the Early Earth. San Francisco: Freeman. Margulis L (1998) Symbiotic Planet: A New View of Evolution. New York: Basic Books. Margulis L, Sagan D (2002) Acquiring Genomes: A Theory of the Origins of Species. New York: Basic Books. Martin MO, Benford G (1996) A Darker Geometry. New York: Baen Books. Martin WF, Muller M (1998) The hydrogen hypothesis for the first eukaryote. Nature 392: 37–41.

Bibliography

487

Maser C, Trappe JM, Nusbaum RA (1978) Fungal-small mammal interrelationships with emphasis on Oregon coniferous forests. Ecology 59: 799–809. Matsuda R (1982) The evolutionary process in talitrid amphipods and salamanders in changing environments with a discussion of “genetic assimilation” and some other evolutionary comments. Can. J. Zool. 60: 733–49. Matsuda R (1987) Animal Evolution in Changing Environments with Special Reference to Abnormal Metamorphosis. New York: Wiley. Maudlin I, Wellburn SC, Mehlitz D (1990) The relationship between rickettsia-like organisms and trypanosome infections in natural populations of tsetse in Liberia. Tropical Medicine and Parasitology 41: 265–267. Maynard Smith J (1978) The Evolution of Sex. Cambridge: Cambridge University Press. Maynard Smith J (1988) Evolutionary progress and levels of selection. In Evolutionary Progress, ed. Nitecki H. Chicago: University of Chicago Press. Maynard Smith J, Szathmáry E (1995) The Major Transitions of Evolution. Oxford: Oxford University Press. Mayr E (1942) Systematics and the Origin of Species. New York: Columbia University Press. Mayr E (1954) Change of genetic environment and evolution. In Evolution as a Process, ed. Huxley J, Hardy A, Ford E. London: Allen and Unwin. Mayr E (1959) Where are we? Cold Spring Harbor Symp. Quant. Biol. 24: 1–14. Mayr E (1960) The emergence of evolutionary novelties. In Evolution after Darwin, volume 1, ed. Tax S. Chicago: University of Chicago Press. Mayr E (1961) Cause and effect in biology. Science 134: 1501–1506. Mayr E (1963) Animal Species and Evolution. Cambridge: Harvard University Press. Mayr E (1972) The nature of the Darwinian revolution. Science 176: 981–987. Mayr E (1980) Prologue: Some thoughts on the history of the evolutionary synthesis. In The Evolutionary Synthesis, ed. Mayr E, Provine W. Cambridge: Harvard University Press. Mayr E (1982) The Growth of Biological Thought. Cambridge: Belknap. Mayr E (1991) One Long Argument: Charles Darwin and the Genesis of Modern Evolutionary Thought. Cambridge: Harvard University Press. McClintock B (1951) Chromosome organization and genic expression. Cold Spring Harbor Symp. Quant. Biol. 16: 13–47. McClintock B (1956) Controlling elements and the gene. Cold Spring Harbor Symp. Quant. Biol. 21: 215–249. McClintock B (1978) Mechanisms that rapidly reorganize the genome. In Stadler Symposium 10, ed. Reder G. New York: Columbia University Press. McClintock B (1980) Modified gene expressions induced by transposable elements. In Mobilization and Reassembly of Genetic Information, ed. Scott W. New York: Academic Press.

488

Bibliography

McDougall W (1927) Modern Materialism and Emergent Evolution. New York: Van Nostrand. McLaren A, Michie D (1958) An effect of the uterine environment upon skeletal morphology in the mouse. Nature 181: 1147–1148 McMenamin M (1998) The Garden of Ediacara. New York: Columbia University Press. McMillan HJ, Wynne-Edwards KE (1998) Evolutionary change in the endocrinology of behavioral receptivity: Divergent roles for progesterone and prolactin within the genus Phodopus. Biology of Reproduction 59: 30–38. McMillan HJ, Wynne-Edwards KE (1999) Divergent reproductive endocrinology of the estrous cycle and pregnancy in dwarf hamsters (Phodopus). Comp. Biochem. Physiol. A 124: 53–67. McNeil M (1987) Under the Banner of Science. Erasmus Darwin and His Age. Manchester: Manchester University Press. Medawar P (1951) Problems of adaptation. New Biol. 11: 10–15. Meir E, Dassow B von, Munro E, Odell GM (2002) Robustness, flexibility, and the role of lateral inhibition in the neurogenic network. Current Biology 12: 778–786. Menge BA, Berlow, EL, Blanchette CA, Navarrete SA, Yamada SB (1994)The keystone species concept: Variation in interaction strength in a rocky intertidal habitat. Ecological Monographs 64: 249–286. Metz CW (1947) Duplication of chromosome parts as a factor in evolution. Am. Nat. 81: 81–103. Meyer-Abich A (1964) The Historico-Philosophical Background of Modern Evolution Biology. Leiden: Brill. Mill J S (1843) A System of Logic, Ratiocinative and Inductive. In The Collected Works of John Stuart Mill, volumes 7 and 8 (Toronto: University of Toronto Press, 1973). Miller J, Fraser SE, McClay D (1995) Dynamics of thin filipodia during sea urchin gastrulation. Development 121: 2501–2511. Milne Edwards H (1834) l’Histoire naturelle des Crustacés, comprenant l’anatomie, la physiologie et la classification de ces animaux. Paris. Milne Edwards H (1834) élémens de zoologie, ou leçons sur l’anatomie, la physiologie, la classification et les moeurs des animaux. Paris. Milne Edwards H (1841) Outlines of Anatomy and Physiology. Boston. Mitchell SD (2003) Biological Complexity and Integrative Pluralism. Cambridge: Cambridge University Press. Mivart SGJ (1871) On the Genesis of Species. London: Macmillan. Monk M (1995) Epigenetic programming of differential gene expression in development and evolution. Developmental Genetics 17: 188–197. Monod J (1971) Chance and Necessity. New York: Knopf. Morgan CL (1891) Animal Life and Intelligence. London: Arnold.

Bibliography

489

Morgan CL (1896) Habit and Instinct. London: Arnold. Morgan CL (1923) Emergent Evolution. London: Williams & Norgate. Morgan CL (1933) The Emergence of Novelty. London: Williams & Norgate. Morgan TH (1903) Evolution and Adaptation. London: Macmillan. Morgan WF, Bodycote J. Fero ML, Hahn PJ, Kapp LN, Pantelias GE, Painter RB (1986) A cytogenetic investigation of DNA replication after hydroxyurea treatment: Implications for gene amplification. Chromosoma 93: 191–196. Morin S, Ghanim M, Sobol I, Czosnek H (2000) The GroEL protein of the whitefly Bemisia tabaci interacts with the coat protein of transmissible and nontransmissible begoviruses in the yeast two-hybrid system. Virology 276: 404–416. Morwood MJ, Soejono RP, Roberts RG, Sutikna T, Turney CSM, Westaway KE, Rink WJ, Ahzo J-X, van den Bergh GD, Rokus AD, Hobbs DR, Moore MW, Bird MI, Fifield LK (2004) Archaeology and age of a new hominin from Flores in eastern Indonesia. Nature 431: 1087–1091. Moxon RE, Wills C (1999) DNA microsatellites: Agents of evolution? Sci. Am. 280, January: 94–99. Muir A, Lever A, Moffett A (2004) Expression and functions of human endogenous retroviruses in the placenta: An update. Placenta 25: S16–S25. Muir W, Howard R (1999) Possible ecological risks of transgenic organism release when transgenes affect mating success: Sexual selection and the Trojan gene hypothesis. Proc. Natl. Acad. Sci. 96: 13853–13856. Müller GB (1991) Developmental mechanisms at the origin of morphological novelty: A sideeffect hypothesis. In Evolutionary Innovations, ed. Nitecki M. Chicago: University of Chicago Press. Müller GB, Newman SA (2003) Origination of Organismal Form: Beyond the Gene in Development and Evolutionary Biology. Cambridge: MIT Press. Müller GB, Streicher J (1989) Ontogeny of the syndesmosis tibiofibularis and the evolution of the bird hindlimb: A caenogenetic feature triggers phenotypic novelty. Anatomy and Embryology 179: 327–339. Müller GB, Wagner GP (1991) Novelty in evolution: Restructuring the concept. Ann. Rev. Ecology and Systematics 22: 229–256. Müller GB, Wagner GP (1996) Homology, Hox genes, and developmental integration. Am. Zool. 36: 4–13. Muller HJ (1930) Types of visible variations induced by X-rays in Drosophila. J. Genet. 22: 299–314. Murphy JJ (1869) Habit and Intelligence. London: Macmillan. Murray MA, Schubinger ML, Palka J (1981) Preferred axon paths in the wing of Drosophila. Soc. Neurosci. Abstracts 7: 347–356.

490

Bibliography

Nagel T (1978) Mortal Questions. Cambridge: Cambridge University Press. Nägeli C von (1884) Mechanisch-Physiologische Theorie der Abstammungslehre. Leipzig: Oldenbourg. Nash MJ (1995) When life exploded. Time 146, no. 23: 38–46. Nelson G (1970) Outline of a theory of comparative biology. Syst. Zool. 19: 373–384. Newcombe RD, Campbell PM, Ollis DL, Cheah E, Russell RJ, Oakshott JG (1997) A single amino acid substitution converts a carboxylesterase to an organophosphorus hydrolase and confers insecticide resistance on a blowfly. Proc. Natl. Acad. Sci. 94: 7464–7468. Newman SA (1993) Is segmentation generic? Bioessays 15: 277–283. Newman SA, Müller GB (2000) Epigenetic mechanisms of character origination. J. Exptl. Zool. (Mol. Dev. Evol.) 288: 304–317. Nicolis G, Prigogine I (1977) Self-Organization in Non-Equilibrium Systems. New York: Wiley. Nielsen C (1995) Animal Evolution: Interrelationships of the Living Phyla. Oxford: Oxford University Press. Niklas KJ (1997) The Evolutionary Biology of Plants. Chicago: University of Chicago Press. Nitecki MH, ed. (1988) Evolutionary Progress. Chicago: University of Chicago Press. Nitecki MH, ed. (1990) Evolutionary Innovations. Chicago: University of Chicago Press. Novick A, Weiner M (1957) Enzyme induction as an all-or-none phenomenon. Proc. Natl. Acad. Sci. 43: 553–566. Novick RP (1980) Plastids. Sci. Am. 243, December: 103–124. Novick RP (1998) Contrasting lifestyles of rolling-circle phages and plasmids. Trends in Biochemical Sciences 23: 434–438. Nowak R (1994) Mining treasures from “junk DNA.” Science 263: 608–610. Nüsslein-Volhard C (1996) Gradients that organize embryo development. Sci. Am. 273, August: 54–61. Nüsslein-Volhard C, Wieschaus E (1980) Mutations affecting segment number and polarity in Drosophila. Nature 287: 795–799. Ohno S (1970) Evolution by Gene Duplication. Berlin: Springer. Oldroyd DR (1980) Darwinian Impacts: An Introduction to the Darwinian Revolution. Atlantic Highlands NJ: Humanities Press. Oparin AI (1938) The Origin of Life. New York: Macmillan. Orgel LE (1994) The origin of life on the earth. Sci. Am. 271, October: 77–83. Osborn HF (1896) A mode of inheritance requiring neither natural selection nor inheritance of acquired character. Trans. New York Acad. Sci. 15: 141–148. Otte D, Endler JA (1989) Speciation and Its Consequences. Sunderland MA: Sinauer.

Bibliography

491

Owen R (1847) Report on the archetype and homologies of the vertebrate skeleton. Rep. Brit. Assoc. Adv. Sci. 169–340. Oyama S (1985) The Ontogeny of Information. Second edition: Durham NC: Duke University Press, 2000. Oyama S (2000) Evolution’s Eye. Durham NC: Duke University Press. Palka J, Schubinger M, Ellison RL (1983) The polarity of axon growth in the wings of Drosophila melanogaster. Devel. Biol. 98: 481–489. Palmeirim I, Henrique D, Ish-Horowicz D, Pourquie O (1997) Avian hairy gene expression identifies a molecular clock linked to vertebrate segmentation and somatogenesis. Cell 91: 639–648. Palmiter RD, Brinster RL, Hammer RE (1982) Dramatic growth of mice from eggs microinjected with metallothionein-growth hormone fusion genes. Nature 300: 611–615. Parrish JE, Oostra BA, Verkerk AJMH, Richards CS, Reynolds J, Spikes AS, Shaffer LG, Nelson DL (1994) Isolation of a GCC repeat showing expansion in FRAXF, a fragile site distal to FRAXA and FRAXE. Nature Genetics 8: 229–235. Paterson HEH (1982) Darwin and the origin of species. South African Journal of Science 78: 272–275. Paterson HEH (1985) The recognition concept of species. In Species and Speciation, ed. Vrba E. Pretoria: Transvaal Museum. Pawson T, Scott JD (1997) Signalling through scaffold, anchoring and adaptor proteins. Science 278: 2075–2080. Pearson RD (1986) Evolutionary theory, regeneration, and cancer. Med. Hypoth. 19: 7–14. Pearson RD (1999) Long-billed European starlings, Sturnus vulgaris. Canadian Field Naturalist 113: 523–524. Pearson RD (2004) The determined embryo: Homeodynamics, hormones and heredity. In Environment, Development and Evolution, ed. Hall B, Pearson R, Müller G. Cambridge: MIT Press. Peirce CS (1898) Reasoning and the Logic of Things. A series of lectures given at Harvard University. First published in Collected Papers of Charles Sanders Peirce; reissued as Reasoning and the Logic of Things (Cambridge: Harvard University Press, 1992). Pelling C (1959) Chromosomal synthesis of ribonucleic acid as shown by the incorporation of uridine labeled with tritium. Nature 184: 655–666. Pennisi E (1998) How the genome readies itself for evolution. Science 281: 113–115. Pepperberg IM (1998) Talking with Alex: Logic and speech in parrots. Scientific American Presents (Exploring Intelligence) 9, no. 4: 60–65. Pepperberg IM (1999) The Alex Studies: Cognitive and Communicative Abilities of Grey Parrots. Cambridge: Harvard University Press.

492

Bibliography

Perkowska E, MacGregor HC, Birnstiel ML (1968) Gene amplification in the oocyte nucleus of mutant and wild-type Xenopus laevis. Nature 217: 649–650. Perru Q (1997) Le concept d’individualité biologique chez Milne-Edwards. Bull. Hist. Epistém. Sc. Vie 4: 147–172. Peters RH (1991) A Critique for Ecology. Cambridge: Cambridge University Press. Pfeffer W (1881) Pflanzenphysiologie. Piaget J (1968) Structuralism. London: Routledge and Kegan Paul. Piaget J (1978) Behavior and Evolution. New York: Random House. Pirkkala L, Nykanen P, Sistonen L (2001) Roles of the heat shock transcription factor in regulation of the heat shock response and beyond. FASEB J. 15, May: 1118–1131. Pirozynski KA, Malloch DW (1975) The origin of land plants: A matter of mycotrophism. Biosystems 6: 153–164. Plate L (1913) Selektionsprinzip und Probleme der Artbildung: Ein handbuch des Darwinismus. Leipzig: Engelmann. Polanyi M (1958) Personal Knowledge. London: Routledge and Kegan Paul. Revised edition: Routledge & Kegan Paul, 1962. Harper Torchbook Edition: 1964. Polanyi M (1968) Life’s irreducible structure. Science 160: 1308–1312. Popper KR (1959) The Logic of Scientific Discovery. New York: Basic Books. Popper KR (1972) Objective Knowledge, an Evolutionary Approach. Oxford: Oxford University Press. Poulton EB (1908) Essays on Evolution 1889–1907. Oxford: Oxford University Press. Pray L (2002) Evolutionists present their 1.3% solution. The Scientist 16: 36–41. Pray L (2004) Epigenetics: Genome, meet your environment. The-scientist.com. Prigogine I, Stengers I (1984) Order out of Chaos. New York: Bantam. Prody CA, Dreyfuss P, Zamir R, Zakut H, Soreq H (1989) De novo amplification within a “silent” human cholinesterase gene in a family subjected to prolonged exposure to organophosphorous insecticides. Proc. Natl. Acad. Sci. 86: 690–694. Prosser CL (1965) Levels of biological organization and their physiological significance. In Ideas in Modern Biology, ed. Moore J. New York: Natural History Press. Provine WB (1971) The Origins of Theoretical Population Genetics. Chicago: University of Chicago Press. Provine WB (1978) The role of mathematical population geneticists in the evolutionary synthesis of the 1930s and 1940s. Studies in History of Biology 2: 167–192. Provine WB (1988) Progress in evolution and meaning in life. In Evolutionary Progress, ed. Nitecki M. Chicago: University of Chicago Press. Provine WB (2001) Afterword. In reissued edition of The Origins of Theoretical Population Genetics. Chicago: University of Chicago Press.

Bibliography

493

Punnett RC (1915) Mimicry in Butterflies. Cambridge: Cambridge University Press. Radinsky L (1984) Ontogeny and phylogeny in horse skull evolution. Evolution 38: 1–15. Raff RA (1996) The Shape of Life. Chicago: University of Chicago Press. Raff RA, Kaufman TC (1983) Embryos, Genes, and Evolution: The Developmental-Genetic Basis of Evolutionary Change. New York: Macmillan. Raff RA, Parr B, Parks A, Wray G (1990) Heterochrony and other mechanisms of radical evolutionary change in early development. In Evolutionary Innovations, ed. Nitecki M. Chicago: University of Chicago Press. Raff RA, Wray G (1989) Heterochrony: Developmental mechanisms and evolutionary results. J. Evol. Biol. 2: 409–434. Raup DM (1972) Taxonomic diversity during the Phanerozoic. Science 177: 1065–1071. Raup DM (1986) The Nemesis Affair. New York: Norton. Raup DM (1991) Extinction: Bad Genes or Bad Luck? New York: Norton. Rebeck J (1994) Synthetic self-replicating molecules. Sci. Am. 271, July: 48–55. Reid RGB (1985a) Evolutionary Theory: The Unfinished Synthesis. Beckenham: Croom Helm. Reid RGB (1985b) Unpredicted factors of evolution: The theory of emergence revisited. Rivista di Biologia 78: 493–512. Reid RGB (1987) The successful monster: Evolution by association. In Towards a New Synthesis of Evolutionary Biology, ed. Mlikovsky J, Novák V. Prague: Czechoslovak Academy of Sciences. Reid RGB (1989) The unwhole organism. Amer. Zool. 29: 1133–1140. Reid RGB (1990) Evolutionary implications of sulphide-oxidizing symbioses in bivalves. In The Bivalvia—Proceedings of a Memorial Symposium in Honour of Sir Charles Maurice Yonge, Edinburgh, 1986, ed. Morton B. Hong Kong: Hong Kong University Press. Reid RGB (1998) Subclass Protobranchia. In Mollusca: The Southern Synthesis, ed. Beasley P, Ross G, Wells A. Melbourne: CSIRO. Reid RGB (2004) Epigenetics and environment: The historical matrix of Matsuda’s pan-environmentalism. In Environment, Development and Evolution, ed. Hall B, Pearson R, Müller G. Cambridge: MIT Press. Reid RGB, Brand D (1986) Sulfide-oxidizing symbioses in lucinaceans: Implications for bivalve evolution. Veliger 29: 3–24. Reimchen TE (2000) Some ecological and evolutionary aspects of bear-salmon interactions in coastal British Columbia. Can. J. Zool. 78: 448–457. Reimchen TE (2001) Salmon nutrients, nitrogen isotopes and coastal forests. Ecoforestry 16: 13–17. Rendel JM (1979) Canalization and selection. In Quantitative Genetic Variation, ed. Thomson J, Thoday J. New York: Academic Press.

494

Bibliography

Rensch B (1959) Evolution above the Species Level. London: Methuen. Riedl R (1978) Order in Living Systems. London: Methuen. Riley D (1978) Developmental psychology, biology and Marxism. Ideology and Consciousness 4: 73–91. Ring RA (1982) Freezing-tolerant insects with low supercooling points. Comp. Biochem. Physiol. 73A: 605–612. Ring RA, Tesar D (1981) Adaptations to cold in Canadian Arctic insects. Cryobiology 18: 199–211. Ritossa F, Malva C, Boncinelli E, Graziani F, Polito L (1971) The first steps of magnification of DNA complementary to ribosomal RNA in Drosophila melanogaster. Proc. Natl. Acad. Sci. 68 1580–1584. Ritter WE, Bailey EW (1928) The organismal concept, its place in science and bearing on philosophy. University of California Publications in Zoology 31: 307–358. Robson GC, Richards OW (1936) The Variation of Animals in Nature. London: Longman. Rollo D (1994) Phenotypes: Their Epigenetics, Ecology, and Evolution. London: Chapman and Hall. Romanes GJ (1892–97) Darwin, and after Darwin: An Exposition of the Darwinian Theory and a Discussion of Post-Darwinian Questions. Chicago: Open Court. Ronshaugen M, McGinnis N, McGinnis W (2002) Hox protein mutation and macroevolution of the insect body plan. Nature 416: 914–917. Roux W (1888) Beitr. Z. Entw. Mech. des Embryo. V. über die künstliche Hervorbringung “halber” Embryonen durch Zerstörung einer der beiden ersten Furchungszellen, soie uber die Nachentwicklung der fehlenden Körperhalfte. Virchows Archiv 14: 419–521. Roux W (1905) Die Enwickelungsmechanik. Engelmann. Rubinsztein DC, Amos W, Leggo J, Goodburn S, Ramesar RS, Old J, Bontrop R, McMahon R, Barton DE, Ferguson-Smith MA (1994) Mutational bias provides a model for the evolution of Huntington disease and predicts a general increase in disease prevalence. Nature Genetics 5: 524–530. Rudman WB (1981) The anatomy of alcyonarian feeding aeolid opisthobranch molluscs and their development of symbiosis with zooxanthellae. Zool. J. Linn. Soc. 72: 219–262. Rumpho ME, Summer EJ, Meinhart JR (1999) Solar-powered sea-slugs: Mollusc-algal chloroplast symbiosis. Plant Physiology 123: 29–38. Rupert JL, Hochachka P (2001) Genetic approaches to understanding human adaptation to altitude in the Andes. J. Exptl. Biol. 204: 3151–3160. Ruse M (1996) Monad to Man. The Concept of Progress in Evolutionary Biology. Cambridge: Harvard University Press. Ruse M (1997) Huxley and evolution as science. In Thomas Henry Huxley’s Place in Science and Letters, ed. Barr A. Atlanta: University of Georgia Press.

Bibliography

495

Rutherford SL, Lindquist S (1998) Hsp90 as a capacitor for morphological evolution. Nature 396: 336–342. Saint-André A von, Blackwell NM, Hall LR, Hoerauf A, Brattig NW, Volkmann L, Taylor MJ, Ford L, Hise AG, Lass JH, Diaconu E, Perlman E (2002) The role of endosymbiotic Wolbachia bacteria in the pathogenesis of riverblindness. Science 295: 1892–1895. Salthe S (1985) Evolving Hierarchical Systems. New York: Columbia University Press. Sapp J (1994) Evolution by Association. Oxford: Oxford University Press. Saunders PT (1984) Development and evolution. In Beyond Neo-Darwinism, ed. Ho M-W, Saunders P. London: Academic Press. Savageau MA (1989) Are there rules governing patterns of gene regulation? In Theoretical Biology, ed. Goodwin B, Saunders P. Edinburgh: Edinburgh University Press. Schalling M, Hudson TJ, Buetow KH, Housman DE (1993) Direct detection of novel expanded trinucleotide repeats in the human genome. Nature Genetics 4: 135–139. Schimper AFW (1885) Untersuchungen über die Chlorophyllkörner und Farbkörper. Jahrbücher für wissenschaftliche Botanik 1: 247–302 Schindewolf OH (1937) Beobachtung und Gedanken zur Deszendenzliehre. Acta Biother. Leiden. 3: 195–212. Schindewolf OH (1962) Neokatastrophismus? Zeits. Deutsch. Geol. Res. 114: 430–435. Schleiden MJ (1838) Beiträge zur Phytogenesis. Müllers Archiv für Anatomie, Physiologie und wissenschaftliche Medecin. Berlin. Schmalhausen II (1949) Factors of Evolution. The Theory of Stabilizing Selection. Philadelphia: Blackiston. Schrödinger I (1944) What Is Life? Cambridge: Cambridge University Press. Schubert M, Holland LZ (2005) The Wnt Gene Family and the Evolutionary Conservation of Wnt Expression. Georgetown TX: Landes/Eurekah Pubmed. Schueller G (1998) Stuff of life. New Scientist 159, no. 2151: 30–35. Schwann T (1839) Mikroskopische Untersuchungen über die übereinstimmung in der Struktur und dem Wachstum der Thiere und Pflanzen. Berlin. Schwartz JH (1999) Sudden Origins: Fossils, Genes, and the Emergence of Species. New York: Wiley. Schwartz M, Vissing J (2002) Paternal inheritance of mitochondrial DNA. N. Engl. J. Med. 347: 576–580. Schwartz RM, Dayhoff MO (1978) Origins of prokaryotes, eukaryotes, mitochondria and chloroplasts. Science 199: 395–403. Segre GV, Goldring SR (1993) Receptors for secretin, calcitonin, parathyroid hormone (PTH)/PTHrelated peptide, vasoactive intestinal peptide, glucagonlike peptide 1, growth hormone-releasing hormone, and glucagon belong to a newly discovered G-protein-linked receptor family. TEM 4: 309–314.

496

Bibliography

Seilacher A, Bose, PK, Pfluger F. 1998. Triploblastic animals more than 1 billion years old: Trace fossil evidence from India. Science 282: 80–83. Severtsov AN (1927) über die Beziehungen zwischen der Ortogenese und der Phylogenese der Tiere. Jena Z. Naturw. 63: 51–180. Severtsov AN (1931) Morphologische Gesetzmässigkeiten der Evolution. Vienna: Fischer. Severtsov AN (1934) Glavnye napravleniia evoliutsionnogo protsessa [Principal directions of the evolutionary process], second edition. Leningrad: Biomedgiz. Severtsov AN (1939) Morfologicheskie zakonomernosti evoliutsii [Morphological laws of evolution], second edition. Moscow: Izdatel’stvo Akademii nauk SSSR. Shapere D (1980) The meaning of the evolutionary synthesis. In The Evolutionary Synthesis, ed. Mayr E, Provine W. Cambridge: Harvard University Press. Shapiro JA (1977) DNA insertion elements and the evolution of chromosome primary structure. Trends in Biochem. Sci. 2: 622–627. Shapiro JA (1984) Observations on the formation of clones containing araB-lacZ cistron fusions. Molecular and General Genetics 194: 79–90. Shapiro JA (1988) Bacteria as multicellular organisms. Sci. Am. 258, June: 82–89. Shapiro JA (1992) Natural genetic engineering in evolution. Genetica 86: 99–111. Shea BT (1988) Heterochrony in primates. In Heterochrony in Evolution, ed. McKinney M. New York: Plenum. Shelley M (1818) Frankenstein: or, the Modern Prometheus. Lackington, Hughes, Harding, Mavor and Jones. Sheppard PM (1958) Natural Selection and Heredity. Hutchinson. Sherman PW, Jarvis JUM, Alexander RD, eds. (1992) The Biology of the Naked Mole-Rat. Princeton: Princeton University Press. Sherrington CS (1906) The Integrative Nature of the Nervous System. New Haven: Yale University Press. Shostak S, Kolluri V (1995) Symbiotic origins of cnidarian cnidocysts. Symbiosis 19: 1–19. Simon JH (1962) The architecture of complexity. Proc. Am. Phil. Soc. 106: 462–482. Simpson GG (1944) Tempo and Mode in Evolution. New York: Columbia University Press. Simpson GG (1953) The Major Features of Evolution. New York: Columbia University Press. Sinha C (1987) Adaptation and representation: The role of ontogenesis in human evolution and development. In Behaviour as One of the Main Factors of Evolution, ed. Leonovicová V, Novak V. Prague: Czechoslovak Academy of Sciences. Skolimowski H (1974) Problems of rationality in biology. In Studies in the Philosophy of Biology, ed. Ayala F, Dobzhansky T. London: Macmillan. Slijper EJ (1942) Biologic-anatomical investigations on the bipedal gait and upright posture in mammals, with special reference to a little goat, born without forelegs. I, II. Proc. Koninkl. Ned. Akad. Wetensch. 45: 407–415.

Bibliography

497

Smith GT, Brenowitz EA, Wingfield JC (1997) Seasonal changes in the size of the song control nucleus HVC revealed by multiple labels. J. Comp. Neurol. 38: 253–261. Smocovitis VB (1996) Unifying Biology: The Evolutionary Synthesis and Evolutionary Biology. Princeton: Princeton University Press. Smolin L (1995) The Life of the Cosmos. New York: Oxford University Press. Smuts JC (1926) Holism and Evolution. London: Macmillan. Snow CP (1959) The Two Cultures. Cambridge: Cambridge University Press. Solé R, Goodwin BC (2000) Signs of Life: How Complexity Pervades Biology. New York: Basic Books. Spemann H (1914) über verzogerte Kernversorgung von Keimteilen. Verh. d. D. Bool. Ges. Freiburg 216–221. Spemann H (1938) Embryonic Development and Induction. New Haven: Yale University Press. Hafner reprint (1962). Spencer H (1852) The Development Hypothesis. The Leader. Spencer H (1864–1866) Principles of Biology. London: Williams & Norgate. Sperry RW (1979) Mental phenomena as causal determinants. In Consciousness and the Brain, ed. Globus G, Maxwell G, Savodnik I. New York: Plenum. Sperry RW (1983) Science and Moral Priority: Merging Mind, Brain, and Human Values. New York: Columbia University Press. Sperry RW (1991) In defence of mentalism and emergent interaction. J. Mind and Behavior 12: 221–245. Spiegelman S (1967) An in vitro analysis of a replicating molecule. American Scientist 55: 221–264. Srivastava RS (1968) Sri Aurobindo and the Theories of Evolution. Sanskrit Office, Chonkhamba. Stanley SM (1979) Macroevolution: Pattern and Process. San Francisco: Freeman. Starr DJ, Cline TW (2002) A host-parasite interaction rescues Drosophila oogenesis defects. Nature 418: 76–79. Stearns SC (1992) The Evolution of Life Histories. Oxford: Oxford University Press. Stebbins GL (1974) Adaptive shifts and evolutionary novelty: A compositional approach. In Studies in the Philosophy of Biology, ed. Ayala F, Dobzhansky T. London: Macmillan. Stebbins GL (1982) Darwin to DNA, Molecules to Humanity. San Francisco: Freeman. Steele EJ (1977) Somatic Selection and Adaptive Evolution. Toronto: Williams & Wallace. Steinberg M (2003) Cell adhesive interactions and tissue self-organization. In Origination of Organismal Form, ed. Müller G, Newman S. Cambridge: MIT Press. Stent G (1969) The Coming of the Golden Age. Garden City NY: Natural History Press.

498

Bibliography

Stephens SG (1951) Possible significance of duplications in evolution. Advances in Genetics 4: 247–265. Stewart C-B, Wilson AC (1987) Sequence convergence and functional adaptation of stomach lysozymes from foregut fermenters. Cold Spring Harbor Symp. Quant. Biol. 52: 891–899. Stewart I, Cohen J (1997) Figments of Reality. Cambridge: Cambridge University Press. Stiassny MIJ (1991) Phylogenetic intrarelationships of the family Cichlidae: An overview. In Cichlid Fishes, ed. Keenleyside M. London: Chapman and Hall. Stiassny MIJ, Meyer A (1999) Cichlids of the rift lakes. Sci. Am. 280, February: 64–69. Suess E (1875) Eine Entstehung der Alpen. Braunmüller. Sulston JED, Schierenberg E, White JG, Thomson JN (1983) The embryonic cell lineage of the nematode Caenorhabditis elegans. Dev. Biol. 100: 64–119. Sumnyj VK, Ruvinskij AO (1989) In memory of Dimitrij Constantinovic Beljaev. In Behaviour as One of the Main Factors of Evolution, ed. Leonovicová V, Novák V. Prague: Czechoslavak Academy of Sciences. Sunquist F (1996) Two species, one design. International Wildlife Magazine. September-October: 10–20. Surani AM (2001) Reprogramming of genome function through epigenetic inheritance. Nature 414: 122–128. Sutherland GR, Haan EA, Kremer E, Lynch M, Pritchard M, Yu S, Richards RI (1991) Hereditary unstable DNA: A new explanation for some old genetic questions? Lancet 338: 289–291. Sutovsky P, Moreno R, Ramalho-Santos J, Schatten G (1999) Ubiquitin tag for sperm mitochondria. Nature 402: 371–372. Sykes B (1999) The molecular genetics of European ancestry. Phil. Trans. Roy. Soc. B 354: 131–139. Symonds N (1994) Directed mutation: A current perspective. J. Theor. Biol. 169: 317–322. Szalay FS (1998) Review of Widamaier EP (1998) Why Geese Don’t Get Obese: How Evolution’s Strategies for Survival Affect Our Everyday Lives. San Francisco: Freeman. Szathmáry E, Maynard Smith J. 1995. The major evolutionary transitions. Nature 374: 227–232. Tartof KD (1974) Unequal mitotic sister chromatid exchange as the mechanism of ribosomal RNA gene magnification. Proc. Natl. Acad. Sci. 71: 1272–1276. Tatewaki M, Provasoli LX, Pintner IJ (1983) Morphogenesis of Monostroma oxypermum (kutz) Doty (Chlorophyceae) in axenic culture, especially in bialgal culture. Phycology 19: 409–416. Taylor JS, Raes J (2005) Small-scale gene duplications. In The Evolution of the Genome, ed. Gregory T. Amsterdam: Elsevier. Teilhard de Chardin P (1955) Le Phénomène humain. Paris: Seuil. Teilhard de Chardin P (1959) The Phenomenon of Man. London: Harper and Row. Harper Torchbook edition, 1961.

Bibliography

499

Tepper S (1998) Six Moon Dance. New York: Avon. Thaler DS (1994) The evolution of genetic intelligence. Science 264: 224–225. Thao ML, Gullan PJ, Baumann P (2002) Secondary (γ-Proteobacteria) endosymbionts infect the primary (β-Proteobacteria) endosymbionts of mealybugs multiple times and coevolve with their hosts. Appl. Environ. Microbiol. 68: 190–197. Thom R (1972) Stabilité structurelle et morphogenèse: Essai d’une théorie général des modèles. Reading: Benjamin. Thompson DW (1917) On Growth and Form. Cambridge: Cambridge University Press. Second edition: 1942. Thompson JN (1989) Concepts of coevolution. Trends in Ecology and Evolution 4, June: 179–183. Thomson KS (1966) The evolution of the tetrapod middle ear in the rhipidistian-tetrapod transition. Am. Zool. 6: 379–397. Thomson KS (1988) Morphogenesis and Evolution. Cambridge: Cambridge University Press. Thomson KS (1992) Macroevolution: The morphological problem. Am. Zool. 32: 106–112. Tiffney BH (1981) Diversity and major events in the evolution of land plants. In Paleobotany, Paleoecology and Evolution, volume 1, ed. Niklas K. Praeger. Ting CN, Rosenberg MP, Snow CM, Samuelson LC, Meisler MH (1992) Endogenous retroviral sequences are required for tissue-specific expression of a human salivary amylase gene. Genes Dev. 6: 1457–1465. Tittiger C, Whyard S, Walker VK (1993) A novel intron site in the triosephosphate isomerase gene in the mosquito Culex tarsaslis. Nature 361: 470–472. Todd NB (1970) Karyotypic fissioning and canid phylogeny. J. Theor. Biol. 26: 445–480. Todd NB (1999) Mammalian evolution: Karyotypic fission theory. In Environmental Evolution, second edition, ed. Margulis L, Matthews C, Haselton A. Cambridge: MIT Press. Todd NB (2001) Kinetochore reproduction underlies karyotypic fission theory: Possible legacy of symbiogenesis in mammalian chromosome evolution. Symbiosis 29: 319–327. Tomarev S I, Callaerts P, Kos L, Zinovieva R, Halder G, Gehring W, Piatigorsky J (1997) Squid Pax6 and eye development. Proc. Natl. Acad. Sci. 94: 2421–2426. Tramontin AD, Smith GT, Breuner CW, Brenowitz EA (1998) Seasonal plasticity and sexual dimorphism in the avian song control system: Stereological measurement of neuron density and number. J. Comp. Neurol. 396: 186–192. Tsarfaty I, Resau JH, Tulong S, Keydar I, Faletto DL, Vandewoude GF (1992) The met rotooncogene receptor and lumen formation. Science 257: 1258–1261. Turner J S (2000) The Extended Organism: The Physiology of Animal-Built Structures. Cambridge: Harvard University Press. Uexküll JJ von (1913) Bausteiner zu einer biologischen Weltanschauung. Munich.

500

Bibliography

Uexküll JJ von (1926) Theoretical Biology. New York: Harcourt Brace. Uvarov B (1966, 1977) Grasshoppers and Locusts. Centre for Overseas Pest Research, London. Uzell T, Spolsky C (1974) Mitochondria and plastids as endosymbionts: A revival of special creation? Am. Sci. 62: 334–343. Valentine JW (1986) Fossil record of the origin of bauplane and its implications. In Patterns and Processes in the History of Life, ed. Raup D, Jablonski D. New York: Springer. Valentine JW (1995) Late Precambrian bilaterians: Grades and clades. In Tempo and Mode in Evolution, ed. Fitch W, Ayala F. Washington: National Academy Press. Van Borm S, Billen J, Boomsma JJ (2002) The diversity of microorganisms associated with Acromyrmex leafcutter ants. BioMed Central Evolutionary Biology 2: 9–26. Van Valen L (1965) Morphological variation and width of the ecological niche. American Naturalist 94: 377–390. Van Valen L (1973) A new evolutionary law. Evol. Theory 1: 1–30. Vendruscolo M, Dobson CM (2005) A glimpse at the organization of the protein universe. Proc. Natl. Acad. Sci. 102: 5641–5642. Vernadsky VI (1926) Biosfera. Leningrad: Nauka. Vernadsky VI (1998) The Biosphere. New York: Copernicus. Verrengia J (1998) Associated Press report. Victoria Times-Colonist, November 26. Virchow R (1858) Die Cellular-pathologie in ihrer Begründung auf physiologische und pathologische Gewebelehre. Berlin. von Neumann J (1966) Theory of Self-Reproducing Automata. University of Illinois Press. Vrba ES (1980) Evolution, species and fossils: How did life evolve? S. Afr. J. Sci. 76: 61–84. Vrba ES (1985a) Environment and evolution: Alternative causes of the temporal distribution of evolutionary events. S. Afr. J. Sci. 81: 229–236. Vrba ES (1985b) The hierarchical expansion of sorting and selection: Sorting and selection cannot be equated. Paleobiology 12: 217–228. Vrba ES (1989) What are the biotic hierarchies of integration and linkage? In Integration and Evolution in Vertebrates, ed. Wake D, Roth G. New York: Wiley. Vrba ES (1995) On the connections between paleoclimate and evolution. In Palaeoclimate and Evolution, ed. Burckle L, Denton G, Partridge T. New Haven: Yale University Press. Vrba ES (2004) Eco-evo-devo and the fossil record. In Environment, Development and Evolution, ed. Hall B, Pearson R, Müller G. Cambridge: MIT Press. Vrba ES, Burckle LH, Denton GH, Partridge TC, eds. (1995) Palaeoclimate and Evolution. New Haven: Yale University Press. Vrba ES, Eldredge N (1984) Individuals, hierarchies and processes: Towards a more complete evolutionary theory. Paleobiology 10: 146–171.

Bibliography

501

Vrba ES, Gould SJ (1986) The hierarchical expansion of sorting and selection: Sorting and selection cannot be equated. Paleobiology 12: 227–228. Waagen W (1869) Die Formenreiche Palaeontologischen Beitragen 2: 179–256.

des

Ammonites

subradiatus.

Geognostisch

Waddington CH (1942) Canalisation of development and the inheritance of acquired characteristics. Nature 150: 563–565. Waddington CH (1952) Genetic assimilation. Nature 169: 278. Waddington CH (1953) The “Baldwin effect,” “genetic assimilation” and “homeostasis.” Evolution 7: 386–387. Waddington CH (1956) The genetic basis of the “assimilated Bithorax stock. J. Genetics 55: 240–245. Waddington CH (1957) The Strategy of the Genes. London: Allen & Unwin. Waddington CH (1961) Genetic assimilation. Advances in Genetics 10: 257–293. Waddington CH (1966) The modular principle and biological form. In Module, Proportion, Symmetry, Rhythm, ed. Kepes G. New York: Braziller. Waddington CH (1968) Paradigm for an evolutionary procession. In Towards a Theoretical Biology, volume 2, ed. Waddington C. Edinburgh: Edinburgh University Press. Wagner GP (1989) The origin of morphological characters and the biological basis of homology. Evolution 43: 1157–1171. Wake D, Roth G (1989a) Complex Organismal Functions: Integration and Evolution in Vertebrates. New York: Wiley. Wake D, Roth G (1989b) The linkage between ontogeny and phylogeny in the evolution of complex systems. In Complex Organismal Functions, ed. Wake and Roth. New York: Wiley. Wake MH (1989) Phylogenesis of direct development and viviparity in vertebrates. In Wake D, Roth G, Complex Organismal Functions. New York: Wiley. Wake MH (2004) Environment, embryonization and viviparity. In Environment, Development and Evolution, ed. Hall B, Pearson R, Müller G. Cambridge: MIT Press. Waldrop MM (1992) Complexity: The Emerging Science at the Edge of Order and Chaos. New York: Simon & Schuster. Wallace AR (1858) On the tendency of varieties to depart indefinitly from the original type. J. Proc. Linn. Soc (Zool.) 3: 53–62. Wallace AR (1870) Contributions to the Theory of Natural Selection. London: Macmillan. Wallis M (1996) The molecular evolution of vertebrate growth hormones: A pattern of near-stasis interrupted by sustained bursts of rapid change. J. Mol. Evol. 43: 93–100. Wallis M (2000) Episodic evolution of protein hormones: Molecular evolution of pituitary prolactin. J. Mol. Evol. 50: 465–473.

502

Bibliography

Watson DMS (1926) The evolution and origin of the Amphibia. Phil. Trans. Roy. Soc. B121: 43–79. Watson JD, Crick FHC (1953) Molecular structure of nucleic acids: A structure for deoxyribose nucleic acid. Nature 171: 737–738. Webster G, Goodwin BC (1984) A structuralist approach to morphology. Rivista di Biologia 77: 503–532. Weele C van der (1999) Images of Development: Environmental Causes in Ontogeny. Albany NY: State University of New York Press. Weismann A (1885) Die Kontinuität des Keimplasmas als Grundläge einer Theorie der Vererbung. Jena: Fischer. Weismann A (1893) The Theory of the Germ Plasm. London: Scott. Weissbach A (1993) A chronicle of DNA methylation (1948–1975). In DNA Methylation, ed. Jost J, Saluz H. Birkhauser. Weitz CA, Garruto RM, Chin CT, Liu JC, He X (2000) Growth of Qinghai Tibetans living at three different high altitudes. Am. J. Phys. Anthrop. 111: 69–88. Weldon WFR (1890) The variations occurring in certain decapod Crustacea. 1. Crangon vulgaris. Proc. Roy. Soc. 47: 445–453. Wells WC (1819) An Account of a Female of the White Race of Mankind, Part of whose Skin Resembles that of a Negro, with some Observations of the Cause of the Differences in Colour and Form between the White and Negro Races of Man. London: Constable. Weng G, Bhalla US, Lyenger R (1999) Complexity in biological signaling systems. Science 284: 92–96. Wesson R (1991) Beyond Natural Selection. Cambridge: MIT Press. West-Eberhard MJ (1989) Phenotypic plasticity and the origins of diversity. Ann. Rev. Ecol. Syst. 20: 249–278. West-Eberhard MJ (2003a) Ryuichi Matsuda: A tribute and a perspective on pan-environmentalism. In Environment, Development and Evolution, ed. Hall B, Pearson R, Müller G. Cambridge: MIT Press. West-Eberhard MJ (2003b) Developmental Plasticity and Evolution. Oxford: Oxford University Press. Wheeler WM (1911) The ant colony as an organism. J. Morph. 22: 307–325. Wheeler WM (1929) Emergent evolution and the development of societies. Scientific Monthly, February. Reprinted in Wheeler, Essays in Philosophical Biology (Russell & Russell, 1939). Whewell W (1840) The Philosophy of the Inductive Sciences. London. Whitehead AN (1925) Science and the Modern World. London: Macmillan. Whyte (1965) Internal Factors in Evolution. New York: Braziller. Social Science Paperbacks edition: Associated Book Publishers, 1968.

Bibliography

503

Willett-Brozeki JE, Savul SA, Richey LE, Baysal BE (2001) Germ line insertion of mtDNA at the breakpoint junction of a reciprocal constitutional translocation. Human Genet. 109: 216–223. Williams GC (1966) Adaptation and Natural Selection: A Critique of Some Current Evolutionary Thought. Princeton: Princeton University Press. Willis JC (1922) Age and Area: A Study of Geographical Distribution and Origin of Species. Cambridge: Cambridge University Press. Willis JC (1940) The Course of Evolution. Cambridge: Cambridge University Press. Willmer E N (1960) Cytology and Evolution. New York: Academic Press. Wilson DS (1975) A general theory of group selection. Proc. Natl. Acad. Sci. 72: 143–146. Wilson EB (1928) The Cell in Development and Heredity, third edition with corrections. London: Macmillan. Wilson EO (1975) Sociobiology. Cambridge: Belknap. Wilson EO (1985) The sociogenesis of insect colonies. Science 228: 1489–1495. Wilson EO (1998a) The biological basis of morality. Atlantic Monthly, April: 53–70. Wilson EO (1998b) Consilience: The Unity of Knowledge. New York: Random House. Wilson LG (1970) Sir Charles Lyell’s Scientific Journals on the Species Question. New Haven: Yale University Press. Wimsatt WC (1979) Reductionism, levels of organization, and the mind-body problem. In Consciousness and the Brain, ed. Globus G, Maxwell G, Savodnik I. New York: Plenum. Wimsatt WC (1980) Reductionistic research strategies and their biases in the units of selection controversy. In Scientific Discovery: Case Studies, ed. Nickles T. Dordrecht: Reidel. Wimsatt WC (1986) Developmental constraints, generative entrenchment, the innate-acquired distinction. In Integrating Scientific Disciplines, ed. Bechtel W. Martinus. The Hague: Nijhoff. Wimsatt WC (1997) Aggregativity: Reductive heuristics for finding emergence. Phil. Sci. 64: S372–S384. Wimsatt WC (1998) Simple systems and phylogenetic diversity. Phil. Sci. 65: 267–275. Wistow G (1993) Identification of lens crystallin: A model system for gene recruitment. Methods in Enzymology 224: 563–575. Woese CR, Fox GE (1977) Phylogenetic structure of the prokaryotic domain: The primary kingdoms. Proc Natl. Acad. Sci. 74: 5088–5090. Wolff CF (1759) Theoria generationis. Berlin. Wolff CF (1764) Theorie von der Generation, in zwo Abhandlungen erklärt und bewiesen. Berlin. Wolffe AP, Matzke MA (1999) Epigenetics: Regulation through repression. Science 286: 481–486. Wood WB (1988) The Nematode Caenorhabditis elegans. Cold Spring Harbor Laboratory Press. Woodger JH (1929) Biological Principles. London: Routledge and Kegan Paul.

504

Bibliography

Wray GA (1998) Promoter logic. Science 279: 1871–1872. Wray GA, Levinton JS, Shapiro LH (1996) Molecular evidence for deep Precambrian divergences among metazoan phyla. Science 247: 568–573. Wray GA, Raff RA (1989) Evolutionary development of cell lineage in the direct-developing sea urchin Heliocidaris erythrogramma. Dev. Biol. 132: 458–470. Wray GA, Raff RA (1990) Novel origins of lineage founder cells in the direct developing sea urchin Heliocidaris erythrogramma. Dev. Biol. 141: 41–54. Wright BE (2000) A biochemical mechanism for non-random mutations and evolution. J. Bacteriol. 182: 2993–3001. Wright S (1931) Evolution in Mendelian populations. Genetics 16: 97–159. Wright S (1932) The roles of mutation, inbreeding and crossbreeding, and selection in evolution. In Proceedings of Sixth International Congress on Genetics, Ithaca. Wurtele H, Little KC, Chartrand P (2003) Illegitimate DNA integration in mammalian cells. Gene Ther. 21: 1791–1799. Wynne-Edwards KE (1998) Evolution of parental care in Phodopus: Conflict between adaptations for survival and adaptations for rapid reproduction. Amer. Zool. 38: 238–250. Wynne-Edwards KE, Surov AV, Telitina AY (1999) Differences in endogenous activity within the genus Phodopus. J. Mammalogy 80: 855–865. Wynne-Edwards VC (1962) Animal Dispersal in Relation to Social Behaviour. Edinburgh: Oliver & Boyd. Yeats WB (1920) The Second Coming. In W. B. Yeats, Selected Poetry, ed. Jeffans A. London: Macmillan. Yoshida N, Oeda K, Watanabe E, Mikami T, Fukita Y, Nishimura K, Komai K, Matsuda K (2001) Chaperonin turned insect toxin. Nature 411: 44. Yuh C-H, Bolouri H, Davidson EH (1998) Genomic cis-regulatory logic: Experimental and computational analysis of a sea urchin gene. Science 279: 1896–1902. Yuh C-H, Bolouri H, Davidson EH (2001) Cis-regulatory logic in the endo16 gene. Development 128: 517–629. Zambryski P (1988) Basic processes underlying Agrobacterium-mediated DNA transfer to plant cells. Ann. Rev. Genet. 22: 1–30. Zambryski P (1989) Agrobacterium-plant cell DNA transfer. In Mobile DNA, ed. Berg D, Howe M. American Society for Microbiology. Zimmer C (1998) At the Water’s Edge: Fish with Fingers. Whales with Legs. New York: Simon and Schuster. Zuckerkandl E (1975) The appearance of new structures and functions in proteins during evolution. J. Mol. Evol. 7: 1–57.

Index

Accommodation epigenetic, 194–196 physiological, 144 Acquisition of heritable characteristics, 355 Adaptability, 13, 68, 82, 294, 315, 403 and adaptation, 140, 141 human, 100, 331 Adaptationism, 37, 39 Adaptiveness, 7, 397 Adaptive radiation, 30, 287, 411 Adaptive suites, 139 Aggregation, 125 Aggregativity, 316, 419 Agrobacterium tumefasciens, 20, 265 Albedo, 367 Alexander, S., 73, 318 Alexandrian schools, 133 Allometry, 47, 219, 267, 274, 275, 278, 279, 356, 357, 365, 367, 374 Allopolyploidy, 98 Ammonite lobate lines, 269 Ammonium ions, 171 Anabaena, 127 Ancon ram, 183 Animated water, 351 Antennapedia, 245 Anticipation diseases, 238, 272, 280 Ant lions, 101 Ants, 101, 102, 129, 359 Aphids, 101 Apoptosis, 235

Arber, A., 407 Archaeopteryx, 187 Archetype, 294, 412, 418 Arctic char, 343 Arctic fox, 55, 56, 151, 152 Ardisia, 122 Aristotelian holism, 82 Aristotle, 72, 126, 128, 181, 317, 322 Aromorphosis, 155 Artemia, 245 Arthropod appendage evolution, 412 Arthur, W., 48, 216, 225, 273 Artificial selection, 53, 56, 371 Ashby, W. R., 301 Associations, terminology of, 97 Autocatalytic molecular systems, 160 Autoevolution, 307 Autonomization, 55, 176 Axolotl, 192 Aymara tribe, 203 Babbage, C., 263 Bacon, F., 62, 322, 329 Baconian and Darwinian epistemology, 361 Baconian induction, 361 Bacteriophages, 127 Bacteroides thetaiotamicron, 123 Baer, K. E. von, 183 Bak, P., 74, 321 Baker, M. W., 239 Balfour, F. M., 185

506

Balon, E. K., 196, 207, 208, 338, 343 Bardou, F., and Jaeger, L., 228 Bateson, W., 4, 54, 64, 185, 341 Beament, J. W., 149, 150 Behavior, 337–339 choices, 360 genes for, 337 hereditary, 367 Behe, M., 12, 249 Belyaev, D., 55, 56, 151 Bemisia tabaci, 102 Bénard cells, 321, 374 Bergson, H., 318 Bernard, C., 138, 145 Berry, C. V., 27 Bertalanffy, L. von, 6, 8 Beurlen, K., 155 Bifurcation, 290 Big Bang, 16, 21, 85, 384, 426, 428 Big Bang of Biology, 3, 25, 85 Biochemical evolution, 156–158 Biochemical pathway switches, 172 Biofilms, 127 Biological structuralism, 307 Biological synthesis, 401 Biosphere, 350 Birdness, 415 Bithorax, 185, 245 Body plans bilaterally symmetrical, 8 radially symmetrical, 8 universal, 344 Bolk, L., 268 Bone morphological protein 2, 279 Bonner, J., 256 Branchiostoma lanceolatum, 246 Brenner, S., 211, 372 Britten, R. J., and Davidson, E. H., 165, 278 Brooks, W. K., 185 Buchnera, 101 Burgess Shale, 6 Burke, A. C., 197 Burkhardt, R., Jr., 265

Index

Burrowing benthic animals, 349 Buss, L., 215 Butler, S., 52 C3 plants, 172 C4 plants, 172 Cactus, 196 Caenorhabditis elegans, 211, 212, 371 Cairns-Smith, G., 161 Callebaut, W., and Raskin-Gutman, D., 301 Callinectes sapidus, 171, 295 Camel, 145 Campbell, J. H., 223, 262 Campbell, N., 307 Canalization, 190 Candida albicans, 207 Cannon, W., 138, 145 Caporale, L., 164, 225, 228 Cartilage formation, in limb buds, 197 Catastrophe, 53, 81,175, 220, 326, 351, 359, 366, 370, 405 Catastrophe insurance, 254 Catastrophic stress, 255 Cathepsins, 169, 170 Caudwell, C., 74, 75 Causal theory of evolution, 23 Causation D. Hume and E. Darwin on, 57 proximate cause of evolution, 58, 406 proximate causes and ultimate effects, 317 ultimate cause of evolution, 58, 294, 295, 406 Cell adhesion molecules, 108 Cell theory, 28 Cellulolytic symbiosis, 113,122, 336 Cellulose, 336, 367 Cenancestor, 162 Cenogenesis, 192, 246 Centriole, 105 Centromere, 107 Cepaea nemoralis, 206 Chaisson, E., 428 Chambers, R., 263, 318 Change of function, 293

Index

Chaperonins, 101 Cheddar man, 12 Cheek pouches, in rodents, 197 Chemoton stage, 162 Child, C. M., 137 Chloroplast, 118, 175, 335, 341 Chromatid break repair, 342 Chromatin marks, 251, 253 Chromosome amplification, 164 and sex, 104 diminution and deletion, 237 mutation, 107 puffs (Balbiani rings), 231 Churchill, F., 38 Chymotrypsin, 233 Cichlid evolution, 416 Circulatory system, 351 Circus clowns, 345 Coacervates, 160 Co-adaptation, 342 Coelurosauravus jaekeli, 217, 320 Co-evolution, 102 Cohen, J., and Stewart I., 67, 71, 311, 360, 387 Commensalism, 99 Competition, 34, 72, 134, 394, 426 Complexification, 115, 389, 402 Complexity, 68–70 Conatus, 46 Concerted evolution, 240, 274 Condition of mutability, 243 Conrad, M., 68, 140, 142, 143, 162 Consilience, 129 Constrained generating procedure, 313 Contingencies, 286, 366 Contingency-dependent emergences, 359 Cope, E. D., 4, 138, 146 Coral reefs, 112 Cormorant, 153 Corning, P., 37, 77 Corpus callosum, 309, 367 Cosmological evolution, 369, 426 Creationism, 2, 9, 31, 375

507

Creodes (cell lineages), 195 Crespi, B., 127 Critical-point emergence, 74, 320, 367 Crocodiles, 173 Croizat, L., 402 Crossroads of evolutionary theory, 421 Crystallins, 170, 248 Cubomedusan, 248 Curate’s egg, 427 Cuvier, G., 2, 370 Cyclic AMP, 168, 229 Cytotaxis, 250 Dachshund, 183 Darwin, C. R., 203, 267, 345, 352, 366, 435 absence of natural selection, 53 artificial selection, 425 laws of growth, 272 monstrosities, 32, 54, 119 natural selection as metaphor, 6 natural selection as consequence of “Struggle for Life,” 9 physiological adaptability, 46 physiology, 138 plasticity of organization, 54, 55 retrogression, 5 species, 417 Darwin, E., 16, 18, 212, 350 Davies, P., 281, 428 Dawkins, R., 22, 60, 61, 130, 249, 430 De Beer, G., 192, 193, 246, 268 Deep homology, 246 De Vries, H., 54, 185 Detoxifying gene duplication, 236 Developmental genetics, 38 Developmental thresholds, 338 Devolution, 395 Dialectical materialism, 361 Dialectical synthesis, 364 Dictyostelium, 358 Differential persistence, 314 Direct filiation theory, 119 Directional exaggerations, 270

508

Disequilibration, 360, 366 Dissostichus mawsoni, 233 Divergent saltations, 365 Diversifying evolution, 15 Djungarian hamster, 131, 151 Dobzhansky, T., 38, 90 Dobzhansky, T., Ayala, F. J., Stebbins, G. L, and Valentine, J. W., 8 Dodo, 346 Dog evolution, 283 Dog whelk morphs, 207 Dohrn, A., 354 Dolly (sheep), 253 D’Orbigny, A., 203 Dosage imbalance, 240 Dose amplification, 165 Dose repetition, 236 Douglas, A., 113, 116 Dover, G., 240, 260, 272, 281 Drummond, H., 73 Dwarf animals, 282 Dwarfism, 187 Dynamic structure, 304 Ecdysone, 231 Ecological release, 52 Ecostasis, 328, 366, 404 Ecosystems and symbiogenesis, 111–115 Edge effect, 149 Ediacarans, 213 Eimer, T., 267, 268 Eldredge, N, 27, 39, 59 Eldredge, N., and Gould, S. J., 50, 357 Eldredge, N., and Vrba, E., 402 Electronic evolution, 372 El-Hani, C., and Pihlström, S., 81 Eliot, T. S., 401, 406 Embryonization, 205 Emergence contingency-dependent, 359 critical-point, 79, 80, 320 and dialectical synthesis, 390 and environmental contingencies, 88 episodic, 366

Index

extrinsic (environmental), 74, 322, 324 formulas, 92 at interfaces, 381 intrinsic (autonomous), 74, 209, 322–324 by natural selection, 90 overview, 319 pressure, 311 and reductionism, 77 saltatory, 79, 80, 319, 359, 366, 367, 380 theses, and selectionist antitheses, 390–398 at thresholds, 196, 359 Through the Looking Glass, 292 Emergence theory empirical tests, 379 preliminary outline, 78 Emergent evolution modes, 80 rate, 410 Emergent properties, 13 constellation of, 145, 286 and resultant properties, 71 Emergentism and causal theory of evolution, 86 definitions of terms, 398, 399 as evolutionary theory, 368 and interactionism, 387 and neo-Darwinism, 93 origins, 72–75 Emperor penguin, 152 Endler, J., 1, 43, 44 Endo16, 263 End of science, 435 Endogenous retroviruses, 103 Endosymbiosis, 98 Entelechy, 189 Enterobacter aerogenes, 101 Epigenesis, 181 Epigenetics accommodatory mechanism, 262, 341 algorithms, 218, 219, 263, 264, 358, 371, 372 burden, 303 deviation, 193 effects of stress, 254

Index

effects of symbioses, 122 effects of transposable elements, 242 exploratory behavior, 200 inheritance systems, 250–257 mechanical stimuli, 217 mechanisms, 258–260 plasticity, 188 thresholds, 320, 343 Epiphenomena, 418 Epistemological monsters, 436 Epitrichal glands, 298 Escherichia coli, 228 Essentialists, 388 Euglena, 103 Eukaryote emergence, 116 Eusociality, 110, 125 Evo-devos, 47 Evolution accelerated in hominins, 10 behavioral, 337–339 by association, 330, 331 developmental, 339–344 in spite of natural selection, 24 neo-Darwinist definition, 36 physiological, 334–337 Evolutionary avalanche, 293 Evolutionary effects of external environment, 348 Evolutionary experiments, 8 Evolutionary “isms,” 3 Evolutionary psychology, 110, 127, 129, 337, 339 Evolutionary stable strategy, 34 Evolvability, 84, 219 Exaptation, 148, 354 Exons, 228, 232 Exon shuffling, 233 Extinctions, 239 Extrinsic emergence, 324 Eyeless, 212 Farmer, D., 382 Fat body, 171

509

Fenchel, T., 411 Fenchel, T., and Riedl, R., 111 Fetalization, 268 Fiber theory, 28, 29, 40 Finches, Darwin’s, 414, 415 Fisher, R. A., 50, 185 Fitness, 33, 395, 423 Fleming, A., and Allison, V. C.,170 Flight, as critical-point emergence, 375 Fondon, J., and Garner, H., 197, 283 Forbes, E., 203 Founder effect, 51 Fragile-X, 280 Frankenstein, 29 Franková, A., 338 Fugu rubripes, 236 Functionswechsel, 354 Gaia, 89, 350 Gaian evolution, 430 Galactosidase, 229 Galen, 72, 133 Galton, F., 49 Gannet, 141 Garstang, W., 192, 209 Gehring, W., 92, 211, 247 Gene amplification, 164 Gene conversion, 240 Gene pool, 11, 45 General theory of biology, 303 Generative conditions, 285, 366, 333, 334 Generative entrenchment, 303, 341 Generative procedures, 385 Generative hypotheses of evolution, 14 Genetic accommodation, 204 Genetic assimilation, 155, 201, 357 Genetic drift, 51 Genetic drive, 278 Genocopying, 208 Genophore, 104 Geophysiology, 350 Gerhart, J., and Kirschner, M., 243, 248, 414 Ghiselin, M., 277, 420

510

Giant animals, 282 Giantism, 187 Gifford lectures, 73 Gilbert, S., 301 Gill arches, 314, 416 Giraffe’s neck, 279 Global warming, 121 Glossinia, 100 Glycolysis, 172 Glycoprotein antifreeze, 234 Goethe, J. W. von, 181 Goldschmidt, R., 43, 51, 186, 208, 224, 268 Goldschmidt toad, 187 Goodrich, E. S., 193 Goodwin, B. C., 1, 23, 71, 91, 214, 215, 248, 249, 432, 433 Gottlieb, G., 146, 147, 338 Gould, S. J., 14, 192, 198, 250, 289, 382, 428 Gould, S. J., and Lewontin, R., 37 Gout, 173, 424 Gradualism, 3, 48 Grant, P., and Grant, R., 414 Grassé, P. -P., 4, 276 Grehan, J., 272 Grehan, J., and Ainsworth, R., 271, 284 Greksa, L. P., 203 Grene, M., 421 Grizzly bears, 387 GroEL, 101 Group, 125 Groups, as wholes, 419 Group selection hypothesis, 318 Growth hormone, 167, 239 Gut functions, 148 Haake, W., 268 Haeckel, E., 198 Haemophilus influenzae, 237 Haines, L. R., 100 Hair, 298 Hair glands, 148 Haldane, J. B. S., 38 Haldane’s aphorism, 431

Index

Hall, B. K., 197, 202, 216, 225, 344 Hall, T. S., 263 Halteres, 245 Hamburger, V., 48 Hardy, A., 193 Hawkes, N., 1 Heat-shock proteins, 235 Hegel, G. W. F., 72 Hegelian doctrine, 361 Helix aspersa, 206 Hemocyanin, 171 Hemoglobin, 169, 174 Hen’s teeth, 250 Heslop-Harrison, J., 272 Heterochrony, 184, 355 Heterorhesis, 190 Hierarchies compositional, 300, 386 control, 386 and emergent levels, 299–303 focal level, 300 generative level, 300 and levels of selection, 299 High table of selectionism, 10 Himmelfarb, G., 42 Hippocratic medicine, 133 His, W., 198 Historical layering, 386 Historical theory of evolution, 23 History of ideas, 291 Ho, M. -W., 179, 180, 208 Ho, M. -W., and Saunders, P., 179, 180 Hoatzin, 346 Holland, J., 22, 67, 73, 76, 312–314, 335, 344, 363, 371, 375, 385, 434 Holobiosis, 98 Holon, 301 Homeoboxes, 244 Homeodynamics, 147 Homeorhesis, 189, 220, 327, 340 Homeostasis, 145, 146, 327, 328, 340, 353, 413 Homeothermy, dinosaur and avian, 298

Index

Homeotic genes, 244 Homo floresiensis, 282 Homologous recombination repair, 235 Homology, 30, 218 Hopeful monster, 186, 283, 422 Horgan, J., 385 Horse, 270, 276 Hox, 164, 244, 320, 344 Hughes, Q. J., and Lambert, D., 407 Human genome project, 11 Huntington’s disease, 238 Hutton, J., 350 Huxley, J., 21, 38, 47, 137, 268, 275, 296, 363, 378, 402 Huxley, T. H., 73, 182, 184 Hydraulic mechanisms, 351 Hydrostatic pressure, 198 Hydrothermal vents, 158 Hylozoism (panpsychism), 48, 319, 384, 424, 427, 428, 429 Hypermorphosis, 367 Imaginal discs, 246 Imprinting, 252 Insects evolution, 412, 413 wings, 245 Insertion sequences, 241 Integrins, 109 Intelligence, 367 Intelligent design, 9, 249, 381 Interactionism, 12, 75, 264 Interference RNA, 230 Internal milieu, 347 Interphene, 198 Interspecific epigenetics, 344 Intolerant abstraction, 70, 157 Intragenic recombination, 232 Introns, 232 Isoenzymes, 165 Jablonka, E., and Lamb, M., 166, 202, 225, 250, 251, 253, 256, 262

511

Jegalian, K., and Lahn, B., 257 Jennings, H. S., 95, 109, 405, 432 Jokes, 434 Junk DNA, 223, 236 Kangaroo, 194 Kant, I., 72 Karyomastigont, 117 Karyotype fission theory, 108 Kauffman, S., 59, 76, 159, 249, 271, 289, 323, 324, 384, 438 Keller, E. F., 199, 271, 410 Key innovations, 83, 285, 298, 373 Kim, J., 77 Kimura, M., 51, 52 Kinetosome, 106 King, D., 238 King, R. C., and Stansfield, W. D., 202 Kingsolver, J. G., 409 Kirschner M., and Gerhart J., 143, 200, 225, 344 Kludges, 5 Koch, A. L., 254 Koch, C., and Laurent, G., 348 Koestler, A., 191, 301, 432, 434 Kollar, E. J., and Fisher, C., 250 Kölliker, R. A. von, 183 Kolnicki, R., 108 K-T (Cretaceous-Tertiary boundary), 3, 276, 412 Kuhn, T. S., 28 Labyrinthodont amphibia, 275 Lac operon, 228 Lactose, 229 Lactose synthetase, 170 Lamarck, J. -B., 3, 17, 29, 46, 130, 151, 177, 220, 266, 347, 363, 402 Lamarckism, 3, 265, 347 Lamprey, 246 Language, 131 Laysan duck, 254 Levit, G., and Hossfeld, U., 269

512

Lewes, G. H., 71 Lewontin, R., 345, 346 Li, W. -H., and Graur, D., 225, 240 Lichens, 99 Light stimulation of development, 348 Lima-de-Faria, A., 94, 307, 383 Limblessness, 187 Linde, A., 429 Livant, W., 11 Lobe fins, 247 Locusts, 176, 177, 188, 349 Logic, 132 Logsdon, J. F., and Doolittle, R. F., 234 Looking Glass logic, 63, 72, 78 Lotka, A., 428 Lovejoy, A., 291, 309 Lovelock, J., 89, 350 Løvtrup, S., 47, 63, 181, 199, 224, 225, 275 Lucinoid bivalves, 411 Luck, 51 Lung capacity, 203 Luther, M., 61 Lwoff, A., 173 Lyell, C., 4 Lymantria, 186 Lysosomes, 169 Lysozyme, 170 Mackie, G., 248 Manx cat, 187 MAP-kinase, 163 Margulis, L., 95, 105, 111, 116, 118 Margulis, L., and Sagan, D., 108, 117 Marxist doctrine, 361 Matsuda, R., 192, 196, 201, 205, 207, 223, 264, 343 Maynard Smith, J., 296 Maynard Smith, J., and Szathmáry, E., 22, 60, 133, 297 Mayr, E., 38, 39, 40, 58, 292–295 Mayr, E., and Provine, W., 47 McClintock, B., 51, 108, 224, 254, 296 McDougall, W., 427

Index

McGinnis, W., 245 McMenamin, M., 213 Mealybug, 102 Medawar, P., 208 Megaceros, 275 Membrane receptor molecules, 109 Meme concept, 120 Metamorphosis, 183, 246, 412 Metaphors biological, 19 emergentistic, 87 religious, 61 Methylation, 251 Methyl transferase, 252 Metz, C. W., 296 Mill, J. S., 72, 374 Milne Edwards, H., 295 Mind, 390, 433 Mitochondria, 117, 124, 174 Mitosis, 106 Mitotic spindle, 105 Mivart, St. G. J., 4, 50, 138, 267, 282, 357, 402 Mix-match principle, 95, 284, 354 Modern synthesis, 2,19, 23, 38, 40, 67, 364, 423 Modularity, 81, 301 Molecular adaptability, 205 Molecular clock, 410 Molecular drive, 240, 274 Monk, M., 250, 252 Monkey parable, 17, 430 Monod, J., 36 Monstroma, 122 Morgan, C. L., 21, 73, 132, 184, 321, 322, 374 Morowitz, H., 318 Morphogenetic fields, 424 Mouse gut epigenesis, 123 Muller, H. J., 19, 158 Müller, G. B., 196, 198, 199, 200, 340 Müller, G. B., and Wagner, G., 179, 180, 309, 310, 340 Multicellularity, 212–219

Index

Multifunctionality of emergents, 147, 286, 314 Murphy, J. J., 137 Mutation theory, 32, 48 Mycorrhizae, 113 Myoglobin, 169 Myotonic dystrophy, 238, 280 Myrmeliontidae, 101 Mystacina tuberculata, 321 Naked gene hypothesis, 158 Naked mole rat, 153 Natural experiment, 19, 80, 87, 297, 322, 351, 421 Natural selection. See also Darwin, C. R. absence of, 371 as barrier to evolution, 407 as “book-keeping,” 67 destabilizing, 55, 379 as differential survival and reproduction, 9, 33 directional, 269 disruptive, 409 as filter, 27 of groups, 420 as hypostasis, 5, 423 internal, 55 “in the wild,” 43 normalizing, 55 pre-Darwinian, 2 as secular creator, 38 of species, 420 as syndrome of causes and effects, 80 stabilizing, 10, 34 units of, 418 Nature philosophy, 181 Navier-Stokes equations, 374 Nelson, G., 402 Nematocyst, 112 Neo-Lamarckism, 3, 31, 177, 220, 266, 347 Neo-Lamarckist process, 253 Neophobism, 415 Neumann, J. von, 158

513

Neural crest cells, 168, 197, 216, 247, 373 New biology, 422 New genes, 169 Newman, S. A., and Müller, G. B., 207, 209, 215, 217, 218, 223, 249, 257, 338, 388 Nicolis, G., and Prigogine, I., 321 Niklas, K., 118 Nitrogen cycles, cellular, 173 Nitrogen fixation, 99,127 Nod factors, 122 Non-homologous recombination, 235 Non-random hypermutability, 164 Norton, H. T. J., 32 Null-hypothesis experiments, 56, 380 Numbers of differentiated cell types, 389 Nüsslein-Volhard, C., 214, 215 Nuttallia obscurata, 144 Octopus, 335 Odell, G., 199, 200, 410 Ohno, S., 165, 301 Ontogenic buffering, 195 Ontogenic plasticity of plants, 335 Oparin, A. I., 160 Operator, 228 Organophosphorus hydrolase, 170 Orgel, L., 159 Origin of life, 157–164, 351 Orthogenesis, 247, 413 and anticipation diseases, 280 and exaggeration of hereditary traits, 273–275 and genetic drive, 278, 279 and laws of growth, 267, 268 Orthoselection, 269 Osborn, H. F., 202, 203 Oscines brains, 294 Osmoregulation, 171, 148 Ostrich calluses, 358 Otter, 187 Overprinting, 233 Owen, R., 181, 295 Oyama S., 12, 75, 208, 263, 290, 308

514

Paedomorphosis, 191, 192, 243, 343 Pan, 226 Panda’s thumb, 198, 309 Panpsychism. See Hylozoism Paradigms, 28 Paraheredity, 98 Parasitism, 103 Parrots African grey, 415 kakapo, 346 kea, 415 Pax, 247 Pearson, R. D., 147, 176 P-element, 281 Peramorphosis, 192 Permease, 229, 250 Persistent patterns, 314, 387 Personal knowledge, 434 Pfeisteria, 127 Phase transitions, 358 Phenocopying, 208 Pheromones, 176 Phocomelic condition, 187 Phodopus campbelli, 151 Phodopus sungorus, 151 Phosphoenolpyruvate, 172 Photosynthetic ecosystems, 112 Physical gill, 150 Physiogenesis, 146, 154, 155, 334 Physiological evolution, 137, 138, 333–337 Phytoplankton, 166 Piaget, J., 338 Pigeon sports, 182 Plant-fungus ecosystems, 112 Plant physiology, 178 Plate, L., 269 Platonic essentialism, 45 Pleiotropic by-product, 293 Point mutation, non-synonymous, 227 Polanyi, M., 308 Pollinators, 103, 336 Polytene chromosomes, 231, 273 Popov, I., 269

Index

Popper, K., 56, 380 Population thinking, 44, 266 Position effects, 107, 224, 243 Post-Lamarckists, 368 Postmodernism, 16, 62–64, 422 Post-prediction, 375 Post-transcriptional gene silencing by RNA, 230 Poulton, E. B., 32 Pray, L., 254 Predictability of emergence, 373–375 Predictable emergences, 376–378 Predictiveness of emergence theory, 373–375 Proctotrupid wasps, 150 Progenesis, 192 Progenote stage, 162 Progress, 17, 60, 403 Progressionism, 17, 406 Progressive evolution, 17, 406 Prolactin, 167, 192, 239 Promoter sequence, 228 Prosser, C. L., 432, 433 Protein domains, 233 storage, 170, 171 tertiary structure, 228 Protobionts, 161 Provine W., 38, 425, 426 Pseudogene, 238 Pteranodon, 218 Pteroid bone, 218 Pug dogs, 273 Punctuated equilibrium, 21, 48, 375, 402 Punnett, R. C., 32 Pytho deplanatus, 351 Qualities, science of, 91 Questiones disputatae, 63 Quetzalocatlus, 218 Quorum sensing, 127 Raff, R., 197, 216, 278, 301 Raff, R., and Kaufman, T. C., 196

Index

Rate of evolution, 20 Reaction channeling, 163 Rebeck, J., 160 Recombinators, 230 Redox interface, 121 Red Queen, 327 Reduction, 11, 69 Reductionism, 11, 70, 77, 423 Redundancy, 8, 376 Regression, physiological, 173 Regulator gene, 229 Reid, C. E., 346, 415 Reid, R. G. B., 75, 271, 308, 341 Reimchen, T., 206 Rensch, B., 38, 47, 269, 276 Repair of double strand breaks in DNA, 238 Repair mechanisms, 234, 235 Repatterning, epigenetic, 199 Repetitive differentiation, 164, 353 Repetitive DNA, 236 Replication slippage, 237 Repressor, 229 Reproduction, 352 Respiration, 174 Retroposition, 241 Retroposon, 242 Retroviral gene acquisition, 355 Retroviruses, 243 Reverse transcriptase, 159, 242 Rhinoceros, 276 Rhizobium, 114, 127 Ribozyme, 159 Riedl, R., 186, 196, 302, 303 Riftia, 111 Ring, R. A., 351 RNA polymerase, 228 RNA world, 159 Robson, G. C., and Richards, O. W., 7 Robustness, epigenetic, 199, 200 physiological, 141 Rollo, D., 139, 302, 324, 376 Romanes, G., 58

515

Rupert, J. L., and Hochachka, P., 203 Ruse, M., 39 Salivary amylase, 243 Saltations, 50, 184, 245, 271, 276, 330 Saltatory evolution, 48 Salthe, S., 300, 306 Sapp, J., 97 Saunders, P., 190 Schaffner, W., 238 Schmalhausen I., 10, 13, 35, 139, 146, 149, 150, 176, 190, 193, 296 Schubert, M., and Holland, L. C., 247 Schwartz, D., 160, 161, 281 Schwartz, J., 39, 316 Scottish covenanters, 61 Second messengers, 168 Segment polarity genes, 244 Segregators, 230 Seilacher, A., Bose, P. K., and Pfluger, F., 215 Selection pressure, 6, 35, 49, 52, 170, 171, 234, 241, 248, 294, 349, 396, 405, 423 Selector genes, 230, 244 Selfish gene, 45 Self-organization, 355 Self-organized criticality, 74, 321 Serial homology, 295 Sesamoid bones, 198 Severtsov, A., 146, 155 Sex, 104 Sex-determining region Y (SRY) gene, 257, 258 Sexual reproduction, 257 Sexual selection, 419 Shapere, D., 47 Shapiro, J., 234, 261, 262 Sherrington, C. S., 138 Simon, J. H., 301 Simpson, G. G., 38, 44, 270 Slijper’s goat, 189 Smocovitis, V. B., 39, 40 Smolin, L., 429, 430 Social cohesiveness, 125

516

Social compartmentalization, 126 Social connectedness, 125 Social differentiation, 126 Social integration, 126 Social permeability, 125 Societies, 109–111, 125 Sodalis glossinidius, 100 Solé, R., and Goodwin, B. C., 128, 311, 369, 371, 374 Somatic mutation, 196 Sonic hedgehog, 279 Soredia, 99 SOS repair system, 235 Specialization bind, 191 Speciation, 85, 243, 417, 418 Species selection, 420 Spemann, H., 185 Spencer, H., 50, 184, 212 Sperry, R., 75, 308 Spiegelman’s monster, 281 Spinoza, B., 46 Spliceosomes, 232 Sponge cell reorganization, 358 Sports, 54, 56, 182 Stable evolutionary strategy, 407 Stadler, L. J., 295 Starlings, long-billed, 187 Stasis, 326 Stebbins, G. L., 296, 415 Step mechanisms, 301 Steroids, 231 St. Hilaire, E. G., 16, 181 St. Hilaire, I., 295 Stiassny, M., and Meyer, A., 416 Streptomyces, 102 Stress, 55, 349 Successful monsters, 368 Suess, E., 350 Sulfur-oxidizing bacteria, 121 Sulfur-oxidizing symbiosis, 121, 411 Supergenes, 224 Super-organism, 110 Symbiocosms, 99

Index

Symbiodinium, 98, 99 Symbiogenesis, 97, 120 Symbiont, 97 Symbioplex, 97 Symbiostasis, 121, 123, 327 Symbiote, 97 Syndactyly, 186 Synergism hypothesis, 77, 317 Synthesis, 40 Systems reduction, 71, 124, 388 Talitrid amphipods, 205 Telemorphosis, 153 Telomerase, 236 Teratology, 183 Terminus genes, 244 Termites, 102 Theory of correlated variation, 185 Theory of organic mechanism, 299 Theory of serial endosymbiosis, 118 Thermal vent, 111 Thermodynamics, second law of, 382 Thermogenesis, 332 Thermoplasma, 116, 117 Thiobios, 111, 411 Thompson, D., 271, 312, 369–371 Thomson, K., 194, 277, 278 Three-ring circus, 5 Threshold transition, 359 Thrombin, 169 Thyroxine, 192 Todd, N., 108 Toleration, physiological, 144 TRAM systems, 260 Transcriptional gene silencing by RNA, 230 Transformation theory, 271 Transgenesis, 243 Transition function, 312, 344 Transposable elements, 241–244 Transposase, 281 Transposition of limb placement, 193 Transposons, 241 Triceratops, 276

Index

Tridacnidae, 112, 123, 412 Trojan genes, 7 Tubulins, 170 Turner, J. S., 346, 349 Turtle carapace, 197 Tyrannosaurus rex, 210 Uexküll, J. J. von, 180, 224, 308, 354 Ultra-Darwinism, 4, 59, 407, 432 Ultramorphosis, 192, 267 Ulva, 122 Undulipod, 106 Uneven crossing over, 237 Urease, 173 Uricase, 173 Uricotelism, 173 Use-It-Or-Lose-It principle, 250 Uvarov, B., 188 Varied repeats of genes, 165, 239, 354 Vasopressin, 167 Vavilov, A. N., 54 Vegetative repetition, 295 Vendiobionts, 213 Venn diagrams, 300, 306 Vernadsky, V., 350 Vesuvian evolution, 242 Vie libre, 175 Viral transduction, 119 Virchow, R., 128 Vital spark, 318 Vitelline protein, 171 Vitellogenesis, 205 Vrba, E., 289, 275, 304 Vrba, E., and Eldredge, N., 303 Vrba’s “rules of emergence,” 305 Waagen, W., 268 Waddington, C. H., 47, 139, 146, 179, 181, 189, 195, 204, 225, 308 Wake, D., and Roth, B., 199 Waldrop, M., 382 Wallace, A. R., 35, 48, 53, 132

517

Water critical point, 351 emergent properties, 115, 374, 375, 427 Water ouzel, 321 Wax layers, in insects, 150 Weak emergences, 375 Weele, C. van der, 206, 207 Wells, W. C., 261 Weng, G., Bhalla, U. S., and Lyengar, R., 163, 347 West-Eberhard, M. J., 177, 189, 204 Whale evolution, 210 Wheeler, W. M., 109 Whewell, W., 129 Whitehead, A. N., 299 Whittaker, R. H., 120 Wholes greater than sums of their parts, 82, 114, 313, 330 Whyte, L. L., 55, 139, 308, 342 Wigglesworthia glossidinia, 100 Williams, G. C., 1, 22, 24, 37, 59 Willmer, E. N., 214 Wilson, E. O., 60, 70, 110, 124, 129 Wimsatt W., 14, 70, 290, 303, 316, 341 Wnt (wingless), 247 Wolbachia, 100, 243 Wolff, C., 295 Wollman, Eugene and Elisabeth, 97, 98 Woodger, J. H., 137, 157, 181 Wood Jones, F., 209 Wood lice, 113 Wright, E. E., 167 Wynne-Edwards, K., 151, 152 Wynne-Edwards, V. C., 318, 420 Xenopus laevis, 236 Y chromosome, 257 Zootermopsis angusticollis, 102