The Extended Phenotype
is the first holder of Oxford's newly endowed Charles Simoeyi Professorship of Public Understanding of Science. Born in Nairobi of British parents, Richard Dawkins was educated at Oxford and did his doctorate under the Nobel prize-winning ethologist Niko Tinbergen. From 1987 to 1969 he was an Assistant Professor at the University of California at Berkeley, then he returned to Oxford as University Lecturer (later Reader) and a Fellow of New College, before taking up his present position in 1995.
Richard Dawkins's bestselling books have played a significant role in the renaissance of science book publishing for a general audience. The Selfish Gene (1976; second edition 1989) was followed by The Extended Phenotype (1982), The Blind Watchmaker (1986), River Out of Eden (1995), Climbing Mount Improbable (1996), and Unweaving the Rainbow (1998). In 1991 he gave the Royal Institution Christmas Lectures. He has won many literary and scientific awards, including the 1987 Royal Society of Literature Award, the 1990 Michael Faraday Award of the Royal Society, the 1994 Nakayama Prize for Human Science, and the 1997 International Cosmos Prize.
The
Extended
Phenotype
The Long Reach of the Gene
Richard Dawkins
Oxford New York
OXFORD UNIVERSITY PRESS
| << | {iv} | >> |
Oxford University Press, Great Clarendon Street, Oxford 0x2 6DP
Oxford New York
Athens Auckland Bangkok Bogota Buenos Aires Calcutta
Cape Town Channai Dar es Salaam Delhi Florence Hong Kong Istanbul
Karachi Kuala Iumpur Mumhai Madrid Melbourne Mexico City
Nairobi Paris Sao Paulo Singapore Taipei Tokyo Toronto Warsaw
and associated companies in Berlin Ibadan
Oxford is a registered trade mark of Oxford University Press
Published in the United States by Oxford University Press Inc., New York
© Richard Dawkins 1989, 1999
Afterword © Daniel Dennett 1999
First published 1982
Raised edition with new Afterword and Further Reading 1999
All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the, prior permission in writing of Oxford University Press. Within the UK, exceptions are allowed in respect of any fair dealing for the purpose of research or private study, or criticism or review, as permitted under the Copyright, Designs and Patents Act, 1988, or in the case of reprographic reproduction in accordance with the terms of the licences issued by the Copyright licensing Agency. Enquiries concerning reproduction outside these terms and in other countries should be sent to the Rights Department, Oxford University Press, at the address above
This book is sold subject to the condition that it shall not, by way of trade or otherwise, be lent, re-sold, hired out or otherwise circulated without the publisher's prior consent in any form of binding or cover other than that in which it is published and without a similar condition including this condition being imposed on the subsequent purchaser
British Library Cataloguing in Publication Data
Data available
Library of Congress Cataloging in Publication Data Data available ISBN 0-19-288051-9
Typeset by Ace Filmsetting Ltd, Frome, Somerset
Printed in Great Britain
on acid-free paper by Cox & Wyman Lid,
Reading, Berkshire
10 9 8 7 6 5 4 3 2 1
| {v} | >> |
The first chapter does some of the work of a Preface, in explaining what the book does and does not set out to accomplish, so I can be brief here. It is not a textbook, nor an introduction to an established field. It is a personal look at the evolution of life, and in particular at the logic of natural selection and the level in the hierarchy of life at which natural selection can be said to act. I happen to be an ethologist, but I hope preoccupations with animal behaviour will not be too noticeable. The intended scope of the book is wider.
The readers for whom I am mainly writing are my professional colleagues, evolutionary biologists, ethologists and sociobiologists, ecologists, and philosophers and humanists interested in evolutionary science, including, of course, graduate and undergraduate students in all these disciplines. Therefore, although this book is in some ways the sequel to my previous book, The Selfish Gene, it assumes that the reader has professional knowledge of evolutionary biology and its technical terms. On the other hand it is possible to enjoy a professional book as a spectator, even if not a participant in the profession. Some laypeople who read this book in draft have been kind enough, or polite enough, to claim to have liked it. It would give me great satisfaction to believe them, and I have added a glossary of technical terms which I hope may help. I have also tried to make the book as near as possible to being enjoyable to read. The resulting tone may possibly irritate some serious professionals. I very much hope not, because serious professionals are the primary audience to whom I wish to speak. It is as impossible to please everybody in literary style as it is in any other matter of taste, and styles that give the most positive pleasure to some people are often the most annoying to others.
Certainly the tone of the book is not conciliatory or apologetic — such is not the way of an advocate that sincerely believes in his case — and I must pack all apology into the Preface. Some of the earlier chapters reply to criticisms of my previous book, which might recur in response to the present {vi} one. I am sorry that this is necessary, and I am sorry if a note of exasperation creeps in from time to time. I trust, at least, that my exasperation remains good humoured. It is necessary to point to past misunderstandings and try to forestall their repetition, but I would not wish to give an aggrieved impression that misunderstanding has been widespread. It has been confined to numerically very limited quarters, but in some cases rather vocal ones. I am grateful to my critics for forcing me to think again about how to express difficult matters more clearly.
I apologize to readers who may find a favourite and relevant work missing from the bibliography. There are those capable of comprehensively and exhaustively surveying the literature of a large field, but I have never been able to understand how they manage it. I know that the examples I have cited are a small subset of those that could have been cited, and are sometimes the writings or recommendations of my friends. If the result appears biased, well, of course it is biased, and I am sorry. I think nearly everybody must be somewhat biased in this kind of way.
A book inevitably reflects the current preoccupations of the author, and these preoccupations are likely to have been among the topics of his most recent papers. When those papers are so recent that it would be an artificial contrivance to change the words, I have not hesitated to reproduce a paragraph almost verbatim here and there. These paragraphs, which will be found in Chapters 4, 5, 6 and 14, are an integral part of the message of this book, and to omit them would be just as artificial as to make gratuitous changes in their wording.
The opening sentence of Chapter 1 describes the book as a work of unabashed advocacy but, well, perhaps I am just a little bit abashed! Wilson (1975, pp. 28–29) has rightly castigated the “advocacy method” in any search for scientific truth, and I have therefore devoted some of my first chapter to a plea of mitigation. I certainly would not want science to adopt the legal system in which professional advocates make the best case they can for a position, even if they believe it to be false. I believe deeply in the view of life that this book advocates, and have done so, at least in part, for a long time, certainly since the time of my first published paper, in which I characterized adaptations as favouring “the survival of the animal's genes...” (Dawkins 1968). This belief — that if adaptations are to be treated as “for the good of something, that something is the gene — was the fundamental assumption of my previous book. The present book goes further. To dramatize it a bit, it attempts to free the selfish gene from the individual organism which has been its conceptual prison. The phenotypic effects of a gene are the tools by which it levers itself into the next generation, and these tools may “extend” far outside the body in which the gene sits, even reaching deep into the nervous systems of other organisms. Since it is not a factual position I am advocating, but a way of seeing facts, I wanted to warn the reader not to expect {vii} “evidence” in the normal sense of the word. I announced that the book was a work of advocacy, because I was anxious not to disappoint the reader, not to lead her on under false pretences and waste her time.
The linguistic experiment of the last sentence reminds me that I wish I had had the courage to instruct the computer to feminize personal pronouns at random throughout the text. This is not only because I admire the current awareness of the masculine bias in our language. Whenever I write I have a particular imaginary reader in mind (different imaginary readers oversee and “filter” the same passage in numerous successive revisions) and at least half my imaginary readers are, like at least half my friends, female. Unfortunately it is still true in English that the unexpectedness of a feminine pronoun, where a neutral meaning is intended, seriously distracts the attention of most readers, of either sex. I believe the experiment of the previous paragraph will substantiate this. With regret, therefore, I have followed the standard convention in this book.
For me, writing is almost a social activity, and I am grateful to the many friends who have, sometimes unwittingly, participated through discussion, argument and moral support. I cannot thank them all by name. Marian Stamp Dawkins has not only provided sensitive and knowledgeable criticism of the whole book in several drafts. She has also kept me going by believing in the project even through the times when I lost my own confidence. Alan Grafen and Mark Ridley, officially my graduate students, really, in their different ways, my mentors and guides through difficult theoretical territory, have influenced the book immeasurably. In the first draft their names seemed to creep in on almost every page, and it was only the pardonable grumblings of a referee that compelled me to banish to the Preface my acknowledgment of debt to them. Cathy Kennedy manages to combine close friendship for me with deep sympathy for my bitterest critics. This has put her in a unique position to advise me, especially over the earlier chapters which attempt to reply to criticism. I fear that she will still not like the tone of these chapters, but such improvement as there may be is largely due to her influence and I am very grateful to her.
I was privileged to have the first draft criticized in its entirety by John Maynard Smith, David C. Smith, John Krebs, Paul Harvey, and Ric Charnov, and the final draft owes much to all of them. In all cases I acted on their advice, even if I did not always take it. Others kindly criticized chapters in their own special fields: Michael Hansell the chapter on artefacts, Pauline Lawrence that on parasites, Egbert Leigh that on fitness, Anthony Hallam the section on punctuated equilibria, W. Ford Doolittle that on selfish DNA, and Diane De Steven botanical sections. The book was finished at Oxford, but begun during a visit to the University of Florida at Gainesville on a sabbatical leave kindly granted by the University of Oxford and the Warden and Fellows of New College. I am grateful to my many {viii} Floridan friends for giving me such a pleasant atmosphere in which to work, especially Jane Brockmaen, who also provided helpful criticism of preliminary drafts, and Donna Gillis, who also did much of the typing. I benefited, too, from a month's exposure to tropical biology as the grateful guest of the Smithsonian Institution in Panama during the writing of the book. Finally, it is a pleasure once again to thank Michael Rodgers, formerly of Oxford University Press and now of W. H. Freeman and Company, a “K-selected” editor who really believes in his books and is their tireless advocate.
Oxford |
Richard Dawkins |
I suppose most scientists — most authors — have one piece of work of which they would say: it doesn't matter if you never read anything else of mine, please at least read this. For me, it is The Extended Phenotype, in particular, the last four chapters constitute the best candidate for the title “innovative” that I have to offer. The rest of the book does some necessary sorting out on the way. Chapters 2 and 3 are replies to criticisms of the now widely accepted “selfish gene” view of evolution. The middle chapters deal with the “units of selection” controversy currently fashionable among philosophers of biology, taking the gene's-eye view; perhaps the most useful contribution here is the “Replicators and Vehicles” distinction. My intention was that this piece of sorting out should put paid to the whole controversy once and for all!
As for the extended phenotype proper, I have never seen any alternative to placing it at the end of the book. Nevertheless, this policy has a disadvantage. The earlier chapters inevitably draw attention to the general “units of selection” issue, and away from the more novel idea of the extended phenotype itself. It is for this reason that I have dropped the original subtitle, “The Gene as the Unit of Selection”, from this edition. The replacement, “The Long Reach of the Gene”, captures the idea of the gene as the centre of a web of radiating power. The book is otherwise unchanged apart from minor corrections.
Oxford |
Richard Dawkins |
| << | {ix} | >> |
| << | {x} | >> |
pp. 141–142 “The Fifth Philosopher's Song” from The Collected Poetry of Aldous Huxley edited by Donald Watt. Copyright © 1971 by Laura Huxley. Reproduced by permission of Mrs. Laura Huxley and Chatto & Windus Ltd, also by permission of Harper & Row, Publishers, Inc. (USA)
p. 16 “McAndrew's Hymn” by Richard Kipling. Extract reproduced by permission of Doubleday & Company Inc.
The Extended Phenotype
| << | {1} | >> |
This is a work of unabashed advocacy. I want to argue in favour of a particular way of looking at animals and plants, and a particular way of wondering why they do the things that they do. What I am advocating is not a new theory, not a hypothesis which can be verified or falsified, not a model which can be judged by its predictions. If it were any of those things, I agree with Wilson (1975, p. 28) that the “advocacy method” would be inappropriate and reprehensible. But it is not any of those things. What I am advocating is a point of view, a way of looking at familiar facts and ideas, and a way of asking new questions about them. Any reader who expects a convincing new theory in the conventional sense of the word is bound to be left, therefore, with a disappointed “so what?” feeling. But I am not trying to convince anyone of the truth of any factual proposition. Rather, I am trying to show the reader a way of seeing biological facts.
There is a well-known visual illusion called the Necker Cube. It consists of a line drawing which the brain interprets as a three-dimensional cube. But there are two possible orientations of the perceived cube, and both are equally compatible with the two-dimensional image on the paper. We usually begin by seeing one of the two orientations, but if we look for several seconds the cube “flips over” in the mind, and we see the other apparent orientation. After a few more seconds the mental image flips back and it continues to alternate as long as we look at the picture. The point is that neither of the two perceptions of the cube is the correct or “true” one. They are equally correct. Similarly the vision of life that I advocate, and label with the name of the extended phenotype, is not provably more correct than the orthodox view. It is a different view and I suspect that, at least in some respects, it provides a deeper understanding. But I doubt that there is any experiment that could be done to prove my claim.
The phenomena that I shall consider — coevolution, arms races, manipulation of hosts by parasites, manipulation of the inanimate world by living things, economic “strategies” for minimizing costs and maximizing benefits — {2} are all familiar enough, and are already the subject of intensive study. Why, then, should the busy reader bother to go on? It is tempting to borrow Stephen Gould's winningly ingenuous appeal at the beginning of a more substantial volume (1977a) and simply say, “Please read the book” and you will find out why it was worth bothering to do so. Unfortunately I do not have the same grounds for confidence. I can only say that, as one ordinary biologist studying animal behaviour, I have found that the viewpoint represented by the label “extended phenotype” has made me see animals and their behaviour differently, and I think I understand them better for it. The extended phenotype may not constitute a testable hypothesis in itself, but it so far changes the way we see animals and plants that it may cause us to think of testable hypotheses that we would otherwise never have dreamed of.
Lorenz's (1937) discovery that a behaviour pattern can be treated like an anatomical organ was not a discovery in the ordinary sense. No experimental results were adduced in its support. It was simply a new way of seeing facts that were already commonplace, yet it dominates modern ethology (Tinbergen 1963), and it seems to us today so obvious that it is hard to understand that it ever needed “discovering”. Similarly, D'Arcy Thompson's (1917) celebrated chapter “On the theory of transformations...” is widely regarded as a work of importance although it does not advance or test a hypothesis. In a sense it is obviously necessarily true that any animal form can be turned into a related form by a mathematical transformation, although it is not obvious that the transformation will be a simple one. In actually doing it for a number of specific examples, D'Arcy Thompson invited a “so what?” reaction from anyone fastidious enough to insist that science proceeds only by the falsifying of specific hypotheses. If we read D'Arcy Thompson's chapter and then ask ourselves what we now know that we did not know before, the answer may well be not much. But our imagination is fired. We go back and look at animals in a new way; and we think about theoretical problems, in this case those of embryology and phytogeny and their interrelations, in a new way. I am, of course, not so presumptuous as to compare the present modest work with the masterpiece of a great biologist. I use the example simply to demonstrate that it is possible for a theoretical book to be worth reading even if it does not advance testable hypotheses but seeks, instead, to change the way we see.
Another great biologist once recommended that to understand the actual we must contemplate the possible: “No practical biologist interested in sexual reproduction would be led to work out the detailed consequences experienced by organisms having three or more sexes; yet what else should he do if he wishes to understand why the sexes are, in fact, always two?” (Fisher 1930a, p. ix). Williams (1975), Maynard Smith (1978a) and others have taught us that one of the commonest, most universal features of life on Earth, sexuality itself, should not be accepted without question. Indeed, its {3} existence turns out to be positively surprising when set against the imagined possibility of asexual reproduction. To imagine asexual reproduction as a hypothetical possibility is not difficult, since we know it is a reality in some animals and plants. But are there other cases where our imagination receives no such prompting? Are there important facts about life that we hardly notice simply because we lack the imagination to visualise alternatives which, like Fisher's three sexes, might have existed in some possible world? I shall try to show that the answer is yes.
Playing with an imaginary world, in order to increase our understanding of the actual world, is the technique of “thought experiment”. It is much used by philosophers. For instance in a collection of essays on The Philosophy of Mind (ed. Glover 1976), various authors imagine surgical operations in which one person's brain is transplanted into another person's body, and they use the thought experiment to clarify the meaning of “personal identity”. At times philosophers' thought experiments are purely imaginary and wildly improbable, but this doesn't matter given the purpose for which they are made. At other times they are informed, to a greater or lesser extent, by facts from the real world, for instance the facts of split-brain experiments.
Consider another thought experiment, this time from evolutionary biology. When I was an undergraduate obliged to write speculative essays on “the origin of the Chordates” and other topics of remote phytogeny, one of my tutors rightly tried to shake my faith in the value of such speculations by suggesting that anything could, in principle, evolve into anything else. Even insects could evolve into mammals, if only the right sequence of selection pressures were provided in the right order. At the time, as most zoologists would, I dismissed the idea as obvious nonsense, and I still, of course, don't believe that the right sequence of selection pressures ever would be provided. Nor did my tutor. But as far as the principle is concerned, a simple thought experiment shows it to be nearly incontrovertible. We need only prove that there exists a continuous series of small steps leading from an insect, say a stag beetle, to a mammal, say a stag. By this I mean that, starting with the beetle, we could lay out a sequence of hypothetical animals, each one as similar to the previous member of the series as a pair of brothers might be, and the sequence would culminate in a red deer stag.
The proof is easy, provided only that we accept, as everyone does, that beetle and deer have a common ancestor, however far back. Even if there is no other sequence of steps from beetle to deer, we know that at least one sequence must be obtained by simply tracing the beetle's ancestors back to the common ancestor, then working forwards down the other line to the deer.
We have proved that there exists a trajectory of stepwise change connecting beetle to deer and, by implication, a similar trajectory from any modern animal to any other modern animal. In principle, therefore, we may {4} presume that a series of selection pressures could be artificially contrived to propel a lineage along one of these trajectories. It was a quick thought experiment along these lines that enabled me to say, when discussing D'Arcy Thompson's transformations, that “In a sense it is obviously necessarily true that any animal form can be turned into a related form by a mathematical transformation, although it is not obvious that the transformation will be a simple one”. In this book I shall make frequent use of the thought-experiment technique. I warn the reader of this in advance, since scientists are sometimes annoyed by the lack of realism in such forms of reasoning. Thought experiments are not supposed to be realistic. They are supposed to clarify our thinking about reality.
One feature of life in this world which, like sex, we have taken for granted and maybe should not, is that living matter comes in discrete packages called organisms. In particular, biologists interested in functional explanation usually assume that the appropriate unit for discussion is the individual organism. To us, “conflict” usually means conflict between organisms, each one striving to maximize its own individual “fitness”. We recognize smaller units such as cells and genes, and larger units such as populations, societies and ecosystems, but there is no doubt that the individual body, as a discrete unit of action, exerts a powerful hold over the minds of zoologists, especially those interested in the adaptive significance of animal behaviour. One of my aims in this book is to break that hold. I want to switch emphasis from the individual body as focal unit of functional discussion. At the very least I want to make us aware of how much we take for granted when we look at life as a collection of discrete individual organisms.
The thesis that I shall support is this. It is legitimate to speak of adaptations as being “for the benefit of” something, but that something is best not seen as the individual organism. It is a smaller unit which I call the active, germ-line replicator. The most important kind of replicator is the “gene” or small genetic fragment. Replicators are not, of course, selected directly, but by proxy; they are judged by their phenotypic effects. Although for some purposes it is convenient to think of these phenotypic effects as being packaged together in discrete “vehicles” such as individual organisms, this is not fundamentally necessary. Rather, the replicator should be thought of as having extended phenotypic effects, consisting of all its effects on the world at large, not just its effects on the individual body in which it happens to be sitting.
To return to the analogy of the Necker Cube, the mental flip that I want to encourage can be characterized as follows. We look at life and begin by seeing a collection of interacting individual organisms. We know that they contain smaller units, and we know that they are, in turn, parts of larger composite units, but we fix our gaze on the whole organisms. Then suddenly the image flips. The individual bodies are still there; they have not moved, {5} but they seem to have gone transparent. We see through them to the replicating fragments of DNA within, and we see the wider world as an arena in which these genetic fragments play out their tournaments of manipulative skill. Genes manipulate the world and shape it to assist their replication. It happens that they have “chosen” to do so largely by moulding matter into large multicellular chunks which we call organisms, but this might not have been so. Fundamentally, what is going on is that replicating molecules ensure their survival by means of phenotypic effects on the world. It is only incidentally true that those phenotypic effects happen to be packaged up into units called individual organisms.
We do not at present appreciate the organism for the remarkable phenomenon it is. We are accustomed to asking, of any widespread biological phenomenon, “What is its survival value?” But we do not say, “What is the survival value of packaging life up into discrete units called organisms?” We accept it as a given feature of the way life is. As I have already noted, the organism becomes the automatic subject of our questions about the survival value of other things: “In what way does this behaviour pattern benefit the individual doing it? In what way does this morphological structure benefit the individual it is attached to?”
It has become a kind of “central theorem” (Barash 1977) of modern ethology that organisms are expected to behave in such a way as to benefit their own inclusive fitness (Hamilton 1964a, b), rather than to benefit anyone, or anything, else. We do not ask in what way the behaviour of the left hind leg benefits the left hind leg. Nor, nowadays, do most of us ask how the behaviour of a group of organisms, or the structure of an ecosystem, benefits that group or ecosystem. We treat groups and ecosystems as collections of warring, or uneasily cohabiting, organisms; and we treat legs, kidneys, and cells as cooperating components of a single organism. I am not necessarily objecting to this focus of attention on individual organisms, merely calling attention to it as something that we take for granted. Perhaps we should stop taking it for granted and start wondering about the individual organism, as something that needs explaining in its own right, just as we found sexual reproduction to be something that needs explaining in its own right.
At this point an accident of the history of biology necessitates a tiresome digression. The prevailing orthodoxy of my previous paragraph, the central dogma of individual organisms working to maximize their own reproductive success, the paradigm of “the selfish organism”, was Darwin's paradigm, and it is dominant today. One might imagine, therefore, that it had had a good run for its money and should, by now, be ripe for revolution, or at least constitute a solid-enough bastion to withstand iconoclastic pinpricks such as any that this book might deliver. Unfortunately, and this is the historical accident I mentioned, although it is true that there has seldom been any {6} temptation to treat units smaller than the organism as agents working for their own benefit, the same has not always been true of larger units. The intervening years since Darwin have seen an astonishing retreat from his individual-centred stand, a lapse into sloppily unconscious group-selectionism, ably documented by Williams (1966), Ghiselin (1974a) and others. As Hamilton (1975a) put it, “...almost the whole field of biology stampeded in the direction where Darwin had gone circumspectly or not at all”. It is only in recent years, roughly coinciding with the belated rise to fashion of Hamilton's own ideas (Dawkins 1979b), that the stampede has been halted and turned. We painfully struggled back, harassed by sniping from a Jesuitically sophisticated and dedicated neo-group-selectionist rearguard, until we finally regained Darwin's ground, the position that I am characterizing by the label “the selfish organism”, the position which, in its modern form, is dominated by the concept of inclusive fitness. Yet it is this hard-won fastness that I may seem to be abandoning here, abandoning almost before it is properly secured; and for what? For a flickering Necker Cube, a metaphysical chimera called the extended phenotype?
No, to renounce those gains is far from my intention. The paradigm of the selfish organism is vastly preferable to what Hamilton (1977) has called “the old, departing paradigm of adaptation for the benefit of the species”. “Extended phenotype” is misunderstood if it is taken to have any connection with adaptation at the level of the group. The selfish organism, and the selfish gene with its extended phenotype, are two views of the same Necker Cube. The reader will not experience the conceptual flip-over that I seek to assist unless he begins by looking at the right cube. This book is addressed to those that already accept the currently fashionable selfish-organism view of life, rather than any form of “group benefit” view.
I am not saying that the selfish organism view is necessarily wrong, but my argument, in its strong form, is that it is looking at the matter the wrong way up. I once overheard an eminent Cambridge ethologist say to an eminent Austrian ethologist (they were arguing about behaviour development): “You know, we really agree. It is just that you say it wrong”. Gentle “individual selectionist”, we really do almost agree, at least in comparison to the group selectionists. It is just that you see it wrong!
Bonner (1958), discussing single-celled organisms, said “...what special use to these organisms are nuclear genes? How did they arise by selection?” This is a good example of the kind of imaginative, radical question that I think we ought to ask about life. But if the thesis of this book is accepted, the particular question should be turned upside down. Instead of asking of what use nuclear genes are to organisms, we should ask why genes chose to group themselves together in nuclei, and in organisms. In the opening lines of the same work, Bonner says: “I do not propose to say anything new or original in these lectures. But I am a great believer in saying familiar, well-known {7} things backwards and inside out, hoping that from some new vantage point the old facts will take on a deeper significance. It is like holding an abstract painting upside down; I do not say that the meaning of the picture will suddenly be clear, but some of the structure of the composition that was hidden may show itself” (p. 1). I came across this after writing my own Necker Cube passage, and was delighted to find the same views expressed by so respected an author.
The trouble with my Necker Cubes, and with Bonner's abstract painting, is that, as analogies, they may be too timid and unambitious. The analogy of the Necker Cube expresses my minimum hope for this book. I am pretty confident that to look at life in terms of genetic replicators preserving themselves by means of their extended phenotypes is at least as satisfactory as to look at it in terms of selfish organisms maximizing their inclusive fitness. In many cases the two ways of looking at life will, indeed, be equivalent. As I shall show, “inclusive fitness” was defined in such a way as to tend to make “the individual maximizes its inclusive fitness” equivalent to “the genetic replicators maximize their survival”. At least, therefore, the biologist should try both ways of thinking, and choose the one he or she prefers. But I said this was a minimum hope. I shall discuss phenomena, “meiotic drive” for instance, whose explanation is lucidly written on the second face of the cube, but which make no sense at all if we keep our mental gaze firmly fixed on the other face, that of the selfish organism. Moving from my minimum hope to my wildest daydream, it is that whole areas of biology, the study of animal communication, animal artefacts, parasitism and symbiosis, community ecology, indeed all interactions between and within organisms, will eventually be illuminated in new ways by the doctrine of the extended phenotype. As is the way with advocates, I shall try to make the strongest case I can, and this means the case for the wilder hopes rather than the more cautious minimum expectations.
If these grandiose hopes are eventually realized, perhaps a less modest analogy than the Necker Cube will be pardoned. Colin Turnbull (1961) took a pygmy friend, Kenge, out of the forest for the first time in his life, and they climbed a mountain together and looked out over the plains. Kenge saw some buffalo “grazing lazily several miles away, far down below. He turned to me and said. "What insects are those?" ... At first I hardly understood, then I realized that in the forest vision is so limited that there is no great need to make an automatic allowance for distance when judging size. Out here in the plains, Kenge was looking for the first time over apparently unending miles of unfamiliar grasslands, with not a tree worth the name to give him any basis for comparison ... When I told Kenge that the insects were buffalo, he roared with laughter and told me not to tell such stupid lies...” (pp. 227–228).
This book as a whole, then, is a work of advocacy, but it is a poor advocate {8} that leaps precipitately to his conclusion when the jury are sceptical. The second face of my Necker Cube is unlikely to click into clear focus until near the end of the book. Earlier chapters prepare the ground, attempt to forestall certain risks of misunderstanding, dissect the first face of the Necker Cube in various ways, point up reasons why the paradigm of the selfish individual, if not actually incorrect, can lead to difficulties.
Parts of some early chapters are frankly retrospective and even defensive. Reaction to a previous work (Dawkins 1976a) suggests that this book is likely to raise needless fears that it promulgates two unpopular “-isms” — “genetic determinism” and “adaptationism”. I myself admit to being irritated by a book that provokes me into muttering “Yes but...” on every page, when the author could easily have forestalled my worry by a little considerate explanation early on. Chapters 2 and 3 try to remove at least two major sources of “yes-buttery” at the outset.
Chapter 4 opens the case for the prosecution against the selfish organism, and begins to hint at the second aspect of the Necker Cube. Chapter 5 opens the case for the “replicator” as the fundamental unit of natural selection. Chapter 6 returns to the individual organism and shows how neither it, nor any other major candidate except the small genetic fragment, qualifies as a true replicator. Rather, the individual organism should be thought of as a “vehicle” for replicators. Chapter 7 is a digression on research methodology. Chapter 8 raises some awkward anomalies for the selfish organism, and Chapter 9 continues the theme. Chapter 10 discusses various notions of “individual fitness”, and concludes that they are confusing, and probably dispensable.
Chapters 11, 12 and 13 are the heart of the book. They develop, by gradual degrees, the idea of the extended phenotype itself, the second face of the Necker Cube. Finally, in Chapter 14, we turn back with refreshed curiosity to the individual organism and ask why, after all, it is such a salient level in the hierarchy of life.
| << | {9} | >> |
Long after his death, tenacious rumours persisted that Adolf Hitler had been seen alive and well in South America, or in Denmark, and for years a surprising number of people with no love for the man only reluctantly accepted that he was dead (Trevor-Roper 1972). In the First World War a story that a hundred thousand Russian troops had been seen landing in Scotland “with snow on their boots” became widely current, apparently because of the memorable vividness of that snow (Taylor 1963). In our own time myths such as that of computers persistently sending householders electricity bills for a million pounds (Evans 1979), or of well-heeled welfare-scroungers with two expensive cars parked outside their government-subsidized council houses, are familiar to the point of cliche. There are some falsehoods, or half-truths, that seem to engender in us an active desire to believe them and pass them on even if we find them unpleasant, maybe in part, perversely, because we find them unpleasant.
Computers and electronic “chips” provoke more than their fair share of such myth-making, perhaps because computer technology advances at a speed which is literally frightening. I know an old person who has it on good authority that “chips” are usurping human functions to the extent not only of “driving tractors” but even of “fertilizing women”. Genes, as I shall show, are the source of what may be an even larger mythology than computers. Imagine the result of combining these two powerful myths, the gene myth and the computer myth! I believe that I may have inadvertently achieved some such unfortunate synthesis in the minds of a few readers of my previous book, and the result was comic misunderstanding. Happily, such misunderstanding was not widespread, but it is worth trying to avoid a repeat of it here, and that is one purpose of the present chapter. I shall expose the myth of genetic determinism, and explain why it is necessary to use language that can be unfortunately misunderstood as genetic determinism.
A reviewer of Wilson's (1978) On Human Nature, wrote: “...although he does not go as far as Richard Dawkins (The Selfish Gene...) in proposing sex-linked {10} genes for "philandering", for Wilson human males have a genetic tendency towards polygyny, females towards constancy (don't blame your mates for sleeping around, ladies, it's not their fault they are genetically programmed). Genetic determinism constantly creeps in at the back door” (Rose 1978). The reviewer's clear implication is that the authors he is criticizing believe in the existence of genes that force human males to be irremediable philanderers who cannot therefore be blamed for marital infidelity. The reader is left with the impression that those authors are protagonists in the “nature or nurture” debate, and, moreover, died-in-the-wool hereditarians with male chauvinist leanings.
In fact my original passage about “philanderer males” was not about humans. It was a simple mathematical model of some unspecified animal (not that it matters, I had a bird in mind). It was not explicitly (see below) a model of genes, and if it had been about genes they would have been sex-limited, not sex-linked! It was a model of “strategies” in the sense of Maynard Smith (1974). The “philanderer” strategy was postulated, not as the way males behave, but as one of two hypothetical alternatives, the other being the “faithful” strategy. The purpose of this very simple model was to illustrate the kinds of conditions under which philandering might be favoured by natural selection, and the kinds of conditions under which faithfulness might be favoured. There was no presumption that philandering was more likely in males than faithfulness. Indeed, the particular run of the simulation that I published culminated in a mixed male population in which faithfulness slightly predominated (Dawkins 1976a, p. 165, although see Schuster & Sigmund 1981). There is not just one misunderstanding in Rose's remarks, but multiple compounded misunderstanding. There is a wanton eagerness to misunderstand. It bears the stamp of snow-covered Russian jackboots, of little black microchips marching to usurp the male role and steal our tractor-drivers' jobs. It is a manifestation of a powerful myth, in this case the great gene myth.
The gene myth is epitomized in Rose's parenthetic little joke about ladies not blaming their mates for sleeping around. It is the myth of “genetic determinism”. Evidently, for Rose, genetic determinism is determinism in the full philosophical sense of irreversible inevitability. He assumes that the existence of a gene “for” X implies that X cannot be escaped. In the words of another critic of “genetic determinism”, Gould (1978, p. 238), “If we are programmed to be what we are, then these traits are ineluctable. We may, at best, channel them, but we cannot change them either by will, education, or culture”.
The validity of the determinist point of view and, separately, its bearing on an individual's moral responsibility for his actions, has been debated by philosophers and theologians for centuries past, and no doubt will be for centuries to come. I suspect that both Rose and Gould are determinists in {11} that they believe in a physical, materialistic basis for all our actions. So am I. We would also probably all three agree that human nervous systems are so complex that in practice we can forget about determinism and behave as if we had free will. Neurones may be amplifiers of fundamentally indeterminate physical events. The only point I wish to make is that, whatever view one takes on the question of determinism, the insertion of the word “genetic” is not going to make any difference. If you are a full-blooded determinist you will believe that all your actions are predetermined by physical causes in the past, and you may or may not also believe that you therefore cannot be held responsible for your sexual infidelities. But, be that as it may, what difference can it possibly make whether some of those physical causes are genetic? Why are genetic determinants thought to be any more ineluctable, or blame-absolving, than “environmental” ones?
The belief that genes are somehow super-deterministic, in comparison with environmental causes, is a myth of extraordinary tenacity, and it can give rise to real emotional distress. I was only dimly aware of this until it was movingly brought home to me in a question session at a meeting of the American Association for the Advancement of Science in 1978. A young woman asked the lecturer, a prominent “sociobiologist”, whether there was any evidence for genetic sex differences in human psychology. I hardly heard the lecturer's answer, so astonished was I by the emotion with which the question was put. The woman seemed to set great store by the answer and was almost in tears. After a moment of genuine and innocent bafflement the explanation hit me. Something or somebody, certainly not the eminent sociobiologist himself, had misled her into thinking that genetic determination is for keeps; she seriously believed that a “yes” answer to her question would, if correct, condemn her as a female individual to a life of feminine pursuits, chained to the nursery and the kitchen sink. But if, unlike most of us, she is a determinist in that strong Calvinistic sense, she should be equally upset whether the causal factors concerned are genetic or “environmental”.
What does it ever mean to say that something determines something? Philosophers, possibly with justification, make heavy weather of the concept of causation, but to a working biologist causation is a rather simple statistical concept. Operationally we can never demonstrate that a particular observed event C caused a particular result R, although it will often be judged highly likely. What biologists in practice usually do is to establish statistically that events of class R reliably follow events of class C. They need a number of paired instances of the two classes of events in order to do so: one anecdote is not enough.
Even the observation that R events reliably tend to follow C events after a relatively fixed time interval provides only a working hypothesis that C events cause R events. The hypothesis is confirmed, within the limits of the statistical method, only if the C events are delivered by an experimenter rather {12} than simply noted by ao observer, and are still reliably followed by R events. It is not necessary that every C should be followed by an R, nor that every R should be preceded by a C (who has not had to contend with arguments such as “smoking cannot cause lung cancer, because I knew a non-smoker who died of it, and a heavy smoker who is still going strong at ninety”?). Statistical methods are designed to help us assess, to any specified level of probabilistic confidence, whether the results we obtain really indicate a causal relationship.
If, then, it were true that the possession of a Y chromosome had a causal influence on, say, musical ability or fondness for knitting, what would this mean? It would mean that, in some specified population and in some specified environment, an observer in possession of information about an individual's sex would be able to make a statistically more accurate prediction as to the person's musical ability than an observer ignorant of the person's sex. The emphasis is on the word “statistically”, and let us throw in an “other things being equal” for good measure. The observer might be provided with some additional information, say on the person's education or upbringing, which would lead him to revise, or even reverse, his prediction based on sex. If females are statistically more likely than males to enjoy knitting, this does not mean that all females enjoy knitting, nor even that a majority do.
It is also fully compatible with the view that the reason females enjoy knitting is that society brings them up to enjoy knitting. If society systematically trains children without penises to knit and play with dolls, and trains children with penises to play with guns and toy soldiers, any resulting differences in male and female preferences are strictly speaking genetically determined differences! They are determined, through the medium of societal custom, by the fact of possession or non-possession of a penis, and that is determined (in a normal environment and in the absence of ingenious plastic surgery or hormone therapy) by sex chromosomes.
Obviously, on this view, if we experimentally brought up a sample of boys to play with dolls and a sample of girls to play with guns, we would expect easily to reverse the normal preferences. This might be an interesting experiment to do, for the result just might turn out to be that girls still prefer dolls and boys still prefer guns. If so, this might tell us something about the tenacity, in the face of a particular environmental manipulation, of a genetic difference. But all genetic causes have to work in the context of an environment of some kind. If a genetic sex difference makes itself felt through the medium of a sex-biased education system, it is still a genetic difference. If it makes itself felt through some other medium, such that manipulations of the education system do not perturb it, it is, in principle, no more and no less a genetic difference than in the former, education-sensitive case: no doubt some other environmental manipulation could be found which did perturb it. {13}
Human psychological attributes vary along almost as many dimensions as psychologists can measure. It is difficult in practice (Kempthorne 1978), but in principle we could partition this variation among such putative causal factors as age, height, years of education, type of education classified in many different ways, number of siblings, birth order, colour of mother's eyes, father's skill in shoeing horses, and, of course, sex chromosomes. We could also examine two-way and multi-way interactions between such factors. For present purposes the important point is that the variance we seek to explain will have many causes, which interact in complex ways. Undoubtedly genetic variance is a significant cause of much phenotypic variance in observed populations, but its effects may be overridden, modified, enhanced or reversed by other causes. Genes may modify the effects of other genes, and may modify the effects of the environment. Environmental events, both internal and external, may modify the effects of genes, and may modify the effects of other environmental events.
People seem to have little difficulty in accepting the modifiability of “environmental” effects on human development. If a child has had bad teaching in mathematics, it is accepted that the resulting deficiency can be remedied by extra good teaching the following year. But any suggestion that the child's mathematical deficiency might have a genetic origin is likely to be greeted with something approaching despair: if it is in the genes “it is written”, it is “determined” and nothing can be done about it; you might as well give up attempting to teach the child mathematics. This is pernicious rubbish on an almost astrological scale. Genetic causes and environmental causes are in principle no different from each other. Some iniuences of both types may be hard to reverse; others may be easy to reverse. Some may be usually hard to reverse but easy if the right agent is applied. The important point is that there is eo general reason for expecting genetic influences to be any more irreversible than environmental ones.
What did genes do to deserve their sinister, juggernaut-like reputation? Why do we not make a similar bogey out of, say, nursery education or confirmation classes? Why are genes thought to be so much more fixed and inescapable in their effects than television, nuns, or books? Don't blame your mates for sleeping around, ladies, it's not their fault they have been inflamed by pornographic literature! The alleged Jesuit boast, “Give me the child for his first seven years, and I'll give you the man”, may have some truth in it. Educational, or other cultural influences may, in some circumstances, be just as unmodifiable and irreversible as genes and “stars” are popularly thought to be.
I suppose part of the reason genes have become deterministic bogeys is a confusion resulting from the well-known fact of the non-inheritance of acquired characteristics. Before this century it was widely believed that experience and other acquisitions of an individual's lifetime were somehow {14} imprinted on the hereditary substance and transmitted to the children. The abandoning of this belief, and its replacement by Weismane's doctrine of the continuity of the germ plasm, and its molecular counterpart the “central dogma”, is one of the great achievements of modern biology. If we steep ourselves in the implications of Weismannian orthodoxy, there really does seem to be something juggernaut-like and inexorable about genes. They march through generations, iniueecing the form and behaviour of a succession of mortal bodies, but, except for rare and non-specific mutagenic effects, they are never influenced by the experience or environment of those bodies. The genes in me came from my four grandparents; they flowed straight through my parents to me, and nothing that my parents achieved, acquired, learned or experienced had any effect on those genes as they flowed through. Perhaps there is something a little sinister about that. But, however inexorable and undeviating the genes may be as they march down the generations, the nature of their phenotypic effects on the bodies they flow through is by no means inexorable and undeviating. If I am homozygous for a gene G, nothing save mutation can prevent my passing G on to all my children. So much is inexorable. But whether or not I, or my children, show the phenotypic effect normally associated with possession of G may depend very much on how we are brought up, what diet or education we experience, and what other genes we happen to possess. So, of the two effects that genes have on the world — manufacturing copies of themselves, and influencing phenotypes — the first is inflexible apart from the rare possibility of mutation; the second may be exceedingly flexible. I think a confusion between evolution and development is, then, partly responsible for the myth of genetic determinism.
But there is another myth complicating matters, and I have already mentioned it at the beginning of this chapter. The computer myth is almost as deep-seated in the modern mind as the gene myth. Notice that both passages I quoted contain the word “programmed”. Thus Rose sarcastically absolved promiscuous men from blame because they are genetically programmed. Gould says that if we are programmed to be what we are then these traits are ineluctable. And it is true that we ordinarily use the word programmed to indicate unthinking inflexibility, the antithesis of freedom of action. Computers and “robots” are, by repute, notoriously inflexible, carrying out instructions to the letter, even if the consequences are obviously absurd. Why else would they send out those famous million pound bills that everybody's friend's friend's cousin's acquaintance keeps receiving? I had forgotten the great computer myth, as well as the great gene myth, or I would have been more careful when I myself wrote of genes swarming “inside gigantic lumbering robots...”, and of ourselves as “survival machines — robot vehicles blindly programmed to preserve the selfish molecules known as genes” (Dawkins 1976a). These passages have been triumphantly quoted, {15} and requoted apparently from secondary and even tertiary sources, as examples of rabid genetic determinism (e.g. 'Nabi' 1981). I am not apologizing for using the language of robotics. I would use it again without hesitation. But I now realize that it is necessary to give more explanation.
From 13 years' experience of teaching it, I know that a main problem with the “selfish-gene survival machine” way of looking at natural selection is a particular risk of misunderstanding. The metaphor of the intelligent gene reckoning up how best to ensure its own survival (Hamilton 1972) is a powerful and illuminating one. But it is all too easy to get carried away, and allow hypothetical genes cognitive wisdom and foresight in planning their “strategy”. At least three out of twelve misunderstandings of kin selection (Dawkins 1979a) are directly attributable to this basic error. Time and again, non-biologists have tried to justify a form of group selection to me by, in effect, imputing foresight to genes: “The long-term interests of a gene require the continued existence of the species; therefore shouldn't you expect adaptations to prevent species extinction, even at the expense of short-term individual reproductive success?” It was in an attempt to forestall errors like this that I used the language of automation and robotics, and used the word “blindly” in referring to genetic programming. But it is, of course, the genes that are blind, not the animals they program. Nervous systems, like man-made computers, can be sufficiently complex to show intelligence and foresight.
Symons (1979) makes the computer myth explicit:
I wish to point out that Dawkins's implication — through the use of words like “robot” and “blindly” — that evolutionary theory favors determinism is utterly without foundation ... A robot is a mindless automaton. Perhaps some animals are robots (we have no way of knowing); however, Dawkins is not referring to some animals, but to all animals and in this case specifically to human beings. Now, to paraphrase Stebbing, “robot” can be opposed to “thinking being” or it can be used figuratively to indicate a person who seems to act mechanically, but there is no common usage of language that provides a meaning for the word “robot” in which it would make sense to say that all living things are robots [p. 41].
The point of the passage from Stebbing which Symons paraphrased is the reasonable one that X is a useless word unless there are some things that are not X. If everything is a robot, then the word robot doesn't mean anything useful. But the word robot has other associations, and rigid inflexibility was not the association I was thinking of. A robot is a programmed machine, and an important thing about programming is that it is distinct from, and done in advance of, performance of the behaviour itself. A computer is {16} programmed to perform the behaviour of calculating square roots, or playing chess. The relationship between a chess-playing computer and the person who programmed it is not obvious, and is open to misunderstanding. It might be thought that the programmer watches the progress of the game and gives instructions to the computer move by move. In fact, however, the programming is finished before the game begins. The programmer tries to anticipate contingencies, and builds in conditional instructions of great complexity, but once the game begins he has to keep his hands off. He is not allowed to give the computer any new hints during the course of the game. If he did he would not be programming but performing, and his entry would be disqualified from the tournament. In the work criticized by Symons, I made extensive use of the analogy of computer chess in order to explain the point that genes do not control behaviour directly in the sense of interfering in its performance. They only control behaviour in the sense of programming the machine in advance of performance. It was this association with the word robot that I wanted to invoke, not the association with mindless inflexibility. As for the mindless inflexibility association itself, it could have been justified in the days when the acme of automation was the rod and cam control system of a marine engine, and Kipling wrote “McAndrew's Hymn”:
From coupler-flange to spindle-guide I see Thy Hand, O God —
Predestination in the stride o'yon connectin'-rod.
John Calvin might ha' forged the same —
But that was 1893 and the heyday of steam. We are now well embarked on the golden age of electronics. If machines ever had associations with rigid inflexibility — and I accept that they had — it is high time they lived them down. Computer programs have now been written that play chess to International Master standard (Levy 1978), that converse and reason in correct and indefinitely complex granimatical English (Winograd 1972), that create elegant and aesthetically satisfying new proofs of mathematical theorems (Hofstadter 1979), that compose music and diagnose illness; and the pace of progress in the field shows no sign of slowing down (Evans 1979). The advanced programming field known as artificial intelligence is in a buoyant, confident state (Boden 1977). Few who have studied it would now bet against computer programs beating the strongest Grand Masters at chess within the next 10 years. From being synonymous in the popular mind with a moronicaily undeviating, jerky-limbed zombie, “robot” will one day become a byword for flexibility and rapid intelligence.
Unfortunately I jumped the gun a little in the passage quoted. When I wrote it I had just returned from an eye-opening and mind-boggling conference on the state of the art of artificial intelligence programming, and {17} I genuinely and innocently in my enthusiasm forgot that robots are popularly supposed to be inflexible idiots. I also have to apologize for the fact that, without my knowledge, the cover of the German edition of The Selfish Gene was given a picture of a human puppet jerking on the end of strings descending from the word gene, and the French edition a picture of little bowler-hatted men with clockwork wind-up keys sticking out of their backs. I have had slides of both covers made up as illustrations of what I was not trying to say.
So, the answer to Symons is that of course he was right to criticize what he thought I was saying, but of course I wasn't actually saying it (Ridley 1980a). No doubt I was partly to blame for the original misunderstanding, but I can only urge now that we put aside the preconceptions derived from common usage (“...most men don't understand computers to even the slightest degree” — Weizenbaum 1976, p. 9), and actually go and read some of the fascinating modern literature on robotics and computer intelligence (e.g. Boden 1977; Evans 1979; Hofstadter 1979).
Once again, of course, philosophers may debate the ultimate determinacy of computers programmed to behave in artificially intelligent ways, but if we are going to get into that level of philosophy many would apply the same arguments to human intelligence (Turing 1950). What is a brain, they would ask, but a computer, and what is education but a form of programming? It is very hard to give a non-supereatural account of the human brain and human emotions, feelings and apparent free will, without regarding the brain as, in some sense, the equivalent of a programmed, cybernetic machine. The astronomer Sir Fred Hoyle (1964) expresses very vividly what, it seems to me, any evolutionist must think about nervous systems:
Looking back [at evolution] I am overwhelmingly impressed by the way in which chemistry has gradually given way to electronics. It is not unreasonable to describe the first living creatures as entirely chemical in character. Although electrochemical processes are important in plants, organized electronics, in the sense of data processing, does not enter or operate in the plant world. But primitive electronics begins to assume importance as soon as we have a creature that moves around ... The first electronic systems possessed by primitive animals were essentially guidance systems, analogous logically to sonar or radar. As we pass to more developed animals we find electronic systems being used not merely for guidance but for directing the animal toward food...
The situation is analogous to a guided missile, the job of which is to intercept and destroy another missile. Just as in our modern world attack and defense become more and more subtle in their methods, so it was the case with animals. And with increasing subtlety, better and {18} better systems of electronics become necessary. What happened in nature has a close parallel with the development of electronics in modern military applications ... I find it a sobering thought that but for the tooth-and-claw existence of the jungle we should not possess our intellectual capabilities, we should not be able to inquire into the structure of the Universe, or be able to appreciate a symphony of Beethoven ... Viewed in this light, the question that is sometimes asked — can computers think? — is somewhat ironic. Here of course I mean the computers that we ourselves make out of inorganic materials. What on earth do those who ask such a question think they themselves are? Simply computers, but vastly more complicated ones than anything we have yet learned to make. Remember that our man-made computer industry is a mere two or three decades old, whereas we ourselves are the products of an evolution that has operated over hundreds of millions of years [pp. 24–26].
Others may disagree with this conclusion, although I suspect that the only alternatives to it are religious ones. Whatever the outcome of that debate, to return to genes and the main point of this chapter, the issue of determinism versus free will is just not affected one way or the other by whether or not you happen to be considering genes as causal agents rather than environmental determinants.
But, it will pardonably be said, there is no smoke without fire. Functional ethologists and “sociobiologists” must have said something to deserve being tarred with the brush of genetic determinism. Or if it is all a misunderstanding there must be some good explanation, because misunderstandings that are so widespread do not come about for no reason, even if abetted by cultural myths as powerful as the gene myth and the computer myth in unholy alliance. Speaking for myself, I think I know the reason. It is an interesting one, and it will occupy the rest of this chapter. The misunderstanding arises from the way we talk about a quite different subject, namely natural selection. Gene selectionism, which is a way of talking about evolution, is mistaken for genetic determinism, which is a point of view about development. People like me are continually postulating genes “for” this and genes for that. We give the impression of being obsessed with genes and with “genetically programmed” behaviour. Take this in conjunction with the popular myths of the Calvieistic determinacy of genes, and of “programmed” behaviour as the hallmark of jactitating Disneyland puppets, and is it any wonder that we are accused of being genetic determinists?
Why, then, do functional ethologists talk about genes so much? Because we are interested in natural selection, and natural selection is differential survival of genes. If we are to so much as discuss the possibility of a behaviour pattern's evolving by natural selection, we have to postulate genetic {19} variation with respect to the tendency or capacity co perform that beoaviour pattern. This is not to say that there necessarily is such genetic variation for any particular behaviour pattern, only that there must have been genetic variation in the past if we are to treat the behaviour pattern as a Darwinian adaptation. Of course the behaviour pattern may not be a Darwinian adaptation, in which case the argument will not apply.
Incidentally, I should defend my usage of “Darwinian adaptation” as synonymous with “adaptation produced by natural selection”, for Gould and Lewontin (1979) have recently emphasized, with approval, the “pluralistic” character of Darwin's own thought. It is indeed true that, especially towards the end of his life, Darwin was driven by criticisms, which we can now see to be erroneous, to make some concessions to “pluralism”: he did not regard natural selection as the only important driving force in evolution. As the historian R. M. Young (1971) has sardonically put it, “...by the sixth edition the book was mistitled and should have read On the Origin of Species by Means of Natural Selection and All Sorts of Other Things”. It is, therefore, arguably incorrect to use “Darwinian evolution” as synonymous with “evolution by natural selection”. But Darwinian adaptation is another matter. Adaptation cannot be produced by random drift, or by any other realistic evolutionary force that we know of save natural selection. It is true that Darwin's pluralism did fleetingly allow for one other driving force that might, in principle, lead to adaptation, but that driving force is inseparably linked with the name of Lamarck, not of Darwin. “Darwinian adaptation” could not sensibly mean anything other than adaptation produced by natural selection, and I shall use it in this sense. In several other places in this book (e.g. in Chapters 3 and 6), we shall resolve apparent disputes by drawing a distinction between evolution in general, and adaptive evolution in particular. The fixation of neutral mutations, for instance, can be regarded as evolution, but it is not adaptive evolution. If a molecular geneticist interested in gene substitutions, or a palaeontologist interested in major trends, argues with an ecologist interested in adaptation, they are likely to find themselves at cross-purposes simply because each of them emphasizes a different aspect of what evolution means.
“Genes for conformity, xenophobia, and aggressiveness are simply postulated for humans because they are needed for the theory, not because any evidence for them exists” (Lewontin 1979b). This is a fair criticism of E. O. Wilson, but not a very damning one. Apart from possible political repercussions which might be unfortunate, there is nothing wrong with cautiously speculating about a possible Darwinian survival value of xenophobia or any other trait. And you cannot begin to speculate, however cautiously, about the survival value of anything unless you postulate a genetic basis for variation in that thing. Of course xenophobia may not vary genetically, and of course xenophobia may not be a Darwinian adaptation, {20} but we can't even discuss the possibility of its being a Darwinian adaptation unless we postulate a genetic basis for it. Lewontin himself has expressed the point as well as anybody: “In order for a trait to evolve by natural selection it is necessary that there be genetic variation in the population for such a trait” (Lewontin 1979b). And “genetic variation in the population for” a trait X is exactly what we mean when we talk, for brevity, of “a gene for” X.
Xenophobia is controversial, so consider a behaviour pattern that nobody would fear to regard as a Darwinian adaptation. Pit-digging in antlions is obviously an adaptation to catch prey. Antlions are insects, neuropteran larvae with the general appearance and demeanour of monsters from outer space. They are “sit and wait” predators who dig pits in soft sand which trap ants and other small walking insects. The pit is a nearly perfect cone, whose sides slope so steeply that prey cannot climb out once they have fallen in. The antlion sits just under the sand at the bottom of the pit, where it lunges with its horror-film jaws at anything that falls in.
Pit-digging is a complex behaviour pattern. It costs time and energy, and satisfies the most exacting criteria for recognition as an adaptation (Williams 1966; Curio 1973). It must, then, have evolved by natural selection. How might this have happened? The details don't matter for the moral I want to draw. Probably an ancestral antlion existed which did not dig a pit but simply lurked just beneath the sand surface waiting for prey to blunder over it. Indeed some species still do this. Later, behaviour leading to the creation of a shallow depression in the sand probably was favoured by selection because the depression marginally impeded escaping prey. By gradual degrees over many generations the behaviour changed so that what was a shallow depression became deeper and wider. This not only hindered escaping prey but also increased the catchment area over which prey might stumble in the first place. Later still the digging behaviour changed again so that the resulting pit became a steep-sided cone, lined with fine, sliding sand so that prey were unable to climb out.
Nothing in tie previous paragraph is contentious or controversial. It will be regarded as legitimate speculation about historical events that we cannot see directly, and it will probably be thought plausible. One reason why it will be accepted as uncontroversial historical speculation is that it makes no mention of genes. But my point is that none of that history, nor any comparable history, could possibly have been true unless there was genetic variation in the behaviour at every step of the evolutionary way. Pit-digging in antlions is only one of the thousands of examples that I could have chosen. Unless natural selection has genetic variation to act upon, it cannot give rise to evolutionary change. It follows that where you find Darwinian adaptation there must have been genetic variation in the character concerned.
Nobody has ever done a genetic study of pit-digging behaviour in antlions (J. Lucas, personal communication). There is no need to do one, if all we {21} want to do is satisfy ourselves of the sometime existence of genetic variation in the behaviour pattern.. It is sufficient that we are convinced that it is a Darwinian adaptation (if you are not convinced that pit-digging is such an adaptation, simply substitute any example of which you are convinced).
I spoke of the sometime existence of genetic variation. This was because it is quite likely that, were a genetic study to be mounted of antlions today, no genetic variation would be found. It is in general to be expected that, where there is strong selection in favour of some trait, the original variation on which selection acted to guide the evolution of the trait will have become used up. This is the familiar “paradox” (it is not really very paradoxical when we think about it carefully) that traits under strong selection tend to have low heritability (Falconer 1960); “...evolution by natural selection destroys the genetic variance on which it feeds” (Lewontin 1979b). Functional hypotheses frequently concern phenotypic traits, like possession of eyes, which are all but universal in the population, and therefore without contemporary genetic variation. When we speculate about, or make models of, the evolutionary production of an adaptation, we are necessarily talking about a time when there was appropriate genetic variation. We are bound, in such discussions, to postulate, implicitly or explicitly, genes “for” proposed adaptations.
Some may balk at treating “a genetic contribution to variation in X” as equivalent to “a gene or genes for X”. But this is a routine genetic practice, and one which close examination shows to be almost inevitable. Other than at the molecular level, where one gene is seen directly to produce one protein chain, geneticists never deal with units of phenotype as such. Rather, they always deal with differences. When a geneticist speaks of a gene “for” red eyes in Drosophila, he is not speaking of the cistron which acts as template for the synthesis of the red pigment molecule. He is implicitly saying: there is variation in eye colour in the population; other things being equal, a fly with this gene is more likely to have red eyes than a fly without the gene. That is all that we ever mean by a gene “for” red eyes. This happens to be a morphological rather than a behavioural example, but exactly the same applies to behaviour. A gene “for” behaviour X is a gene “for” whatever morphological and physiological states tend, to produce that behaviour.
A related point is that the use of single-locus models is just a conceptual convenience, and this is true of adaptive hypotheses in exactly the same way as it is true of ordinary population genetic models. When we use single-gene language in our adaptive hypotheses, we do not intend to make a point about single-gene models as against multi-gene models. We are usually making a point about gene models as against non-gene models, for example as against “good of the species” models. Since it is difficult enough convincing people that they ought to think in genetic terms at all rather than in terms of, say, the good of the species, there is no sense in making things even more {22} difficult by trying to handle the complexities of many loci at the outset. What Lloyd (1979) calls the OGAM (one gene analysis model) is, of course, not the last word in genetic accuracy. Of course we shall eventually have to face up to multi-locus complexity. But the OGAM is vastly preferable to modes of adaptive reasoning that forget about genes altogether, and this is the only point I am trying to make at present.
Similarly we may find ourselves aggressively challenged to substantiate our “claims” of the existence of “genes for” some adaptation in which we are interested. But this challenge, if it is a real challenge at all, should be directed at the whole of the neo-Darwinian “modern synthesis” and the whole of population genetics. To phrase a functional hypothesis in terms of genes is to make no strong claims about genes at all: it is simply to make explicit an assumption which is inseparably built into the modern synthesis, albeit it is sometimes implicit rather than explicit.
A few workers have, indeed, flung just such a challenge at the whole neo-Darwinian modern synthesis, and have claimed not to be neo-Darwinians. Goodwin (1979) in a published debate with Deborah Charlesworth and others, said, “...neo-Darwinism has an incoherence in it ... we are not given any way of generating phenotypes from genotypes in neo-Darwinism. Therefore the theory is in this respect defective”. Goodwin is, of course, quite right that development is terribly complicated, and we don't yet understand much about how phenotypes are generated. But that they are generated, and that genes contribute significantly to their variation are incontrovertible facts, and those facts are all we need in order to make neo-Darwinism coherent. Goodwin might just as well say that, before Hodgkin and Huxley worked out how the nerve impulse fired, we were not entitled to believe that nerve impulses controlled behaviour. Of course it would be nice to know how phenotypes are made but, while embryologists are busy finding out, the rest of us are entitled by the known facts of genetics to carry on being neo-Darwinians, treating embryonic development as a black box. There is no” competing theory that has even a remote claim to be called coherent.
It follows from the fact that geneticists are always concerned with phenotypic differences that we need not be afraid of postulating genes with indefinitely complex phenotypic effects, and with phenotypic effects that show themselves only in highly complex developmental conditions. Together with Professor John Maynard Smith, I recently took part in a public debate with two radical critics of'sociobiology”, before an audience of students. At one time in the discussion we were trying to establish that to talk of a gene “for X” is to make no outlandish claim, even where X is a complex, learned behaviour pattern. Maynard Smith reached for a hypothetical example and came up with a “gene for skill in tying shoelaces”. Pandemonium broke loose at this rampant genetic determinism! The air was thick with the unmistakable sound of worst suspicions being gleefully confirmed. Delightedly {23} sceptical cries drowned the quiet and patient explanation of just what a modest claim is being made whenever one postulates a gene for, say, skill in tying shoelaces. Let me explain the point with the aid of an even more radical-sounding yet truly innocuous thought experiment (Dawkins 1981).
Reading is a learned skill of prodigious complexity, but this provides no reason in itself for scepticism about the possible existence of a gene for reading. All we would need in order to establish the existence of a gene for reading is to discover a gene for not reading, say a gene which induced a brain lesion causing specific dyslexia. Such a dyslexic person might be normal and intelligent in all respects except that he could not read. No geneticist woiild be particularly surprised if this type of dyslexia turned out to breed true in some Mendelian fashion. Obviously, in this event, the gene would only exhibit its effect in an environment which included normal education. In a prehistoric environment it might have had no detectable effect, or it might have had some different effect and have been known to cave-dwelling geneticists as, say, a gene for inability to read animal footprints. In our educated environment it would properly be called a gene “for” dyslexia, since dyslexia would be its most salient consequence. Similarly, a gene which caused total blindness would also prevent reading, but it would not usefully be regarded as a gene for not reading. This is simply because preventing reading would not be its most obvious or debilitating phenotypic effect.
Returning to our gene for specific dyslexia, it follows from the ordinary conventions of genetic terminology that the wild-type gene at the same locus, the gene that the rest of the population has in double dose, would properly be called a gene “for reading”. If you object to that, you must also object to our speaking of a gene for tallness in Mendel's peas, because the logic of the terminology is identical in the two cases. In both cases the character of interest is a difference, and in both cases the difference only shows itself in some specified environment. The reason why something so simple as a one gene difference can have such a complex effect as to determine whether or not a person can learn to read, or how good he is at tying shoelaces, is basically as follows. However complex a given state of the world may be, the difference between that state of the world and some alternative state of the world may be caused by something extremely simple.
The point I made using antlions is a general one. I could have used any real or purported Darwinian adaptation whatsoever. For further emphasis I shall use one more example. Tinbergen et al. (1962) investigated the adaptive significance of a particular behaviour pattern in black-headed gulls (Lams ridibundus), eggshell removal. Shortly after a chick hatches, the parent bird grasps the empty eggshell in the bill and removes it from the vicinity of the nest. Tinbergen and his colleagues considered a number of {24} possible hypotheses about the survival value of this behaviour pattern. For instance they suggested that the empty eggshells might serve as breeding grounds for harmful bacteria, or the sharp edges might cut the chicks. But the hypothesis for which they ended up finding evidence was that the empty eggshell serves as a conspicuous visual beacon summoning crows and other predators of chicks or eggs to the nest. They did ingenious experiments, laying out artificial nests with and without empty eggshells, and showed that eggs accompanied by empty eggshells were, indeed, more likely to be attacked by crows than eggs without empty eggshells by their side. They concluded that natural selection had favoured eggshell removal behaviour of adult gulls, because past adults who did not do it reared fewer children.
As in the case of antlion digging, nobody has ever done a genetic study of eggshell removal behaviour in black-headed gulls. There is no direct evidence that variation in tendency to remove empty eggshells breeds true. Yet clearly the assumption that it does, or once did, is essential for the Tinbergen hypothesis. The Tinbergen hypothesis, as normally phrased in gene-free language, is not particularly controversial. Yet it, like all the rival functional hypotheses that Tinbergen rejected, rests fundamentally upon the assumption that once upon a time there must have been gulls with a genetic tendency to remove eggshells, and other gulls with a genetic tendency not to remove them, or to be less likely to remove them. There must have been genes for removing eggshells.
Here I must enter a note of caution. Suppose we actually did a study of the genetics of eggshell removal behaviour in modern gulls. It would be a behaviour-geneticist's dream to find a simple Mendelian mutation which radically altered the behaviour pattern, perhaps abolished the behaviour altogether. By the argument given above, this mutant would truly be a gene “for” not removing eggshells, and, by definition, its wild-type allele would have to be called a gene for eggshell removal. But now comes the note of caution. It most definitely does not follow that this particular locus “for” eggshell removal was one of the ones upon which natural selection worked during the evolution of the adaptation. On the contrary, it seems much more probable that a complex behaviour pattern like eggshell removal must have been built up by selection, on a large number of loci, each having a small effect in interaction with the others. Once the behaviour complex had been built up, it is easy to imagine a single major mutation arising, whose effect is to destroy it. Geneticists perforce must exploit the genetic variation available for them to study. They also believe that natural selection must have worked on similar genetic variation in wreaking evolutionary change. But there is no reason for them to believe that the loci controlling modern variation in an adaptation were the very same loci at which selection acted in building up the adaptation in the first place.
Consider the most famous example of single gene control of complex {25} behaviour, the case of Rothenbuhler's (1964) hygienic bees. The point of using this example is that it illustrates well how a highly complex behaviour difference can be due to a single gene difference. The hygienic behaviour of the Brown strain of honeybees involves the whole neuromuscular system, but the fact that they perform the behaviour whereas Van Scoy bees do not is, according to Rothenbuhler's model, due to differences at two loci only. One locus determines the uncapping of cells containing diseased brood, the other locus determines the removing of diseased brood after uncapping. It would be possible, therefore, to imagine a natural selection in favour of uncapping behaviour and a natural selection in favour of removing behaviour, meaning selection of the two genes versus their respective alleles. But the point I am making here is that, although that could happen, it is not likely to be very interesting evolutionarily. The modern uncapping gene and the modern removing gene may very well not have been involved in the original natural selection process that steered the evolutionary putting together of the behaviour.
Rothenbuhler observed that even Van Scoy bees sometimes perform hygienic behaviour. They are just quantitatively much less likely to do so than are Brown bees. It is likely, therefore, that both Brown and Van Scoy bees have hygienic ancestors, and both have in their nervous systems the machinery of uncapping and removing behaviour: it is just that Van Scoy bees have genes that prevent them from turning the machinery on. Presumably if we went back even further in time we should find an ancestor of all modern bees which not only was not hygienic itself but had never had a hygienic ancestor. There must have been an evolutionary progression building up the uncapping and removing behaviour from nothing, and this evolutionary progression involved the selection of many genes which are now fixed in both the Brown and the Van Scoy strains. So, although the uncapping and the removing genes of the Brown strain really are rightly called genes for uncapping and removing, they are defined as such only because they happen to have alleles whose effect is to prevent the behaviour from being performed. The mode of action of these alleles could be boringly destructive. They might simply cut some vital link in the neural machinery. I am reminded of Gregory's (1961) vivid illustration of the perils of making inferences from ablation experiments on the brain; “...the removal of any of several widely spaced resistors may cause a radio set to emit howls, but it does not follow that howls are immediately associated with these resistors, or indeed that the causal relation is anything but the most indirect. In particular, we should not say that the function of the resistors in the normal circuit is to inhibit howling. Neurophysiologists, when faced with a comparable situation, have postulated "suppressor regions"”.
This consideration seems to me to be a reason for caution, not a reason for rejecting the whole genetic theory of natural selection! Never mind if living {26} geneticists are debarred from studying the particular loci at which selection in the past gave rise to the original evolution of interesting adaptations. It is too bad if geneticists usually are forced to concentrate on loci that are convenient rather than evolutionarily important. It is still true that the evolutionary putting together of complex and interesting adaptation consisted in the replacement of genes by their alleles.
This argument can contribute tangentially to the resolution of a fashionable contemporary dispute, by helping to put the issue in perspective. It is now highly, indeed passionately, controversial whether there is significant genetic variation in human mental abilities. Are some of us genetically brainier than others? What we mean by “brainy” is also highly contentious, and rightly so. But I suggest that, by any meaning of the term, the following propositions cannot be denied, (1) There was a time when our ancestors were less brainy than we are. (2) Therefore there has been an increase in-braininess in our ancestral lineage. (3) That increase came about through evolution, probably propelled by natural selection. (4) Whether propelled by selection or not, at least part of the evolutionary change in phenotype reflected an underlying genetic change: allele replacement took place and consequently mean mental ability increased over generations. (5) By definition therefore, at least in the past, there must have been significant genetic variation in braininess within the human population. Some people were genetically clever in comparison with their contemporaries, others were genetically relatively stupid.
The last sentence may engender a frisson of ideological disquiet, yet none of my five propositions could be seriously doubted, nor could their logical sequence. The argument works for brain size, but it equally works for any behavioural measure of cleverness we care to dream up. It does not depend on simplistic views of human intelligence as being a one-dimensional scalar quantity. The fact that intelligence is not a simple scalar quantity, important as that fact is, is simply irrelevant. So is the difficulty of measuring intelligence in practice. The conclusion of the previous paragraph is inevitable, provided only that we are evolutionists who agree to the proposition that once upon a time our ancestors were less clever (by whatever criterion) then we are. Yet in spite of all that, it still does not follow that there is any genetic variation in mental abilities left in the human population today: the genetic variance might all have been used up by selection. On the other hand it might not, and my thought experiment shows at least the inadvisability of dogmatic and hysterical opposition to the very possibility of genetic variation in human mental abilities. My own opinion, for what it is worth, is that even if there is such genetic variation in modern human populations, to base any policy on it would be illogical and wicked.
The existence of a Darwinian adaptation, then, implies the sometime existence of genes for producing the adaptation. This is not always made {27} explicit. It is always possible to talk about the natural selection of a behaviour pattern in two ways. We can either talk about individuals with a tendency to perform the behaviour pattern being “fitter” than individuals with a less strongly developed tendency. This is the now fashionable phraseology, within the paradigm of the “selfish organism” and the “central theorem of sociobiology”. Alternatively, and equivalently, we can talk directly of genes for performing the behaviour pattern surviving better than their alleles. It is always legitimate to postulate genes in any discussion of Darwinian adaptation, and it will be one of my central points in this book that it is often positively beneficial to do so. Objections, such as I have heard made, to the “unnecessary geneticizing” of the language of functional ethology, betray a fundamental failure to face up to the reality of what Darwinian selection is all about.
Let me illustrate this failure by another anecdote. I recently attended a research seminar given by an anthropologist. He was trying to interpret the incidence among various human tribes of a particular mating system (it happened to be polyandry) in terms of a theory of kin selection. A kin selection theorist can make models to predict the conditions under which we would expect to find polyandry. Thus, on one model applied to Tasmanian native hens (Maynard Smith & Ridpath 1972), the population sex ratio would need to be male-biased, and partners would need to be close kin, before a biologist would predict polyandry. The anthropologist sought to show that his polyandrous human tribes lived under such conditions, and, by implication, that other tribes showing the more normal patterns of monogamy or polygyny lived under different conditions.
Though fascinated by the information he presented, I tried to warn him of some difficulties in his hypothesis. I pointed out that the theory of kin selection is fundamentally a genetic theory, and that kin-selected adaptations to local conditions had to come about through the replacement of alleles by other alleles, over generations. Had his polyandrous tribes been living, I asked, under their current peculiar conditions for long enough — enough generations — for the necessary genetic replacement to have taken place? Was there, indeed, any reason to believe that variations in human mating systems are under genetic control at all?
The speaker, supported by many of his anthropological colleagues in the seminar, objected to my dragging genes into the discussion. He was not talking about genes, he said, but about a social behaviour pattern. Some of his colleagues seemed uncomfortable with the very mention of the four-letter word “gene”. I tried to persuade him that it was he who had “dragged genes in” to the discussion although, to be sure, he had not mentioned the word gene in his talk. That is exactly the point I am trying to make. You cannot talk about kin selection, or any other form of Darwinian selection, without dragging genes in, whether you do so explicitly or not. By even speculating {28} about Me selection as an explanation of differences in tribal mating systems, my anthropologist friend was implicitly dragging genes into the discussion. It is a pity he did not make it explicit, because he would then have realized what formidable difficulties lay in the path of his kin selection hypothesis: either his polyandrous tribes had to have been living, in partial genetic isolation, under their peculiar conditions for a large number of centuries, or natural selection had to have favoured the universal occurrence of genes programming some complex “conditional strategy”. The irony is that, of all the participants in that seminar on polyandry, it was I who was advancing the least “genetically deterministic” view of the behaviour under discussion. Yet because I insisted on making the genetic nature of the kin selection hypothesis explicit, I expect I appeared to be characteristically obsessed with genes, a “typical genetic deteraiinist”. The story illustrates well the main message of this chapter, that frankly facing up to the fundamental genetic nature of Darwinian selection is all too easily mistaken for an unhealthy preoccupation with hereditarian interpretations of ontogenetic development.
The same prejudice against explicit mention of genes where one can get away with an individual-level circumlocution is common among biologists. The statement, “genes for performing behaviour X are favoured over genes for not performing X” has a vaguely naive and unprofessional ring to it. What evidence is there for such genes? How dare you conjure up ad hoc genes simply to satisfy your hypothetical convenience! To say “individuals that perform X are fitter than individuals that do not perform X” sounds much more respectable. Even if it is not known to be true, it will probably be accepted as a permissible speculation. But the two sentences are exactly equivalent in meaning. The second one says nothing that the first does not say more clearly. Yet if we recognize this equivalence and talk explicitly about genes “for” adaptations, we run the risk of being accused of “genetic determinism”. I hope I have succeeded in showing that this risk results from nothing more than misunderstanding. A sensible and unexceptionable way of thinking about natural selection — “gene selectionism” — is mistaken for a strong belief about development — “genetic determinism”. Anyone who thinks clearly about the details of how adaptations come into being is almost bound to think, implicitly if not explicitly, about genes, albeit they may be hypothetical genes. As I shall show in this book, there is much to be said for making the genetic basis of Darwinian functional speculations explicit rather than implicit. It is a good way of avoiding certain tempting errors of reasoning (Lloyd 1979). In doing this we may give the impression, entirely for the wrong reason, of being obsessed with genes and all the mythic baggage that genes carry in the contemporary journalistic consciousness. But determinism, in the sense of an inflexible, tramline-following ontogeny, is, or should be, a thousand miles from our thoughts. Of course, individual sociobiologists may or may not be genetic determieists. They may be {29} Rastafarians, Shakers or Marxists. But their private opinions on genetic determinism, like their private opinions on religion, have nothing to do with the fact that they use the language of “genes for behaviour” when talking about natural selection.
A large part of this chapter has been based on the assumption that a biologist might wish to speculate on the Darwinian “function” of behaviour patterns. This is not to say that all behaviour patterns necessarily have a Darwinian function. It may be that there is a large class of behaviour patterns which are selectively neutral or deleterious to their performers, and cannot usefully be regarded as the products of natural selection. If so, the arguments of this chapter do not apply to them. But it is legitimate to say “I am interested in adaptation. I don't necessarily think all behaviour patterns are adaptations, but I want to study those behaviour patterns that are adaptations”. Similarly, to express a preference for studying vertebrates rather than invertebrates does not commit us to the belief that all animals are vertebrates. Given that our field of interest is adaptive behaviour, we cannot talk about the Darwinian evolution of the objects of interest without postulating a genetic basis for them. And to use “a gene for X” as a convenient way of talking about “the genetic basis of X”, has been standard practice in population genetics for over half a century.
The question of how large is the class of behaviour patterns that we can consider to be adaptations is an entirely separate question. It is the subject of the next chapter.
| << | {30} | >> |
In one way or another, this book is largely preoccupied with the logic of Darwinian explanations of function. Bitter experience warns that a biologist who shows a strong interest in functional explanation is likely to be accused, sometimes with a passion that startles those more accustomed to scientific than ideological debate (Lewontin 1977), of believing that all animals are perfectly optimal — accused of being an “adaptationist” (Lewontin 1979a, b; Gould & Lewontin 1979). Adaptationism is defined as “that approach to evolutionary studies which assumes without further proof that all aspects of the morphology, physiology and behavior of organisms are adaptive optimal solutions to problems” (Lewontin 1979b). In the first draft of this chapter I expressed doubts that anyone was truly an adaptationist in the extreme sense, but I have recently found the following quotation from, ironically enough, Lewontin himself: “That is the one point which I think all evolutionists are agreed upon, that it is virtually impossible to do a better job than an organism is doing in its own environment” (Lewontin 1967). Lewontin has since, it seems, travelled his road to Damascus, so it would be unfair to use him as my adaptationist spokesman. indeed together with Gould he has, in recent years, been one of the most articulate and forceful critics of adaptationism. As my representative adaptationist I take A. J. Cain, who has remained (Cain 1979) consistently true to the views expressed in his trenchant and elegant paper on “The perfection of animals”.
Writing as a taxonomist, Cain (1964) is coecereed to attack the traditional dichotomy between “functional” characters, which by implication are not reliable taxonomic indicators, and “ancestral” characters which are. Cain argues forcefully that ancient “groundplan” characters, like the pentadactyl limb of tetrapods and the aquatic phase of amphibians, are there because they are functionally useful, rather than because they are inescapable historical legacies as is often implied. If one of two groups “is in any way more primitive than the other, then its primitiveness must in itself be an {31} adaptation to some less specialized mode of life which it can pursue successfully; it cannot be merely a sign of inefficiency” (p. 57). Cain makes a similar point about so-called trivial characters, criticizing Darwin for being too ready, under the at first sight surprising influence of Richard Owen, to concede functionlessness: “No one will suppose that the stripes on the whelp of a lion, or the spots on the young blackbird, are of any use to these animals...” Darwin's remark must sound foolhardy today even to the most extreme critic of adaptationism. Indeed, history seems to be on the side of the adaptationists, in the sense that in particular instances they have confounded the scoffers again and again. Cain's own celebrated work, with Sheppard and their school, on the selection pressures maintaining the banding polymorphism in the snail Cepaea nemoralis may have been partly provoked by the fact that “it had been confidently asserted that it could not matter to a snail whether it had one band on its shell or two” (Cain, p. 48). “But perhaps the most remarkable functional interpretation of a "trivial" character is given by Manton's work on the diplopod Polyxenus, in which she has shown that a character formerly described as an "ornament" (and what could sound more useless?) is almost literally the pivot of the animal's life” (Cain, p. 51).
Adaptationism as a working hypothesis, almost as a faith, has undoubtedly been the inspiration for some outstanding discoveries, von Frisch (1967), in defiance of the prestigious orthodoxy of von Hess, conclusively demonstrated colour vision in fish and in honeybees by controlled experiments. He was driven to undertake those experiments by his refusal to believe that, for example, the colours of flowers were there for no reason, or simply to delight men's eyes. This is, of course, not evidence for the validity of adaptationist faith. Each question must be tackled afresh, on its merits.
Wenner (1971) performed a valuable service in questioning von Frisch's dance language hypothesis, since he provoked J. L. Gould's (1976) brilliant confirmation of von Frisch's theory. If Wenner had been more of an adaptationist Gould's research might never have been done, but Wenner would also not have allowed himself to be so blithely wrong. Any adaptationist, while perhaps conceding that Wenner had usefully exposed lacunae in von Frisch's original experimental design, would instantly have jumped, with Lindauer (1971), on the fundamental question of why bees dance at all. Wenner never denied that they dance, nor that the dance contained all the information about the direction and distance of food that von Frisch claimed. All he denied was that other bees used the dance information. An adaptationist could not have rested happy with the idea of animals performing such a time-consuming, and above all complex and statistically improbable, activity for nothing. Adaptationism cuts both ways, however. I am now delighted that Gould did his clinching experiments, and it is entirely to my discredit that, even in the unlikely event of my having {32} been ingenious enough to think of them, I would have been too adaptationist to have bothered. I just knew Wenner was wrong (Dawkins 1969)!
Adaptationist thinking, if not blind conviction, has been a valuable stimulator of testable hypotheses in physiology. Barlow's (1961) recognition of the overwhelming functional need in sensory systems to reduce redundancy in input led him to a uniquely coherent understanding of a variety of facts about sensory physiology. Analogous functional reasoning can be applied to the motor system, and to hierarchical systems of organization generally (Dawkins 1976b; Hailman 1977). Adaptationist conviction cannot tell us about physiological mechanism. Only physiological experiment can do that. But cautious adaptationist reasoning can suggest which of many possible physiological hypotheses are most promising and should be tested first.
I have tried to show that adaptationism can have virtues as well as faults. But this chapter's main purpose is to list and classify constraints on perfection, to list the main reasons why the student of adaptation should proceed with caution. Before coming to my list of six constraints on perfection, I should deal with three others that have been proposed, but which I find less persuasive. Taking, first, the modern controversy among biochemical geneticists about “neutral mutations”, repeatedly cited in critiques of adaptationism, it is simply irrelevant. If there are neutral mutations in the biochemists' sense, what this means is that any change in polypeptide structure which they induce has no effect on the enzymatic activity of the protein. This means that the neutral mutation will not change the course of embryonic development, will have no phenotypic effect at all, as a whole-organism biologist would understand phenotypic effect. The biochemical controversy over neutralism is concerned with the interesting and important question of whether all gene substitutions have phenotypic effects. The adaptationism controversy is quite different. It is concerned with whether, given that we are dealing with a phenotypic effect big enough to see and ask questions about, we should assume that it is the product of natural selection. The biochemist's “neutral mutations” are more than neutral. As far as those of us who look at gross morphology, physiology and behaviour are concerned, they are not mutations at all. It was in this spirit that Maynard Smith (1976b) wrote: “I interpret "rate of evolution" as a rate of adaptive change. In this sense, the substitution of a neutral allele would not constitute evolution...” If a whole-organism biologist sees a genetically determined difference among phenotypes, he already knows he cannot be dealing with neutrality in the sense of the modern controversy among biochemical geneticists.
He might, nevertheless, be dealing with a neutral character in the sense of an earlier controversy (Fisher & Ford 1950; Wright 1951). A genetic difference could show itself at the phenotypic level, yet still be selectively {33} neutral. But mathematical calculations such as those of Fisher (1930b) and Haldane (1932a) show how unreliable human subjective judgement can be on the “obviously trivial” nature of some biological characters. Haldane, for example, showed that, with plausible assumptions about a typical population, a selection pressure as weak as 1 in 1000 would take only a few thousand generations to push an initially rare mutation to ixation, a small time by geological standards. It appears that, in the controversy referred to above, Wright was misunderstood (see below). Wright (1980) was embarrassed at finding the idea of evolution of nonadaptive characters by genetic drift labelled the “Sewall Wright effect”, “not only because others had previously advanced the same idea, but because I myself had strongly rejected it from the first (1929), stating that pure random drift leads "inevitably to degeneration and extinction"”. I have. attributed apparent nonadaptive taxonomic differences to pleiotropy, where not merely ignorance of an adaptive significance”. Wright was in fact showing how a subtle mixture of drift and selection can produce adaptations superior to the products of selection alone (see pp. 39–40).
A second suggested constraint on perfection concerns allometry (Huxley 1932): “In cervine deer, antler size increases more than proportionately to body size ... so that larger deer have more than proportionately large antlers. It is then unnecessary to give a specifically adaptive reason for the extremely large antlers of large deer” (Lewontin 1979b). Well, Lewontin has a point here, but I would prefer to rephrase it. As it stands it suggests that the allometric constant is constant in a God-given immutable sense. But constants on one time scale can be variables on another. The allometric constant is a parameter of embryonic development. Like any other such parameter it may be subject to genetic variation and therefore it may change over evolutionary time (Clutton-Brock & Harvey 1979). Lewontin's remark turns out to be analogous to the following: all primates have teeth; this is just a plain fact about primates, and it is therefore unnecessary to give a specifically adaptive reason for the presence of teeth in primates. What he probably meant to say is something like the following.
Deer have evolved a developmental mechanism such that growth of antlers relative to body size is allometric with a particular constant of allometry. Very probably the evolution of this allometric system of development occurred under the influence of selection pressures having nothing to do with the social function of antlers: probably it was conveniently compatible with pre-existing developmental processes in a way which we shall not understand until we know more about the biochemical and cellular details of embryology. Maybe ethological consequences of the extra large antlers of large deer exert a selective effect, but this selection pressure is likely to be swamped in importance by other selection pressures concerned with concealed internal embryological details. {34}
Williams (1966, p. 16) invoked allometry in the service of a speculation about the selection pressures leading to increased brain size in man. He suggested that the prime focus of selection was on early teachability, at an elementary level, of children. “The resulting selection for acquiring verbal facility as early as possible might have produced, as an allometric effect on cerebral development, populations in which an occasional Leonardo might arise”. Williams, however, did not see allometry as a weapon against the use of adaptive explanations. One feels that he was rightly less loyal to his particular theory of cerebral hypertrophy than to the general principle enunciated in his concluding rhetorical question: “Is it not reasonable to anticipate that our understanding of the human mind would be aided greatly by knowing the purpose for which it was designed?”
What has been said of allometry applies also to pleiotropy, the possession by one gene of more than one phenotypic effect. This is the third of the suggested constraints on perfection that I want to get out of the way before embarking on my main list. It has already been mentioned in my quotation from Wright. A possible source of confusion here is that pleiotropy has been used as a weapon by both sides in this debate, if indeed it is a real debate. Fisher (1930b) reasoned that it was unlikely that any one of a gene's phenotypic effects was neutral, so how much more unlikely was it that all of a gene's pleiotropic effects could be neutral. Lewontin (1979b), on the other hand, remarked that “many changes in characters are the result of pleiotropic gene action, rather than the direct result of selection on the character itself. The yellow color of the Malpighian tubules of an insect cannot itself be the subject of natural selection since that color can never be seen by any organism. Rather it is the pleiotropic consequence of red eye pigment metabolism, which may be adaptive”. There is no real disagreement here. Fisher was talking of the selective effects on a genetic mutation, Lewontie of selective effects on a phenotypic character; it is the same distinction, indeed, as I was making in discussing neutrality in the biochemical geneticists' sense.
Lewontin's point about pleiotropy is related to another one which I shall come on to below, about the problem of defining what he calls the natural “suture lines”, the “phenotypic units” of evolution. Sometimes the dual effects of a gene are in principle inseparable; they are different views of the same thing, just as Everest used to have two names depending on which side it was seen from. What a biochemist sees as an oxygen-carrying molecule may be seen by an ethologist as red coloration. But there is a more interesting kind of pleiotropy in which the two phenotypic effects of a mutation are separable. The phenotypic effect of any gene (versus its alleles) is not a property of the gene alone, but also of the embryological context in which it acts. This allows abundant opportunities for the phenotypic effects of one mutation to be modified by others, and is the basis of such respected ideas as Fisher's {35} (1930a) theory of the evolution of dominance, the Medawar (1952) and Williams (1957) theories of senescence, and Hamilton's (1967) theory of Y-chrotnosome inertness. In the present connection, if a mutation has one beneficial effect and one harmful one, there is no reason why selection should not favour modifier genes that detach the two phenotypic effects, or that reduce the harirful effect while enhancing the beneficial one. As in the case of allometry, Lewontin took too static a view of gene action, treating pleiotropy as if it was a property of the gene rather than of the interaction between the gene and its (modifiable) embryological context.
This brings me to my own critique of naive adaptationism, my own list of constraints on perfection, a list which has much in common with those of Lewontin and Cain, and those of Maynard Smith (1978b), Oster and Wilson (1978), Williams (1966), Curio (1973) and others. There is, indeed, much more agreement than the polemical tone of recent critiques would suggest. I shall not be concerned with particular cases, except as examples. As Cain and Lewontin both stress, it is not of general interest to challenge our ingenuity in dreaming up possible advantages of particular strange things that animals do. Here we are interested in the more general question of what the theory of natural selection entitles us to expect. My first constraint on perfection is an obvious one, mentioned by most writers on adaptation.
The animal we are looking at is very probably out of date, built under the influence of genes that were selected in some earlier era when conditions were different. Maynard Smith (1976b) gives a quantitative measure of this effect, the iag load”. He (Maynard Smith 1978b) cites Nelson's demonstration that gannets, who normally lay only one egg, are quite capable of successfully incubating and rearing two if an extra one is experimentally added. Obviously an awkward case for the Lack hypothesis on optimal clutch size, and Lack himself (1966) was not slow to use the “time-lag” escape route. He suggested, entirely plausibly, that the gannet clutch size of one. egg evolved during a time when food was less plentiful, and that there had not yet been time for them to evolve to meet the changed conditions.
Such post hoc rescuing of a hypothesis in trouble is apt to provoke accusations of the sin of unfalsifiability, but I find such accusations rather unconstructive, almost nihilistic. We are not in Parliament or a court of law, with advocates of Darwinism scoring debating points against opponents, and vice versa. With the exception of a few genuine opponents of Darwinism, who are unlikely to be reading this, we are all in this together, all Darwinians who substantially agree on how we interpret what is, after all, the only workable theory we have to explain the organized complexity of life. We should all sincerely want to know why gannets lay only one egg when they could lay two, rather than treating the fact as a debating point. Lack's {36} invoking of the “time-lag” hypothesis may have been post hoc, but it is still thoroughly plausible, and it is testable. No doubt there are other possibilities which, with luck, may also be testable. Maynard Smith is surely right that we should leave aside the “defeatist” (Tinbergen 1965) and uetestable “natural selection has bungled again” explanation as a last resort, as a matter of simple research strategy if nothing else. Lewontin (1978b) says much the same: “In a sense, then, biologists are forced to the extreme adaptationist program because the alternatives, although they are undoubtedly operative in many cases, are untestable in particular cases”.
Returning to the time-lag effect itself, since modern man has drastically changed the environment of many animals and plants over a time-scale that is negligible by ordinary evolutionary standards, we can expect to see anachronistic adaptations rather often. The hedgehog antipredator response of rolling up into a ball is sadly inadequate against motor cars.
Lay critics frequently bring up some apparently maladaptive feature of modern human behaviour — adoption, say, or contraception — and fling down a challenge to “explain that if you can with your selfish genes”. Obviously, as Lewontin, Gould and others have rightly stressed, it would be possible, depending on one's ingenuity, to pull a “sociobiological” explanation out of a hat, a “just-so story”, but I agree with them and Cain that the answering of such challenges is a trivial exercise; indeed it is likely to be positively harmful. Adoption and contraception, like reading, mathematics, and stress-induced illness, are products of an animal that is living in an environment radically different from the one in which its genes were naturally selected. The question, about the adaptive significance of behaviour in an artificial world, should never have been put; and although a silly question may deserve a silly answer, it is wiser to give no answer at all and to explain why.
A useful analogy here is one that I heard from R. D. Alexander. Moths fly into candle flames, and this does nothing to help their inclusive fitness. In the world before candles were invented, small sources of bright light in darkness would either have been celestial bodies at optical infinity, or they might have been escape holes from caves or other enclosed spaces. The latter case immediately suggests a survival value for approaching light sources. The former case also suggests one, but in a more indirect sense (Fraenkel & Gune 1940). Many insects use celestial bodies as compasses. Since these are at optical infinity, rays from them are parallel, and an insect that maintains a fixed orientation of, say, 30° to them will go in a straight line. But if the rays do not come from infinity they will not be parallel, and an insect that behaves in this way will spiral in to the light source (if steering an acute-angled course) or spiral away (if steering an obtuse-angled course) or orbit the source (if steering a course of exactly 90° to the rays). Self-immolation by insects in candle flames, then, has no survival value in itself: according to {37} this hypothesis, it is a byproduct of the useful habit of steering by means of sources of light which are “assumed” to be at infinity, That assumption was once safe. It now is safe no longer, and it may be that selection is even now working to modify the insects” behaviour. (Not necessarily, however. The overhead costs of making the necessary improvements may outweigh the benefits they might bring: moths that pay the costs of discriminating candles from stars may be less successful, on average, than moths that do not attempt the costly discrimination and accept the low risk of self-immolation — see next chapter.)
But now we have reached a problem which is more subtle than the simple time-lag hypothesis itself. This is the problem, already mentioned, about what characteristics of animals we choose to recognize as units which require explanation. As Lewontin (1979b) puts it, “What are the "natural" suture lines for evolutionary dynamics? What is the topology of phenotype in evolution? What are the phenotypic units of evolution?” The candle flame paradox arose only because of the way in which we chose to characterize the moth's behaviour. We asked “Why do moths fly into candle flames?” and were puzzled. If we had characterized the behaviour differently and asked “Why do moths maintain a fixed angle to light rays (a habit which incidentally causes them to spiral into the light source if the rays happen not to be parallel)?”, we should not have been so puzzled.
Consider human male homosexuality as a more serious example. On the face of it, the existence of a substantial minority of men who prefer sexual relations with their own sex rather than with the opposite sex constitutes a problem for any simple Darwinian theory. The rather discursive title of a privately circulated homosexualist pamphlet, which the author was kind enough to seed me, summarizes the problem: “Why are there "gays" at all? Why hasn't evolution eliminated "gayness" millions of years ago?” The author, incidentally, thinks the problem so important that it seriously undermines the whole Darwinian view of life. Trivers (1974), Wilson (1975, 1978), and especially Weinrich (1976) have considered various versions of the possibility that homosexuals may, at some time in history, have been functionally equivalent to sterile workers, foregoing personal reproduction the better to care for other relatives. I do not find this idea particularly plausible (Ridley & Dawkins, 1981), certainly no more so than a “sneaky male” hypothesis. According to this latter idea, homosexuality represents an “alternative male tactic” for obtaining matings with females. In a society with harem defence by dominant males, a male who is known to be homosexual is more likely to be tolerated by a dominant male than a known heterosexual male, and an otherwise subordinate male may be able, by virtue of this, to obtain clandestine copulations with females.
But I raise the “sneaky male” hypothesis not as a plausible possibility so much as a way of dramatizing how easy and inconclusive it is to dream up {38} explanations of this kind (Lewontin, 1979b, used the same didactic trick in discussing apparent homosexuality in Drosophila). The main point I wish to make is quite different and much more important. It is again the point about how we characterize the phenotypic feature that we are trying to explain.
Homosexuality is, of course, a problem for Darwinians only if there is a genetic component to the difference between homosexual and heterosexual individuals. While the evidence is controversial (Weinrich 1976), let us assume for the sake of argument that this is the case. Now the question arises, what does it mean to say there is a genetic component to the difference, in common parlance that there is a gene (or genes) “for” homosexuality? It is a fundamental truism, of logic more than of genetics, that the phenotypic “effect” of a gene is a concept that has meaning only if the context of environmental influences is specified, environment being understood to include all the other genes in the genome. A gene “for” A in environment X may well turn out to be a gene for B in environment Y. It is simply meaningless to speak of an absolute, context-free, phenotypic effect of a given gene.
Even if there are genes which, in today's environment, produce a homosexual phenotype, this does not mean that in another environment, say that of our Pleistocene ancestors, they would have had the same phenotypic effect. A gene for homosexuality in our modern environment might have been a gene for something utterly different in the Pleistocene. So, we have the possibility of a special kind of “time-lag effect” here. It may be that the phenotype which we are trying to explain did not even exist in some earlier environment, even though the gene did then exist. The ordinary time-lag effect which we discussed at the beginning of this section was concerned with changes in the environment as manifested in changed selection pressures. We have now added the more subtle point that changes in the environment may change the very nature of the phenotypic character we set out to explain.
The jet engine superseded the propeller engine because, for most purposes, it was superior. The designers of the first jet engine started with a clean drawing board. Imagine what they would have produced if they had been constrained to “evolve” the first jet engine from an existing propeller engine, changing one component at a time, nut by nut, screw by screw, rivet by rivet. A jet engine so assembled would be a weird contraption indeed. It is hard to imagine that an aeroplane designed in that evolutionary way would ever get off the ground. Yet in order to complete the biological analogy we have to add yet another constraint. Not only must the end product get off the ground; so must every intermediate along the way, and each intermediate must be superior to its predecessor. When looked at in this light, far from {39} expecting animals to be perfect we may wonder that anything about them works at all.
Examples of the Heath Robinson (or Rube Goldberg — Gould 1978) character of animals are harder to be confident of than the previous paragraph might lead us to expect. A favourite example, suggested to me by Professor J. D. Currey, is the recurrent laryngeal nerve. The shortest distance from the brain to the larynx in a mammal, especially a giraffe, is emphatically not via the posterior side of the aorta, yet that is the route taken by the recurrent laryngeal. Presumably there once was a time in the remote ancestry of the mammals when the straight line from origin to end organ of the nerve did run posterior to the aorta. When, in due course, the neck began to lengthen, the nerve lengthened its detour posterior to the aorta, but the marginal cost of each step in the lengthening of the detour was not great. A major mutation might have re-routed the nerve completely, but only at a cost of great upheaval in early embryonic processes. Perhaps a prophetic, God-like designer back in the Devonian could have foreseen the giraffe and designed the original embryonic routing of the nerve differently, but natural selection has no foresight. As Sydney Brenner has remarked, natural selection could not be expected to have favoured some useless mutation in the Cambrian simply because “it might come in handy in the Cretaceous”.
The Picasso-like face of a flatfish such as a sole, grotesquely twisted to bring both eyes round to the same side of the head, is another striking demonstration of a historical constraint on perfection. The evolutionary history of these fish is so clearly written into their anatomy, that the example is a good one to thrust down the throats of religious fundamentalists. Much the same could be said of the curious fact that the retina of the vertebrate eye appears to be installed backwards. The light-sensitive “photocells” are at the back of the retina, and light has to pass through the connecting circuitry, with some inevitable attenuation, before it reaches them. Presumably it would be possible to write down a very long sequence of mutations which would eventually lead to the production of an eye whose retina was “the right way round” as it is in cephalopods, and this might be, in the end, slightly more efficient. But the cost in embryological upheaval would be so great that the intermediate stages would be heavily disfavoured by natural selection in comparison with the rival, patched-up job which does, after all, work pretty well. Pittendrigh (1958) has well said of adaptive organization that it is “a patchwork of makeshifts pieced together, as it were, from what was available when opportunity knocked, and accepted in the hindsight, not the foresight, of natural selection” (see also Jacob, 1977, on “tinkering”).
Sewall Wright's (1932) metaphor, which has become known under the name of the “adaptive landscape”, conveys the same idea that selection in favour of local optima prevents evolution in the direction of ultimately {40} superior, more global optima. His somewhat misunderstood (Wright 1980) emphasis on the role of genetic drift in allowing lineages to escape from the pull of local optima, and thereby attain a closer approximation to what a human might recognize as “the” optimal solution, contrasts interestingly with Lewontin's (1979b) invoking of drift as an “alternative to adaptation”. As in the case of pleiotropy, there is no paradox here. Lewontin is right that “the finiteness of real populations results in random changes in gene frequency so that, with a certain probability, genetic combinations with lower reproductive fitness will be fixed in a population”. But on the other hand it is also true that, to the extent that local optima constitute a limitation on the attainment of design perfection, drift will tend to provide an escape (Lande 1976). Ironically, then, a weakness in natural selection can theoretically enhance the likelihood of a lineage attaining optimal design! Because it has no foresight, unalloyed natural selection is in a sense an awft-perfection mechanism, hugging, as it will, the tops of the low foot-hills of Wright's landscape. A mixture of strong selection interspersed with periods of relaxation of selection and drift may be the formula for crossing the valleys to the high uplands. Clearly if “adaptationism” is to become an issue where debating points are scored, there is scope for both sides to have it both ways!
My own feeling is that somewhere here may lie the solution to the real paradox of this section on historical constraints. The jet engine analogy suggested that animals ought to be risible monstrosities of lashed-up improvisation, top-heavy with grotesque relics of patched-over antiquity. How can we reconcile this reasonable expectation with the formidable grace of the hunting cheetah, the aerodynamic beauty of the swift, the scrupulous attention to deceptive detail of the leaf insect? Even more impressive is the detailed agreement between different convergent solutions to common problems, for instance the multiple parallels that exist between the mammal radiations of Australia, South America and the Old World. Cain (1964) remarks that, “Up to now it has usually been assumed, by Darwin and others, that convergence will never be so good as to mislead us” but he goes on to give examples where competent taxonomists have been fooled. More and more groups which had hitherto been regarded as decently monophy-letic, are now being suspected of polyphyletic origin.
The citation of example and counter-example is mere idle fact-dropping. What we need is constructive work on the relation between local and global optima in an evolutionary context. Our understanding of natural selection itself needs to be supplemented by a study of “escapes from specialization” to use Hardy's (1954) phrase. Hardy himself was suggesting neoteny as an escape from specialization, while in this chapter, following Wright, I have emphasized drift in this role.
Mullerian mimicry in butterflies may prove to be a useful case-study here. Turner (1977) remarks that “among the long-winged butterflies of the {41} tropical American rainforests (ithomiids, heliconids, danaids, pierids, pericopids) there are six distinct warning patterns, and although all the warningly colored species belong to one of these mimicry "rings" the rings themselves coexist in the same habitats through most of the American tropics and remain very distinct. ... Once the difference between two patterns is too great to be jumped by a single mutation, convergence becomes virtually impossible, and the mimicry rings will coexist indefinitely”. This is one of the only cases where “historical constraints” may be close to being understood in full genetic detail. It may provide a worthwhile opportunity also for the study of the genetic details of “valley-crossing”, which in the present case would consist in the detachment of a type of butterfly from the orbit of one mimicry ring, and its eventual “capture” by the “pull” of another mimicry ring. Though he does not invoke drift as an explanation in this case, Turner tantalizingly indicates that “In southern Europe Amaia phegea ... has ... captured Zygenea ephialies from the Miillerian mimicry ring of zygaenids, homopterans, etc. to which it still belongs outside the range of A. phegea in northern Europe...”
At a more general theoretical level, Lewontin (1978) notes that “there may often be several alternative stable equilibriums of genetic composition even when the force of natural selection remains the same. Which of these adaptive peaks in the space of genetic composition is eventually reached by a population depends entirely on chance events at the beginning of the selective process ... For example, the Indian rhinoceros has one horn and the African rhinoceros has two. Horns are an adaptation for protection against predators but it is not true that one horn is specifically adaptive under Indian conditions as opposed to two horns on the African plains. Beginning with two somewhat different developmental systems, the two species responded to the same selective forces in slightly different ways”. The point is basically a good one, although it is worth adding that Lewontie's uncharacteristically “adaptationist” blunder about the functional significance of rhinoceros horns is not trivial. If horns really were an adaptation against predators it would indeed be hard to imagine how a single horn could be more useful against Asian predators while two horns were of more help against African predators. However if, as seems much more likely, rhinoceros horns are an adaptation for intraspecific combat and intimidation, it could well be the case that a one-horned rhino would be at a disadvantage in one continent while a two-horned rhino would suffer in the other. Whenever the name of the game is intimidation (or sexual attraction as Fisher taught us long ago), mere conformity to the majority style, whatever that majority style may happen to be, can have advantages. The details of a threat display and its associated organs may be arbitrary, but woe betide any mutant individual that departs from established custom (Maynard Smith & Parker 1976). {42}
No matter how strong a potential selection pressure may be, no evolution will result unless there is genetic variation for it to work on. “Thus, although I might argue that the possession of wings in addition to arms and legs might be advantageous to some vertebrates, none has ever evolved a third pair of appendages, presumably because the genetic variation has never been available” (Lewontin 1979b). One could reasonably dissent from this opinion. It may be that the only reason pigs have no wings is that selection has never favoured their evolution. Certainly we must be careful before we assume, on human-centred common-sense grounds, that it would obviously be handy for any animal to have a pair of wings even if it didn't use them very often, and that therefore the absence of wings in a given lineage must be due to lack of available mutations. Female ants can sprout wings if they happen to be nurtured as queens, but if nurtured as workers they do not express their capacity to do so. More strikingly, the queens in many species use their wings only once, for their nuptial flight, and then take the drastic step of biting or breaking them off at the roots in preparation for the rest of their life underground. Evidently wings have costs as well as benefits.
One of the most impressive demonstrations of the subtlety of. Charles Darwin's mind is given by his discussion of winglessness and the costs of having wings in the insects of oceanic islands. For present purposes, the relevant point is that winged insects may risk being blown out to sea, and Darwin (1859, p. 177) suggested that this is why many island insects have reduced wings. But he also noted that some island insects are far from wingless; they have extra large wings.
This is quite compatible with the action of natural selection. For when a new insect first arrived on the island, the tendency of natural selection to enlarge or to reduce the wings, would depend on whether a greater number of individuals were saved by successfully battling with the winds, or by giving up the attempt and rarely or never flying. As with mariners ship-wrecked near a coast, it would have been better for the good swimmers if they had been able to swim still further, whereas it would have been better for the bad swimmers if they had not been able to swim at all and had stuck to the wreck.
A neater piece of evolutionary reasoning would be hard to find, although one can almost hear the baying chorus of “Unfalsifiable! Tautological! Just-so story!”
Returning to the question of whether pigs ever could develop wings, Lewontie is undoubtedly right that biologists interested in adaptation cannot afford to ignore the question of the availability of mutational variation. It is certainly true that many of us, with Maynard Smith (1978a) though without {43} his and Lewontie's authoritative knowledge of genetics, tend to assume “that genetic variance of an appropriate kind will usually exist”. Maynard Smith's grounds are that “with rare exceptions, artificial selection has always proved effective, whatever the organism or the selected character”. A notorious case, fully conceded by Maynard Smith (1978b), where the genetic variation necessary to an optimality theory often seems to be lacking, is that of Fisher's (1930a) sex ratio theory. Cattle breeders have had no trouble in breeding for high milk yield, high beef production, large size, small size, hornlessness, resistance to various diseases, and fierceness in fighting bulls. It would obviously be of immense benefit to the dairy industry if cattle could be bred with a bias towards producing heifer calves rather than bull calves. All attempts to do this have singularly failed, apparently because the necessary genetic variation does not exist. It may be the measure of how misled is my own biological intuition that I find this fact rather astonishing, indeed worrying. I would like to think that it is an exceptional case, but Lewontin is certainly right that we need to pay more attention to the problem of the limitations of available genetic variation. From this point of view, a compilation of the amenability or resistance to artificial selection of a wide variety of characters would be of great interest.
Meanwhile, there are certain common-sense things that can be said. Firstly, it may make sense to invoke lack of available mutation to explain why animals do not have some adaptation which we think reasonable, but it is harder to apply the argument the other way round. For instance, we might indeed think that pigs would be better off with wings and suggest that they lack them only because their ancestors never produced the necessary mutations. But if we see an animal with a complex organ, or a complex and time-consuming behaviour pattern, we would seem to be on strong grounds in guessing that it must have been put together by natural selection. Habits such as dancing in bees as already discussed, “anting” in birds, “rocking” in stick insects, and egg-shell removal in gulls are positively time-consuming, energy-consuming and complex. The working hypothesis that they must have a Darwinian survival value is overwhelmingly strong. In a few cases it has proved possible to find out what that survival value is (Tinbergen 1963).
The second common-sense point is that the hypothesis of “no available mutations” loses some of its force if a related species, or the same species in other contexts, has shown itself capable of producing the necessary variation. I shall mention below a case where the known capabilities of the digger wasp Ammophila campestris were used to illuminate the lack of similar capabilities in the related species Sphex ichneumoneus. A more subtle version of the same argument can be applied within any one species. For instance, Maynard Smith (1977, see also Daly 1979) concludes a paper with an up-beat question: Why do male mammals not lactate? We need not go into the details of why he thought they ought to; he may have been wrong, his model {44} may have been wrongly set up, and the real answer to his question may be thai it would not pay male mammals to lactate. The point here is that this is a slightly different kind of question from “Why don't pigs have wings?”; We know that male mammals contain the genes necessary for lactation, because all the genes in a female mammal have passed through male ancestors and may be handed on to male descendants. Genetic male mammals treated with hormones, indeed, can develop as lactating females. This all makes it less plausible that the reason male mammals don't lactate is simply that they haven't “thought of it” mutationally speaking. (Indeed, I bet I could breed a race of spontaneously lactating males by selecting for increased sensitivity to progressively reduced dosages of injected hormone, an interesting practical application of the Baldwin/Waddington Effect.)
The third common-sense point is that if the variation that is being postulated consists in a simple quantitative extension of already existing variation it is more plausible than a radical qualitative innovation. It may be implausible to postulate a mutant pig with wing rudiments, but it is not implausible to postulate a mutant pig with a curlier tail than existing pigs. I have elaborated this point elsewhere (Dawkins 1980).
In any case, we need a more subtle approach to the question of what is the evolutionary impact of differing degrees of mutability. It is not good enough to ask, in an all or none way, whether there is or is not genetic variation available to respond to a given selection pressure. As Lewontin (1979) rightly says, “Not only is the qualitative possibility of adaptive evolution constrained by available genetic variation, but the relative rates of evolution of different characters are proportional to the amount of genetic variance for each”. I think this opens up an important line of thought when combined with the notion of historical constraints treated in the previous section. The point can be illustrated with a fanciful example.
Birds fly with wings made of feathers, bats with wings consisting of flaps of skin. Why do they not both have wings made in the same way, whichever way is “superior”? A confirmed adaptationist might reply that birds must be better off with feathers and bats better off with skin flaps. An extreme anti-adaptationist might say that very probably feathers would actually be better than skin-flaps for both birds and bats, but bats never had the good fortune to produce the right mutations. But there is an intermediate position, one which I find more persuasive than either extreme. Let us concede to the adaptationist that, given enough time, the ancestors of bats probably could have produced the sequence of mutations necessary for them to sprout feathers. The operative phrase is “given enough time”. We are not making an all-or-none distinction between impossible and possible mutational variation, but simply stating the undeniable fact that some mutations are quantitatively more probable than others. In this case, ancestral mammals {45} might have produced both mutants with rudimentary feathers and mutants with rudimentary skin flaps. But the proto-feather mutants (they might have had to go through an intermediate stage of small scales) were so slow in making their appearance in comparison with the skin-flap mutants, that skin-flap wings had long ago appeared and led to the evolution of passably efficient wings.
The general point is akin to the one already made about adaptive landscapes. There we were concerned with selection preventing lineages from escaping the clutches of local optima. Here we have a lineage faced with two alternative routes of evolution, one leading to, say, feathered wings, the other to skin-flap wings. The feathered design may be not only a global optimum but the present local optimum as well. The lineage, in other words, may be sitting exactly at the foot of the slope leading to the feathered peak of the Sewall Wright landscape. If only the necessary mutations were available it would climb easily up the hill. Eventually, according to this fanciful parable, those mutations might have come, but — and this is the important point — they were too late. Skin-flap mutations had come before them, and the lineage had already climbed too far up the slopes of the skin-flap adaptive hill to turn back. As a river takes the line of least resistance downhill, thereby meandering in a route that is far from the most direct one to the sea, so a lineage will evolve according to the effects of selection on the variation available at any given moment. Once a lineage has begun to evolve in a given direction, this may in itself close options that were formerly available, sealing off” access to a global optimum. My point is that lack of available variation does not have to be absolute in order to become a significant constraint on perfection. It need only be a quantitative brake to have dramatic qualitative effects. In spirit, then, I agree with Gould and Calloway (1980) when they say, citing Vermeij's (1973) stimulating paper on the mathematics of morphological versatility that, “Some morphologies can be twisted, bent and altered in a variety of ways, and others cannot”. But I would prefer to soften “cannot”, to make it a quantitative constraint, not an absolute barrier.
McCleery (1978), in an agreeably comprehensible introduction to the McFarland school of ethological optimality theory, mentions H. A. Simon's concept of'satisficing” as an alternative to optimizing. If optimizing systems are concerned with maximizing something, satisficing systems get away with doing just enough. In this case, doing enough means doing enough to stay alive. McCleery contents himself with complaining that such “adequacy” concepts have not generated much experimental work. I think evolutionary theory entitles us to be a bit more negative a priori. Living things are not selected for their capacity simply to stay alive; they are staying alive in competition with other such living things. The trouble with satisficing as a {46} concept is that it completely leaves out the competitive element which is fundamental to all life. In Gore Vidal's words: “It is not enough to succeed. Others must fail”.
On the other hand “optimizing” is also an unfortunate word because it suggests the attainment of what an engineer would recognize as the best design in a global sense. It tends to overlook the constraints on perfection which are the subject of this chapter. In many ways the word “meliorizing” expresses a sensible middle way between optimizing and satisficing. Where optimus means best, melior means better. The points we have been considering about historical constraints, about Wright's adaptive landscapes and about rivers following the line of immediate least resistance, are all related to the fact that natural selection chooses the better of present available alternatives. Nature does not have the foresight to put together a sequence of mutations which, for all that they may entail temporary disadvantage, set a lineage on the road to ultimate global superiority. It cannot refrain from favouring slightly advantageous available mutations now, so as to take better advantage of superior mutations which may arrive later. Like a river, natural selection blindly meliorizes its way down successive lines of immediately available least resistance. The animal that results is not the most perfect design conceivable, nor is it merely good enough to scrape by. It is the product of a historical sequence of changes, each one of which represented, at best, the better of the alternatives that happened to be around at the time.
“If there were no constraints on what is possible, the best phenotype would live for ever, would be impregnable to predators, would lay eggs at an infinite rate, and so on” (Maynard Smith 1978b). “An engineer, given carte blanche on his drawing board could design an "ideal" wing for a bird, but he would demand to know the constraints under which he must work. Is he constrained to use feathers and bones, or may he design the skeleton in ikanium alloy? How much is he allowed to spend on the wings, and how much of the available economic investment must be diverted into, say, egg production?” (Dawkins & Brockmann 1980). In practice, an engineer will normally be given a specification of minimum performance such as “The bridge must bear a load of the tons ... The aeroplane wing must not break until it receives a stress three times what would be expected in worst-case turbulent conditions; now go ahead and build it as cheaply as you can”. The best design is the one that satisfies (“satisfices”) the criterion specification at the least cost. Any design that achieves “better” than the specified criterion performance is likely to be rejected, because presumably the criterion could be achieved more cheaply.
The particular criterion specification is an arbitrary working rule. There is {47} nothing magic about a safety margin of three times the expected worst-case conditions. Military aircraft may be designed with more risky safety margins than civilian ones. In effect, the engineer's optimization instructions amount to a monetary evaluation of human safety, speed, convenience, pollution of the atmosphere, etc. The price put on each of these is a matter of judgement, and is often a matter of controversy.
In the evolutionary design of animals and plants, judgement does not enter into it, nor does controversy except among the human spectators of the show. In some way, however, natural selection must provide the equivalent of such judgement: risks of predation must be evaluated against risks of starving and benefits of mating with an extra female. For a bird, resources spent on making breast muscles for powering wings are resources that could have been spent on making eggs. An enlarged brain would permit a finer tuning of behaviour to environmental details, past and present, but at a cost of an enlarged head, which means extra weight at the front end of the body, which in turn necessitates a larger tail for aerodynamic stability, which in turn ... Winged aphids are less fecund than wingless ones of the same species (J. S. Kennedy, personal communication). That every evolutionary adaptation must cost something, costs being measured in lost opportunities to do other things, is as true as that gem of traditional economic wisdom, “There is no such thing as a free lunch”.
Of course the mathematics of biological currency-conversion, of evaluating the costs of wing muscle, singing time, predator-vigilance time, etc., in some common currency such as “gonad equivalents”, are likely to be very complex. Whereas the engineer is allowed to simplify his mathematics by working to an arbitrarily chosen minimum threshold of performance, the biologist is granted no such luxury. Our sympathy and admiration must go out to those few biologists who have attempted to grapple with these problems in detail (e.g. Oster & Wilson 1978; McFarland & Houston 1981).
On the other hand, although the mathematics may be formidable, we don't need mathematics to deduce the most important point, which is that any view of biological optimization that denies the existence of costs and trade-offs is doomed. An adaptationist who looks at one aspect of an animal's body or behaviour, say the aerodynamic performance of its wings, while forgetting that efficiency in the wings can only be bought at a cost which will be felt somewhere else in the animal's economy, would deserve all the criticism he gets. It has to be admitted that too many of us, while never actually denying the importance of costs, forget to mention them, perhaps even forget to think about them, when we discuss biological function. This has probably provoked some of the criticism that has come our way. In an earlier section I quoted Pittendrigh's remark that adaptive organization was a “patchwork of makeshifts”. We must also not forget that it is a tangle of compromises (Tinbergen 1965). {48}
In principle, it would seem a valuable heuristic procedure to assume that an animal is optimizing something under a given set of constraints, and to try to work out what those constraints are. This is a restricted version of what McFarland and his colleagues call the “reverse optimality” approach (e.g. McCleery 1978). As a case study I shall take some work with which I happen to be familiar.
Dawkins and Brockmann (1980) found that the digger wasps (Sphex ichneumoneus) studied by Brockmann behaved in a way that a naive human economist might have criticized as maladaptive. Individual wasps appeared to commit the “Concorde Fallacy” of valuing a resource according to how much they had already spent on it, rather than according to how much they could get out of it in the future. Very briefly, the evidence is as follows. Solitary females provision burrows with stung and paralysed katydids which are to serve as food for their larvae (see Chapter 7). Occasionally two females find themselves provisioning the same burrow, and they usually end up fighting over it. Each fight goes on until one wasp, thereby defined as the loser, flees from the area, leaving the winner in control of the burrow and all the katydids caught by both wasps. We measured the “real value” of a burrow-as the number of katydids which it contained. The “prior investment” by each wasp in the burrow was measured as the number of katydids which she, as an individual, had put into it. The evidence suggested that each wasp fought for a time proportional to her own investment, rather than proportional to the “true value” of the burrow.
Such a policy has great human psychological appeal. We too tend to fight tenaciously for property which we have put great effort into acquiring. The fallacy gets its name from the fact that, at a time when sober economic judgement of future prospects counselled abandoning the developing of the Concorde airliner, one of the arguments in favour of continuing with the half completed project was retrospective: “We have already spent so much on it that we cannot back out now”. A popular argument for prolonging wars gave rise to the other name for the fallacy, the “Our boys shall not have died in vain” fallacy.
When Dr. Brockmane and I first realized that digger wasps behaved in like manner, I was, it has to be confessed, a little disconcerted, possibly because of my own past investment of effort (Dawkins & Carlisle 1976; Dawkins 1976a) in persuading my colleagues that the psychologically appealing Concorde Fallacy was, indeed, a fallacy! But then we started to think more seriously about cost constraints. Could it be that what appeared to be maladaptive was better interpreted as an optimum, given certain constraints? The question then became: is there a constraint such that the wasps' Coecordiae behaviour is the best they can achieve under it?
In fact the question was more complicated than that, because it was necessary to substitute Maynard Smith's (1974) concept of evolutionary {49} stability (“ESS” — see Chapter 7) for that of simple optimality, but the principle remains that a reverse optimaiity approach might be heuristically valuable. If we can show that an animal's behaviour is what would be produced by an optimizing system working under constraint X, maybe we can use the approach to learn something of the constraints under which animals actually do work.
In the present case it seemed that the relevant constraint might be one of sensory capacity. If the wasps, for some reason, cannot count katydids in the burrow, but can metre some aspect of their own hunting efforts, there is an asymmetry of information possessed by the two combatants. Each one “knows” that the burrow contains at least b katydids, where b is the number she herself has caught. She may “estimate” that the true number in the burrow is larger than b, but she does not know how much larger. Under such conditions Grafen (in preparation) has shown that the expected ESS is approximately the one originally calculated by Bishop and Cannings (1978) for the so-called “generalized war of attrition”. The mathematical details can be left aside; for present purposes what matters is that the behaviour expected by the extended war of attrition model would look very like the Concordian behaviour actually shown by the wasps.
If we were interested in testing the general hypothesis that animals optimize, this kind of post hoc rationalization would be suspect. By post hoc modification of the details of the hypothesis, one is bound to lied a version which fits the facts. Maynard Smith's (1978b) reply to this kind of criticism is very relevant: “...in testing a model we are not testing the general proposition that nature optimizes, but the specific hypotheses about constraints, optimization criteria, and heredity”. In the present case we are making a general assumption that nature does optimize within constraints, and testing particular models of what those constraints might be.
The particular constraint suggested — inability of the wasp's sensory system to assess the contents of a burrow — is in accordance with independent evidence from the same population of wasps (Brockmann, Grafen & Dawkins 1979; Brockmann & Dawkins 1979). There is eo reason to regard it as an irrevocably binding limitation for all time. Probably the wasps could evolve the capacity to assess nest contents, but only at a cost. Digger wasps of the related species Ammophila campesiris “have long been known to make an assessment of the contents of each of their nests every day (Baerends 1941). Unlike Sphex which provisions one burrow at a time, lays an egg, then fills the burrow in with soil and leaves the larva to eat the provision on its own, Ammophila campesiris is a progressive provisioned of several burrows concurrently. A female tends two or three growing larvae, each in a separate burrow, at the same time. The ages of her various larvae are staggered, and their food needs are different. Every morning she assesses the current contents of each burrow on a special early morning “inspection round”. By {50} experimentally changing the contents of burrows, Baerends showed that the female adjusts her whole day's provisioning of each burrow according to what it contained at the time of her morning inspection. The contents of the burrow at any other time of day have no effect on her behaviour, even though she is provisioning it all day. She appears, therefore, to use her assessment faculty sparingly, switching it off for the rest of the day after the morning inspection, almost as though it was a costly, power-consuming instrument. Fanciful as that analogy may be, it surely suggests that the assessment faculty, whatever it is, may have overhead running costs, even if (G. P. Baerends, personal communication) these consist only in the time consumed.
Sphex ichneumoneus, not being a progressive provisioner, and tending only one burrow at a time, presumably has less need than Ammophila for a burrow-assessment faculty. By not attempting to count prey in the burrow, it can save itself not only the running expenditure that Ammophila seems so careful to ration; it can also save itself the initial manufacturing costs of the necessary neural and sensory apparatus. Probably it could benefit slightly from having an ability to assess burrow contents, but only on the comparatively rare occasions when it finds itself competing for a burrow with another wasp. It is easy to believe that the costs outweigh the benefits, and that selection has therefore never favoured the evolution of assessment apparatus. I think this is a more constructive and interesting hypothesis than the alternative hypothesis that the necessary mutational variation has never arisen. Of course we have to admit that the latter might be right, but I would prefer to keep it as a hypothesis of last resort.
One of the main topics to be tackled in this book is that of the level at which natural selection acts. The kind of adaptations we should see if selection acted at the level of the group would be quite different from the adaptations we should expect if selection acts at the level of the individual. It follows that a group selectionist might well see as imperfections, features which an individual selectionist would see as adaptations. This is the main reason why I regard as unfair Gould and Lewontin's (1979) equating of modern adaptationism with the naive perfectionism that Haldane named after Voltaire's Dr Pangloss. With reservations due to the various constraints on perfection, an adaptationist may believe that all aspects of organisms are “adaptive optimal solutions to problems”, or that “it is virtually impossible to do a better job than an organism is doing in its given environment”. Yet the same adaptationist may be extremely fussy about the kind of meaning he allows to words like “optimal” and “better”. There are many kinds of adaptive, indeed Panglossian, explanations, for example most {51} group-selectionist ones, which would be utterly ruled out by the modern adaptationist.
For the Panglossian the demonstration that something is “beneficial” (to whom or to what is often not specified) is a sufficient explanation for its existence. The neo-Darwinian adaptationist, on the other hand, insists upon knowing the exact nature of the selective process that has led to the evolution of the putative adaptation. In particular, he insists on precise language about the level at which natural selection is supposed to have acted. The Panglossian looks at a one-to-one sex ratio and sees that it is good: does it not minimize the wastage of the population's resources? The neo-Darwinian adaptationist considers in detail the fates of genes acting on parents to bias the sex ratio of their offspring, and calculates the evolutionarily stable state of the population (Fisher 1930a). The Panglossian is disconcerted by 1:1 sex ratios in polygynous species, in which a minority of males hold harems and the rest sit about in bachelor herds consuming almost half the population's food resources yet contributing not at all to the population's reproduction. The neo-Darwinian adaptationist takes this in his stride. The system may be hideously uneconomical from the population's point of view, but, from the point of view of the genes influencing the trait concerned, there is no mutant that could do better. My point is that neo-Darwinian adaptationism, is not a catch-all, blanket faith in all being for the best. It rules out of court most of the adaptive explanations that readily occur to the Panglossian.
Some years ago, a colleague received an application from a prospective graduate student wishing to work on adaptation, who was brought up a religious fundamentalist and did not believe in evolution. He believed in adaptations, but thought they were designed by God, designed for the benefit of ... ah, but that is just the problem! It might be thought that it did not matter whether the student believed adaptations were produced by natural selection or by God. Adaptations are “beneficial” whether because of natural selection or because of beneficient design, and could not a fundamentalist student be usefully employed in uncovering the detailed ways in which they were beneficial? My point is that this argument will not do, because what is beneficial to one entity in the hierarchy of life is harmful to another, and creationism gives us no grounds for supposing that one entity's welfare will be preferred to another's. In passing, the fundamentalist student might pause to wonder at a God who goes to great trouble to provide predators with beautiful adaptations to catch prey, while with the other hand giving prey beautiful adaptations to thwart them. Perhaps He enjoys the spectator sport. Returning to the main point, if adaptations were designed by God, He might have designed them to benefit the individual animal (its survival or — not the same thing — its inclusive fitness), the species, some other species such as mankind (the usual view of religious fundamentalists), the “balance of nature”, or some other inscrutable purpose {52} known only to Him. These are frequently incompatible alternatives. It really matters for whose benefit adaptations are designed. Facts such as the sex ratio in harem-forming mammals are inexplicable on certain hypotheses and easily explicable on others. The adaptationist working within the framework of a proper understanding of the genetical theory of natural selection countenances only a very restricted set of the possible functional hypotheses which the Panglossian might admit.
One of the main messages of this book is that, for many purposes, it is better to regard the level at which selection acts as neither the organism, oor the group or any larger unit, but the gene or small genetic fragment. This difficult topic will be debated in later chapters. For the present, it is sufficient to note that selection at the level of the gene can give rise to apparent imperfections at the level of the individual. I shall discuss “meiotic drive” and related phenomena in Chapter 8, but the classic example is the case of heterozygous advantage. A gene may be positively selected because of its beneficial effects when heterozygous, even though it has harmful effects when homozygous. As a consequence of this, a predictable proportion of the individual organisms in the population will have defects. The general point is this. The genome of an individual organism in a sexual population is the product of a more or less random shuffling of the genes in the population. Genes are selected over their alleles because of their phenotypic effects, averaged over all the individual bodies in which they are distributed, over the whole population, and through many generations. The effects that a given gene has will usually depend upon the other genes with which it shares a body: heterozygous advantage is just a special case of this. A certain proportion of bad bodies seems an almost inevitable consequence of selection for good genes, where good refers to the average effects of a gene on a statistical sample of bodies in which it finds itself permuted with other genes.
Inevitable, that is, as long as we accept the Mendelian shuffle as given and inescapable. Williams (1979), disappointed at finding no evidence for adaptive fine-adjustment of the sex ratio, makes the perceptive point that
Sex is only one of many offspring characters that would seem adaptive for a parent to control. For instance, in human populations affected by sickle-cell anaemia, it would be advantageous for a heterozygous woman to have her A eggs fertilized only by a-bearing sperm, and vice versa, or even to abort all homozygous embryos. Yet if mated to another heterozygote she will reliably submit to the Mendelian lottery, even though this means markedly lowered fitness for half her children ... The really fundamental questions in evolution may be answerable only by regarding each gene as ultimately in conflict with every other gene, even those at other loci in the same cell. A really {53} valid theory of natural selection must be based ultimately on selfish replicators, genes and all other entities capable of the biased accumulation of different variant forms.
Amen!
However well adapted an animal may be to environmental conditions, those conditions must be regarded as a statistical average. It will usually be impossible to cater for every conceivable contingency of detail, and any given animal will therefore frequently be observed to make “mistakes”, mistakes which can easily be fatal. This is not the same point as the time-lag problem already mentioned. The time-lag problem arises because of non-stationarities in the statistical properties of the environment: average conditions now are different from the average conditions experienced by the animal's ancestors. The present point is more inescapable. The modern animal may be living in identical average conditions to those of an ancestor, yet the detailed moment to moment occurrences facing either of them are not the same from day to day, and are too complex for precise prediction to be possible.
It is particularly in behaviour that such mistakes are seen. The more static attributes of an animal, its anatomical structure for instance, are obviously adapted only to long-term average conditions. An individual is either big or small, it cannot change size from minute to minute as the need arises. Behaviour, rapid muscular movement, is that part of an animal's adaptive repertoire which is specifically concerned with high speed adjustment. The animal can be now here, now there, now up a tree, now underground, rapidly accommodating to environmental contingencies. The number of such possible contingencies, when defined in all their detail, is like the number of possible chess, positions, virtually infinite. Just as chess-playing computers (and chess-playing people) learn to classify chess positions into a manageable number of generalized classes, so the best that an adaptationist can hope for is that an animal will have been programmed to behave in ways appropriate to a manageable number of general contingency classes. Actual contingencies will fit these general classes only approximately, and apparent mistakes are cherefore bound to be made.
The animal that we see up a tree may come from a long line of tree-dwelling ancestors. The trees in which the ancestors underwent natural selection were, in general, much the same as the trees of today. General rules of behaviour which worked then, such as “Never go out on a limb that is too thin”, still work. But the details of any one tree are inevitably different from {54} the details of another. The leaves are in slightly different places, the breaking strain of the branches is only approximately predictable from their diameter, and so on. However strongly adaptationist our beliefs may be, we can only expect animals to be average statistical optimizers, never perfect anticipators of every detail.
So far we have considered the environment as statistically complex and therefore hard to predict. We have not reckoned on its being actively malevolent from our animal's point of view. Tree boughs surely do not deliberately snap out of spite when monkeys venture on to them. But a “tree bough” may turn out to be a camouflaged python, and our monkey's last mistake is then no accident but is, in a sense, deliberately engineered. Part of a monkey's environment is non-living or at least indifferent to the monkey's existence, and the monkey's mistakes can be put down to statistical unpredictability. But other parts of the monkey's environment consist of living things that are themselves adapted to profit at the expense of monkeys. This portion of the monkey's environment may be called malevolent.
Malevolent environmental influences may themselves be hard to predict for the same reasons as indifferent ones, but they introduce an added hazard; an added opportunity for the victim to make “mistakes”. The mistake made by a robin in feeding a cuckoo in its nest is presumably in some sense a maladaptive blunder. This is not an isolated, unpredictable occurrence such as arises because of the statistical unpredictability of the non-malevolent part of the environment. It is a recurrent blunder, afflicting generation after generation of robins, even the same robin several times in its life. Examples of this kind always make us wonder at the compliance, in evolutionary time, of the organisms that are manipulated against their best interests. Why doesn't selection simply eliminate the susceptibility of robins to the deception of cuckoos? This kind of problem is one of many which I believe will one day become the stock in trade of a new subdiscipline of biology — the study of manipulation, arms races and the extended phenotype. Manipulation and arms races form the subject of the next chapter, which in some ways can be regarded as an expansion of the theme of the final section of this chapter. {55}
| << | {55} | >> |
One of my purposes in this book is to question the “central theorem” that it is useful to expect individual organisms to behave in such a way as to maximize their own inclusive fitness, or in other words to maximize the survival of copies of the genes inside them. The end of the previous chapter suggests one way in which the central theorem might be violated. Organisms might consistently work in the interests of other organisms rather than of themselves. That is, they might be “manipulated”.
The fact that animals frequently cause other animals to perform some action that is against their own best interests is, of course, well known. Obviously it happens every time an angler fish catches prey, every time a cuckoo is fed by its foster mother. I shall make use of both these examples in this chapter, but I shall also emphasize two points that have not always been stressed. Firstly, it is natural to assume that even if a manipulator gets away with it temporarily, it is only a matter of evolutionary time before the lineage of manipulated organisms comes up with a counter-adaptation. In other words, we tend to assume that manipulation only works because of the “time-lag” constraint on perfection. In this chapter I shall point out that, on the contrary, there are conditions under which we should expect manipulators to succeed consistently and for indefinite lengths of evolutionary time. I shall discuss this later under the catch-phrase of “arms races”.
Secondly, until the last decade or so, most of us have paid insufficient attention to the likelihood of intraspecific manipulation, especially exploitative manipulation within the family. I attribute this deficiency to a residuum of group-selectionist intuition which often lurks in the depths of the biologist's mind even after group selection has been rejected at the surface level of reason. I think that a minor revolution has taken place in the way we think about social relationships. “Genteel” (Lloyd 1979) ideas of vaguely benevolent mutual cooperation are replaced by an expectation of stark, ruthless, opportunistic mutual exploitation (e.g. Hamilton 1964a, b, 1970, 1971a; Williams 1966; Trivers 1972, 1974; Ghiselin 1974a; Alexander {56} 1974). This revolution is popularly associated with the name “sociobiology”, although the association is somewhat ironic since, as I have suggested before, Wilson's (1975) great book of that name is in many respects pre-revolutionary in attitude: not the new synthesis, but the last and greatest synthesis of the old, benevolent regime (e.g. his Chapter 5).
I can exemplify the changed view by quoting from one recent paper, Lloyd's (1979) entertaining review of sexual dirty tricks in insects.
Selection for haste in males, and coyness in females, results in what amounts to competition between the sexes. Males may be selected to bypass any choice that the females attempt to exercise, and then females selected to maintain their options, to not be misled or to have their choices subverted. If males subdue and seduce females with true aphrodisiacs [Lloyd gives evidence elsewhere in the paper], females may be expected to escape sooner or later in evolutionary time. And after sperm has been placed in a female, she should manipulate it: store, transfer (from chamber to chamber), use, eat, or dissolve it, as she makes additional observations on males. Females may accept and store sperm from a male for insurance that they will get a mate, and then become choosy ... It is possible for sperm to be manipulated in the female (e.g. sex determination in Hymenoptera). Female reproductive morphology often includes sacs, valves, and tubes that could have evolved in this context. In fact, it is possible that some reported examples of sperm competition are actually cases of sperm manipulation ... Given that females, to one extent or another, subvert male interests by the internal manipulation of ejaculate, it is not inconceivable that males will have evolved little openers, snippers, levers and syringes that put sperm in the places females have evolved (“intended”) for sperm with priority usage — collectively, a veritable Swiss Army Knife of gadgetry!
This kind of unsentimental, dog eat dog, language would not have come easily to biologists a few years ago, but nowadays I am glad to say it dominates the textbooks (e.g. Alcock 1979).
Such dirty tricks, as often as not, involve direct action, the muscles of one individual moving to molest the body of another. The manipulation which is the subject of this chapter is more indirect and more subtle. An individual induces the effectors of another individual to work against its own best interests, and in favour of the interests of the manipulator. Alexander (1974) was one of the first to emphasize the importance of such manipulation. He generalized his concept of queen domination in the evolution of social insect worker behaviour, to produce a wide-ranging theory of “parental manipulation” (see also Ghiselin 1974a). He suggested that parents are in such a {57} commanding position over their offspring that offspring may be forced to work in the interests of their parents' genetic fitness, even where this conflicts with their own. West-Eberhard (1975) follows Alexander in dignifying parental manipulation as one of three general ways in which individual “altruism” can evolve, the others being kin selection and reciprocal altruism. Ridley and I make the same point, but do not restrict ourselves to parental manipulation (Ridley & Dawkins 1981).
The argument is as follows. Biologists define behaviour as altruistic if it favours other individuals at the expense of the altruist himself. A problem arises, incidentally, over how benefit and expense are to be defined. If they are defined in terms of individual survival, acts of altruism are expected to be very common, and will be taken to include parental care. If they are defined in terms of individual reproductive success, parental care no longer counts as altruism, but altruistic acts towards other relatives are predicted by neo-Darwinian theory. If benefit and expense are defined in terms of individual inclusive fitness, neither parental care nor care for other genetic relatives counts as altruism, and indeed a naive version of the theory expects that altruism should not really exist. Any of the three definitions can be justified, although if we must talk of altruism at all I prefer the first definition, the one that allows parental care as altruistic. But my point here is that, whichever definition we favour, it is met if the “altruist” is forced — manipulated — by the beneficiary into donating something to him. For instance, the definition (any of the three) allows us no choice but to regard the feeding of a cuckoo nestling by its foster parent as altruistic behaviour. Maybe this means we need a new kind of definition, but that is another issue. Krebs and I take the argument to its logical conclusion, and interpret all of animal communication as manipulation of signal-receiver by signal-sender (Dawkins & Krebs 1978).
Manipulation is, indeed, pivotal to the view of life expounded in this book, and it is a trifle ironic that I am one of those who has criticized Alexander's concept of parental manipulation (Dawkins 1976a, pp. 145–148; Blick 1977; Parker & Macnair 1978; Krebs & Davies 1978; Stamps & Metcalf 1980), and, in turn, been criticized for doing so (Sherman 1978; Harpending 1979; Daly 1980). Notwithstanding these defenders, Alexander (1980, pp. 38–39) himself has conceded that his critics were right.
Clearly some clarification is necessary. Neither I, nor any of the later critics mentioned, doubted that selection would favour parents who succeeded in manipulating their offspring, over parents who did not. Nor did we doubt that parents would indeed, in many cases, “win the arms race” against offspring. All we were objecting to was the logic of suggesting that parents enjoy a built-in advantage over their children, simply because all children aspire to become parents. That is no more true than that children enjoy a built-in advantage simply because all parents were once children. {58} Alexander had suggested that tendencies to selfishness in children, tendencies to act against the interests of their parents, could not spread because, when the child grew up, its own children's inherited selfishness against it would detract from its own reproductive success. For Alexander this sprang from his conviction that “the entire parent-offspring interaction has evolved because it benefited one of the two individuals — the parent. No organism can evolve parental behaviour, or extend its parental care, unless its own reproduction is thereby enhanced” (Alexander 1974, p. 340). Alexander was, therefore, thinking firmly within the paradigm of the selfish organism. He upheld the central theorem that animals act in the interests of their own inclusive fitness, and he understood this to preclude the possibility of offspring acting against their parent's interests. But the lesson I prefer to learn from Alexander is that of the central importance of manipulation itself, which I regard as a violation of the central theorem.
I believe animals exert strong power over other animals, and that frequently an animal's actions are most usefully interpreted as working in the interests of another individual's inclusive fitness, rather than its own. Later in this book we shall dispense with the use of the concept of inclusive fitness altogether, and the principles of manipulation will be subsumed under the umbrella of the extended phenotype. But for the rest of this chapter it will be convenient to discuss manipulation at the level of the individual organism.
There is an inevitable overlap here with a paper which I wrote jointly with J. R. Krebs on animal signals as manipulation (Dawkins & Krebs 1978). Before proceeding, I feel obliged to acknowledge that this paper has been severely criticized by Hinde (1981). Some of his criticisms, which do not affect the parts of the paper I wish to use here, are answered by Caryl (1982), who had also been criticized by Hinde. Hinde took us to task for unfairness in quoting apparently group-selectionist statements by Tinbergen (1964) and others whom we labelled, for want of a better title, classical ethologists. I have sympathy with this historical criticism. The “good of the species” passage which we quoted from Tinbergen was a genuine quotation, but I agree that it was not typical of Tinbergen's (e.g. 1965) thinking at that time. The following much earlier quotation from Tinbergen would, perhaps, have been a fairer one for us to have chosen: “...just as the functioning of hormones and the nervous system implies not only the sending out of signals but also a specific responsiveness in the reacting organ, so the releaser system involves a specific responsiveness to particular releasers in the reacting individual as well as a specific tendency to send out the signals in the initiator. The releaser system ties individuals into units of a super-individual order and renders them higher units subject to natural selection” (Tinbergen 1954). Even if outright group-selectionism was strongly opposed by Tinbergen and most of his pupils in the early 1960s, I still think that nearly all of us thought of animal signals in terms of a vague notion of “mutual {59} benefit”: if signals were not actually “for the good of the species” (as in the unrepresentative quotation from Tinbergen), they were “for the mutual benefit of both signaller and receiver”. The evolution of ritualized signals was regarded as mutual evolution: enhanced signalling power on one side was accompanied by increased sensitivity to the signals on the other side.
Today we would recognize that if receiver sensitivity increases, signal strength does not need to increase but, instead, is more likely to decrease owing to the attendant costs of conspicuous or loud signals. This might be called the Sir Adrian Boult principle. Once, in rehearsal, Sir Adrian turned to the violas and told them to play out more. “But Sir Adrian”, protested the principal viola, “you were indicating less and less with your baton”. “The idea”, retorted the maestro, “is that I should do less and less, and you should do more and more!” in those cases where animal signals really are of mutual benefit, they will tend to sink to the level of a conspiratorial whisper: indeed this may often have happened, the resulting signals being too inconspicuous for us to have noticed them. If signal strength increases over the generations this suggests, on the other hand, that there has been increasing sales resistance on the side of the receiver (Williams 1966).
As mentioned earlier, an animal will not necessarily submit passively to being manipulated, and an evolutionary “arms race” is expected to develop. Arms races were the subject of a second joint paper (Dawkins & Krebs 1979). We were not, of course, the first to make the points that follow in this chapter, but nevertheless it will be convenient to quote from our two joint papers. With permission from Dr Krebs, I shall do so without breaking the flow with repeated bibliographic citations and quotation marks.
An animal often needs to manipulate objects in the world around it. A pigeon carries twigs to its nest. A cuttlefish blows sand from the sea bottom to expose prey. A beaver fells trees and, by means of its dam, manipulates the entire landscape for miles around its lodge. When the object an animal seeks to manipulate is non-living, or at least when it is not self-mobile, the animal has no choice but to shift it by brute force. A dung beetle can move a ball of dung only by forcibly pushing it. But sometimes an animal may benefit by moving an “object” which happens, itself, to be another living animal. The object has muscles and limbs of its own, controlled by a nervous system and sense organs. While it may still be possible to shift such an “object” by brute force, the goal may often be more economically engineered by subtler means. The object's internal chain of command — sense organs, nervous system, muscles — may be infiltrated and subverted. A male cricket does not physically roll a female along the ground and into his burrow. He sits and sings, and the female comes to him under her own power. From his point of view this communication is energetically more efficient than trying to take her by force.
A question immediately arises. Why should the female stand for it? Since {60} she is in control of her own muscles and limbs, why would she approach the male unless it is in her genetic interests to do so? Surely the word manipulation is appropriate only if the victim is unwilling? Surely the male cricket is simply informing the female of a fact that is useful to her, that over here is a ready and willing male of her own species. Having given her this information, doesn't he then leave it up to her to approach him or not as she pleases, or as natural selection has programmed her?
Well and good when males and females happen to have identical interests, but examine the premise of the last paragraph. What entitles us to assert that the female is “in control of her own muscles and limbs”? Doesn't this beg the very question we are interested in? By advancing a manipulation hypothesis we are, in effect, suggesting that the female may not be in control of her own muscles and limbs, and that the male may be. This example could, of course, be reversed, and the female be said to manipulate the male. The point being made has no specific connection with sexuality. I could have used the example of plants, lacking muscles of their owe, using insect muscles as effector organs to transport their pollen, and fuelling those muscles with nectar (Heinrich 1979). The general point is that an organism's limbs may be manipulated to work in the interests of the genetic fitness of another organism. This statement cannot be made convincing until later in the book when we have introduced the idea of the extended phenotype. For this chapter we are still working within the paradigm of the selfish organism, albeit we are starting to stretch it so that it creaks ominously at the edges.
The example of male and female cricket may have been poorly chosen, because, as I said above, many of us have only recently become used to the idea of sexual relations as a battle. Many of us have yet to absorb into our consciousness the fact that “selection can act in opposition on the two sexes. Commonly, for a given type of encounter, males will be favoured if they do mate and females if they don't” (Parker 1979; see also West-Eberhard 1979). I shall return to this, but for the moment let us use a starker example of manipulation. As starkly ruthless as any battle in nature is that between predator and prey. There are various techniques a predator may use to catch his prey. He may run after them and attempt to outpace, outstay, or outflank them. He may sit in one place and ambush or trap them. Or he may do as angler fish and “femmes fatales” fireflies (Lloyd 1975, 1981) do, and manipulate the prey's own nervous system so that it actively approaches its own doom. An angler fish sits on the sea bottom and is highly camouflaged except for a long rod projecting from the top of the head, on the end of which is the “lure”, a flexible piece of tissue which resembles some appetizing morsel such as a worm. Small fish, prey of the angler, are attracted by the lure which resembles their own prey. When they approach it the angler “plays” them down into the vicinity of his mouth, then suddenly opens his jaws and {61} the prey are engulfed in the inrush of water. Instead of using massive body and tail muscles in active pursuit of prey, the angler uses the small economical muscles controlling his rod, to titillate the prey's nervous system via its eyes. Finally it is the prey fish's own muscles that the angler uses to close the gap between them. Krebs and I informally characterized animal “communication” as a means by which one animal makes use of another animal's muscle power. This is roughly synonymous with manipulation.
The same question arises as before. Why does the victim of manipulation stand for it? Why does the prey fish rush literally into the jaws of death? Because it “thinks” it is really rushing to get a meal itself. More formally, natural selection has acted on its ancestors, favouring tendencies to approach small wriggling objects, because small wriggling objects are usually worms. Since they are not always worms but sometimes angler fish lures, there may well be some selection on prey fish to be cautious, or to sharpen up their powers of discrimination. To the extent that lures are good mimics of worms, we may surmise that selection has acted on the angler fishes' ancestors to improve them, in the face of improved discrimination by their prey. Some prey are caught, and angler fish do make a living, so some manipulation is going on successfully.
It is convenient to use the metaphor of an arms race whenever we have progressive improvements in adaptations in one lineage, as an evolutionary response to progressive counter-improvements in an enemy lineage. It is important to realize who are the parties that are “racing” against one another. They are not individuals but lineages. To be sure, it is individuals who attack and defend, individuals who kill or resist being killed. But the arms race takes place on the evolutionary time-scale, and individuals do not evolve. It is lineages that evolve, and lineages that exhibit progressive trends in response to the selection pressures set up by the progressive improvements in other lineages.
One lineage will tend to evolve adaptations to manipulate the behaviour of another lineage, then the second lineage will evolve counter-adaptations. We must obviously be interested in any general rules governing whether one or the other lineage can “win”, or have a built-in advantage. It was just such a built-in advantage that Alexander attributed to parents over their children. Apart from his major theoretical argument which, as we have seen, is now not favoured, he also plausibly suggested various practical advantages of parents over offspring: “...the parent is bigger and stronger than the offspring, hence in a better position to impose its will” (Alexander 1974). This is true, but we must not forget the lesson of the previous paragraph, that an arms race is run in evolutionary time. In any one generation the muscles of a parent are stronger than those of its offspring, and whoever controls those muscles must have the upper hand. But the {62} question at issue is, who controls the parents' muscles? As Trivers (1974) says, “An offspring cannot fling its mother to the ground at will and nurse ... the offspring is expected to employ psychological rather than physical tactics”.
Krebs and I suggested that animal signals could be thought of as employing psychological tactics in rather the same way as human advertisements. Advertisements are not there to inform, or to misinform, they are there to persuade. The advertiser uses his knowledge of human psychology, of the hopes, fears and secret motives of his targets, and he designs an advertisement which is effective in manipulating their behaviour. Packard's (1957) expose of the deep psychological techniques of commercial advertisers makes fascinating reading for the ethologist. A supermarket manager is quoted as saying “People like to see a lot of merchandise. When there are only three of four cans of an item on a shelf, they just won't move”. The obvious analogy with lek birds does not lose its value merely because the physiological mechanism of the effect will probably prove to be different in the two cases. Hidden cine-cameras recording the eye-blinking rate of housewives in a supermaket indicated that in some cases the effect of the multiplicity of bright-coloured packages was to induce a mild hypnoidal trance.
K. Nelson once gave a talk at a conference, entitled “Is bird song music? Well, then, is it language? Well, then, what is it?” Perhaps bird song is more akin to hypnotic persuasion, or to a form of drugging. The nightingale's song induced in John Keats a drowsy numbness “...as though of hemlock I had drunk”. Might it not have an even more powerful effect on the nervous system of another nightingale? If nervous systems are susceptible to drug-like influences via the normal sense organs, should we not positively expect that natural selection would have favoured the exploitation of such possibilities, would have favoured the development of visual, olfactory, or auditory “drugs”?
A neurophysiologist, presented with an animal with a complex nervous system and asked to manipulate the animal's behaviour, may insert electrodes in sensitive points of the brain and stimulate electrically, or make pinpoint lesions. An animal normally does not have direct access to the brain of another animal, although I shall mention one example, the so-called brainworm, in Chapter 12. But the eyes and ears are also entry ports to the nervous system, and there might be patterns of light or sound which, if properly deployed, could be as effective as direct electrical stimulation. Grey Walter (1953) vividly illustrates the power of flashing lights tuned to the frequency of human EEG rhythms: in one case a man felt “an irresistible urge to strangle the person next to him”.
If we were asked to imagine what an “auditory drug” would sound like, say for making the soundtrack of a science fiction film, what would we say? The incessant rhythm of an African drum; the eerie trilling of the tree cricket {63} Oecanthus, of which it has been said that if moonlight could be heard that is how it would sound (quoted in Bennet-Clark 1971); or the song of the nightingale? All three strike me as worthy candidates, and I believe all three have been in some sense designed for the same purpose: the manipulation of one nervous system by another, in a way that is not in principle different from manipulation of a nervous system by a neurophysiologist's electrodes. Of course if an animal sound powerfully affects a human nervous system this is probably incidental. The hypothesis being advanced is that selection has shaped animal sounds to manipulate some nervous system, not necessarily a human one. The snort of a pig-frog Rana grylio may affect another pig-frog as the nightingale affected Keats, or the skylark Shelley. The nightingale happened to be a better example for me to choose, because human nervous systems can be moved to deep emotion by nightingale song, whereas they are usually moved to laugh at pig-frog snorts.
Consider that other celebrated songster, the canary, since much happens to be known about the physiology of its reproductive behaviour (Hinde & Steel 1978). If a physiologist wants to bring a female canary into reproductive condition, increase the size of her functional ovary and cause her to start nest-building and other reproductive behaviour patterns, there are various things he can do. He can inject her with gonadotropins or oestrogens. He can use electric light to increase the day-length that she experiences. Or, most interestingly from our point of view, he can play her a tape recording of male canary song. It apparently has to be canary song; budgerigar song will not do, although budgerigar song has a similar effect on female budgerigars.
Now suppose a male canary wanted to bring a female into “reproductive condition, what could he do? He does not have a syringe to inject hormones. He cannot switch on artificial lights in the female's environment. Of course what he does is sing. The particular pattern of sounds that he makes enters the female's head through her ears, is translated into nerve impulses, and bores insidiously into her pituitary. The male does not have to synthesize and inject gonadotropins; he makes the female's pituitary work to synthesize them for him. He stimulates her pituitary by means of nerve impulses. They are not “his” nerve impulses, in the sense that they all occur within the female's own nerve cells. But they are his in another sense. It is his particular sounds which are subtly fashioned to make the female's nerves work on her pituitary. Where a physiologist might pump gonadotropins into the female's breast muscle, or electric current into her brain, the male canary pours song into her ear. And the effect is the same. Schleidt (1973) discusses other examples of such “tonic” effects of signals on the physiology of the receiver.
“Bores insidiously into her pituitary” may be too much for some readers. Certainly it begs important questions. The obvious point that will be made is that the female may benefit from the transaction as much as the male, in {64} which case “insidiously” and “manipulation” are inappropriate words. Alternatively, if there is any insidious manipulation it may be by the female on the male. Maybe the female “insists” upon an exhausting performance of song by her mate before she will come into reproductive condition, thereby selecting only the most robust male for a mate. I think something along those lines is quite probably the explanation of the extraordinarily high rate of copulation observed in large cats (Eaton 1978). Schaller (1972) followed a sample male lion for 55 hours, during which he copulated 157 times, with an average inter-copulation interval of 21 minutes. Ovulation in cats is induced by copulation. It seems plausible that the prodigious copulation rates shown by male lions are the end product of a runaway arms race, in which females insisted on progressively more copulations before ovulating, and males were selected for ever-increased sexual stamina. Lion copulation stamina might be regarded as the behavioural equivalent of a peacock's tail. A version of this hypothesis is compatible with that of Bertram (1978), that females have devalued the currency of copulation as a means of reducing the occurrence of disruptive male fights.
To continue the quotation from Trivers on psychological manipulation tactics that children might use against their parents:
Since an offspring will often have better knowledge of its real needs than will its parents, selection should favour parental attentiveness to signals from its offspring that apprize the parent of the offspring's condition ... But once such a system has evolved, the offspring can begin to employ it out of context. The offspring can cry not only when it is famished but also when it merely wants more food than the parent is selected to give ... Selection will then of course favor parental ability to discriminate the two uses of the signals, but still subtler mimicry and deception by the offspring are always possible [Trivers 1974].
We have arrived again at the main question raised by arms races. Are there any generalizations we can make about which side is likely to “win” an arms race?
First, what does it mean to speak of one side or the other “winning”? Does the “loser” eventually go extinct? At times this may happen. A bizarre possibility is suggested by Lloyd and Dybas (1966) for periodical cicadas (Simon, 1979, gives an entertaining recent account). Periodical cicadas are three species of the genus Magicicada. All three species have two varieties, a 17-year and a 13-year variety. Each individual spends 17 (or 13) years feeding as an underground nymph, before emerging for a few weeks of adult reproductive life, after which it dies. In any particular area all life cycles are synchronized, with the result that each area experiences a cicada plague {65} every 17 (or 13) years. The functional significance of this is presumably that would-be predators or parasites are swamped in plague years and starved in intervening years. The risk to an individual who remains synchronized is therefore less than the risk to one who breaks synchrony and emerges, say, a year early. But given the admitted advantage of synchrony, why didn't the cicadas settle on a shorter life cycle than 17 (or 13) years, thereby reducing the unfortunate delay before reproduction? Lloyd and Dybas's suggestion is that the long life cycle is the end result of an “evolutionary race through time” with a now extinct predator (or parasitoid). “This hypothetical parasitoid presumably had a life cycle that was almost synchronized with, and nearly equal in length to (but always slightly less than), the ancestral protoperiodical cicada. As the theory goes, the cicadas finally outran their parasitoid pursuer and the poor specialized beast went extinct” (Simon 1979). Ingeniously, Lloyd and Dybas go further and suggest that the arms race culminated in life cycles lasting a prime number of years (13 or 17) because otherwise a predator with a shorter life cycle might synchronize with the cicadas every second or every third time around!
There is really no need for the “loser” lineage of an arms race to go extinct, as suggested for the cicada parasitoid. It may be that the “winner” is such a rare species that it constitutes a relatively negligible risk to individuals of the “loser” species. The winner only wins in the sense that its adaptations against the loser are not effectively countered. This is good for individuals of the winner lineage, but it may not be very bad for individuals of the loser lineage who, after all, are running other races simultaneously against other lineages, possibly very successfully.
This “rare-enemy effect” is an important example of an asymmetry in selection pressures acting on the two sides of an arms race. Although we gave no formal models, Krebs and I considered one such asymmetry qualitatively under the catchphrase heading of the “life/dinner principle”, named after a fable of Aesop which was called to our attention by M. Slatkin: “The rabbit runs faster than the fox, because the rabbit is running for his life while the fox is only running for his dinner”. The general point here is that for an animal on one side of the arms race the penalty of failure is more severe than for an animal on the other side of the arms race. Mutations that make foxes run more slowly than rabbits might therefore survive in the fox gene-pool longer than mutations that cause rabbits to run slowly can expect to survive in the rabbit gene-pool. A fox may reproduce after being outrun by a rabbit. No rabbit has ever reproduced after being outrun by a fox. Foxes may therefore be better able than rabbits to divert resources away from adaptations for running fast and into other adaptations; rabbits therefore appear to “win” the arms race as far as running speed is concerned. The point is really one about asymmetries in strengths of selection pressure.
Probably the simplest asymmetry arises from what I have just called the {66} rare-enemy effect, where one side in the arms race is rare enough to exert a relatively negligible influence on any given individual on the other side. This can be illustrated by the angler fish example again, and the example has the additional merit of showing that there is no necessary reason why prey (“rabbits”) should come out “ahead” of predators (“foxes”). Assume that angler fish are rather rare, so however unpleasant it may be for an individual prey to be caught by an angler fish, the risk of this happening to a randomly designated individual is not high. Any adaptation costs something. For a small fish to discriminate angler fish lures from real worms it needs sophisticated visual processing apparatus. To make this apparatus may cost it resources which could otherwise have been put into, say gonads (see Chapter 3). Then again, using the apparatus takes time, which could have been spent on courting females or defending a territory, or indeed chasing the prey if it is judged to be a real worm. Finally, a fish that is very cautious about approaching worm-like objects may cut its risk of being eaten, but it also increases its risk of starving. This is because it may avoid many perfectly good worms because they might have been angler fish lures. It may well be that the balance of costs and benefits favours complete recklessness on the part of some prey animals. Individuals that always try to eat small wriggling objects, and damn the consequences, may, on average, do better than individuals that pay the costs of attempting to discriminate real worms from angler fish lures. William James made much the same point in 1910: “There are more worms unattached to hooks than impaled upon them; therefore, on the whole, says Nature to her fishy children, bite at every worm and take your chances” (quoted by Staddon 1981).
Now look at it from the point of view of the angler fish. It, too, needs to spend resources on apparatus to outwit its opponents in the arms race. The resources that go into making a lure could have been put into gonads. The time spent sitting motionless waiting patiently for prey could have been spent in actively searching for a mate. But the angler fish absolutely depends on the success of its lure for survival. An angler fish that is not well equipped to lure prey will starve. A prey fish that is not well equipped to avoid being lured may run only a small risk of being eaten, and may more than compensate for the risk by saving the cost of making and using the equipment.
In this example the angler wins the arms race against its prey, simply because its side of the arms race constitutes a relatively negligible threat to any given individual of the other side, through being rare. This does not, by the way, mean that selection will favour predators that take steps to be rare, or to constitute a small threat to their prey! Within the angler fish population selection will favour those individuals who are the most efficient killers and the most prolific contributors to reducing the rarity of their species. But the angler fish lineage may “keep ahead” in the arms race against their {67} prey, simply as an incidental consequence of being rare for some other reason, rare in spite of all their efforts.
In addition, there are likely to be frequency-dependent effects for various reasons (Slatkin & Maynard Smith 1979). For instance, as individuals on one side in an arms race, say angler fish, become rarer, they will constitute a steadily weaker threat to individuals on the other side of the arms race. Therefore individuals on the other side will be selected to shift precious resources away from adaptations concerned with this particular arms race. In the present example, prey fish are selected to shift resources away from anti-angler adaptations and into other things such as gonads, possibly to the extent of total recklessness as suggested above. This will make life easier for individual anglers, who will therefore become more numerous. Anglers will then come to constitute a severer threat to the prey fish, who will then be selected to shift resources back into anti-angler adaptations. As usual in such arguments, we do not have to imagine an oscillation. There may, instead, be an evolutionarily stable endpoint to the arms race; stable, that is, until an environmental change shifts the relevant costs and benefits.
Obviously, until we know a great deal more about costs and benefits in the field, we cannot predict the outcome of particular arms races. For present purposes this does not matter. All we need is to satisfy ourselves that in any particular arms race individuals on one side may have more to lose than individuals on the other side. The rabbit has his life to lose, the fox only has his dinner. The incompetent angler fish dies; the incompetent avoider of angler fish only runs a tiny risk of dying, and may, by saving costs, end up better off than the “competent” prey fish.
We only have to accept the general plausibility of such asymmetries in order to answer the question raised by our discussion of manipulation. We agreed that if one organism could get away with manipulating the nervous system of another, and exploiting its muscle power, selection would favour such manipulation. But we were brought up short by the reflection that selection would also favour resistance to being manipulated. Should we, then, really expect to see effective manipulation in nature? The life/dinner principle, and other such principles as the rare-enemy effect, provide us with an answer. If the individual manipulator has more to lose by failing to manipulate than the individual victim has to lose by failing to resist manipulation, we should expect to see successful manipulation in nature. We should expect to see animals working in the interests of other animals' genes.
Brood parasitism provides what may be the most striking example. A parent bird such as a reed warbler sweeps a copious flow of food from a large catchment area into the narrow funnel of its nest. There is a living to be made by any creature that evolves the necessary adaptations to insert itself in the funnel and intercept the flow. That is what cuckoos and other brood parasites have done. But the reed warbler is not an uncomplaining {68} cornucopia of free food. It is an active, complex machine, with sense organs, muscles and a brain. The brood parasite must not only have its body inserted in the host's nest. It must also infiltrate the defences of the, host's nervous system, and its ports of entry are the host's sense organs. The cuckoo uses key stimuli to unlock the host's machinery of parental care and subvert it.
The advantages of the brood-parasite way of life are so manifest that today we are taken aback to find Hamilton and Orians needing, in 1965, to defend the proposition that it has been favoured by natural selection, against theories of “degenerative breakdown” of normal breeding behaviour. Hamilton and Orians went on to provide a satisfying discussion of the probable evolutionary origins of brood parasitism, the preadaptations which preceded its evolution, and the adaptations which have accompanied its evolution.
One of these adaptations is egg mimicry. The perfection of egg mimicry, in at least some “gentes” of cuckoo, shows that foster parents are potentially capable of keen-eyed discrimination against interlopers. This only underlines the mystery of why cuckoo hosts seem to be so poor at discriminating against cuckoo nestlings. Hamilton and Orians (1965) express the problem vividly: “Young brown-headed cowbirds and European Cuckoos, as they reach their maximum dimension, dwarf their foster parents. Consider the ludicrous sight of a tiny Garden Warbler ... [Latin name leaves exact species intended unclear] standing atop a cuckoo to reach the mouth of the gaping parasite. Why does not the Garden Warbler take the adaptive measure of abandoning the nestling prematurely, especially when to the human observer it is so clearly identifiable?” When a parent is a small fraction of the size of the nestling it is feeding, the most rudimentary eyesight should suffice to show that something has gone seriously wrong with the normal parental process. Yet, at the same time, the existence of egg mimicry shows that hosts are capable of fastidious discrimination, making use of keen eyesight. How are we to explain this paradox (see also Zahavi 1979)?
One fact that helps to reduce the mystery is that there must be stronger selection pressure on hosts to discriminate against cuckoos at the egg stage than against cuckoos at the nestling stage, simply because eggs occur earlier. The benefit of detecting a cuckoo egg is the potential gaining of an entire breeding cycle in the future. The benefit of detecting a nearly iedged cuckoo is the saving of only a few days, at a time when it may be too late to breed again anyway. Another mitigating circumstance in the case of Cuculus canorus (Lack 1968) is that the host's own young is usually not there for simultaneous comparison, having been tipped out by the baby cuckoo. It is well known that discrimination is easier if there is a model actually present for comparison.
Various authors have invoked the “supernormal stimulus”, in one form or another. Thus Lack remarks (p. 88) that “the young cuckoo, with its huge {69} gape and loud begging call, has evidently evolved in exaggerated form the stimuli which elicit the feeding response of parent passerine birds. So much is this so that there are many records of adult passerine birds feeding a iedged young C. canorus raised by a different host species; this, like lipstick in the courtship of mankind, demonstrates successful exploitation by means of a "super-stimulus"”. Wickler (1968) makes a similar point, quoting Heinroth as having referred to foster parents as behaving like “addicts”, and to the cuckoo nestling as a “vice of its foster parents”. As it stands, this kind of suggestion will strike many critics as unsatisfying, because it immediately prompts a question at least as big as the one it answers. Why doesn't selection eliminate from the host species the tendency to be “addicted” to “supernormal stimuli”?
This, of course, is where the arms race concept comes in again. When a human behaves in a way that is manifestly bad for him, for instance when he continually takes poison, we may explain his behaviour in at least two ways. He may not realize that the substance that he is drinking is poison, so closely does it resemble a genuinely nutritous substance. This corresponds to the host bird's being fooled by the cuckoo's egg mimicry. Or he may be unable to save himself because of some direct subverting influence of the poison on his nervous system. Such is the case of the heroin addict who knows the drug is killing him, but who cannot stop taking it because the drug itself controls his nervous system. We have already seen that the cuckoo nestling's lipstick-like gape is regarded as a supernormal stimulus, and that foster parents have been described as apparently “addicted” to the supernormal stimulus. Could it be that the host can no more resist the supernormal manipulative power of the cuckoo nestling than the junkie can resist his fix, or than the brainwashed prisoner can resist the orders of his captor, however much it would benefit him to do so? Perhaps cuckoos have put their adaptive emphasis on mimetic deception at the egg stage, but on positive manipulation of the host's nervous system at the late nestling stage.
Any nervous system can be subverted if treated in the right way. Any evolutionary adaptation of the host nervous system to resist manipulation by cuckoo nestlings lays itself open to counter-adaptation by the cuckoos. Selection acting on cuckoos will work to find whatever chinks there may be in the hosts' newly evolved psychological armour. Host birds may be very good at resisting psychological manipulation, but cuckoos might become even better at manipulating. All we need to postulate is that, for some reason such as that suggested by the life/dinner principle or the rare-enemy effect, cuckoos have won the arms race: a cuckoo in the nest has got to manipulate its host successfully or it will surely die; its individual foster-parent will benefit somewhat if it resists manipulation, but it still has a good chance of future reproductive success in other years even if it fails to resist this particular cuckoo. Moreover, cuckoos might be sufficiently rare that the risk {70} of an individual of the foster species being parasitized is low; conversely the “risk” of an individual cuckoo's being a parasite is 100 per cent, no matter how common or rare either party to the arms race may be. The cuckoo is descended from a line of ancestors, every single one of whom has successfully fooled a host. The host is descended from a line of ancestors, many of whom may never have encountered a cuckoo in their lives, or may have reproduced successfully after being parasitized by a cuckoo. The arms race concept completes the classical supernormal stimulus explanation, by providing a functional account of the host's maladaptive behaviour, instead of leaving it as an unexplained limitation of the nervous system.
In one respect my treatment of cuckoos as manipulators may be found unsatisfying. The cuckoo is, after all, only diverting the normal parental behaviour of its host. It has not succeeded in building into the host's behavioural repertoire a whole new behaviour pattern that was not there, in some form, before. Some might find analogies with drugs, hypnotism and electrical stimulation of the brain more persuasive if an example of this more extreme kind of manipulation could be found. A possible case is the “preening invitation” display of another brood parasite, the brown-headed cowbird Molothrus ater (Rothstein 1980). Allopreening, the preening of one individual by another, is not uncommon within various species of birds. It is not particularly surprising, therefore, that cowbirds should succeed in getting other birds of several species to preen them. Again, this can be seen as a simple diversion of intraspecific allopreening, with the cowbird providing a supernormal exaggeration of the normal eliciting stimuli for allopreening. What is rather surprising is that cowbirds manage to get themselves preened by species that never engage in intraspecific allopreening.
The drug analogy is especially apt for insect “cuckoos” that use chemical means to coerce their hosts into acts that are profoundly damaging to their own inclusive fitness. Several species of ant have no workers of their own. The queens invade nests of other species, dispose of the host queen, and use the host workers to bring up their own reproductive young. The method of disposing of the host queen varies, in some species, such as the descriptively named Bothriomyrmex regicidus and B. decapitans, the parasite queen rides about on the back of the host queen and then, in Wilson's (1971) delightful description, “begins the one act for which she is uniquely specialized: slowly cutting off the head of her victim” (p. 363).
Monomorium santschii achieves the same result by more subtle means. The host workers have weapons wielded by strong muscles, and nerves attached to the muscles; why should the parasite queen exert her own jaws if she can subvert the nervous systems controlling the numerous jaws of the host workers? It does not seem to be known how she achieves it, but she does: the host workers kill their own mother and adopt the usurper. A chemical {71} secreted by the parasite queen seems the likely weapon, in which case it might be labelled a pheromone, but it is probably more illuminating to think of it as a formidably powerful drug. In line with this interpretation, Wilson (1971, p. 413) writes of symphylic substances as being “more than just elementary nutritive substances or even analogues of the natural host pheromones. Several authors have spoken of a narcotizing effect of symphylic substances”. Wilson also uses the word “intoxicant” and quotes a case in which worker ants under the influence of such a substance became temporarily disoriented and less sure of their footing.
Those who have never been brainwashed or addicted to a drug find it hard to understand their fellow men who are driven by such compulsions. In the same naive way we cannot understand a host bird's being compelled to feed an absurdly oversized cuckoo, or worker ants wantonly murdering the only being in the whole world that is vital to their genetic success. But such subjective feelings are misleading, even where the relatively crude achievements of human pharmacology are concerned. With natural selection working on the problem, who would be so presumptions as to guess what feats of mind control might not be achieved? Do not expect to see animals always behaving in such a way as to maximize their own inclusive fitness. Losers in an arms race may behave in some very odd ways indeed. If they appear disoriented and unsure of their footing, this may be only the beginning.
Let me stress again what a feat of mind-control the Monomorium santschii queen achieves. To a sterile worker ant, her mother is a kind of genetic goldmine. For a worker ant to kill her own mother is an act of genetic madness. Why do the workers do it? I am sorry I can do no more than, once again, vaguely talk about arms races. Any nervous system is vulnerable to manipulation by a clever-enough pharmacologist. There is no difficulty in believing that natural selection acting on M. santschii would seek out the weak points in the host workers' nervous system, and insert a pharmacological key in the lock. Selection on the host species would soon have plugged those weak points, whereupon selection on the parasite would improve the drug, and the arms race was under way. If M. santschii is sufficiently rare, it is easy to see that it might “win” the arms race, even though regicide is such a disastrous act for each host colony whose workers succumb to it. The overall risk of parasitization by M. santschii could be very low even though the marginal cost of regicide, given that an M. santschii queen has entered, is disastrously high. Each individual M. santschii queen is descended from a line of ancestors every one of whom has succeeded in manipulating host workers into regicide.. Each host worker is descended from a line of ancestors whose colonies may seldom have been within 10 miles of an M. santschii queen. The costs of “bothering” to be equipped to {72} resist manipulation by an occasional M. santschii queen may outweigh the benefits. Reflections such as this lead me to believe that the hosts might well lose the arms race.
Other species of parasitic ants use a different system. Instead of sending out queens to implant their eggs in host nests and using host labour there, they transport host labour back to their own nests. These are the so-called slavemaking ants. Slavemaking species have workers, but these workers devote part, or in some cases all, of their energy to going on slaving expeditions. They raid nests of other species and carry off larvae and pupae. These subsequently hatch in the slavers' nest, where they work normally, foraging and tending brood, not “realizing” that they are, in effect, slaves. The advantage of the slavemaking way of life is presumably that most of the cost of feeding the workforce in the larval stage is saved. That cost is borne by the home colony from which the slave pupae were taken.
The slavemaking habit is interesting from the present point of view, because it raises an unusual arms race asymmetry. Presumably there is an arms race between slavemaking species and slave species. Adaptations to counter slavery, for instance enlarged soldier jaws for driving off slave-raiders, should be expected in species that are victims of slave raids. But surely the more obvious countermeasure the slaves could take would simply be to withhold their labour in the slavers' nest, or to kill slavemaker brood instead of feeding them? It seems the obvious countermeasure, but there are formidable obstacles to its evolution. Consider an adaptation to “go on strike”, to refuse to work in the slavemakers' nest. The slave workers would of course have to have some means of recognizing that they had hatched in a foreign nest, but that should not be difficult in principle. The problem arises when we think in detail of how the adaptations would be passed on.
Since workers don't reproduce, all worker adaptations, in any social insect species, have to be passed on by reproductive relatives of the workers. This normally presents no insuperable problems, because workers directly assist their own reproductive relatives, so genes giving rise to worker adaptations directly assist copies of themselves in reproductives. But take the example of a mutant gene causing slave workers to go on strike. It may very effectively sabotage the slavemakers' nest, possibly wipe it out altogether. And to what effect? The area now contains one less slavemaking nest, presumably a good thing for all potential victim nests in the area, not just the eest from which the rebel slaves came, but nests containing non-striking genes as well. The same kind of problem arises in the general case of the spread of “spiteful” behaviour (Hamilton 1970; Knowlton & Parker 1979).
The only easy way the genes for striking can be preferentially passed on is for striking to benefit, selectively, the strikers' own home nest, the nest they left behind and in which their own reproductive relatives are being reared. This could happen if slavemakers habitually returned to make repeat raids {73} on the same nest, but otherwise we must conclude that anti-slavery adaptations must be confined to the period before the slave pupae have left their home nest. Once the slaves have arrived in the slavemakers' nest they effectively drop out of the arms race since they no longer have any power to influence the success of their reproductive relatives. The slavemakers can develop manipulative adaptations of any degree of sophistication, physical or chemical, pheromones or powerful drugs, and the slaves cannot evolve countermeasures.
Actually, the very fact that the slaves cannot evolve countermeasures will tend to reduce the likelihood that the manipulative techniques evolved by the slavemakers will be very sophisticated: the fact that the slaves cannot retaliate, in an evolutionary sense, means that the slavemakers do not need to spend costly resources on elaborate and sophisticated manipulation adaptations, because simple and cheap ones will do. The example of slavery in ants is rather a special one, but it illustrates a particularly interesting sense in which one side in an arms race can be said to lose completely.
A case could be made for drawing an analogy here with the hybrid frog Rana esculenta (White 1978). This common European frog, the edible frog of French restaurants, is not a species in the normal sense of the word. Individuals of the “species” are really various kinds of hybrids between two other species, Rana ridibunda and R. lessonae. There are two different diploid forms and two different triploid forms of R. esculenta. For simplicity I shall consider only one of the diploid forms, but the argument holds for all the varieties. These frogs coexist with R. lessonae. Their diploid karyotype consists of one set of lessonae chromosomes and one set of ridibunda chromosomes. At meiosis they discard the lessonae chromosomes and produce pure ridibunda gametes. They mate with lessonae individuals, thereby restoring the hybrid genotype in the next generation. In this race of Rana esculenta bodies, therefore, ridibunda genes are germ-liee replicators, lessonae genes dead-end ones. Dead-end replicators can exert phenotypic effects. They can even be naturally selected. But the consequences of that natural selection are irrelevant to evolution (see Chapter 5). To make the next paragraph easier to follow, I shall call R. esculenta H (for hybrid), R. ridibunda G (for germ-line) and R. lessonae D (for dead-end, although it should be remembered that “D” genes are dead-end replicators only when in H frogs; when in “D” frogs they are normal germ-line replicators).
Now, consider a gene in the D gene-pool which exerts an effect on D bodies to make them refuse to mate with H. Such a gene should be favoured by natural selection over its H-tolerating allele, since the latter will tend to end up in H bodies of the next generation, and will be discarded at meiosis. G genes are not discarded at meiosis, and they will tend to be selected if they influence H bodies so that they overcome the reluctance of D individuals to mate with them. We should therefore see an arms race between G genes {74} acting on H bodies, and D genes acting on D bodies. In those respective bodies, both sets of genes are germ-liee replicators. But what about D genes acting on H bodies? They should have just as powerful an influence over H phenotypes as G genes, since they constitute exactly half of the H genome. Naively, we might expect them to carry their arms race against G genes over into the H bodies which they share. But in H bodies, those D genes are in the same position as ants that have been taken as slaves. Any adaptation that they mediate in the H body cannot be passed on to the next generation; for the gametes produced by the H individual, regardless of how his developing phenotype, and indeed his survival, may have been influenced by D genes, are strictly G gametes. Just as would-be slave ants can be selected to resist being taken into captivity while they are still in their home nest, but cannot be selected to subvert the slavemaking nest once they are in it, so D genes can be selected to influence D bodies so that they resist being incorporated in H genomes in the first place, but once so incorporated they are no longer under selection, even though they can still have phenotypic effects. They lose the arms race because they are a dead-end. A similar argument could be made for fish of the hybrid “species” Poeciliopsis monacha-occidentalis (Maynard Smith 1978a).
The inability of slaves to evolve counter-adaptations was originally invoked by Trivers and Hare (1976), in their theory of the sex-ratio arms race in social Hymenoptera. This is one of the best known of recent discussions of a particular arms race, and it is worth considering further. Elaborating on ideas of Fisher (1930a) and Hamilton (1972), Trivers and Hare reasoned that the evolutionarily stable sex ratio in ant species with one singly mated queen per nest cannot be simply predicted. If the queen is assumed to have all power over the sex of reproductive offspring (young queens and males), the stable ratio of economic investment in male and female reproductives is 1:1. If, on the other hand, non-laying workers are assumed to hold all power over investment in young, the stable ratio will be 3:1 in favour of females, ultimately because of the haplodiploid genetic system. There is, therefore, a potential conflict between queen and workers. Trivers and Hare reviewed the, admittedly imperfect, available data, and reported a good average fit to the 3:1 prediction, from which they concluded that they had found evidence for worker power winning the battle against queen power. It was a clever attempt to use real data to test a hypothesis of a kind that is often criticized as uetestable, but like other innovative first attempts it is easy to find fault with it. Alexander and Sherman (1977) complained about Trivers and Hare's handling of the data, and also suggested an alternative explanation for the female-biased sex ratio common in ants. Their explanation (“local mate competition”), like that of Trivers and Hare, was originally derived from Hamilton, in this case his paper on extraordinary sex ratios (1967).
This controversy has had the good effect of stimulating further work. {75} Especially illuminating in the context of arms races and manipulation is the paper by Charnov (1978), which is concerned with the origins of eusociality, and which introduces a potentially important version of the life/dinner principle. His argument works for diploid as well as haplodiploid organisms, and I shall consider the diploid case first. Consider a mother whose elder children have still not left the nest when the next brood hatches. When the time comes for them to leave the nest and begin their own reproduction, the young have the option, instead, of staying behind and helping to rear their young siblings. As is now well known, all other things being equal such a fledgling should be genetically indifferent between rearing offspring and rearing full siblings (Hamilton 1964a, b). But suppose the old mother could exert any manipulative power over the decision of her elder children: would she “prefer” that they leave and rear families of their own, or stay and rear her next brood? Obviously that they should stay and rear her next brood, since grandchildren are half as valuable to her as children. (The argument as it stands is incomplete. If she manipulated all her children for the whole of her life into rearing yet more non-producing child labourers, her germ-line would peter out. We must assume that she manipulates some offspring of the same genetic type into developing into reproductives and others into developing as workers.) Selection will, then, favour such manipulative tendencies in parents.
Normally, when we postulate selection in favour of manipulation we dutifully pay lip service to counter-selection on the victim to resist manipulation. The beauty of Charnov's point is that in this case there will be no counter-selection. The “arms race” is a walk-over because one side, so to speak, doesn't even try. The offspring being manipulated are, as we have already seen, indifferent to whether they rear young siblings or offspring of their own (again assuming all other things equal). Therefore, although we may postulate reverse manipulation by offspring of parents, this is bound, at least in the simple example visualized by Charnov, to be outweighed by parental manipulation of offspring. This is an asymmetry to be added to the list of parental advantages offered by Alexander (1974), but I find it more generally convincing than any others on the list.
At first sight it might appear that Charnov's argument does not apply to haplodiploid animals, and that would be a pity since most social insects are haplodiploid. But this view is mistaken. Charnov himself shows this for the special assumption that the population has an unbiased sex ratio, in which case even in haplodiploid species females are indifferent between rearing siblings (r = the average of 3/4 and 1/4) and rearing offspring (r = 1/2). But Craig (1980) and Grafen (in preparation) independently show that Charnov did not even need to assume an unbiased sex ratio. The potential worker is still indifferent between rearing siblings and rearing offspring at any conceivable population sex ratio. Thus suppose the population sex ratio is female-biased, {76} even suppose it conforms to Trivers and Hare's predicted 3:1. Since the worker is more closely related to her sister than to her brother or her offspring of either sex, it might seem that she would “prefer” to rear siblings over offspring given such a female-biased sex ratio: is she not gaining mostly valuable sisters (plus only a few relatively worthless brothers) when she opts for siblings? But this reasoning neglects the relatively great reproductive value of males in such a population as a consequence of their rarity. The worker may not be closely related to each of her brothers, but if males are rare in the population as a whole each one of those brothers is correspondingly highly likely to be an ancestor of future generations.
The mathematics confirm that Charnov's conclusion is even more general than he suggested. In both diploid and haplodiploid species, at any population sex ratio, an individual female is theoretically indifferent whether she herself rears offspring or younger siblings. She is not, however, Indifferent whether her offspring rear their own children or their siblings: she prefers them to rear their siblings (her offspring) over their offspring (her grandchildren). Therefore if there is any question of manipulation Ie this situation, parental manipulation of offspring is more likely than offspring manipulation of parents.
It might appear that Charnov's, Craig's and Grafen's conclusions radically contradict those of Trivers and Hare on sex ratios in social Hymenoptera. The statement that, at any sex ratio, a female hymenopteran is indifferent between rearing siblings and rearing offspring, sounds tantamount to saying that she is also indifferent to what the sex ratio in her nest Is. But this is not so. It is still true that, given the assumption of worker control over investment in male and female reproductives, the resulting evolutionarily stable sex ratio will not be necessarily the same as the evolutionarily stable sex ratio given queen control. In this sense a worker is not Indifferent to the sex ratio: she may well work to shift the sex ratio away from what the queen is “trying” to achieve.
Trivers and Hare's analysis of the exact nature of the conflict between queen and workers over the sex ratio can be extended in ways that further illuminate the concept of manipulation (e.g. Oster & Wilson 1978). The following account is derived from Grafen (in preparation). I shall not anticipate his conclusions in detail, but wish to emphasize one principle which is explicit in his analysis as well as implicit in that of Trivers and Hare. The question is not “Has the "best" sex ratio been successfully achieved?” On the contrary, we make a working assumption that natural selection has produced a result, given some constraints, and then ask what those constraints are (see Chapter 3). In the present case we follow Trivers and Hare in recognizing that the evolutionarily stable sex ratio depends crucially upon which parties to the arms race have practical power, but we recognize a wider range of possible dispositions of power than they did. In {77} effect, Trivers and Hare deduced the consequences of two alternative assumptions about practical power; firstly the assumption that the queen exerts all the power, and secondly the assumption that the workers exert all the power. But many other possible assumptions could be made, and each gives rise to a different prediction of the evolutionarily stable sex ratio. In other parts of their paper, indeed, Trivers and Hare consider some of these, for instance the assumption that workers are able to lay their own male eggs.
Grafen, like Bulmer and Taylor (in preparation), has explored the consequences of assuming that power is divided as follows: the queen has absolute power over the sex of the eggs that she lays; the workers have absolute power over feeding the larvae. The workers can thus determine how many of the available female eggs shall develop into queens and how many into workers. They have the power to starve the young of one sex or the other, but they have to work within the constraint of what the queen gives them in the way of eggs. Queens have the power to lay eggs in any sex ratio they choose, including withholding, totally, eggs of one sex or the other. But, once laid, those eggs are at the mercy of the workers. A queen might, for instance, play the strategy (in the game theory sense) of laying only male eggs in a given year. Reluctant as we might expect them to be, the workers have no option but to rear their brothers. The queen, in this case, can preempt certain worker strategies, such as “preferentially feed sisters”, simply because she “plays” first. But there are other things workers can do.
Using game theory, Grafen shows that only certain queen strategies are evolutionarily stable replies to particular worker strategies, and only certain worker strategies are evolutionarily stable replies to particular queen strategies. The interesting question is, What are the evolutionarily stable combinations of worker and queen strategies? It turns out that there is more than one answer, and there can be as many as three evolutionarily stable states for a given set of parameters. Grafen's particular conclusions are not my concern here, although I will remark that they are interestingly “counterintuitive”. What is my concern is that the evolutionarily stable state of the model population depends upon the assumptions we make about power. Trivers and Hare contrasted two possible absolute assumptions (absolute worker power versus absolute queen power). Grafen investigated one plausible division of power (queens have power over eggs, workers over larval feeding). But, as I have already noted, numerous other assumptions about power could be made. Each assumption generates different predictions about evolutionarily stable sex ratios, and tests of the predictions can therefore be regarded as providing evidence about the disposition of power in the nest.
For instance, we might focus our research attention on the exact moment when a queen “decides” whether to fertilize a given egg or not. It is plausible to assume that, since the event takes place within the queen's own body, that {78} particular decision is likely to have been selected to benefit the queen's genes. Plausible it may be, but it is precisely this kind of assumption that the doctrine of the extended phenotype is going to call in question. For the moment, we simply note the possibility that workers might manipulate the queen's nervous system, by pheromonal or other means, so as to subvert her behaviour in their genetic interests. Similarly, it is worker nerves and muscles that are immediately responsible for feeding the larvae, but we are not, therefore, necessarily entitled to assume that worker limbs move only in the interests of worker genes. As is well known, there is massive pheromonal traffic flow from queen to workers, and it is easy to imagine powerful manipulation of worker behaviour by queens. The point is that each assumption about power which we might make yields a testable prediction about sex ratios, and it is for this insight that we have to thank Trivers and Hare, not for the particular model whose predictions they happened to test.
It is even conceivable, in some Hymenoptera, that males might exert power. Brockmann (1980) is making an intensive study of mud-daubing wasps Trypoxylon politum. These are “solitary” (as opposed to truly social) wasps, but they are not always totally alone. As in other sphecids, each female builds her own nest (in this case out of mud), provisions it with paralysed prey (spiders), lays one egg on the prey, then seals up the nest and begins the cycle again. In many Hymenoptera, the female carries a lifetime's supply of sperm from one brief period of insemination early in life. T. politum females, however, copulate frequently throughout adult life. Males haunt female nests, losing no opportunity to copulate with the female on each of her returns to the nest. A male may spend hours at a time sitting passively in the nest, probably helping to guard it against parasites, and fighting with other males who attempt to enter. Unlike most male Hymenoptera then, the male T. politum is present at the scene of the action. Might he not, therefore, be potentially in a position to influence the sex ratio, in the same kind of way as has been postulated for worker ants?
If males did exert power, whipt would we expect the consequences to be? Since a male passes all his genes on to his daughters, and none to his mate's sons, genes tending to make males favour daughters over sons would be favoured. If males exerted total power, completely determining the sex ratio of their mates' offspring, the consequence would be odd. No males would be born in the first generation of male power. As a result, in the following year all eggs laid would be unfertilized and therefore male. The population would therefore oscillate violently and then go extinct (Hamilton 1967). If males exerted a limited amount of power, less drastic consequences would follow, the situation being formally analogous to that of the “driving X chromosome” in the normal diploid genetic system (Chapter 8). In any case a male hymenopteran, if he found himself in a position to influence the sex ratio of his mate's children, would be expected to try to do so in a female direction. {79} He might do this by trying to influence his mate's decision whether to release sperm from her spermatheca. It is not obvious how he might actually do this, but it is known that honeybee queens take longer over laying a female egg than a male one, perhaps using the extra time to achieve fertilization. It would be interesting to try experimentally interrupting a queen in the middle of egg-laying, to see if the delay increased the chance of a female egg emerging.
Do male T. politum show any behaviour that we might suspect of being an attempt at such manipulation, for example do they behave as if trying to prolong egg-laying? Brockmann describes a curious behaviour pattern called “holding”. This is seen alternating with copulation during the final minutes before the female lays her egg. In addition to brief copulations throughout the provisioning phase of the nest, the final egg-laying and sealing up of the nest is heralded by a prolonged bout of repeated copulations, which lasts many minutes. The female goes head first into the vertical, organ pipe-shaped mud nest, and pushes her head up into the cluster of paralysed spiders lodged in the top of the nest. Her abdomen is facing the entrance at the bottom of the nest, and in this position the male copulates with her. The female then turns round so that she is head downwards facing out of the nest, and probes the spiders with the tip of her abdomen, as if about to lay an egg. The male meanwhile “holds” her head in his forelegs for about half a minute, grabs her antennae and pulls her downwards away from the spiders. She then turns around and they copulate again. She again turns to probe the spiders with her abdomen, the male again holds her head and drags her down. The whole cycle repeats some half dozen times. Finally, after one especially long bout of head-holding, the female lays her egg.
Once the egg is laid, its sex is determined. We have already considered the hypothetical possibility of worker ants manipulating their mother's nervous system, forcing her to change her fertilizing decision in their genetic interest. Brockmann's suggestion is that male T. politum might attempt similar subversion, and that the head-holding and dragging behaviour may be a manifestation of their manipulation technique. When the male seizes the female's antennae and drags her away from the spiders which she is probing with her abdomen, is he forcing a postponement of egg-laying as a means of increasing the chance of the egg's being fertilized in the oviduct? The plausibility of this suggestion might depend on exactly where the egg is in the female's body during the time of holding. Or is he, as Dr W. D. Hamilton has suggested to us, blackmailing the female by, in effect, threatening to bite her head off unless she postpones egg-laying until after further copulation? Perhaps he gains by repeated copulations, simply by flooding the female's internal passages with his sperm, thereby raising the chance that the egg will encounter a sperm without one being deliberately released from the spermatheca by the female. Clearly these are just suggestions for further {80} research, and Brockmann, together with Grafen and others, is following them up. Preliminary indications, I understand, do not support the hypothesis that males actually succeed in exerting power over the sex ratio.
This chapter is intended to begin the process of undermining the reader's confidence in the central theorem of the selfish organism. That theorem states that individual animals are expected to work for the good of their own inclusive fitness, for the good of copies of their own genes. The chapter has shown that animals are quite likely to work hard and vigorously for the good of some other individual's genes, and to the detriment of their own. This is not necessarily just a temporary departure from the central theorem, a brief interlude of manipulative exploitation before counterselection on the victim lineage redresses the balance. I have suggested that fundamental asymmetries such as the life/dinner principle, and the rare-enemy effect, will see to it that many arms races reach a stable state in which animals on one side permanently work for the benefit of animals on the other side, and to their own detriment; work hard, energetically, wantonly against their own genetic interests. When we see the members of a species consistently behaving in a certain way, “anting” in birds or whatever it is, we are apt to scratch our heads and wonder how the behaviour benefits the animals' inclusive fitness. How does it benefit a bird to allow ants to run all through its feathers? Is it using the ants to clean it of parasites, or what? The conclusion of this chapter is that we might instead ask whose inclusive fitness the behaviour is benefiting! is it the animal's own, or that of some manipulator lurking behind the scenes? In the case of “anting” it does seem reasonable to speculate about advantages to the bird, but perhaps we should give at least a sideways glance at the possibility that it is an adaptation for the good of the ants!
| << | {81} | >> |
In 1957, Beezer argued that “the gene” could no longer continue as a single, unitary concept. He split it into three: the muton was the minimum unit of mutational change; the recon was the minimum unit of recombination; and the cistron was defined in a way that was directly applicable only to microorganisms, but it was effectively equivalent to the unit responsible for synthesizing one polypeptide chain. I have suggested adding a fourth unit, the optimon, the unit of natural selection (Dawkins 1978b). Independently, E. Mayr (personal communication) coined the term “selecton” to serve the same purpose. The optimon (or selecton) is the “something” to which we refer when we speak of an adaptation as being for the good of something. The question is, what is that something; what is the optimon?
The question of what is the “unit of selection” has been debated from time to time in the literature both of biology (Wynne-Edwards 1962; Williams 1966; Lewontin 1970a; Leigh 1977; Dawkins 1978a; Alexander & Borgia 1978; Wright 1980) and of philosophy (Hull 1980a, b; Wimsatt in preparation). At first sight it seems a rather uselessly theological argument. Hull, indeed, explicitly regards it as “metaphysical” (though none the worse for that). I must justify my interest in it. Why does it matter what we consider to be the unit of selection? There are various reasons, but I shall give only one. I agree with Williams (1966), Curio (1973) and others that there is a need to develop a serious science of adaptation — teleonomy as Pittendrigh (1958) called it. The central theoretical problem of teleonomy will be that of the nature of the entity for whose benefit adaptations may be said to exist. Are they for the benefit of the individual organism, for the benefit of the group or species of which it is a member, or for the benefit of some smaller unit inside the individual organism? As already emphasized in Chapter 3, this really matters. Adaptations for the good of a group will look quite different from adaptations for the good of an individual. {82}
Gould (1977b) sets out what, at first sight, appears to be the issue:
The identification of individuals as the unit of selection is a central theme in Darwin's thought... Individuals are the unit of selection; the “struggle for existence” is a matter among individuals ... In the last fifteen years challenges to Darwin's focus on individuals have come from above and from below. From above, Scottish biologist V. C. Wynne-Edwards raised orthodox hackles fifteen years ago by arguing that groups, not individuals, are units of selection, at least for the evolution of social behaviour. From below, English biologist Richard Dawkins has recently raised my hackles with his claim that genes themselves are units of selection, and individuals merely their temporary receptacles.
Gould is invoking the idea of a hierarchy of levels in the organization of life. He sees himself as perched on an intermediate rung of a ladder, with group selectionists above and gene selectionists below. The present chapter and the next will show that this kind of analysis is false. There is, of course, a hierarchy of levels of biological organization (see next chapter), but Gould is applying it incorrectly. The conventional dispute between group selection and individual selection is different in category from the apparent dispute between individual selection and gene selection. It is wrong to think of the three as arranged on a single-dimensional ladder, such that words like “above” and “below” have transitive meaning. I shall show that the well-aired dispute between group and individual is concerned with what I shall call “vehicle selection” and can be regarded as a factual biological dispute about units of natural selection. The attack “from below”, on the other hand, is really an argument about what we ought to mean when we talk about a unit of natural selection.
To anticipate the conclusion of these two chapters, there are two ways in which we can characterize natural selection. Both are correct; they simply focus on different aspects of the same process. Evolution is the external and visible manifestation of the differential survival of alternative replicators (Dawkins 1978a). Genes are replicators; organisms and groups of organisms are best not regarded as replicators; they are vehicles in which replicators travel about. Replicator selection is the process by which some replicators survive at the expense of other replicators. Vehicle selection is the process by which some vehicles are more successful than other vehicles in ensuring the survival of their replicators. The controversy about group selection versus individual selection is a controversy about the rival claims of two suggested kinds of vehicle. The controversy about gene selection versus individual (or group) selection is a controversy about whether, when we talk about a unit of selection, we ought to mean a vehicle at all, or a replicator. Much the same {83} point has been realized by the philosopher D. L. Hull (1980 a, b), but after some thought I prefer to persist with my own terminology rather than adopt his “interactors” and “evolvors”.
I define a replicator as anything in the universe of which copies are made. Examples are a DNA molecule, and a sheet of paper that is xeroxed. Replicators may be classified in two ways. They may be “active” or “passive”, and, cutting across this classification, they may be “germ-line” or “dead-end” replicators.
An active replicator is any replicator whose nature has some influence over its probability of being copied. For example a DNA molecule, via protein synthesis, exerts phenotypic effects which influence whether it is copied; this is what natural selection is all about. A passive replicator is a replicator whose nature has no influence over its probability of being copied. A xeroxed sheet of paper at first sight seems to be an example, but some might argue that its nature does influence whether it is copied, and therefore that it is active: humans are more likely to xerox some sheets of paper than others, because of what is written on them, and these copies are, in their turn, relatively likely to be copied again. A section of DNA that is never transcribed might be a genuine example of a passive replicator (but see Chapter 9 on “selfish DNA”).
A germ-line replicator (which may be active or passive) is a replicator that is potentially the ancestor of an indefinitely long line of descendant replicators. A gene in a gamete is a germ-line replicator. So is a gene in one of the germ-line cells of a body, a direct mitotic ancestor of a gamete. So is any gene in Amoeba pro tens. So is an RNA molecule in one of Orgel's (1979) test-tubes. A dead-end replicator (which also may be active or passive) is a replicator which may be copied a finite number of times, giving rise to a short chain of descendants, but which is definitely not the potential ancestor of an indefinitely long line of descendants. Most of the DNA molecules in our bodies are dead-end replicators. They may be the ancestors of a few dozen generations of mitotic replication, but they will definitely not be long-term ancestors.
A DNA molecule in the germ-line of an individual who happens to die young, or who otherwise fails to reproduce, should not be called a dead-end replicator. Such germ-lines are, as it turns out, terminal. They fail in what may metaphorically be called their aspiration to immortality. Differential failure of this kind is what we mean by natural selection. But whether it succeeds in practice or not, any germ-line replicator is potentially immortal. It “aspires” to immortality but in practice is in danger of failing. All the DNA molecules in fully sterile social insect workers, however, are true dead-end replicators. They do not even aspire to replicate indefinitely. Workers lack a germ-line, not as a matter of misfortune but as a matter of design. In this respect they resemble human liver cells rather than the spermatogonia of a human who happens to be celibate. There may be awkward intermediate {84} cases, for Instance the “sterile” worker who becomes facultatively fertile if her mother dies, and the Sireptocarpus leaf which is not expected to propagate a new plant but which can do so if planted as a cutting. But this is getting theological: let us not worry precisely how many angels can dance on a pin's head.
As I said, the active/passive distinction cuts across the germ-liee/dead-end distinction. All four combinations are conceivable. Particular interest attaches to one of the four, the active germ-line replicator, for it is, I suggest, the “optimon”, the unit for whose benefit adaptations exist. The reason active germ-line replicators are important units is that, wherever in the universe they may be found, they are likely to become the basis for natural selection and hence evolution. If replicators exist that are active, variants of them with certain phenotypic effects tend to oyt-replicate those with other phenotypic effects. If they are also germ-line replicators, these changes in relative frequency can have long-term, evolutionary impact. The world tends automatically to become populated by germ-line replicators whose active phenotypic effects are such as to ensure their successful replication. It is these phenotypic effects that we see as adaptations to survival. When we ask whose survival they are adapted to ensure, the fundamental answer has to be not the group, nor the individual organism, but the relevant replicators themselves.
I have previously summed up the qualities of a successful replicator “in a slogan reminiscent of the French Revolution: Longevity, Fecundity, Fidelity” (Dawkins 1978a). Hull (1980b) explains the point clearly.
Replicators need not last forever. They need only last long enough to produce additional replicators [fecundity] that retain their structure largely intact [fidelity]. The relevant longevity concerns the retention of structure through descent. Some entities, though structurally similar, are not copies because they are not related by descent. For example, although atoms of gold are structurally similar, they are not copies of one another because atoms of gold do not give rise to other atoms of gold. Conversely, a large molecule can break down into successively smaller molecules as its quaternary, tertiary, and secondary bonds are severed. Although descent is present, these successively smaller molecules cannot count as copies because they lack the requisite structural similarity.
A replicator may be said to “benefit” from anything that increases the number of its descendant (“germ-line”) copies. To the extent that active germ-line replicators benefit from the survival of the bodies in which they sit, we may expect to see adaptations that can be interpreted as for bodily survival. A large number of adaptations are of this type. To the extent that {85} active germ-line replicators benefit from the survival of bodies other than those in which they sit, we may expect to see “altruism”, parental care, etc. To the extent that active germ-line replicators benefit from the survival of the group of individuals in which they sit, over and above the two effects just mentioned, we may expect to see adaptations for the preservation of the group. But all these adaptations will exist, fundamentally, through differential replicator survival. The basic beneficiary of any adaptation is the active, germ-line replicator, the optimon.
It is important not to forget the “germ-line” proviso in the specification of the optimon. This is the point of Hull's gold atom analogy. Krebs (1977) and I (Dawkins 1979a) have previously criticized Barash (1977) for suggesting that sterile worker insects care for other workers because they share genes with them. I would not harp on this again, had the error not been repeated twice recently in print (Barash 1978; Kirk 1980). It would be more correct to say that workers care for their reproductive siblings who carry germ-line copies of the caring genes. If they care for other workers, it is because those other workers are likely to work on behalf of the same reproductlves (to whom they also are kin), not because the workers are kin to each other. Worker genes may be active, but they are dead-end, not germ-line replicators.
No copying process is infallible. It is no part of the definition of a replicator that its copies must all be perfect. It is fundamental to the idea of a replicator that when a mistake or “mutation” does occur it is passed on to future copies: the mutation brings Into existence a new kind of replicator which “breeds true” until there is a further mutation. When a sheet of paper is xeroxed, a blemish may appear on the copy which was not present on the original. If the xerox copy itself is now copied, the blemish is incorporated into the second copy (which may also introduce a new blemish of its own). The important principle is that in a chain of replicators errors are cumulative.
I have previously used the word “gene” in the same sense as I would now use “genetic replicator”, to refer to a genetic fragment which, for all that it serves as a unit of selection, does not have rigidly fixed boundaries. This has not met with uniform approval. The eminent molecular biologist Gunther Stent (1977) wrote that “One of the great triumphs of 20th century biology was the eventual unambiguous Identification of the Mendellan hereditary factor, or gene ... as that unit of genetic material ... in which the amino acid sequence of a particular protein is encoded”. Stent therefore violently objected to my adopting the equivalent of Williams's (1966) definition of a gene as “that which segregates and recombines with appreciable frequency”, describing it as a “heinous terminological sin”.
Such motherly protectiveness towards a rather recently usurped technical term is not universal among molecular biologists, for one of the greatest of them has recently written that “the theory of the "selfish gene" will have to {86} be extended to any stretch of DNA” (Crick 1979). And, as we saw at the beginning of this chapter, another molecular biologist of the first rank, Seymour Benzer (1957), recognized the shortcomings of the traditional gene concept, but rather than grab the traditional word gene itself for one particular molecular usage, he chose the more modest course of coining a useful set of new terms — muton, recon and cistron, to which we may add the optimon. Benzer recognized that all three of his units had claims to be regarded as equivalent to the gene of the earlier literature. Stent's uncompromising elevation of the cistron to the honour is arbitrary, though admittedly it is quite common. A more balanced view was given by the lamented W. T. Keeton (1980): “It may seem strange that geneticists continue to employ different definitions of the gene for different purposes. The fact is that, at the present stage of knowledge, one definition is more useful in one context and another in another context; a rigid terminology would only hamper the formulation of current ideas and research aims”. Lewontin (1970b), too, gets it right when he says that “it is only the chromosomes that obey Mendel's Law of Independent Assortment, and only the nucleotide base that is indivisible. The codons and the genes [cistrons] lie in between, being neither unitary nor independent in their behavior at meiosis”.
But let us not become worked up over terminology. Meanings of words are important, but not important enough to justify the ill-feeling they sometimes provoke, as in the present case of Stent (and also as in Stent's passionate and apparently sincere denunciation of my. following the standard modern fashion of redefining “selfishness” and “altruism” in non-subjective senses — see Dawkins, 1981, for a reply to a similar criticism). I am happy to replace “gene” with “genetic replicator” where there is any doubt.
Heinous terminological sins aside, Stent makes the more important point that my unit is not precisely delimited in the way that the cistron is. Well, perhaps I should say “in the way that the cistron once seemed to be”, for the recent discovery of “embedded” cistrons in virus ÔX174, and of “exons” surrounding “introns” must be causing a little discomfort to anyone who likes his units rigid. Crick (1979) expresses the sense of novelty well: “In the last 2 years there has been a mini-revolution in molecular genetics. When I came to California, in September 1976, I had no idea that a typical gene might be split into several pieces and I doubt if anyone else had”. Crick significantly adds a footnote to the word gene: “Throughout this article I have deliberately used the word "gene" in a loose sense since at this time any precise definition would be premature”. My unit of selection, whether I called it gene (Dawkins 1976a) or replicator (1978a) never had any pretensions to unitariness anyway. For the purposes for which it was defined, unitariness is not an important consideration, although I readily see that it might be important for other purposes. {87}
The word replicator is purposely defined in a general way, so that it does not even have to refer to DNA. I am, indeed, quite sympathetic towards the idea that human culture provides a new milieu in which an entirely different kind of replicator selection can go on. In the next chapter we shall look briefly at this matter, and also at the claims of species gene-pools to be regarded as replicators in a large-scale selection process governing “macroevolutionary” trends. But in the rest of this chapter we shall be concerned only with genetic fragments, and “replicator” will be used as an abbreviation for “genetic replicator”.
In principle, we may consider any portion of chromosome as a potential candidate for the title of replicator. Natural selection may usually be safely regarded as the differential survival of replicators relative to their alleles. The word allele is nowadays customarily used of cistrons but it is clearly easy, and in the spirit of this chapter, to generalize it to any portion of chromosome. If we look at a portion of chromosome five cistrons long, its alleles are the alternative sets of five cistrons that exist at the homologous loci of all the chromosomes in the population. An allele of an arbitrary sequence of twenty-six codons is an alternative homologous sequence of twenty-six codons somewhere in the population. Any stretch of DNA, beginning and ending at arbitrarily chosen points on the chromosome, can be considered to be competing with allelomorphic stretches for the region of chromosome concerned. It further follows that we can generalize the terms homozygous and heterozygous. Having picked out an arbitrary length of chromosome as our candidate replicator, we look at the homologous chromosome in the same diploid individual. If the two chromosomes are identical over the whole length of the replicator, the individual is homozygous for that replicator, otherwise it is heterozygous.
When I said “arbitrarily chosen portion of chromosome”, I really meant arbitrarily. The twenty-six codons that I chose might well span the border between two cistrons. The sequence still potentially fits the definition of a replicator, it is still possible to think of it as having alleles, and it still may be thought of as homozygous or heterozygous to the corresponding portion of the homologous chromosome in a diploid genotype. This, then, is our candidate replicator. But a candidate should be regarded as an actual replicator only if it possesses some minimum degree of longevity/fecundity/fidelity (there may be trade-offs among the three). Other things being equal, it is clear that larger candidates will have lower longevity/fecundity/fidelity than smaller ones because they are more vulnerable to being broken by recombination events. So, how large and how small a portion of chromosome is it useful to treat as a replicator?
This depends on the answer to another question: “useful for what?” The reason a replicator is interesting to Darwinians is that it is potentially immortal, or at least very long-lived in the form of copies. A successful {88} replicator is one that succeeds in lasting, in the form of copies, for a very long time measured in generations, and succeeds in propagating many copies of itself. An unsuccessful replicator is one that potentially might have been long-lived, but in fact failed to survive, say because it caused the successive bodies in which it found itself to be sexually unattractive. We may apply the terms “successful” and “unsuccessful” to any arbitrarily defined portion of chromosome. Its success is measured relative to its alleles, and, if there is heterozygosity at the replicator locus in the population, natural selection will change the relative frequencies of the allelomorphic replicators in the population. But if the arbitrarily chosen portion of chromosome is very long it is not even potentially long-lived in its present form, for it is likely to be split apart by crossing-over in any given generation, regardless of how successful it may be in making a body survive and reproduce. To go to an extreme, if the potential replicator we consider is a whole chromosome, the difference between a “successful” and an unsuccessful chromosome is of no significance, since both are almost bound to be split by crossing-over before the next generation in any case: their “fidelity” is zero.
This can be put in another way. An arbitrarily defined length of chromosome, or potential replicator, may be said to have an expected half-life, measured in generations. Two kinds of factor will affect this half-life. Firstly, replicators whose phenotypic effects render them successful at their business of propagating themselves will tend to have a long half-life. Replicators with longer half-lives than their alleles will come to predominate in the population, and this is the familiar process of natural selection. But if we set selection pressures on one side, we can say something about the half-life of a replicator on the basis of its length alone. If the stretch of chromosome we choose to define as our replicator of interest is long, it will tend to have a shorter half-life than a shorter replicator, simply because it is more likely to be broken by crossing-over. A very long portion of chromosome ceases to deserve the title of replicator at all.
A corollary is that a long portion of chromosome, even if successful in terms of its phenotypic effects, will not be represented in many copies in the population. The Y chromosome excepted, and depending upon crossover rates, it seems unlikely that I share any whole chromosomes with any other individual. I assuredly share many small portions of chromosomes with others, and if we choose our portions small enough their likelihood of being shared becomes very high indeed. It is not normally useful to speak of inter-chromosome selection, therefore, since each chromosome is probably unique. Natural selection is the process by which replicators change in frequency in the population relative to their alleles. If the replicator under consideration is so large that it is probably unique, it cannot be said to have a “frequency” to change. We must choose our arbitrary portion of chromosome so that it is small enough to last, at least potentially, for many generations before being {89} split by crossing-over; small enough to have a “frequency” that can be changed by natural selection. Is it possible to choose it too small? I shall return to this question below, after approaching it from another direction.
I shall make no attempt to specify exactly how long a portion of chromosome can be permitted to be before it ceases to be usefully regarded as a replicator. There is no hard and fast rule, and we don't need one. It depends on the strength of the selection pressure of interest. We are not seeking an absolutely rigid definition, but “a kind of fading-out definition, like the definition of "big" or "old"”. If the selection pressure we are discussing is very strong, that is if one replicator makes its possessors very much more likely to survive and reproduce than its alleles do, the replicator can be quite large and still be usefully regarded as a unit that is naturally selected. If, on the other hand, the difference in survival consequences between a putative replicator and its alleles is almost negligible, the replicators under discussion would have to be quite small if the difference in their survival values is to make itself felt. This is the rationale behind Williams's (1966, p. 25) definition: “In evolutionary theory, a gene could be defined as any hereditary information for which there is a favorable or unfavorable selection bias equal to several or many times its rate of endogenous change”.
The possibility of strong linkage disequilibrium (Clegg 1978) does not weaken the case. It simply increases the size of the chunk of genome that we can usefully treat as a replicator. If, which seems doubtful, linkage disequilibrium is so strong that populations contain “only a few gametic types” (Lewontin 1974, p. 312), the effective replicator will be a very large chunk of DNA. When what Lewontin calls lc, the “characteristic length” (the “distance over which coupling is effective”), is only “a fraction of the chromosome length, each gene is out of linkage equilibrium only with its neighbors but is assorted essentially independently of other genes farther away. The characteristic length is, in some sense, the unit of evolution since genes within it are highly correlated. The concept is a subtle one, however. It does not mean that the genome is broken up into discrete adjacent chunks of length lc. Every locus is the center of such a correlated segment and evolves in linkage with the genes near it” (Lewontin 1974).
Similarly, Slatkin (1972) wrote that “It is clear that when permanent linkage disequilibrium is maintained in a population, the higher order interactions are important and the chromosome tends to act as a unit. The degree to which this is true in any given system is a measure of whether the gene or the chromosome is the unit of selection, or, more accurately, what parts of the genome can be said to be acting in unison”. And Templeton et al. (1976) wrote that “...the unit of selection is a function in part of the intensity of selection: the more intense the selection, the more the whole genome tends to hold together as a unit”. It was in this spirit that I playfully {90} contemplated titling an earlier work The slightly selfish big bit of chromosome and the even more selfish little bit of chromosome (Dawkins 1976a, p. 35).
It has often been suggested to me that a fatal objection to replicator selectionism is the existence of within-cistron crossing-over. If chromosomes were like bead necklaces, the argument runs, with crossing-over always breaking the necklace between beads and not within them, you might hope to define discrete replicators in the population, containing an integral number of cistrons. But since crossover can occur anywhere (Watson 1976), not just between beads, all hope of defining discrete units disappears.
This criticism underestimates the elasticity that the replicator concept is permitted by the purpose for which it was coined. As I have just shown, we are not looking for discrete units, but for pieces of chromosome of indeterminate length which become more or less numerous than alternatives of exactly the same length. Moreover, as Mark Ridley reminds me, most within-cistron crossovers are, in any case, indistinguishable in their effects from between-cistron crossovers. Obviously, if the cistron concerned happens to be homozygous, paired at meiosis with an identical allele, the two sets of genetic material exchanged in a crossover will be identical, and the crossover might just as well never have happened. If the cistrons concerned are heterozygous, differing at one nucleotide locus, any within-cistron crossover that occurs “north” of the heterozygous nucleotide will be indistinguishable from one at the northern boundary of the cistron; any within-cistron crossover “south” of the heterozygous nucleotide will be indistinguishable from one at the southern boundary of the cistron. Only if the cistrons differ at two loci, and the crossover occurs between them, will it be identifiable as a within-cistron crossover. The general point is that it does not particularly matter where crossovers occur in relation to cistron boundaries. What matters is where crossovers occur in relation to heterozygous nucleotides. If, for instance, a sequence of six adjacent cistrons happens to be homozygous throughout an entire breeding population, a crossover anywhere within the six will be exactly equivalent in effect to a crossover at either end of the six.
Natural selection can cause changes in frequency only at nucleotide loci that are heterozygous in the population. If there are large intervening chunks of nucleotide sequences that never differ among individuals, these cannot be subject to natural selection, for there is nothing to choose between them. Natural selection must focus its attention on heterozygous nucleotides. It is changes at the single nucleotide level that are responsible for evolutionarily significant phenotypic changes, although of course the unvarying remainder of the genome is necessary to produce a phenotype at all. Have we, then, arrived at an absurdly reductionistic reductio ad absurdum? Shall we write a book called The Selfish Nucleotide? Is adenine engaged, in a remorseless struggle against cytosine for possession of locus number 30004?
At the very least, this is not a helpful way to express what is going on. It {91} becomes downright misleading if it suggests to the student that adenine at one locus is, in some sense, allied with adenine at other loci, pulling together for an adeeine team. If there is any sense in which purines and pyrimidiees compete with each other for heterozygous loci, the struggle at each locus is insulated from the struggle at other loci. The molecular biologist may, for his own important purposes (Ghargaff cited in Judson 1979), count adenines and cytosines in the genome as a whole, but to do so is an idle exercise for the student of natural selection. If they are competitors at all, they are competitors for each locus separately. They are indifferent to the fate of their exact replicas at other loci (see also Chapter 8).
But there is a more interesting reason for rejecting the concept of the selfish nucleotide, in favour of some larger replicating entity. The whole purpose of our search for a “unit of selection” is to discover a suitable actor to play the leading role in our metaphors of purpose. We look at an adaptation and want to say, “It is for the good of...”. Our quest in this chapter is for the right way to complete that sentence. It is widely admitted that serious error follows from the uncritical assumption that adaptations are for the good of the species. I hope I shall be able to show, in this book, that yet other theoretical dangers, albeit lesser ones, attend the assumption that adaptations are for the good of the individual organism. I am suggesting here that, since we must speak of adaptations as being for the good of something, the correct something is the active, germ-like replicator. And while it may not be strictly wrong to say that an adaptation is for the good of the nucleotide, i.e. the smallest replicator responsible for the phenotypic differences concerned in the evolutionary change, it is not helpful to do so.
We are going to use the metaphor of power. An active replicator is a chunk of genome that, when compared to its alleles, exerts phenotypic power over its world, such that its frequency increases or decreases relative to that of its alleles. While it is undoubtedly meaningful to speak of a single nucleotide as exerting power in this sense, it is much more useful, since the nucleotide only exerts a given type of power when embedded in a larger unit, to treat the larger unit as exerting power and hence altering the frequency of its copies. It might be thought that the same argument could be used to justify treating an even larger unit, such as the whole genome, as the unit that exerts the power. This is not so,-at least for sexual genomes.
We reject the whole sexual genome as a candidate replicator, because of its high risk of being fragmented at meiosis. The single nucleotide does not suffer from this problem but, as we have just seen, it raises another problem. It cannot be said to have a phenotypic effect except in the context of the other nucleotides that surround it in its cistron. It is meaningless to speak of the phenotypic effect of adenine. But it is entirely sensible to speak of the phenotypic effect of substituting adenine for cytosine at a named locus within a named cistron. The case of a cistron within a genome is not analogous. {92} Unlike a nucleotide, a cistron is large enough to have a consistent phenotypic effect, relatively, though not completely, independently of where it lies on the chromosome (but not regardless of what other genes share its genome). For a cistron, its sequential context vis-a-vis other cistroes is not overwhelmingly important in determining its phenotypic effect in comparison with its alleles. For a nucleotide's phenotypic effect, on the other hand, its sequential context is everything.
Bateson (1981) expresses the following misgiving about “replicator selection”.
A winning character is defined in relation to another one while genetic replicators are thought about in absolute and atomistic terms. The difficulty is brought home if you ask yourself, what exactly is Dawkins' replicator? You might answer: “That bit of genetic material making the difference between the winning and losing characters”. You would have stated that a replicator must be defined in relation to something else. Alternatively, your reply might be: “A replicator consists of all the genes required for the expression of the surviving character”. In that case you are saddled with a complex and unwieldy concept. Either way your answer would show how misleading it is to think of replicators as the atoms of evolution.
I certainly would join Bateson in rejecting the second of his two alternative answers, the unwieldly one. The first of his alternatives, on the other hand, expresses exactly my position, and I do not share Bateson's misgivings about it. For my purposes a genetic replicator is defined by reference to its alleles, but this is not a weakness of the concept. Or, if it is deemed to be a weakness, it is a weakness that afflicts the whole science of population genetics, not just the particular idea of genetic units of selection. It is a fundamental truth, though it is not always realized, that whenever a geneticist studies a gene “for” any phenotypic character, he is always referring to a difference between two alleles. This is a recurring theme throughout this book.
To prove the pudding, let me show how easy it is to use the gene as a conceptual unit of selection, while admitting that it is only defined by comparison with its alleles. It is now accepted that a particular major gene for dark coloration in the peppered moth Biston betularia has increased in frequency in industrial areas because it produces phenotypes that are superior in industrial areas (Kettlewell 1973). At the same time, we have to admit that this gene is only one of thousands that are necessary in order for the dark coloration to show itself. A moth cannot have dark wings unless it has wings, and it cannot have wings unless it has hundreds of genes and hundreds of equally necessary environmental factors. But this is all {93} irrelevant. The difference between the car banana and the typica phenotype can still be due to a difference at one locus, even though the phenotypes themselves could not exist without the participation of thousands of genes. And it is the same difference that is the basis of the natural selection. Both geneticists and natural selection are concerned with differences! However complex the genetic basis of features that all members of a species have in common, natural selection is concerned with differences. Evolutionary change is a limited set of substitutions at identifiable loci.
Further difficulties will be dealt with in the next chapter. Meanwhile, I end this chapter with a small diversion which may be helpful in illustrating the replicator or “gene's-eye” view of evolution. An appealing aspect of the view appears if we look backwards in time. The replicators in frequent existence today constitute a relatively successful subset of those that have existed in the past. A given replicator in me could, in theory, be traced backwards through a straight line of ancestors. These ancestors, and the environments that they provided for the replicator, can be regarded as the replicator's “past experience”.
The past experience of autosomal genetic fragments in a species is, statistically speaking, similar. It consists of an ensemble of typical species bodies, approximately 50 per cent male and 50 per cent female bodies, bodies which grew through a wide spectrum of ages at least up to reproductive age; and it includes a good random shuffling of “companion” genes at other loci. The genes that exist today tend to be the ones that are good at surviving in that statistical ensemble of bodies, and in company with that statistical ensemble of companion genes. As we shall see, it is selection in favour of the qualities needed for survival in company with other, similarly selected, genes, that gives rise to the appearance of “coadapted genomes”. I shall show in Chapter 13 that this is a much more illuminating interpretation of the phenomenon of coadaptation than its alternative, that “the coadapted genome is the true unit of selection”.
Probably no two genes in an organism have identical past experiences, though a linked pair may come close, and, mutants aside, all the genes on a Y chromosome have travelled together through the same set of bodies for a large number of generations. But the exact nature of a gene's past experience is of less interest than the generalizations one can make about the past experiences of all genes that exist today. For instance, however variable the set of my ancestors may be, they all had in common that they survived at least to reproductive age, they copulated heterosexually and were fertile. The same generalization cannot be made about the historical set of bodies that were not my ancestors. The bodies that provided the past experience of existing genes are a non-random subset of all the bodies that have ever existed.
The genes that exist today reflect the set of environments that they have {94} experienced in the past. This includes the internal environments provided by the bodies the genes have inhabited, and also external environments, desert, forest, seashore, predators, parasites, social companions, etc. This is, of course, not because the environments have imprinted their qualities on the genes — that would be Lamarckism (see Chapter 9) — but because the genes that exist today are a selected set, and the qualities that made them survive reflect the qualities of the environments in which they survived.
I said that a gene's experience consists of time spent in approximately 50 per cent male and 50 per cent female bodies, but this is not, of course, true of genes on sex chromosomes. In mammals, assuming no Y-chromosome crossing-over, a gene on a Y chromosome has experienced only male bodies, and a gene on an X chromosome has spent two-thirds of its history in female bodies and one-third in male bodies. In birds, Y-chromosome genes have experienced only female bodies, and in particular cases such as cuckoos we can say something further. Female Cuculus canorus are divided into “gentes”, each gens parasitizing a different species of host (Lack 1968). Apparently each female learns the qualities of her own foster parents and their nest, and returns when adult to parasitize the same species. Males do not seem to discriminate as to gens in their choice of mate, and therefore they act as vehicles for gene flow between gentes. Of the genes in a female cuckoo, therefore, those on autosomes and the X chromosome have probably had recent experience of all gentes in the population, and been “reared” by foster parents belonging to all the species parasitized by the cuckoo population. But the Y chromosome, uniquely, is confined for long sequences of generations to one gens and one foster species. Of all the genes sitting in a robin's nest, one subset — robin genes and cuckoo Y-chromosome genes (and robin flea genes) — has sat in robins' nests for many generations back. Another subset — cuckoo autosomal and X-chromosome genes — has experienced a mixture of nests. Of course the first subset share only part of their experience, a long series of robin nests. In other aspects of their experience, cuckoo Y-chromosome genes will have more in common with other cuckoo genes than with robin genes. But as far as certain particular selection pressures found in nests are concerned, cuckoo Y-chromosome genes have more in common with robin genes than with cuckoo autosome genes. It is natural, then, that cuckoo Y chromosomes should have evolved to reflect their peculiar experience, while other cuckoo genes simultaneously evolved to reflect their more general experience a kind of incipient ietragenomic “speciation” at the chromosomal level. It is, indeed, widely assumed for this reason that genes for foster species-specific egg mimicry must be carried on the Y chromosome, while genes for general parasitic adaptations might be carried on any chromosome.
I am not sure whether the fact has significance, but this backwards way of looking shows X chromosomes, too, to have a peculiar history. A gene on an {95} autosome in a female cuckoo is as likely to have come from the father as from the mother, in which latter case it experiences the same host species for two generations running. A gene on an X chromosome in a female cuckoo is bound to have come from the father, and therefore is not especially likely to experience the same host species two generations running. A statistical “runs test” on the sequence of foster species experienced by an autosomal-gene would therefore reveal a slight runs effect, greater than that for a gene on an X chromosome, and much less than that for a gene on a Y chromosome.
In any animal, an inverted portion of chromosome may resemble a Y chromosome in being unable to cross over. The “experience” of any part of the “inversion supergene” therefore repeatedly includes the other parts of the supergene and their phenotypic consequences. A habitat selection gene anywhere in such a supergene, say a gene that makes individuals choose dry microclimates, would then provide a consistent habitat “experience” for successive generations of the whole supergene. Thus a given gene may “experience” consistently dry habitats for the same kind of reason as a gene on a cuckoo's Y chromosome consistently experiences meadow-pipit nests. This will provide a consistent selection pressure bearing upon that locus, favouring alleles that are adapted to a dry habitat in the same way as alleles for mimicking meadow-pipit eggs are favoured in Y chromosomes of female cuckoos of the meadow-pipit gens. This particular inversion supergene will tend to be found for generations in dry habitats, even though the rest of the genome may be randomly shufHed over the whole range of habitats available to the species. Many different loci on the inverted portion of chromosome may therefore come to be adapted to a dry climate, and again something akin to intragenomic incipient speciation may go on. I find this backwards way of looking at the past “experience” of genetic replicators helpful.
Germ-line replicators, then, are units that actually survive or fail to survive, the difference constituting natural selection. Active replicators have some effect on the world, which influences their chances of surviving. It is the effects on the world of successful active germ-line replicators that we see as adaptations. Fragments of DNA qualify as active germ-line replicators. Where there is sexual reproduction, these fragments must not be defined too large if they are to retain the property of self-duplication. And they must not be defined too small if they are to be usefully regarded as active.
If there were sex but no crossing-over, each chromosome would be a replicator, and we should speak of adaptations as being for the good of the chromosome. If there is no sex we can, by the same token, treat the entire genome of an asexual organism as a replicator. But the organism itself is not a replicator. This is for two quite distinct reasons which should not be confused with each other. The first reason follows from the arguments developed in this chapter, and applies only where there is sexual reproduction and meiosis: meiosis and sexual fusion see to it that not even our genomes are {96} replicators, so we ourselves are not replicators either. The second reason applies to asexual as well as sexual reproduction. It will be explained in the next chapter, which goes on to discuss what organisms, and also groups of organisms, are, given that they are not replicators.
| << | {97} | >> |
I have made so much of the fragmenting effects of meiosis as a reason for not regarding sexually reproduced organisms as replicators, that it is tempting to see this as the only reason. If this were true, it should follow that asexually reproduced organisms are true replicators, and that where reproduction is asexual we could legitimately speak of adaptations as “for the good of the organism”. But the fragmenting effect of meiosis is not the only reason for denying that organisms are true replicators. There is a more fundamental reason, and it applies to asexual organisms as much as to sexual ones.
To regard an organism as a replicator, even an asexual organism like a female stick insect, is tantamount to a violation of the “central dogma” of the non-inheritance of acquired characteristics. A stick insect looks like a replicator, in that we may lay out a sequence consisting of daughter, granddaughter, great-granddaughter, etc., in which each appears to be a replica of the preceding one in the series. But suppose a flaw or blemish appears somewhere in the chain, say a stick insect is unfortunate enough to lose a leg. The blemish may last for the whole of her lifetime, but it is not passed on to the next link in the chain. Errors that affect stick insects but not their genes are not perpetuated. Now lay out a parallel series consisting of daughter's genome, granddaughter's genome, great-granddaughter's genome, etc. If a blemish appears somewhere along this series it will be passed on to all subsequent links in the chain. It may also be reflected in the bodies of all subsequent links in the chain, because in each generation there are causal arrows leading from genes to body. But there is no causal arrow leading from body to genes. No part of the stick insect's phenotype is a replicator. Nor is her body as a whole. It is wrong to say that “just as genes can pass on their structure in gene lineages, organisms can pass on their structure in organism lineages”.
I am sorry if I am about to labour this argument, but I fear it was my failure to be clear about it before that led to an unnecessary disagreement {98} with Bateson, a disagreement which it is worth going to some trouble to sort out. Bateson (1978) made the point that genetic determinants of development are necessary but not sufficient. A gene may “program” a particular bit of behaviour “without it being the only thing to do so”. He goes
Dawkins accepts all this but then reveals his uncertainty about which language he is using by immediately giving special status back to the gene as the programmer. Consider a case in which the ambient environmental temperature during development is crucial for the expression of a particular phenotype. If the temperature changes by a few degrees the survival machine is beaten by another one. Would not that give as much status to a necessary temperature value as to a necessary gene? The temperature value is also required for the expression of a particular phenotype. It is also stable (within limits) from one generation to the next. It may even be transmitted from one generation to the next if the survival machine makes a nest for its offspring. Indeed, using Dawkins' own style of ideological argument one could claim that the bird is the nest's way of making another nest [Bateson 1978].
I replied to Bateson, but did so too briefly, picking on the last remark about birds' nests and saying, “A nest is not a true replicator because a [non-genetic] "mutation" which occurs in the construction of a nest, for example the accidental incorporation of a pine needle instead of the usual grass, is not perpetuated in future "generations of nests". Similarly, protein molecules are not replicators, nor is messenger RNA” (Dawkins 1978a). Bateson had taken the catchphrase about a bird being a gene's way of making another gene, and inverted it, substituting “nest” for “gene”. But the parallel is not a valid one. There is a causal arrow going from gene to bird, but none in the reverse direction. A changed gene may perpetuate itself better than its unmutated allele. A changed nest will do no such thing unless, of course, the change is due to a changed gene, in which case it is the gene that is perpetuated, not the nest. A nest, like a bird, is a gene's way of making another gene.
Bateson is worried that I seem to give “special status” to genetic determinants of behaviour. He fears that an emphasis on the gene as the entity for whose benefit organisms labour, rather than the other way around, leads to an undue emphasis on the importance of genetic as opposed to environmental determinants of development. The answer to this is that when we are talking about development it is appropriate to emphasize non-genetic as well as genetic factors. But when we are talking about units of selection a different emphasis is called for, an emphasis on the properties of replicators. {99} The special status of genetic factors rather than non-genetic factors is deserved for one reason only: genetic factors replicate themselves, blemishes and all, but non-genetic factors do not.
Let us grant with both hands that the temperature in the nest housing a developing bird is important both for its immediate survival and for the way it develops and therefore for its long-term success as an adult. The immediate effects of gene products on the biochemical springs of development may, indeed, closely resemble the effects of temperature changes (Waddington 1957). We could even imagine the enzyme products of genes as little Bunsen burners, selectively applied at crucial nodes of the branching biochemical tree of embryonic causation, controlling development by selective control of biochemical reaction rates. An embryologist rightly sees no fundamental distinction between genetic and environmental causal factors, and he correctly regards each as necessary but not sufficient. Bateson was putting the embryologist's point of view, and no ethologist is better qualified to do so. But I was not talking embryology. I was not concerned with the rival claims of determinants of development. I was talking about replicators surviving in evolutionary time, and Bateson certainly agrees that neither a nest, nor the temperature inside it, nor the bird that built it, are replicators. We can quickly see that they are not replicators by experimentally altering one of them. The change may wreak havoc on the animal, on its development and its chances of survival, but the change will not be passed on to the next generation. Now make a similar mutilation (mutation) to a gene in the germ-line: the change may or may not affect the bird's development and its survival, but it can be passed on to the next generation; it can be replicated.
As is so often the case, an apparent disagreement turns out to be due to mutual misunderstanding. I thought that Bateson was denying proper respect to the Immortal Replicator. Bateson thought that I was denying proper respect to the Great Nexus of complex causal factors interacting in development. In fact, each of us was laying legitimate stress on considerations which are important for two different major fields of biology, the study of development and the study of natural selection.
An organism, then, is not a replicator, not even (despite Lewontin 1970a — see Dawkins 1982) a crude replicator with poor copying fidelity. It is therefore better not to speak of adaptations as being for the good of the organism. What about larger units, groups of organisms, species, communities of species, etc.? Some of these larger groupings are clearly subject to a version of the “internal fragmentation destroys copying fidelity” argument. The fragmenting agent in this case is not the recombining effects of meiosis, but immigration and emigration, the destruction of the integrity of groups by the movement of individuals into and out of them. As I have put it before, they are like clouds in the sky or dust-storms in the desert. They are temporary aggregations or federations. They are not stable through {100} evolutionary time. Populations may last a long while, but they are constantly bleeding with other populations and so losing their identity. They are also subject to evolutionary change from within. A population is not a discrete enough entity to be a unit of natural selection, not stable and unitary enough to be “selected” in preference to another population. But, just as the “fragmentation” argument applied only to a subset of organisms, sexual ones, so it also applies only to a subset at the group level. It applies to groups capable of interbreeding, but it does not apply to reproductively isolated species.
Let us, then; examine whether species behave sufficiently like coherent entities, multiplying and giving rise to other species, to deserve to be called replicators. Note that that is not the same as Ghiselin's (1974b) logical claim that species are “individuals” (see also Hull 1976). Organisms, too, are individuals in Ghiselin's sense, and I hope I have established that organisms are not replicators. Do species, or, to be more precise, do reproductively isolated gene-pools, really answer to the definition of replicators?
It is important to remember that mere immortality is not a sufficient qualification. A lineage, such as a sequence of parents and offspring from the long-unchanged brachiopod genus Lingula, is unending in the same sense, and to the same extent, as a lineage of genes. Indeed, for this example we perhaps need not have chosen a “living fossil” like Lingula. Even a rapidly evolving lineage can, in a sense, be treated as an entity which is either extinct or extant at any moment in geological time. Now, certain kinds of lineage may be more likely to go extinct than others, and we may be able to discern statistical laws of extinction. For example, lineages whose females reproduce asexually may be systematically more or less likely to go extinct than lineages whose females stick to sex (Williams 1975; Maynard Smith 1978a). It has been suggested that ammonite and bivalve lineages with a high rate of evolving larger size (i.e. with a high rate of obeying Cope's Rule) are more likely to go extinct than more slowly evolving lineages (Hallam 1975). Leigh (1977) makes some excellent points about differential lineage extinction, and its relationship to lower levels of selection: “...those species are favored where selection within populations works more nearly for the good of the species”. Selection “...favors species that have, for whatever reason, evolved genetic systems where a gene's selective advantage more nearly matches its contribution to fitness”. Hull (1980a, b) is particularly clear about the logical status of the lineage, and about its distinction from the replicator and the interactor (Hull's name for what I am calling the “vehicle”).
Differential lineage extinction, though technically a form of selection, is not enough in itself to generate progressive evolutionary change. Lineages may be “survivors”, but this does not make them replicators. Grains of sand {101} are survivors. Hard grains, made of quartz or diamond, will last longer than soft grains made of chalk. But nobody has ever invoked hardness selection among sand grains as the basis for an evolutionary progression. The reason, fundamentally, is that grains of sand do not multiply. One grain may survive a long time, but it does not multiply and make copies of itself. Do species, or other groups of organisms, multiply? Do they replicate?
Alexander and Borgia (1978) assert that they do, and that they are therefore true replicators: “Species give rise to species; species multiply”. The best case I can make for regarding species, or rather their gene-pools, as multiplying replicators arises from the theory of “species selection” associated with the palaeontological idea of “punctuated equilibria” (Eldredge & Gould 1972; Stanley 1975, 1979; Gould & Eldredge 1977; Gould 1977c, 1980a, b; Levinton & Simon 1980). I will take some time to, discuss this body of theory, since “species selection” is very relevant to this chapter. Another reason for taking the time is that I regard the suggestions of Eldredge and Gould as of great interest to biology generally, but I am anxious that they should not be oversold as more revolutionary than they actually are. Gould and Eldredge (1977, p. 117) are themselves conscious of this danger, though for different reasons.
My fear stems from the growing influence of a vigilant corps of lay critics of Darwinism, either religious fundamentalists or Shavian/Koestlerian Lamarckists who, for reasons that have nothing to do with science, eagerly seize upon anything that, with imperfect understanding, can be made to sound anti-Darwinian. Journalists are often only too ready to pander to the unpopularity of Darwinism in some lay circles. One of Britain's least disreputable daily newspapers (The Guardian, 21 November 1978) served up a journalistically garbled but still just recognizable version of the Eldredge/Gould theory in a leading article, as evidence that all is not well with Darwinism. Predictably, this elicited some uncomprehending fundamentalist glee in the letter columns of the paper, some of it from disquietingly influential sources, and the public could well have been left with the impression that even “the scientists” themselves now have doubts about Darwinism. Dr Gould informs me that The Guardian did not favour him with a reply to his letter of protest. Another British newspaper, The Sunday Times (8 March 1981), in a much longer article called “The new clues that challenge Darwin”, sensationally exaggerated the difference between the Eldredge/Gould theory and other versions of Darwinism. The British Broadcasting Corporation also got in on the act around the same time, in two separate programmes made by rival production teams. They were called The Trouble with Evolution and Did Darwin get it Wrong? and they differed hardly at all except that one had Eldredge and the other had Gould! The second programme actually went to the lengths of digging up some {102} fundamentalists to comment on the Eldredge/Gould theory: not surprisingly, the ill-understood appearance of dissension within the ranks of Darwinists was meat and drink to them.
Journalistic standards are not unknown in learned periodicals too. Science (Vol. 210, pp. 883–887, 1980) reported on a recent conference on macroevolution under the dramatic heading, “Evolutionary theory under fire”, and with the equally sensational subtitle, “An historic conference in Chicago challenges the four-decade long dominance of the Modern Synthesis” (see criticisms by Futuyma et al. 1981). As Maynard Smith is quoted as saying at the same conference, “You are in danger of preventing understanding by suggesting that there is intellectual antagonism where none exists” (see also Maynard Smith 1981). In the face of all the ballyhoo, I am anxious to be very clear about, to quote one of their own section headings, exactly “What Eldredge and Gould Did Not (And Did) Say”.
The theory of punctuated equilibria suggests that evolution does not consist in continuous smooth change, “stately unfolding”, but goes in jerks, punctuating long periods of stasis. Lack of evolutionary change in fossil lineages should not be written off as “no data”, but should be recognized as the norm, and as what we should really expect to see if we take our modern synthesis, particularly its embedded ideas of allopatric speciation, seriously: “As a consequence of the allopatric theory, new fossil species do not originate in the place where their ancestors lived. It is extemely improbable that we shall be able to trace the gradual splitting of a lineage merely by following a certain species up through a local rock column” (Eldredge & Gould 1972, p. 94). Of course microevolution by ordinary natural selection, what I would call genetic replicator selection, goes on, but it is largely confined to brief bursts of activity around the crisis times known as speciation events. These bursts of microevolution are usually completed too fast for palaeontologists to track them. All we can see is the state of the lineage before and after the new species is formed. It follows that “gaps” in the fossil record between species, far from being the embarrassment Darwinians have sometimes taken them to be, are exactly what we should expect.
The palaeontological evidence can be argued about (Gingerich 1976; Gould & Eldredge 1977; Hallam 1978), and I am not qualified to judge it. Approaching from a non-palaeontological direction, indeed in lamentable ignorance of the whole Eldredge/Gould theory, I once found the Wrightian/Mayrian idea of buffered gene-pools resisting change, but occasionally succumbing to genetic revolutions, satisfyingly compatible with one of my own enthusiasms, Maynard Smith's (1974) “evolutionarily stable strategy” concept:
The gene pool will become an evolutionarily stable set of genes, defined as a gene pool which cannot be invaded by any new gene. Most new {103} genes which arise, either by mutation or reassortment or immigration, are quickly penalized by natural selection: the evolutionarily stable set is restored. Occasionally ... there is a transitional period of instability, terminating in a new evolutionarily stable set ... a population might have more than one alternative stable point, and it might occasionally flip from one to another. Progressive evolution may be not so much a steady upward climb as a series of discrete steps from stable plateau to stable plateau [Dawkins 1976a, p. 93].
I am also quite impressed with Eldredge and Gould's reiterated point about time-scales: “How can we view a steady progression yielding a 10% increase in a million years as anything but a meaningless abstraction? Can this varied world of ours possibly impose such minute selection pressures so uninterruptedly for so long?” (Gould & Eldredge 1977). “...to see gradualism at all in the fossil record implies such an excruciatingly slow rate of per-generation change that we must seriously consider its invisibility to natural selection in the conventional mode — changes that confer momentary adaptive advantages” (Gould 1980a). I suppose the following analogy might be made. If a cork floats from one side of the Atlantic to the other, travelling steadily without deviating or going backwards, we might invoke the Gulf Stream or Trade Winds in explanation. This will seem plausible if the time the cork takes to cross the ocean is of the right order of magnitude, say a few weeks or months. But if the cork should take a million years to cross the ocean, again not deviating or going backwards but steadily inching its way across, we should not be satisfied with any explanation in terms of currents and winds. Currents and winds just don't move that slowly, or if they do they will be so weak that the cork will be overwhelmingly buffeted by other forces, backwards as much as forwards. If we found a cork steadily moving at such an extremely slow rate, we would have to seek a wholly different kind of explanation, an explanation commensurate with the time-scale of the phenomenon observed.
Incidentally, there is a mildly interesting historical irony here. One of the early arguments used against Darwin was that there wasn't enough time for the proposed amount of evolution to have happened. It seemed hard to imagine that selection pressures were strong enough to achieve all that evolutionary change in the short time then thought to be available. The argument of Eldredge and Gould just given is almost the exact opposite: it is hard to imagine a selection pressure weak enough to sustain such a slow rate of unidirectional evolution over such a long period! Perhaps we should take warning from this historical twist. Both arguments resort to the “hard to imagine” style of reasoning that Darwin so wisely cautioned us against.
Although I find Eldredge and Gould's time-scale point somewhat plausible, I am less confident about it than they are, since I do fear the {104} limitations of my own imagination. After all, the gradualist theory does not really need to assume unidirectional evolution for long periods. In terms of my cork analogy, what if the wind is so weak that the cork takes a million years to cross the Atlantic? Waves and local currents may indeed send it backwards almost as much as forwards. But when all is added up, the net statistical direction of the cork may still be determined by a slow and relentless wind.
I also wonder whether Eldredge and Gould give sufficient attention to the possibilities opened up by “arms races” (Chapter 4). In characterizing the gradualist theory that they attack, they write: “The postulated mechanism for gradual uni-directional change is "orthoselection", usually viewed as a constant adjustment to a uni-directional change in one or more features of the physical environment” (Eldredge & Gould 1972). If the winds and currents of the physical environment pressed steadily in the same direction over geological time-scales, it might indeed seem likely that animal lineages would reach the other side of their evolutionary ocean so fast that palaeontologists could not track their passage.
But change “physical” to “biological” and things may look different. If each small adaptive step in one lineage calls forth a counteradaptation in another lineage — say its predators — slow, directional orthoselection looks rather more plausible. The same is true of intraspecific competition, where the optimum size, say, for an individual can be slightly larger than the present population mode, whatever the present population mode may be. “...in the population as a whole there is a constant tendency to favor a size slightly above the mean. The slightly larger animals have a very small but in the long run, in large populations, decisive advantage in competition ... Thus, populations that are regularly evolving in this way are always well adapted as regards size in the sense that the optimum is continuously included in their normal range of variation, but a constant asymmetry in the centripetal selection favors a slow upward shift in the mean” (Simpson 1953, p. 151). Alternatively (and shamelessly having it both ways!), if evolutionary trends really are punctuated and stepped, perhaps this in itself could be explained by the arms race concept, given time-lags between adaptive advances on one side of the arms race and responses by the other.
But let us provisionally accept the theory of punctuated equilibria as an excitingly different way of looking at familiar phenomena, and turn to the other side of Gould and Eldredge's (1977) equation, “punctuated equilibria + Wright's rule = species selection”. Wright's Rule (not his own coining) is “the proposition that a set of morphologies produced by speciation events is essentially random with respect to the direction of evolutionary trends within a clade” (Gould & Eldredge 1977). For example, even if there is an overall trend towards, larger size in a set of related lineages, Wright's Rule suggests that there is no systematic tendency for newly branched-off species to be {105} larger than their parent species. The analogy with the “randomness” of mutation is clear, and this leads directly to the right-hand side of the equation. If new species differ from their predecessors randomly with respect to major trends, major trends themselves must be due to differential extinction among those new species — “species selection”, to use Stanley's (1975) term.
Gould (1980a) regards “the testing of Wright's rule as a major task for macroevolutionary theory and paleobiology. For the theory of species selection, in its pure form, depends upon it. Consider, for example, a lineage displaying Cope's rule of increasing body size — horses, for example. If Wright's rule be valid, and new species of horses arise equally often at sizes smaller and larger than their ancestors, then the trend is powered by species selection. But if new species arise preferentially at sizes larger than their ancestors, then we don't require species selection at all, since random extinction would still yield the trend”. Gould here simultaneously sticks his neck out and hands Occam's Razor to his opponents! He could easily have claimed that even if his mutation-analogue (speciation) was directed, the trend might still be reinforced by species selection (Levinton & Simon 1980). Williams (1966, p. 99) in an interesting discussion which I have not seen cited in the literature on punctuated equilibria, considers a form of species selection acting in opposition to, and perhaps outweighing, the overall trend of evolution within species. He again uses the horse example, and the fact that earlier fossils tend to be smaller than later ones:
From this observation, it is tempting to conclude that, at least most of the time and on the average, a larger than mean size was an advantage to an individual horse in its reproductive competition with the rest of its population. So the component populations of the Tertiary horse-fauna are presumed to have been evolving larger size most of the time and on the average. It is conceivable, however, that precisely the opposite is true. It may be that at any given moment during the Tertiary, most of the horse populations were evolving a smaller size. To account for the trend towards larger size it is merely necessary to make the additional assumption that group selection favoured such a tendency. Thus, while only a minority of the populations may have been evolving a larger size, it could have been this minority that gave rise to most of the populations of a million years later”.
I do not find it hard to believe that some of the major macroevoiutionary trends, of the Cope's Rule type (but see Hallam 1978), observed by palaeontologists are due to species selection, in the sense of this passage from Williams, which I think is the same as Eldredge and Gould's sense. As I believe all {106} three authors would agree, this is a very different matter from accepting group selection as an explanation for individual self-sacrifice: adaptations for the good of the species. There we are talking about another kind of group selection model, where the group is really seen not as a replicator but as a vehicle for replicators. I shall come to this second kind of group selection later. Meanwhile I shall argue that a belief in the power of species selection to shape simple major trends is not the same as a belief in its power to put together complex adaptations such as eyes and brains.
The major trends of palaeontology, simple increases in absolute size, or of relative sizes of different parts of the body, are important and interesting, but they are, above all, simple. Accepting Eldredge and Gould's belief that natural selection is a general theory that can be phrased on many levels, the putting together of a certain quantity of evolutionary change demands a certain minimum number of selective replicator-eliminations. Whether the replicators that are selectively eliminated are genes or species, a simple evolutionary change requires only a few replicator substitutions. A large number of replicator substitutions, however, are needed for the evolution of a complex adaptation. The minimum replacement cycle time when we consider the gene as replicator is one individual generation, from zygote to zygote. It is measured in years or months, or smaller time units. Even in the largest organisms it is measured in only tens of years. When we consider the species as replicator, on the other hand, the replacement cycle time is the interval from speciation event to speciation event, and may be measured in thousands of years, tens of thousands, hundreds of thousands. In any given period of geological time, the number of selective species extinctions that can have taken place is many orders of magnitude less than the number of selective allele replacements that can have taken place.
Again it may be just a limitation of the imagination, but while I can easily see species selection shaping a simple size trend such as the Tertiary elongation of horse legs, I cannot see such slow replicator-elimination putting together a suite of adaptations such as those of whales to aquatic life. Now, it may be said, surely that is unfair. However complex the aquatic adaptations of whales may be when taken as a whole, can they not be broken down into a set of simple size trends with allometric constants of varying magnitude and sign in different parts of the body? If you can stomach one-dimensional elongation in horses' legs as due to species selection, why not a whole set of equally simple size trends advancing in parallel, and each one driven by species selection? The weakness in this argument is a statistical one. Suppose there are ten such parallel trends, which is surely a highly conservative estimate for the evolution of aquatic adaptations in whales. If Wright's Rule is to apply to all ten, in any one speciation event each of the ten trends is as likely to be reversed as to be advanced. The chance that all ten will be advanced in any one speciation event is one half to the power ten, {107} which is less than one in a thousand. If there are twenty parallel trends, the chance of any one speciation event advancing all twenty simultaneously is less than one in a million.
Admittedly, some advancement towards complex multidimensional adaptation might be achieved through species selection even if not all ten (or twenty) trends were advanced together in any one selective event. After all, much the same critical point can be made about the selective deaths of individual organisms: it is rare to find one individual animal that is optimal in all the different dimensions of measurement. The argument finally returns to the difference in cycle times. We shall have to make a quantitative judgement taking into account the vastly greater cycle time between replicator deaths in the species selection case than in the gene selection case, and also taking into account the combinatorial problem raised above. I have neither the data nor the mathematical skills to undertake this quantitative judgement, though I have a dim feeling for the kind of methodology that would be involved in setting up an appropriate null hypothesis: it would fall under the general heading which I like to think of as “What D'Arcy Thompson might have done with a computer”, and programs of the appropriate type have already been written (Raup et al. 1973). My provisional guess is that species selection will not be found to be a generally satisfactory explanation of complex adaptation.
Consider another line of argument that the species selectionist might adopt. He might protest that it is unreasonable of me to base my combinatorial calculation on the assumption that the ten major trends in the evolution of whales are independent. Surely the number of combinations we have to consider will be drastically reduced by correlations among the different trends? It is important here to distinguish two different sources of correlation, which may be called incidental correlation and adaptive correlation. An incidental correlation is an intrinsic consequence of the facts of embryology. For example, elongation of the left foreleg is unlikely to be independent of elongation of the right foreleg. Any mutation that achieves one is intrinsically likely to achieve the other simultaneously. Slightly less obviously, the same might be partially true of elongation of the forelegs and the hindlegs. There are probably many similar cases that are less obvious still.
An adaptive correlation, on the other hand, does not follow directly from the mechanics of embryology. A lineage that moves from a terrestrial to an aquatic existence is likely to need changes to its locomotory system and its respiratory system, and there is no obvious reason to expect any intrinsic connection between them. Why should trends that convert walking limbs to flippers be intrinsically correlated with trends to boost the efficiency of the lungs in extracting oxygen? Of course it is possible that two such coadapted trends might be correlated as an incidental consequence of embryological {108} mechanisms, but the correlation is no more likely to be positive than negative. We are back to our combinatorial calculation, albeit we must exercise care in counting our separate dimensions of change.
Finally, the species selectionist may retreat and invoke ordinary low-level natural selection to weed out ill-coadapted combinations of change, so that speciation events only serve up already tried and proved combinations to the sieve of species selection. But this “species selectionist” is, by Gould's lights, no species selectionist at all! He has conceded that all the interesting evolutionary change results from inter-allele selection and not from inter-species selection, albeit it may be concentrated in brief bursts, punctuating stasis. He has conceded the violation of Wright's Rule. And if “Wright's Rule” now seems unfairly easy to violate, that is what I meant when I said that Gould had stuck his neck out. I should repeat that Wright himself is not responsible for naming the rule.
The theory of species selection, growing out of that of punctuated equilibria, is a stimulating idea which may well explain some single dimensions of quantitative change in macroevolution. I would be very surprised if it could be used to explain the sort of complex multidimensional adaptation that I find interesting, the “Paley's watch”, or “Organs of extreme Perfection and Complication”, kind of adaptation that seems to demand a shaping agent at least as powerful as a deity. Replicator selection, where the replicators are alternative alleles, may well be powerful enough. If the replicators are alternative species, however, I doubt if it is powerful enough, because it is too slow. Eldredge and Cracraft (1980, p. 269) appear to agree: “The concept of natural selection (fitness differences, or differential reproduction of individuals within populations) appears to be a corroborated, within-population phenomenon, and constitutes the best available explanation for the origin, maintenance, and possible modification of adaptations”. If this is indeed the opinion of “punctuationists” and “species selectionists” generally, it is hard to see what all the fuss is about.
For simplicity I have discussed the theory of species selection as one in which the species is treated as a replicator. The reader will have noticed, however, that this is rather like speaking of an asexually reproducing organism as a replicator. Earlier in this chapter, we saw that the test of mutilation forces us strictly to limit the title of replicator to the genome of, say, a stick insect, not the stick insect itself. Similarly, in the species selection model, it is not the species that is the replicator but the gene-pool. It is tempting, now, to say, “In that case, why not go the whole hog and regard the gene as replicator rather than some larger unit, even in the Eldredge/Gould model?” The anwer is that if they are right about a gene-pool being a coadapted unit, homeostatically buffered against change, it might have the same kind of right to be treated as a single replicator as has the genome of a stick insect. The gene-pool has this right, however, only if it {109} is reproductively isolated, just as the genome has that right only if it is reproduced asexually. Even then the right is a tenuous one.
Earlier in this chapter, we established that an organism is definitely not a replicator, although its genome may be if it is asexually reproduced. We have now seen that there may be a case for regarding the gene-pool of a reproductively isolated group, such as a species, as a replicator. If we provisonally accept the logic of this case, we can visualize evolution directed by selection among such replicators, but I have just concluded that this kind of selection is unlikely to explain complex adaptation. Apart from the small genetic fragment, which we discussed in the previous chapter, are there any other plausible candidates for the title of replicator?
I have previously supported the case for a completely non-genetic kind of replicator, which flourishes only in the environment provided by complex, communicating brains. I called it the “meme” (Dawkins 1976a). Unfortunately, unlike Cloak (1975) but, if I understand them aright, like Lumsden and Wilson (1980), I was insufficiently clear about the distinction between the meme itself, as replicator, on the one hand, and its “phenotypic effects” or “meme products” on the other. A meme should be regarded as a unit of information residing in a brain (Cloak's “i-culture”). It has a definite structure, realized in whatever physical medium the brain uses for storing information. If the brain stores information as a pattern of synaptic connections, a meme should in principle be visible under a microscope as a definite pattern of synaptic structure. If the brain stores information in “distributed” form (Pribram 1974), the meme would not be localizable on a microscope slide, but still I would want to regard it as physically residing in the brain. This is to distinguish it from its phenotypic effects, which are its consequences in the outside world (Cloak's “m-culture”).
The phenotypic effects of a meme may be in the form of words, music, visual images, styles of clothes, facial or hand gestures, skills such as opening milk bottles in tits, or panning wheat in Japanese macaques. They are the outward and visible (audible, etc.) manifestations of the memes within the brain. They may be perceived by the sense organs of other individuals, and they may so imprint themselves on the brains of the receiving individuals that a copy (not necessarily exact) of the original meme is graven in the receiving brain. The new copy of the meme is then in a position to broadcast its phenotypic effects, with the result that further copies of itself may be made in yet other brains.
Returning, for clarification, to DNA as our archetypal replicator, its consequences on the world are of two important types. Firstly, it makes copies of itself, making use of the cellular apparatus of replicases, etc. Secondly, it has effects on the outside world, which influence the chances of its copies' surviving. The first of these two effects corresponds to the meme's use of the apparatus of inter-individual communication and imitation to {110} make copies of itself. If individuals live in a social climate in which imitation is common, this corresponds to a cellular climate rich in enzymes for copying DNA.
But what about the second kind of effect of DNA, the kind conventionally called “phenotypic”? How do a meme's phenotypic effects contribute to its success or failure in being replicated? The answer is the same as for the genetic replicator. Any effect that a meme has on the behaviour of a body bearing it may influence that meme's chance of surviving. A meme that made its bodies run over cliffs would have a fate like that of a gene for making its bodies run over cliffs. It would tend to be eliminated from the meme-pool. But just as promoting bodily survival is only part of what constitutes success in genetic replicators, so there are many other ways in which memes may work phenotypically for their own preservation. If the phenotypic effect of a meme is a tune, the catchier it is the more likely it is to be copied. If it is a scientific idea, its chances of spreading through the world's scientific brains will be influenced by its compatibility with the already established corpus of ideas. If it is a political or religious idea, it may assist its own survival if one of its phenotypic effects is to make its bodies violently intolerant of new and unfamiliar ideas. A meme has its own opportunities for replication, and its own phenotypic effects, and there is no reason why success in a meme should have any connection whatever with genetic success.
This is regarded by many of my biological correspondents as the weakest point in the whole meme theory (Greene 1978; Alexander 1980, p. 78; Staddon 1981). I don't see the problem; or, rather, I do see the problem but I don't think it is any greater for memes as replicators than it is for genes. Time and again, my sociobiological colleagues have upbraided me as a turncoat, because I will not agree with them that the ultimate criterion for the success of a meme must be its contribution to Darwinian “fitness”. At bottom, they insist, a “good meme” spreads because brains are receptive to it, and the receptiveness of brains is ultimately shaped by (genetic) natural selection. The very fact that animals imitate other animals at all must ultimately be explicable in terms of their Darwinian fitness.
But there is nothing magic about Darwinian fitness in the genetic sense. There is no law giving it priority as the fundamental quantity that is maximized. Fitness is just a way of talking about the survival of replicators, in this case genetic replicators. If another kind of entity arises, which answers to the definition of an active germ-line replicator, variants of the new replicator that work for their own survival will tend to become more numerous. To be consistent, we could invent a new kind of “individual fitness”, which measured the success of an individual in propagating his memes.
It is, of course, true that “Memes are utterly dependent upon genes, but genes can exist and change quite independently of memes” (Bonner 1980). {111}
But this does not mean that the ultimate criterion for success in meme selection is gene survival. It does not mean that success goes to those memes that favour the genes of the individuals bearing them. To be sure, this will sometimes be so. Obviously a meme that causes individuals bearing it to kill themselves has a grave disadvantage, but not necessarily a fatal one. Just as a gene for suicide sometimes spreads itself by a roundabout route (e.g. in social insect workers, or parental sacrifice), so a suicidal meme can spread, as when a dramatic and well-publicized martyrdom inspires others to die for a deeply loved cause, and this in turn inspires others to die, and so on (Vidal 1955).
It is true that the relative survival success of a meme will depend critically on the social and biological climate in which it finds itself, and this climate will certainly be influenced by the genetic make-up of the population. But it will also depend on the memes that are already numerous in the meme-pool. Genetic evolutionists are already happy with the idea that the relative success of two alleles can depend upon which genes at other loci dominate the gene-pool, and I have already mentioned this in connection with the evolution of “coadapted genomes”. The statistical structure of the gene-pool sets up a climate or environment which affects the success of any one gene relative to its alleles. Against one genetic background one allele may be favoured; against another genetic background its allele may be favoured. For example, if the gene-pool is dominated by genes that make animals seek dry places, this will set up selection pressures in favour of genes for an impermeable skin. But alleles for a more permeable skin will be favoured if the gene-pool happens to be dominated by genes for seeking damp places. The point is the obvious one that selection at any one locus is not independent of selection at other loci. Once a lineage begins evolving in a particular direction, many loci will fall into step, and the resulting positive feedbacks will tend to propel the lineage in the same direction, in spite of pressures from the outside world. An important aspect of the environment which selects between alleles at any one locus will be the genes that already dominate the gene-pool at other loci.
Similarly, an important aspect of selection on any one meme will be the other memes that already happen to dominate the meme-pool (Wilson 1975). If the society is already dominated by Marxist, or Nazi memes, any new meme's replication success will be influenced by its compatibility with this existing background. Positive feedbacks will provide a momentum which can carry meme-based evolution in directions unconnected with, or even contradictory to, the directions that would be favoured by gene-based evolution. I agree with Pulliam and Dunford (1980) that cultural evolution “owes its origin and its rules to genetic evolution, but it has a momentum all its own”.
There are, of course, significant differences between meme-based and gene-based selection processes (Cavalli-Sforza & Feldman 1973, 1981). {112} Memes are not strung out along linear chromosomes, and it is not clear that they occupy and compete for discrete “loci”, or that they have identifiable “alleles”. Presumably, as in the case of genes, we can strictly only talk about phenotypic effects in terms of differences, even if we just mean the difference between the behaviour produced by a brain containing the meme and that of a brain not containing it. The copying process is probably much less precise than in the case of genes: there may be a certain “mutational” element in every copying event, and this, by the way, is also true of the “species selection” discussed earlier in the chapter. Memes may partially blend with each other in a way that genes do not. New “mutations” may be “directed” rather than random with respect to evolutionary trends. The equivalent of Weismannism is less rigid for memes than for genes: there may be “Lamarckian” causal arrows leading from phenotype to replicator, as well as the other way around. These differences may prove sufficient to render the analogy with genetic natural selection worthless or even positively misleading. My own feeling is that its main value may lie not so much in helping us to understand human culture as in sharpening our perception of genetic natural selection. This is the only reason I am presumptuous enough to discuss it, for I do not known enough about the existing literature on human culture to make an authoritative contribution to it.
Whatever the claims of memes to be regarded as replicators in the same sense as genes, the first part of this chapter established that individual organisms are not replicators. Nevertheless, they are obviously functional units of great importance, and it is now necessary to establish exactly what their role is. If the organism is not a replicator, what is it? The answer is that it is a communal vehicle for replicators. A vehicle is an entity in which replicators (genes and memes) travel about, an entity whose attributes are affected by the replicators inside it, an entity which may be seen as a compound tool of replicator propagation. But individual organisms are not the only entities that might be regarded as vehicles in this sense. There is a hierarchy of entities embedded in larger entities, and in theory the concept of vehicle might be applied to any level of the hierarchy.
The concept of hierarchy is a generally important one. Chemists believe that matter is made of about a hundred different kinds of atoms, interacting with each other by means of their electrons. Atoms are gregarious, forming huge assemblages which are governed by laws at their own level. Without contradicting the laws of chemistry, therefore, we find it convenient to ignore atoms when we are thinking about large lumps of matter. When explaining the workings of a motor car we forget atoms and van der Waal's forces as units of explanation, and prefer to talk of cylinders and sparking plugs. This lesson applies not just to the two levels of atoms and cylinder heads. There is a hierarchy, ranging from fundamental particles below the atomic level up {113} through molecules and crystals to the macroscopic chunks which our unaided sense organs are built to appreciate.
Living matter introduces a whole new set of rungs to the ladder of complexity: macromolecules folding themselves into their tertiary forms, intracellular membranes and organelles, cells, tissues, organs, organisms, populations, communities and ecosystems. A similar hierarchy of units embedded in larger units epitomizes the complex artificial products of living things — semiconductor crystals, transistors, integrated circuits, computers and embedded units that can only be understood in terms of “software”. At every level the units interact with each other following laws appropriate to that level, laws which are not conveniently reducible to laws at lower levels.
This has all been said many times before, and is so obvious as to be almost platitudinous. But one sometimes has to repeat platitudes in order to prove that one's heart is in the right place! Especially if one wishes to emphasize a slightly unconventional sort of hierarchy, for this may be mistaken for a “reductionist” attack on the idea of hierarchy itself. Reductionism is a dirty word, and a kind of “holistier than thou” self-righteousness has become fashionable. I enthusiastically follow this fashion when talking about mechanisms within individual bodies, and have advocated “neuro-economic” and “software” explanations of behaviour in preference to conventional neurophysiological ones (Dawkins 1976b). I would favour an analogous approach to individual development. But there are times when holistic preaching becomes an easy substitute for thought, and I believe the dispute about units of selection provides examples of this.
The neo-Weismannist view of life which this book advocates lays stress on the genetic replicator as a fundamental unit of explanation. I believe it has an atom-like role to play in functional, teleonomic explanation. If we wish to speak of adaptations as being “for the good of something, that something is the active, germ-line replicator. This is a small chunk of DNA, a single “gene” according to some definitions of the word. But I am of course not suggesting that small genetic units work in isolation from each other, any more than a chemist thinks that atoms do. Like atoms, genes are highly gregarious. They are often strung together along chromosomes, chromosomes are wrapped up in groups in nuclear membranes, enveloped in cytoplasm and enclosed in cell membranes. Cells too are normally not isolated, but cloned to form the huge conglomerates we know as organisms. We are now plugged into the familiar embedded hierarchy, and need go no further. Functionally speaking, too, genes are gregarious. They have phenotypic effects on bodies, but they do not do so in isolation. I stress this over and over again in this book.
The reason I may sound reductionistic is that I insist on an atomistic view of units of selection, in the sense of the units that actually survive or fail to {114} survive, while being whole-heartedly Interactlonlst when it comes to the development of the phenotypic means by which they survive:
Of course it is true that the phenotypic effect of a gene is a meaningless concept outside the context of many, or even all, of the other genes in the genome. Yet, however complex and intricate the organism may be, however much we may agree that the organism is a unit of function, I still think it is misleading to call it a unit of selection. Genes may interact, even "blend", in their effects on embryonic development, as much as you please. But they do not blend when it comes to being passed on to future generations. I am not trying to belittle the importance of the individual phenotype in evolution. I am merely trying to sort out exactly what its role is. It is the all important Instrument of replicator preservation: it is not that which is preserved [Dawkins 1978a, p. 69].
In this book I am using the word “vehicle” for an Integrated and coherent “instrument of replicator preservation”.
A vehicle is any unit, discrete enough to seem worth naming, which houses a collection of replicators and which works as a unit for the preservation and propagation of those replicators. I repeat, a vehicle is not a replicator. A replicator's success is measured by its capacity to survive in the form of copies. A vehicle's success is measured by its capacity to propagate the replicators that ride inside it. The obvious and archetypal vehicle is the individual organism, but this may not be the only level in the hierarchy of life at which the title is applicable. We can examine as candidate vehicles chromosomes and cells below the organism level, groups and communities above it. At any level, if a vehicle is destroyed, all the replicators inside it will be destroyed. Natural selection will therefore, at least to some extent, favour replicators that cause their vehicles to resist being destroyed. In principle this could apply to groups of organisms as well as to single organisms, for if a group is destroyed all the genes inside it are destroyed too.
Vehicle survival is only part of the story, however. Replicators that work for the “reproduction” of vehicles at various levels might tend “to do better than rival replicators that work merely for vehicle survival. Reproduction at the organism level is familiar enough to need no further discussion. Reproduction at the group level is more problematic. In principle a group may be said to have reproduced if it sends off a “propagule”, say a band of young organisms who go out and found a new group. The idea of a nested hierarchy of levels at which selection might take place — vehicle selection in my terms — is emphasized in Wilson's (1975) chapter on group selection (e.g. his figure 5-1).
I have previously given reasons for sharing in the general scepticism about {115} “group selection” and selection at other high levels, and nothing in the recent literature tempts me to change my mind. But that is not the point at issue here. The point here is that we must be clear about the difference between those two distinct kinds of conceptual units, replicators and vehicles. I have suggested that the best way of understanding Eldredge and Gould's theory of “species selection” is in terms of species as replicators. But the majority of models ordinarily called “group selection”, including all those reviewed by Wilson (1975), and most of those reviewed by Wade (1978), are implicitly treating groups as vehicles. The end result of the selection discussed is a change in gene frequencies, for example an increase of “altruistic genes” at the expense of “selfish genes”. It is still genes that are regarded as the replicators which actually survive (or fail to survive) as a consequence of the (vehicle) selection process.
As for group selection itself, my prejudice is that it has soaked up more theoretical ingenuity than its biological interest warrants. I am informed by the editor of a leading mathematics journal that he is continually plagued by ingenious papers purporting to have squared the circle. Something about the fact that this has been proved to be impossible is seen as an irresistible challenge by a certain type of intellectual dilettante. Perpetual motion machines have a similar fascination for some amateur inventors. The case of group selection is hardly analogous: it has never been proved to be impossible, and never could be. Nevertheless, I hope I may be forgiven for wondering whether part of group selection's enduring romantic appeal stems from the authoritative hammering the theory has received ever since Wynne-Edwards (1962) did us the valuable service of bringing it out into the open. Anti-group selectionism has been embraced by the establishment as orthodox, and, as Maynard Smith (1976a) notes, “It is in the nature of science that once a position becomes orthodox it should be subjected to criticism...” This is, no doubt, healthy, but Maynard Smith drily goes on: “It does not follow that, because a position is orthodox, it is wrong...” More generous recent treatments of group selection are given by Gilpin (1975), E. O. Wilson (1975), Wade (1978), Boorman and Levitt (1980), and D. S. Wilson (1980, but see criticism by Grafen 1980).
I am not going to go again into the debate over group selection versus individual selection. This is because the main purpose of this book is to draw attention to the weaknesses of the whole vehicle concept, whether the vehicle is an individual organism or a group. Since even the staunchest group selectionist would agree that the individual organism is a far more coherent and important “unit of selection”, I shall concentrate my attack on the individual organism as my representative vehicle, rather than on the group. The case against the group should be strengthened by default.
It may seem that I have invented my own concept, the vehicle, as an Aunt Sally to be easily knocked down. This is not so. I simply use the name vehicle {116} to give expression to a concept which is fundamental to the predominant orthodox approach to natural selection. It is admitted that, in some fundamental sense, natural selection consists in the differential survival of genes (or larger genetic replicators). But genes are not naked, they work through bodies (or groups, etc.). Although the ultimate unit of selection may indeed be the genetic replicator, the proximal unit of selection is usually regarded as something larger, usually an individual organism. Thus Mayr (1963) devotes a whole chapter to demonstrating the functional coherence of the whole genome of an individual organism. I shall discuss Mayr's points in detail in Chapter 13. Ford (1975, p. 185) disdainfully writes off the “error” that “the unit of selection is the gene, whereas it is the individual”. Gould (1977b) says:
Selection simply cannot see genes and pick among them directly. It must use bodies as an intermediary. A gene is a bit of DNA hidden within a cell. Selection views bodies. It favors some bodies because they are stronger, better insulated, earlier in their sexual maturation, fiercer in combat, or more beautiful to behold ... If, in favoring a stronger body, selection acted directly upon a gene for strength, then Dawkins might be vindicated. If bodies were unambiguous maps of their genes, then battling bits of DNA would display their colors externally and selection might act upon them directly. But bodies are no such thing... Bodies cannot be atomized into parts, each constructed by an individual gene. Hundreds of genes contribute to the building of most body parts and their action is channeled through a kaleidoscopic series of environmental influences: embryonic and postnatal, internal and external.
Now if this were really a good argument, it would be an argument against the whole of Mendelian genetics, just as much as against the idea of the gene as the unit of selection. The Lamarckian fanatic H. G. Cannon, indeed, explicitly uses it as such: “A living body is not something isolated, neither is it a collection of parts as Darwin envisaged it, like, as I have said before, so many marbles in a box. That is the tragedy of modern genetics. The devotees of the neo-Mendelian hypothesis regard the organism as so many characters controlled by so many genes. Say what they like about polygenes — that is the essence of their fantastic hypothesis” (Cannon 1959, p. 131).
Most people would agree that this is not a good argument against Mendelian genetics, and no more is it a good argument against treating the gene as the unit of selection. The mistake which both Gould and Cannon make is that they fail to distinguish genetics from embryology. Mendelism is a theory of particulate inheritance, not particulate embryology. The argument of Cannon and Gould is a valid argument against particulate {117} embryology and in favour of blending embryology. I myself give similar arguments elsewhere in this book (e.g. the cake analogy in the section of Chapter 9 called “The poverty of preformationism”). Genes do indeed bleed, as far as their effects on developing phenotypes are concerned. But, as I have already emphasized sufficiently, they do not blend as they replicate and recombine down the generations. It is this that matters for the geneticist, and it is also this that matters for the student of units of selection.
Gould goes on:
So parts are not translated genes, and selection doesn't even work directly on parts. It accepts or rejects entire organisms because suites of parts, interacting in complex ways, confer advantages. The image of individual genes, plotting the course of their own survival, bears little relationship to developmental genetics as we understand it. Dawkins will need another metaphor: genes caucusing, forming alliances, showing deference for a chance to join a pact, gauging probable environments. But when you amalgamate so many genes and tie them together in hierarchical chains of action mediated by environments, we call the resultant object a body.
Gould comes closer to the truth here, but the truth is subtler, as I hope to show in Chapter 13. The point was alluded to in the previous chapter. Briefly, the sense in which genes may be said to “caucus” and form “alliances” is the following. Selection favours those genes which succeed in the presence of other genes, which in turn succeed in the presence of them. Therefore mutually compatible sets of genes arise in gene-pools. This is more subtle and more useful than to say that “we call the resultant object a body”.
Of course genes are not directly visible to selection. Obviously they are selected by virtue of their phenotypic effects, and certainly they can only be said to have phenotypic effects in concert with hundreds of other genes. But it is the thesis of this book that we should not be trapped into assuming that those phenotypic effects are best regarded as being neatly wrapped up in discrete bodies (or other discrete vehicles). The doctrine of the extended phenotype is that the phenotypic effect of a gene (genetic replicator) is best seen as an effect upon the world at large, and only incidentally upon the individual organism — or any other vehicle — in which it happens to sit.
| << | {118} | >> |
This is a chapter about practical research methodology. There will be those who accept the thesis of this book at a theoretical level but who will object that, in practice, field workers find it useful to focus their attention on individual advantage. In a theoretical sense, they will say, it is right to see the natural world as a battleground of replicators, but in real research we are obliged to measure and compare the Darwinian fitness of individual organisms. I want to discuss a particular piece of research in detail to show that this is not necessarily the case. Instead of comparing the success of individual organisms, it is often in practice more useful to compare the success of “strategies” (Maynard Smith 1974) or “programs” or “subroutines”, averaged across the individuals that use them. Of the many pieces of research that I could have discussed, for instance the work on “optimal foraging” (Pyke, Pulliam & Charnov 1977; Krebs 1978), Parker's (1978a) dungflies, or any of the examples reviewed by Davies (1982), I choose Brockmann's study of digger wasps purely because I am very familiar with it (Brockmann, Grafen & Dawkins 1979; Brockmann &Dawkins 1979; Dawkins & Brockmann 1980).
I shall use the word “program” in exactly the same sense as Maynard Smith uses “strategy”. I prefer “program” to “strategy” because experience warns that “strategy” is quite likely to be misunderstood, in at least two different ways (Dawkins 1980). And incidentally, following the Oxford English Dictionary and standard American usage, I prefer “program” to “programme” which appears to be a nineteenth-century affectation imported from the French. A program (or strategy) is a recipe for action, a set of notional instructions that an animal seems to be “obeying”, just as a computer obeys its program. A computer programmer writes out his program in a language such as Algol or Fortran, which may look rather like imperative English. The machinery of the computer is so set up that it behaves as if {119} obeying these quasi-English instructions. Before it can run, the program is translated (by computer) into a set of more elementary “machine language” instructions, closer to the hardware and further from easy human comprehension. In a sense it is these machine language instructions that are “actually” obeyed rather than the quasi-English program, although in another sense both are obeyed, and in yet another sense neither is!
A person watching and analysing the behaviour of a computer whose program had been lost might, in principle, be able to reconstruct the program or its functional equivalent. The last four words are crucial. For his own convenience he will write the reconstructed program in some particular language — Algol, Fortran, a flow chart, some particular rigorous subset of English. But there is no way of knowing in which, if any, of these languages the program was originally written. It may have been directly written in machine language, or “hard-wired” into the machinery of the computer at manufacture. The end result is the same in any case: the computer performs some useful task such as calculating square roots, and a human can usefully treat the computer as if it was “obeying” a set of imperative instructions written out in a language that is convenient for humans to understand. I think that for many purposes such “software explanations” of behaviour mechanisms are just as valid and useful as the more obvious “hardware explanations” favoured by neurophysiologists.
A biologist looking at an animal is in somewhat the same position as an engineer looking at a computer running a lost program. The animal is behaving in what appears to be an organized, purposeful way, as if it was obeying a program, an orderly sequence of imperative instructions. The animal's program has not actually been lost, for it never was written out. Rather, natural selection cobbled together the equivalent of a hard-wired machine code program, by favouring mutations that altered successive generations of nervous systems to behave (and to learn to change their behaviour) in appropriate ways. Appropriate means, in this case, appropriate for the survival and propagation of the genes concerned. Nevertheless, although no program was ever written down, just as in the case of the computer running a program which has been lost, it is convenient for us to think of the animal as “obeying” a program “written” in some easily understood language such as English. One of the things we can then do is to imagine alternative programs or subroutines which might “compete” with each other for “computer time” in the nervous systems of the population. Though we must treat the analogy with circumspection, as I shall show, we can usefully imagine natural selection as acting directly on a pool of alternative programs or subroutines, and treat individual organisms as temporary executors and propagators of these alternative programs.
In a particular model of animal fighting, for example, Maynard Smith (1972, p. 19) postulated five alternative “strategies” (programs): {120}
1 Fight conventionally; retreat if opponent proves to be stronger or if opponent escalates.
2 Fight at escalated level. Retreat only if injured.
3 Start conventionally. Escalate only if opponent escalates.
4 Start conventionally. Escalate only if opponent continues to fight conventionally.
5 Fight at escalated level. Retreat before getting hurt if opponent does likewise.
For the purpose of computer simulation it was necessary to define these five “strategies” more rigorously, but for human understanding the simple imperative English notation is preferable. The important point for this chapter is that the five strategies were thought of as if they (rather than individual animals) were competing entities in their own right. Rules were set up in the computer simulation for the “reproduction” of successful strategies (presumably individuals adopting successful strategies reproduced and passed on genetic tendencies to adopt those same strategies, but the details of this were ignored). The question asked was about strategy success, not individual success.
A further important point is that Maynard Smith was seeking the “best” strategy in only a special sense. In fact he was seeking an “evolutionarily stable strategy” or ESS. The ESS has been rigorously defined (Maynard Smith 1974), but it can be crudely encapsulated as a strategy that is successful when competing with copies of itself. This, may seem an odd property to single out, but the rationale is really very powerful. If a program or strategy is successful, this means that copies of it will tend to become more numerous in the population of programs and will ultimately become almost universal. It will therefore come to be surrounded by copies of itself. If it is to remain universal, therefore, it must be successful when competing against copies of itself, successful compared with rare different strategies that might arise by mutation or invasion. A program that was not evolutionarily stable in this sense would not last long in the world, and would therefore not present itself for our explanation.
In the case of the five strategies listed above, Maynard Smith wanted to know what would happen in a population containing copies of all five programs. Was there one of the five which, if it came to predominate, would retain its numerical superiority against all comers? He concluded that program number 3 is an ESS: when it happens to be very numerous in the population no other program from the list does better than program 3 itself (actually there is a problem over this particular example — Dawkins 1980, p. 7 — but I shall ignore it here). When we talk of a program as “doing better” or as being “successful” we are notionally measuring success as capacity to propagate copies of the same program in the next generation: in reality this {121} is likely to mean that a successful program is one which promotes the survival and reproduction of the animal adopting it.
What Maynard Smith, together with Price and Parker (Maynard Smith & Price 1973; Maynard Smith & Parker 1976), has done is to take the mathematical theory of games and work out the crucial respect in which that theory must be modified to suit the Darwinist's purpose. The concept of the ESS, a strategy that does relatively well against copies of itself, is the result. I have already made two attempts at advocating the importance of the ESS concept and explaining its broad applicability in ethology (Dawkins 1976a, 1980), and I shall not repeat myself unnecessarily here. Here my purpose is to develop the relevance of this way of thinking for the subject of the present book, the debate about the level at which natural selection acts. I shall begin by recounting a specific piece of research that used the ESS concept. All the facts I shall give are from the field observations of Dr Jane Brockmann, reported in detail elsewhere and briefly mentioned in Chapter 3. I shall have to give a brief account of the research itself before I can relate it to the message of this chapter.
Sphex ichneumoneus is a solitary wasp, solitary in the sense that there are no social groups and no sterile workers, although the females do tend to dig their nests in loose aggregations. Each female lays her own eggs, and all labour on behalf of the young is completed before the egg is laid — the wasps are not “progressive provisioners”. The female lays one egg in an underground nest which she has previously provisioned with stung and paralysed katydids (long-horned grasshoppers). Then she closes up that nest, leaving the larva to feed on the katydids, while she herself starts work on a new nest. The life of an adult female is limited to about six summer weeks. If one wished to measure it, the success of a female could be approximated as the number of eggs that she successfully lays on adequate provisions during this time.
The thing that especially interested us was that the wasps seemed to have two alternative ways of acquiring a nest. Either a female would dig her own nest in the ground, or she would attempt to take over an existing nest which another wasp had dug. We called these two behaviour patterns digging and entering, respectively. Now, how can two alternative ways of achieving the same end, in this case two alternative ways of acquiring a nest, coexist in one population? Surely one or the other ought to be more successful, and the less successful one should be removed from the population by natural selection? There are two general reasons why this might not happen, which I will express in the jargon of ESS theory: firstly, digging and entering might be two outcomes of one “conditional strategy”; secondly, they might be equally successful at some critical frequency maintained by frequency-dependent selection — part of a “mixed ESS” (Maynard Smith 1974, 1979). If the first possibility were correct, all wasps would be programmed with the same {122} conditional rule: “If X is true, dig, otherwise enter”; for instance, “If you happen to be a small wasp, dig, otherwise use your superior size to take over another wasp's burrow”. We failed to find any evidence for a conditional program of this or any other kind. Instead, we convinced ourselves that the second possibility, the “mixed ESS”, fitted the facts.
There are in theory two kinds of mixed ESS, or rather two extremes with a continuum between. The first extreme is a balanced polymorphism. In this case, if we want to use the initials “ESS”, the final S should be thought of as standing for state of the population rather than strategy of the individual. If this possibility obtained, there would be two distinct kinds of wasps, diggers and enterers, who would tend to be equally successful. If they were not equally successful, natural selection would tend to eliminate the less successful one from the population. It is too much to hope that, by sheer coincidence, the net costs and benefits of digging would exactly balance the net costs and benefits of entering. Rather, we are invoking frequency-dependent selection. We postulate a critical equilibrium proportion of diggers, p*, at which the two kinds of wasps are equally successful. Then, if the proportion of diggers in the population falls below the critical frequency, diggers will be favoured by selection, while if it rises above the critical frequency enterers will be favoured. In this way the population would hover around the equilibrium frequency.
It is easy to think of plausible reasons why benefit might be frequency-dependent in this way. Clearly, since new burrows come into existence only when diggers dig them, the fewer diggers there are in the population the stronger will be the competition among enterers for burrows, and the lower the benefit to. a typical enterer. Conversely, when diggers are very numerous, available burrows abound and enterers tend to prosper. But, as I said, frequency-dependent polymorphism is only one end of a continuum. We now turn to the other end.
At the other end of the continuum there is no polymorphism among individuals. In the stable state all wasps obey the same program, but that program is itself a mixture. Every wasp is obeying the instruction, “Dig with probability p, enter with probability 1–p”; for instance, “Dig on 70 per cent of occasions, enter on 30 per cent of occasions”. If this is regarded as a “program”, we could perhaps refer to digging and entering themselves as “subroutines”. Every wasp is equipped with both subroutines. She is programmed to choose one or the other on each occasion with a characteristic probability, p.
Although there is no polymorphism of diggers and enterers here, something mathematically equivalent to frequency-dependent selection can go on. Here is how it would work. As before, there is a critical population frequency of digging, p*, at which entering would yield exactly the same “pay-off” as digging, p* is, then, the evolutionarily stable digging probability. {123} If the stable probability is 0.7, programs instructing wasps to follow a different rule, say “Dig with probability 0.75”, or “Dig with probability 0.65”, would do less well. There is a whole family of “mixed strategies” of the form “Dig with probability p, enter with probability 1–p” and only one of these is the ESS.
I said that the two extremes were joined by a continuum. I meant that the stable population frequency of digging, p* (70 per cent or whatever it is), could be achieved by any of a large number of combinations of pure and mixed individual strategies. There might be a wide distribution of values in individual nervous systems in the population, including some pure diggers and pure enterers. But, provided the total frequency of digging in the population is equal to the critical value p*, it would still be true that digging and entering were equally successful, and natural selection would not act to change the relative frequency of the two subroutines in the next generation. The population would be in an evolutionarily stable state. The analogy with Fisher's (1930a) theory of sex ratio equilibrium will be clear.
Proceeding from the conceivable to the actual, Brockmann's data show conclusively that these wasps are not in any simple sense polymorphic. Individuals sometimes dug and sometimes entered. We could not even detect any statistical tendency for individuals to specialize in digging or entering. Evidently, if the wasp population is in a mixed evolutionarily stable state, it lies away from the polymorphism end of the continuum. Whether it is at the other extreme — all individuals running the same stochastic program — or whether there is some more complex mixture of pure and mixed individual programs, we do not know. It is one of the central messages of this chapter that, for our research purpose, we did not need to know. Because we refrained from talking about individual success, but thought instead about subroutine success averaged over all individuals, we were able to develop and test a successful model of the mixed ESS which left open the question of where along the continuum our wasps lay. I shall return to this point after giving some pertinent facts and after outlining the model itself.
When a wasp digs a burrow, she may either stay with it and provision it, or she may abandon it. Reasons for abandoning nests were not always obvious, but they included invasion by ants and other undesirable aliens. A wasp who moves into a burrow that another has dug may find that the original owner is still in residence. In this case she is said to have joined the previous owner, and the two wasps usually work on the same nest for a while, both independently bringing katydids to it. Alternatively, the entering wasp may be fortunate enough to hit upon a nest that has been abandoned by its original owner, in which case she has it to herself. The evidence indicates that entering wasps cannot distinguish a nest that has been abandoned from one that is still occupied by its previous owner. This fact is not so surprising as it may seem, since both wasps spend most of their time out hunting, so two {124} wasps “sharing” the same nest seldom meet. When they do meet they fight, and in any case only one of them succeeds in laying an egg in the nest under dispute.
Whatever had precipitated the abandoning of a nest by the original owner seemed usually to be a temporary Inconvenience, and an abandoned nest constituted a desirable resource which was soon used by another wasp. A wasp who enters an abandoned burrow saves herself the costs associated with digging one. On the other hand, she runs the risk that the burrow she enters has not been abandoned. It may still contain the original owner, or it may contain another entering wasp who got there first. In either of these cases the entering wasp lets herself in for a high risk of a costly fight, and a high risk that she will not be the one to lay the egg at the end of a costly period of provisioning the nest.
We developed and tested a mathematical model (Brockmann, Grafen & Dawkins 1979) which distinguished four different “outcomes” or fates that could befall a wasp in any particular nesting episode.
1 She could be forced to abandon the nest, say by an ant raid.
2 She could end up alone, in sole charge of the nest.
3 She could be joined by a second wasp.
4 She could join an already incumbent wasp.
Outcomes 1 to 3 could result from an initial decision to dig a burrow. Outcomes 2 to 4 could result from an initial decision to enter. Brockmann's data enabled us to measure, in terms of probability of laying an egg per unit time, the relative “payoffs” associated with each of these four outcomes. For instance, in one study population in Exeter, New Hampshire, Outcome 4, joining”, had a payoff score of 0.35 eggs per 100 hours. This score was obtained by averaging over all occasions when wasps ended up in that outcome. To calculate it we simply added up the total number of eggs laid by wasps who, on the occasion concerned, had joined an already Incumbent wasp, and divided by the total time spent by wasps on nests that they had joined. The corresponding score for wasps who began alone but were subsequently joined was 1.06 eggs per 100 hours, and that for wasps who remained alone was 1.93 eggs per 100 hours.
If a wasp could control which of the four outcomes she ended up in, she should “prefer” to end up alone since this outcome carried the highest payoff rate, but how might she achieve this? It was a key assumption of our model that the four outcomes did not correspond to decisions that were available to a wasp. A wasp can “decide” to dig or to enter. She cannot decide to be joined or to be alone any more than a man can decide not to get cancer. These are outcomes that depend on circumstances beyond the Individual's control. In this case they depend on what the other wasps in the population {125} do. But just as a man may statistically reduce his chances of getting cancer, by taking the decision to stop smoking, so a wasp's “task” is to make the only decision open to her — dig or enter — in such a way as to maximize her chances of ending up in a desirable outcome. More strictly, we seek the stable value of p, p*, such that when p* of the decisions in the population are digging decisions, no mutant gene leading to the adoption of some other value of p will be favoured by natural selection.
The probability that an entering decision will lead to some particular outcome, such as the desirable “alone” outcome, depends on the overall frequency of entering decisions in the population. If a large amount of entering is going on in the population, the number of available abandoned burrows goes down, and the chances go up that a wasp that decides to enter will find herself in the undesirable position of joining an incumbent. Our model enables us to take any given value of p, the overall frequency of digging in the population, and predict the probability that an individual that decides to dig, or an individual that decides to enter, will end that episode in each of the four outcomes. The average payoff to a wasp that decides to dig, therefore, can be predicted for any named frequency of digging versus entering in the population as a whole. It is simply the sum, over the four outcomes, of the expected payoff yielded by each outcome, multiplied by the probability that a wasp that digs will end up in that outcome. The equivalent sum can be worked out for a wasp that decides to enter, again for any named frequency of digging versus entering in the population. Finally, making certain plausible additional assumptions which are listed in the original paper, we solve an equation to find the population digging frequency at which the average expected benefit to a wasp that digs is exactly equal to the average expected benefit to a wasp that enters. That is our predicted equilibrium frequency which we can compare with the observed frequency in the wild population. Our expectation is that the real population should either be sitting at the equilibrium frequency or else in the process of evolving towards the equilibrium frequency. The model also predicts the proportion of wasps ending up in each of the four outcomes at equilibrium, and these figures too can be tested against the observed data. The model's equilibrium is theoretically stable in that it predicts that deviations from equilibrium will be corrected by natural selection.
Brockmann studied two populations of wasps, one in Michigan and one in New Hampshire. The results were different in the two populations. In Michigan the model failed to predict the observed results, and we concluded that it was quite inapplicable to the Michigan population, for unknown reasons as discussed in the original paper (the fact that the Michigan population has now gone extinct is probably fortuitous!). The New Hampshire population, on the other hand, gave a convincing fit to the predictions of the model. The predicted equilibrium frequency of entering {126} was 0.44, and the observed frequency was 0.41. The model also successfully predicted the frequency of each of the four “outcomes” in the New Hampshire population. Perhaps most importantly, the average payoff resulting from digging decisions did not differ significantly from the average payoff resulting from entering decisions.
The point of my telling this story in the present book has now finally arrived. I want to claim that we would have found it difficult to do this research if we had thought in terms of individual success, rather than in terms of strategy (program) success averaged over all individuals. If the mixed ESS had happened to lie at the balanced polymorphism end of the continuum, it would, indeed, have made sense to ask something like the following. Is the success of wasps that dig equal to the success of wasps that enter? We would have classified wasps as diggers or enterers, and compared the total lifetime's egg-laying success of individuals of the two types, predicting that the two success scores should be equal. But as we have seen these wasps are not polymorphic. Each individual sometimes digs and sometimes enters.
It might be thought that it would have been easy to do something like the following. Classify all individuals into those that entered with a probability less than 0.1, those that entered with a probability between 0.1 and 0.2, those with a probability between 0.2 and 0.3, between 0.3 and 0.4, 0.4 and 0.5, etc. Then compare the lifetime reproductive successes of wasps in the different classes. But supposing we did this, exactly what would the ESS theory predict? A hasty first thought is that those wasps with a p value close to the equilibrium p* should enjoy a higher success score than wasps with some other value of p: the graph of success against p should peak at an “optimum” at p*. But p* is not really an optimum value, it is an evolutionarily stable value. The theory expects that, when p* is achieved in the population as a whole, digging and entering should be equally successful. At equilibrium, therefore, we expect no correlation between a wasp's digging probability and her success. If the population deviates from equilibrium in the direction of too much entering, the “optimum” choice rule becomes “always dig” (not “dig with probability p*”). If the population deviates from equilibrium in the other direction, the “optimum” policy is “always enter”. If the population fluctuates at random around the equilibrium value, analogy with sex ratio theory suggests that in the long run genetic tendencies to adopt exactly the equilibrium value, p*, will be favoured over tendencies to adopt any other consistent value of p (Williams 1979). But in any one year this advantage is not particularly likely to be evident. The sensible expectation of the theory is that there should be no significant difference in success rates among the classes of wasps.
In any case this method of dividing wasps up into classes presupposes that there is some consistent variation among wasps in digging tendency. The {127} theory gives us no particular reason to expect that there should be any such variation. Indeed, the analogy with sex ratio theory just mentioned gives positive grounds for expecting that wasps should not vary in digging probability. In accordance with this, a statistical test on the actual data revealed no evidence of inter-individual variation in digging tendency. Even if there were some individual variation, the method of comparing the success of individuals with different p values would have been a crude and insensitive one for comparing the success rates of digging and entering. This can be seen by an analogy.
An agriculturalist wishes to compare the efficacy of two fertilizers, A and B. He takes ten fields and divides each of them into a large number of small plots. Each plot is treated, at random, with either A or B, and wheat is sown in all the plots of all the fields. Now, how should he compare the two fertilizers? The sensitive way is to take the yields of all the plots treated with A and compare them with the yields of all the plots treated with B, across all ten fields. But there is another, much cruder way. It happens that in the random allocation of fertilizers to plots, some of the ten fields received a relatively large amount of fertilizer A, while others happened to be given a relatively large amount of B. The agriculturalist could, then, plot the overall yield of each of the ten fields against the proportion of the field that was treated with fertilizer A rather than B. If there is a very pronounced difference in quality between the two fertilizers, this method might just show it up, but far more probably the difference would be masked. The method of comparing the yields of the ten fields would be efficient only if there is very high between-field variance, and there is no particular reason to expect this.
In the analogy, the two fertilizers stand for digging and entering. The fields are the wasps. The plots are the episodes of time that individual wasps devote either to digging or to entering. The crude method of comparing digging and entering is to plot the lifetime success of individual wasps against their proportionate digging tendency. The sensitive way is the one we actually employed.
We made a detailed and exhaustive inventory of the time spent by each wasp on each burrow with which she was associated. We divided each individual female's adult lifetime into consecutive episodes of known duration, each episode being designated a digging episode if the wasp concerned began her association with the burrow by digging it. Otherwise, it was designated an entering episode. The end of each episode was signalled by the wasp's leaving the nest for the last time. This instant was also treated as the start of the next burrow episode, even though the next burrow site had not, at the time, been chosen. That is to say, in our time accounting, the time spent searching for a new burrow to enter, or searching for a place to dig a new burrow, was designated retroactively as time “spent” on that new {128} burrow. It was added to the time subsequently spent provisioning the burrow with katydids, fighting other wasps, feeding, sleeping, etc., until the wasp left the new burrow for the last time.
At the end of the season, therefore, we could add up the total number of wasp hours spent on dug-burrow episodes, and also the total number of wasp hours spent on entered-burrow episodes. For the New Hampshire study these two figures were 8518.7 hours and 6747.4 hours, respectively. This is regarded as time spent, or invested, for a return, and the return is measured in numbers of eggs. The total number of eggs laid at the end of the dug-burrow episodes (i.e. by wasps that had dug the burrow concerned) in the whole New Hampshire population during the year of study was 82. The corresponding number for entered-burrow episodes was 57 eggs. The success rate of the digging subroutine was, therefore, 82/8518.7 = 0.96 eggs per 100 hours. The success rate of the entering subroutine was 57/6747.4 = 0.84 eggs per 100 hours. These success scores are averaged across all the individuals who used the two subroutines. Instead of counting the number of eggs laid in her lifetime by an individual wasp — the equivalent of measuring the wheat yield of each of the ten fields in the analogy — we count the number of eggs laid “by” the digging (or entering) subroutine per unit “running time” of the subroutine.
There is another respect in which it would have been difficult for us to have done this analysis if we had insisted on thinking in terms of individual success. In order to solve the equation to predict the equilibrium entering frequency, we had to have empirical estimates of the expected payoffs of each of the four “outcomes” (abandons, remains alone, is joined, joins). We obtained payoff scores for the four outcomes in the same way as we obtained success scores for each of the two strategies, dig and enter. We averaged over all individuals, dividing the total number of eggs laid in each outcome by the total time spent on episodes that ended up in that outcome. Since most individuals experienced all four outcomes at different times, it is not clear how we could have obtained the necessary estimates of outcome payoffs if we had thought in terms of individual success.
Notice the important role of time in the computation of the “success” of the digging and entering subroutines (and of the payoff given by each outcome). The total number of eggs laid “by” the digging subroutine is a poor measure of success until it has been divided by the time spent on the subroutine. The number of eggs laid by the two subroutines might be equal, but if digging episodes are on average twice as long as entering episodes, natural selection will presumably favour entering. In fact rather more eggs were laid “by” the digging subroutine than by the entering one, but correspondingly more time was spent on the digging subroutine so the overall success rates of the two were approximately equal. Notice too that we do not specify whether the extra time spent on digging is accounted for by a greater number of wasps {129} digging, or by each digging episode lasting longer. The distinction may be important for some purposes, but it doesn't matter for the kind of economic analysis we undertook.
It was clearly stated in the original paper (Brockmann, Grafen & Dawkins 1979), and must be repeated here, that the method we used depended upon some assumptions. We assumed, for instance, that a wasp's choice of subroutine on any particular occasion did not affect her survival or success rate after the end of the episode concerned. Thus the costs of digging were assumed to be reflected totally in the time spent on digging episodes, and the costs of entering reflected in the time spent on entering episodes. If the act of digging had imposed some extra cost, say a risk of wear and tear to the limbs, shortening life expectation, our simple time-cost accounting would need to be amended. The success rates of the digging and the entering subroutines would have to be expressed, not in eggs “per hour, but” in eggs per “opportunity cost”. Opportunity cost might still be measured in units of time, but digging time would have to be scaled up in a costlier currency than entering time, because each hour spent in digging shortens the expectation of effective life of the individual. Under such circumstances it might be necessary, in spite of all the difficulties, to think in terms of individual success rather than subroutine success.
It is for this kind of reason that Clutton-Brock et al. (1982) are probably wise in their ambition to measure the total lifetime reproductive success rates of their individual red deer stags. In the case of Brockmann's wasps, we have reason to think that our assumptions were correct, and that we were justified in ignoring individual success and concentrating on subroutine success. Therefore what N. B. Davies, in a lecture, jocularly called the “Oxford method” (measuring subroutine success) and the “Cambridge method” (measuring individual success) may each be justified in different circumstances. I am not saying that the Oxford method should always be used. The very fact that it is sometimes preferable is sufficient to answer the claim that field workers interested in measuring costs and benefits always have to think in terms of individual costs and benefits.
When computer chess tournaments are held, a layman might imagine that one computer plays against another. It is more pertinent to describe the tournament as being between programs. A good program will consistently beat a poor program, and it doesn't make any difference which physical computer either program is running on. Indeed the two programs could swap physical computers every other game, each one running alternately in an IBM and an ICL computer, and the result at the end of the tournament will be the same as if one program consistently ran in the IBM and the other consistently ran in the ICL. Similarly, to return to the analogy at the beginning of this chapter, the digging subroutine “runs” in a large number of different physical wasp nervous systems. Entering is the name of a rival {130} subroutine which also rues in many different wasp nervous systems, including some of the same physical nervous systems as, at other times, run the digging subroutine. Just as a particular IBM or ICL computer functions as the physical medium through which any of a variety of chess programs can act out their skills, so one individual wasp is the physical medium through which sometimes the digging subroutine, at other times the entering subroutine, acts out its characteristic behaviour.
As already explained, I call digging and entering “subroutines”, rather than programs, because we have already used “program” for the overall lifetime choosing rules of an individual. An individual is regarded as being programmed with a rule for choosing the digging or the entering subroutine with some probability p. in the special case of a polymorphism, where each individual is either a lifelong digger or a lifelong enterer, p becomes 1 or 0, and the categories program and subroutine become synonymous. The beauty of calculating the egg-laying success rates of subroutines rather than of individuals is that the procedure we adopt is the same regardless of where on the mixed strategy continuum our animals are. Anywhere along the continuum we still predict that the digging subroutine should, at equilibrium, enjoy a success rate equal to that of the entering subroutine.
It is tempting, though rather misleading, to push this line of thought to what appears to be its logical conclusion, and think in terms of selection acting directly on subroutines in a subroutine pool. The population's nervous tissue, its distributed computer hardware, is inhabited by many copies of the digging subroutine and many copies of the entering subroutine. At any given time the proportion of running copies of the digging subroutine is p. There is a critical value of p, called p*, at which the success rate of the two subroutines is equal. If either of the two becomes too numerous in the subroutine pool, natural selection penalizes it and the equilibrium is restored.
The reason this is misleading is that selection really works on the differential survival of alleles in a gene pool. Even with the most liberal imaginable interpretation of what we mean by gene control, there is no useful sense in which the digging subroutine and the entering subroutine could be thought of as being controlled by alternative alleles. If for no other reason, this is because the wasps, as we have seen, are not polymorphic, but are programmed with a stochastic rule for choosing to dig or enter on any given occasion. Natural selection must favour genes that act on the stochastic program of individuals, in particular controlling the value of p, the digging probability. Nevertheless, although it is misleading if taken too literally, the model of subroutines competing directly for running time in nervous systems provides some useful short cuts to getting the right answer.
The idea of selection in a notional pool of subroutines also leads us to think about yet another time-scale on which an analogue of frequency-dependent {131} selection might occur. The present model allows that from day to day the observed number of running copies of the digging subroutine might change, as individual wasps obeying their stochastic programs switch their hardware from one subroutine to another. So far I have implied that a given wasp is born with a built-in predilection to dig with a certain characteristic probability. But it is also theoretically possible that wasps might be equipped to monitor the population around them with their sense organs, and choose to dig or enter accordingly. In ESS jargon focusing on the individual level, this would be regarded as a conditional strategy, each wasp obeying an “if clause” of the following form: “If you see a large amount of entering going on around you, dig, otherwise enter”. More practically, each wasp might be programmed to follow a rule of thumb such as: “Search for a burrow to enter; if you have not found one after a time I, give up and dig your own”. As it happens our evidence goes against such a “conditional strategy” (Brockmann & Dawkins 1979), but the theoretical possibility is interesting. From the present point of view what is particularly interesting is this. We could still analyse the data in terms of a notional selection between subroutines in a subroutine pool, even though the selection process leading to the restoration of the equilibrium when perturbed would not be natural selection on a generational time-scale. It would be a developmentally stable strategy or DSS (Dawkins 1980) rather than an ESS, but the mathematics could be much the same (Harley 1981).
I must warn that analogical reasoning of this sort is a luxury that we dare not indulge unless we are capable of clearly seeing the limitations of the analogy. There are real and important distinctions between Darwinian selection and behavioural assessment, just as there were real and important distinctions between a balanced polymorphism and a true mixed evolutionarily stable strategy. Just as the value of p, the individual's digging probability, was considered to be adjusted by natural selection, so, in the behavioural assessment model, t, the individual's criterion for responding to the frequency of digging in the population, is presumably influenced by natural selection. The concept of selection among subroutines in a subroutine pool blurs some important distinctions while pointing up some important similarities: the weaknesses of this way of thinking are linked to its strengths. What I do remember is that, when we were actually wrestling with the difficulties of the wasp analysis, one of our main leaps forward occurred when, under the influence of A. Grafen, we kicked the habit of worrying about individual reproductive success and switched to an imaginary world where “digging” competed directly with “entering”; competed for “running time” in future nervous systems.
This chapter has been an interlude, a digression. I have not been trying to argue that “subroutines”, or “strategies” are really true replicators, true units of natural selection. They are not. Genes and fragments of genomes are true {132} replicators. Subroutines and strategies can be thought of for certain purposes as if they were replicators, but when those purposes have been served we must return to reality. Natural selection really has the effect of choosing between alleles in wasp gene-pools, alleles which influence the probability that individual wasps will enter or dig. We temporarily laid this knowledge aside and entered an imaginary world of “inter-subroutiee selection” for a specific methodological purpose. We were justified in doing this because we were able to make certain assumptions about the wasps, and because of the already demonstrated mathematical equivalence between the various ways in which a mixed evolutionarily stable strategy can be put together.
As in the case of Chapter 4, the purpose of this chapter has been to undermine our confidence in the individual-centred view of teleonomy, in this case by showing that it is not always useful, in practice, to measure individual success if we are to study natural selection in the field. The next two chapters discuss adaptations which, by their very nature, we cannot even begin to understand if we insist on thinking in terms of individual benefit.
| << | {133} | >> |
Natural selection is the process whereby replicators out-propagate each other. They do this by exerting phenotypic effects on the world, and it is often convenient to see those phenotypic effects as grouped together in discrete “vehicles” such as individual organisms. This gives substance to the orthodox doctrine that each individual body can be thought of as a unitary agent maximizing one quantity — “fitness”, various notions of which will be discussed in Chapter 10. But the idea of individual bodies maximizing one quantity relies on the assumption that replicators at different loci within a body can be expected to “cooperate”. In other words we must assume that the allele that survives best at any given locus tends to be the one that is best for the genome as a whole. This is indeed often the case. A replicator that ensures its own survival and propagation down the generations by conferring on its successive bodies resistance to a dangerous disease, say, will thereby benefit all the other genes in the successive genomes of which it is a member. But it is also easy to Imagine cases where a gene might promote its own survival while harming the survival chances of most of the rest of the genome. Following Alexander and Borgia (1978) I shall call such genes outlaws.
I distinguish two main classes of outlaws. An “allelic outlaw” is defined as a replicator that has a positive selection coefficient at its own locus but for which, at most other loci, there is selection in favour of reducing its effect at its own locus. An example is a “segregation distorter” or “melotic drive” gene. It is favoured at its own locus through getting Itself Into more than 50 per cent of the gametes produced. At the same time, genes at other loci whose effect is to reduce the segregation distortion will be favoured by selection at their respective loci. Hence the segregation distorter is an outlaw. The other main class of outlaw, the “laterally spreading outlaw”, is less familiar. It will be discussed in the next chapter.
From the viewpoint of this book there is a sense in which we expect all genes to be potential outlaws, so much so that the very term might seem {134} superfluous. On the other hand it can be argued, firstly, that outlaws are unlikely to be found in nature because at any given locus the allele that is best at surviving will nearly always turn out to be the allele that is best at promoting the survival and reproduction of the organism as a whole. Secondly various authors, following Leigh (1971), have argued that, even if outlaws did arise and were temporarily favoured by selection, they are likely, in Alexander and Borgia's words, “to have their effects nullified, at least to the extent that they are outnumbered by the other genes in the genome”. It does, indeed, follow from the definition of an outlaw that this should tend to happen. The suggestion being made is that, whenever an outlaw arises, selection will favour modifier genes at so many other loci that no trace will remain of the phenotypic effects of the outlaw. Outlaws will therefore, the suggestion goes, be transient phenomena. This would not, however, make them negligible: if genomes are riddled with outlaw-suppressing genes, this, in itself, is an important effect of outlaw genes, even if no trace remains of their original phenotypic effects. I shall discuss the relevance of modifier genes in a later section.
There is a sense in which a “vehicle” is worthy of the name in inverse proportion to the number of outlaw replicators that it contains. The idea of a discrete vehicle maximizing a unitary quantity — fitness — depends on the assumption that the replicators that it serves all stand to gain from the same properties and behaviour of their shared vehicle. If some replicators would benefit from the vehicle's doing act X, while other replicators would benefit from its doing act Y, that vehicle is correspondingly less likely to behave as a coherent unit. It will have the attributes of a human organization that is governed by a quarrelsome committee — pulled this way and that, and unable to show decisiveness and consistency of purpose.
There is a superficial analogy here with group selection. One of the problems of the theory that groups of organisms function as effective gene vehicles is that outlaws (from the group's point of view) are very likely to arise and be favoured by selection. If we are hypothesizing the evolution of individual restraint through group selection, a gene that makes individuals behave selfishly in an otherwise altruistic group is analogous to an outlaw. It is the near inevitability of such “outlaws” arising that has dashed the hopes of many a group selection modeller.
The individual body is a much more persuasive gene vehicle than the group because, among other reasons, outlaw replicators within the body are not very likely to be strongly favoured over their alleles. The fundamental reason for this is the drilled formality of the mechanisms of individual reproduction, the “gavotte of chromosomes” as Hamilton (1975b) called it. If all replicators “know” that their only hope of getting into the next generation is via the orthodox bottleneck of individual reproduction, all will have the same “interests at heart”; survival of the shared body to reproductive age, {135} successful courtship and reproduction of the shared body, and a successful outcome to the parental enterprise of the shared body. Enlightened self-interest discourages outlawry when all replicators have an equal stake in the normal reproduction of the same shared body.
Where reproduction is asexual the stake is equal and total, for all replicators have the same 100 per cent chance of finding themselves in every child produced by their joint efforts. Where reproduction is sexual the corresponding chance for each replicator is only half as great, but the ritualized courtliness of meiosis, Hamilton's “gavotte”, largely succeeds in guaranteeing to each allele an equal chance of reaping the benefits of the success of the joint reproductive enterprise. It is, of course, another question why the gavotte of the chromosomes is so courtly. It is an immensely important question which I shall evade on a simple plea of cowardice. It is one of a set of questions about the evolution of genetic systems with which better minds than mine have wrestled more or less unsuccessfully (Williams 1975, 1980; Maynard Smith 1978a), a set of questions which moved Williams to remark that “there is a kind of crisis at hand in evolutionary biology”. I don't understand why meiosis is the way it is, but given that it is much follows. In particular the organized fair-dealing of meiosis helps to account for the coherence and harmony which unites the parts of an individual organism. If, at the level of the group of individuals as potential vehicle, the privilege of reproduction was granted with the same scrupulous probity in a similarly well-disciplined “gavotte of the organisms”, group selection might become a more plausible theory of evolution. But, with the possible exception of the very special case of the social insects, group “reproduction” is anarchical and favourable to individual outlawry. Even social insect colonies will never seem fully harmonious again after Trivers and Hare's ingenious analysis of sex ratio conflicts (see Chapter 4).
This consideration tells us where we should look first if we want to discover outlaws within the individual body as vehicle. Any replicator that managed to subvert the rules of meiosis so that it enjoyed more than the ordained 50 per cent chance of ending up in a gamete would, other things being equal, tend to be favoured over its alleles in natural selection. Such genes are known to geneticists under the name of meiotic drive genes or segregation distorters. I have already used them as an example to illustrate my definition of an outlaw.
The account of segregation distorters that I shall mainly follow is that of Crow (1979), who uses language congenial to the spirit of this book. His paper is called “Genes that violate Mendel's rules”, and it ends as follows: “The Mendelian system works with maximum efficiency only if it is scrupulously fair to all genes. It is in constant danger, however, of being {136} upset by genes that subvert the meiotic process to their own advantage ... There are many refinements of meiosls and sperm formation whose purpose is apparently to render such cheating unlikely. And yet some genes have managed to beat the system”.
Crow suggests that segregation distorters may be much more common than we ordinarily realize, for the methods of geneticists are not well geared to detecting them, especially if they produce only slight, quantitative effects. The SD genes in Drosophila are particularly well studied, and here there is some indication as to the actual mechanism of distortion. “While the homologous chromosomes are still paired up during meiosis, the SD chromosome might do something to its normal partner (and rival) that later causes a dysfunction of the sperm receiving the normal chromosome ... SD might actually break the other chromosome” (Crow 1979, my macabre emphasis). There is evidence that, in SD-heterozygous individuals, sperms not containing the SD chromosome have abnormal, and presumably faulty, tails. It might be thought, then, that the faulty tail results from some sabotage to the non-SD chromosome in the sperms containing it. This cannot be the whole story, as Crow points out, because sperms have been shown to be capable of developing normal tails without any chromosomes at all. Indeed, the whole sperm phenotype seems usually to be under the control of the diploid genotype of the father, not its own haploid genotype (Beatty & Gluecksohn-Waelsch 1972; see below). “The effect of the SD chromosome on its homologue cannot, then, be simply to inactivate some function, because no function is required. SD must somehow induce its partner to commit a positive act of sabotage”.
Segregation distorters prosper when rare, because the chances are then good that their victims are alleles, not copies. When common, the distorter tends to occur homozygously, and therefore sabotages copies of itself, making the organism virtually sterile. The story is more complicated than this, but computer simulations described by Crow suggest that a stable proportion of segregation distorter genes will be maintained at a frequency somewhat greater than would be accounted for by recurrent mutation alone. There is some evidence that this-is so in real life.
In order to qualify as an outlaw, a segregation distorter must do harm to most of the rest of the genome, not just to its alleles. Segregation distorters could have this effect by reducing the total gamete count of the individual. Even if they did not do this, there is a more general reason for expecting there to be selection at other loci in favour of suppressing them (Crow 1979). The argument needs to be developed stage by stage. Firstly many genes, when compared with their alleles, have several pleiotropic effects. Lewontin (1974) goes so far as to speak of “...the undoubted truth that every gene affects every character...”. While to call this an “undoubted truth” may be, {137} to put it mildly, an enthusiastic exaggeration, for my purposes I need only assume that most new mutations have several pleiotropic effects.
Now it is reasonable to expect that most such pleiotropic effects will be deleterious — mutational effects usually are. If a gene is favoured by selection resulting from one beneficial effect, this will be because the advantages of its beneficial effect quantitatively outweigh the disadvantages of its other effects. Normally, by “beneficial” and “deleterious”, we mean beneficial and deleterious to the whole organism. In the case of a segregation distorter, however, the beneficial effect we are talking about is beneficial to the gene alone. Any pleiotropic effects that it may have on the body are pretty likely to be deleterious to the whole body's survival and reproduction. Segregation distorters are therefore, on the whole, likely to be outlaws: we expect that selection will favour genes at other loci whose phenotypic effect is to reduce the segregation distortion. This brings us to the topic of modifiers.
The classic proving ground for the theory of modifier genes was R. A. Fisher's explanation of the evolution of dominance. Fisher (1930a, but see Charlesworth 1979) suggested that the beneficial effects of a given gene tend to become dominant through the selection of modifiers, while its deleterious effects tend to become recessive. He noted that dominance and recessiveness are not properties of genes themselves, but properties of their phenotypic effects. Indeed a given gene can be dominant in one of its pleiotropic effects and recessive in another. A phenotypic effect of a gene is the joint product of itself and its environment, an environment which includes the rest of the genome. This interactive view of gene action, which Fisher had to argue for at length in 1930, had become so well accepted by 1958 that he was able to take it for granted in the second edition of his great book. It follows from it that dominance or recessiveness, like any other phenotypic effect, may itself evolve, through the selection of other genes elsewhere in the genome, and this was the basis of Fisher's theory of dominance. Although these other genes are known as modifiers, it is now realized that there is no separate category of modifier genes as distinct from major genes. Rather, any gene may serve as modifier of the phenotypic effects of any other gene. Indeed, any given gene's phenotypic effects are susceptible to modification by many other genes in the genome, genes which themselves may have many other major and minor effects (Mayr 1963). Modifiers have been invoked for various other theoretical purposes, for instance in the Medawar/Williams/Hamilton progression of theories of the evolution of senescence (Kirkwood & Holliday 1979).
The relevance of modifiers to the subject of outlaw genes has already been alluded to. Since any gene's phenotypic effects may be subject to {138} modification by genes at other loci, and since outlaws, by definition, work to the detriment of the rest of the genome, we should expect selection to favour genes that happen to have the effect of neutralizing the outlaw's deleterious effects on the body as a whole. Such modifiers would be favoured over alleles that did not influence the outlaw's effects. Hickey and Craig (1966), studying a sex-ratio distorting gene (see below) in the yellow-fever mosquito Aedes aegypti, found evidence for evolutionary diminution of the distorting effect which could be interpreted as resulting from the selection of modifiers (though their own interpretation was slightly different). If outlaws do, in general, call forth the selection of suppressing modifiers, there will presumably be an arms race between each outlaw and its modifiers.
As in any other arms race (Chapter 4), we now ask whether there is any general reason to expect one side to prevail over the other. Leigh (1971, 1977), Alexander and Borgia (1978), Kurland (1979, 1980), Hartung (in press), and others suggest that there is just such a general reason. Since, for any single outlaw, suppressing modifiers may arise anywhere in the genome, the outlaw will be outnumbered. As Leigh (1971) puts it, “It is as if we had to do with a parliament of genes: each acts in its own self-interest, but if its acts hurt the others, they will combine together to suppress it ... However, at loci so closely linked to a distorter that the benefits of "riding its coattails" outweigh the damage of its disease, selection tends to enhance the distortion effect. Thus a species must have many chromosomes if, when a distorter arises, selection at most loci is to favor its suppression. Just as too small a parliament may be perverted by the Cabals of a few, a species with only one, slightly linked chromosome is an easy prey to distorters” (Leigh 1971, p. 249). I am not sure what I think about Leigh's point about chromosome numbers, but his more general point that there is some sense in which outlaws may be “outnumbered” (Alexander & Borgia 1978, p. 458) by their modifiers seems to me to have promise.
I suppose “outnumbering” could, in practice, work in two main ways. Firstly, if different modifiers each cause a quantitative diminution of the outlaw's effect, several modifiers might combine additively. Secondly, if any one of several modifiers would suffice to neutralize the outlaw, the chance of effective neutralization goes up with the number of available modifier loci. Alexander and Borgia's metaphor of “outnumbering”, and Leigh's metaphor of the power of the collective in a “parliament” of the many, could be given meaning in either or both of these two ways. It is important for the argument that segregation distorters at different loci could not, in any obvious sense, “pool their efforts”. They are not working for some common end of “general segregation distortion”. Rather, each one is working to distort segregation in favour of itself, and this will hurt other segregation distorters just as much as it hurts non-distorters. Suppressors of segregation distorters, on the other hand, could, in a sense, pool their efforts. {139}
The parliament of genes is one of those metaphors which, if we are not careful, tricks us into thinking that it explains more than it does. Like all humans, but unlike genes, human Members of Parliament are highly sophisticated computers capable of using foresight and language to conspire and reach agreements. Outlaws may seem to be suppressed by agreement of a collective in the parliament of genes, but what is really going on is the selection of modifier genes in preference to non-modifying alleles at their respective loci. Needless to say, Leigh and the other advocates of the “parliament of genes” hypothesis are well aware of this. I now want to extend the list of outlaws.
If a segregation distorter occurs on a sex chromosome, it is not only an outlaw in conflict with the rest of the genome and therefore subject to suppression by modifiers: it also, incidentally, threatens the whole population with extinction. This is because, in addition to ordinary detrimental side effects, it also distorts the sex ratio, and may even eliminate one sex from the population altogether. In one of Hamilton's (1967) computer simulations, a single mutant male with a “driving Y” chromosome causing males to have only sons and no daughters, was introduced into a population of 1000 males and 1000 females. It took only fifteen generations to drive the model population extinct for lack of females. Something like this effect has been demonstrated in the laboratory (Lyttle 1977). The possibility of using a driving Y gene in the control of serious pests like the yellow-fever mosquito did not escape Hickey and Craig (1966). It is a method with sinister elegance because it is so cheap: all the work of spreading the pest control agent is done by the pests themselves together with natural selection. It is like “germ warfare” except that the lethal “germ” is not an extraneous virus but a gene in the species's own gene-pool. Perhaps the distinction is not a fundamental one anyway (Chapter 9).
X-linked drive is likely to have the same sort of detrimental effect on populations as Y-linked drive, but tends to take more generations to extinguish the population (Hamilton 1967). The driving gene on an X chromosome causes males to have daughters rather than sons (except in birds, Lepidoptera, etc.). As we saw in Chapter 4, if haploid male Hymenoptera could iniuence the amount of care devoted to their spouses' offspring, they would favour daughters rather than sons, since males pass no genes on to sons. The mathematics of this situation are analogous to the case of X-linked drive, the whole genome of the male hymenopteran functioning like an X chromosome (Hamilton 1967, p. 481 and footnote 18).
It is often the case that X chromosomes cross over with each other but not with Y chromosomes. It follows that all genes on X chromosomes could stand to gain from the presence in the gene-pool of a driving X gene which distorts {140} gametogenesis in the heterogametic sex in favour of X gametes and against Y gametes. Genes on X chromosomes are, in a sense, united against Y genes, in a kind of “anti linkage group”, simply because they have no chance of finding themselves on a Y chromosome. Modifiers to suppress X-linked meiotic drive in the heterogametic sex might well not be favoured if they arose at other loci on X chromosomes. They would be favoured if they arose on autosomes. This is different from the case of segregation distorters on autosomes: here there might well be selection in favour of suppressing them by modifiers at other loci even on the same chromosome. X-linked distorters affecting gamete production in the heterogametic sex are, then, outlaws from the point of view of the autosomal part of the gene-pool, but not from the point of view of the rest of the X-chromosomal part of the gene-pool. This potential “solidarity” among the genes on sex chromosomes suggests that the concept of the outlaw gene may be too simple. It conveys the image of a single rebel standing out against the rest of the genome. At times we might, instead, do better to think in terms of wars between rival gangs of genes, for instance the X-chromosome genes against the rest. Cosmides and Tooby (1981) coin the useful term “coreplicon” for such a gang of genes that replicate together and therefore tend to work for the same ends. In many cases neighbouring coreplicons will blur into each other.
Ganging up by Y-chromosome genes is even more to be expected. As long as Y chromosomes do not cross over, it is clear that all genes on a Y chromosome stand to gain from the presence of a Y-linked distorter every bit as much as the distorter gene itself. Hamilton (1967) made the interesting suggestion that the reason for the well-known inertness of Y chromosomes (hairy ears seems to be the only conspicuous Y-linked trait in man) is that Y-suppressing modifiers have been positively selected elsewhere in the genome. It is not obvious how a modifier might go about suppressing the phenotypic activity of an entire chromosome, since the various phenotypic effects of a single chromosome are usually so heterogeneous. (Why would selection not suppress only the effects of the driving genes, leaving other Y-linked effects intact?) I suppose it might do it by physically deleting large chunks of Y chromosome, or by contriving to quarantine the Y chromosome from the cell's transcription machinery.
A weird example of a driving replicator which is probably not a gene in the ordinary sense of the word is given by Werren, Skinner and Charnov (1981). They studied the parasitoid wasp Nasonia vitripennis, in which there is a variety of males called D1, or “daughterless”. Wasps being haplodiploid, males pass their genes only to daughters: a male's mate may have sons, but those sons are haploid and fatherless. When D1 males mate with females, they cause them to produce all male broods. Most of the sons of females mated to D1 males are themselves D1 males. Although no nuclear genes pass from father to son, therefore, the D1 factor somehow does pass {141} from father to “son”. The D1 factor rapidly spreads, in exactly the same way as a driving Y chromosome would. It is not known what the D1 factor physically consists of. It is certainly not nuclear genetic material, and it is theoretically possible that it is not even composed of nucleic acid, although Werren et al. suspect that it probably is cytoplasmically borne nucleic acid. Theoretically, any kind of physical or chemical influence of a D1 male on his mate, which causes her to have D1 sons, would spread like a driving Y chromosome, and would qualify as an active germ-line replicator in the sense of Chapter 5. It is also an outlaw par excellence, for it spreads itself at the total expense of all the nuclear genes in the males that bear it.
With some exceptions, all the diploid cells of an organism are genetically identical, but the haploid gametes it produces are all different. Only one out of many sperms in an ejaculate can fertilize an egg, and there is therefore potential for competition among them. Any gene that found phenotypic expression when in the haploid state in a sperm cell could be favoured over its alleles if it improved the competitive ability of the sperm. Such a gene would not necessarily be sex-linked: it could be found on any chromosome. If it was sex-linked it would have the effect of biasing the sex ratio, and would be an outlaw. If it was on an autosome it would still qualify as an outlaw for the general kind of reason already given for any segregation distorter: “...If there were genes affecting sperm-cell function there would be competition among sperm cells, and a gene that improved the ability to fertilize would Increase in the population. If such a gene happened to cause, say, malfunction of the liver, that would be just too bad; the gene would increase anyway, since selection for good health is much less effective than selection by competition among sperm cells” (Crow 1979). There is, of course, no particular reason why a sperm competition gene should happen to cause malfunction of the liver but, as already pointed out, most mutations are deleterious, so some undesirable side effect is pretty likely.
Why does Crow assert that selection for good health is much less effective than selection by competition among sperm cells? There must inevitably be a quantitative trade-off involving the magnitude of the effect on health. But, that aside, and even allowing for the controversial possibility that only a minority of sperms are viable (Cohen 1977), the argument appears to have force because the competition between sperm cells in an ejaculate would seem to be so fierce.
A million million spermatozoa,
All of them alive:
Out of their cataclysm but one poor Noah
Dare hope to survive. {142}
And of that billion minus one
Might have chanced to be
Shakespeare, another Newton, a new Donne —
But the One was Me.
Shame to have ousted your betters thus.
Taking ark while the others remained outside!
Better for all of us, froward Homunculus,
If you'd quietly died!
Aldous Huxley
One might imagine that a mutant gene that expressed itself when in the haploid genotype of a spermatozoon, causing increased competitive ability, say an improved swimming tail or the secretion of a spermicide to which the sperm itself was immune, would be immediately favoured by a selection pressure gigantic enough to outweigh all but the most catastrophic of deleterious side effects on the diploid body. But although it may be true that only one in hundreds of millions of sperms “dare hope to survive”, the calculation looks very different from the point of view of a single gene. If we forget linkage groups and brand-new mutations for a moment, however rare a gene may be in the gene-pool, if a given male has it in his diploid genotype, at least 50 per cent of his sperms must have it. If one sperm has received a gene giving it competitive advantage, 50 per cent of its rivals in the same ejaculate will have received the same gene. Only if the mutation has arisen de novo during the genesis of a single sperm will the selection pressure be astronomical in magnitude. Usually it will be a more modest selection pressure, not of millions to one but only two to one. If we take linkage effects into consideration the calculation is more complicated, and the selection pressure in favour of competitive sperms will increase somewhat.
In any case, it is a strong enough pressure for us to expect that, if genes expressed themselves when in the haploid genotype of the sperm, outlaws would be favoured, to the detriment of the rest of the genes in the diploid father's genome. It seems, to say the least, fortunate that sperm phenotypes are, as a matter of fact, usually not under the control of their own haploid genotypes (Beatty & Gluecksohe-Waelsch 1972). Of course sperm phenotypes must be under some genetic control, and natural selection has doubtless worked on the genes controlling sperm phenotypes to perfect sperm adaptation. But those genes seem to express themselves when in the diploid genotype of the father, not when in the haploid genotype of the sperm. When in the sperm they are passively carried.
The passivity of their genotypes may be an immediate consequence of the lack of cytoplasm in spermatozoa: a gene cannot achieve phenotypic {143} expression except via cytoplasm. This is a proximal explanation. But it is at least worth toying with reversing the proposition to obtain an ultimate functional explanation: sperms are made small, as an adaptation to prevent the phenotypic expression of the haploid genotype. On this hypothesis we are proposing an arms race between (haploid-expressed) genes for increased competitive ability among spermatozoa on the one hand, and genes expressing themselves when in the diploid genotype of the father on the other hand, causing sperms to become smaller and therefore unable to give phenotypic expression to their own haploid genotypes. This hypothesis does not explain why eggs are larger than sperms; it assumes the basic fact of anisogamy, and therefore does not aspire to be an alternative to theories of the origins of anisogamy (Parker 1978b; Maynard Smith 1978a; Alexander & Borgia 1979). Moreover not all sperms are small, as Sivinski (1980) reminds us in a most intriguing review. But the present explanation still deserves consideration as an ancillary to others. It is analogous to Hamilton's (1967) explanation for the inertness of Y chromosomes, to which I have already alluded.
Some of the outlaws I have been considering have been realistic, and are actually known to geneticists. Some of the suggested outlaws that I shall now come to are, frankly, pretty improbable. I make no apology for this. I see them as thought experiments. They play the same role in helping me to think straight about reality as imaginary trains travelling at nearly the speed of light do for physicists.
So, in this spirit of thought experiment, imagine a gene on a Y chromosome which makes its possessor kill his daughters and feed them to his sons. This is clearly a behavioural version of a driving Y-chromosome effect. If it arose it would tend to spread for the same reason, and it would be an outlaw in the same sense that its phenotypic effect would be detrimental to the rest of the male's genes. Modifiers, on any chromosome other than the Y chromosome, which tended to reduce the phenotypic effect of the daughter-killing gene would be favoured over their alleles. In a sense the outlaw gene is using the sex of the male's children as a convenient label for the presence or absence of itself: all sons are labelled as definite possessors of the gene, all daughters as definite non-possessors of it.
A similar argument can be made for X chromosomes. Hamilton (1972, p. 201) pointed out that in normal diploid species a gene on an X chromosome in the homogametic sex has three-quarters of a chance of being identical by descent with a gene in a sibling of the homogametic sex. Thus the “X-chromosome relatedness” of human sisters is as high as the overall relatedness of hymenopteran sisters, and higher than the overall relatedness of human sisters. Hamilton went so far as to wonder whether an X-chromosome effect {144} might account for the fact that helpers at the nest in birds seem, usually, to be elder brothers, rather than sisters, of the nestlings (the male sex is homogametic in birds). He noted that the X chromosome in birds accounts for some 10 per cent of the whole genome, and that it is therefore not too improbable that the genetic basis for brotherly care might lie on the X chromosome. If so, brotherly care might be favoured by the same kind of selection pressure as Hamilton had earlier suggested for sisterly care in Hymenoptera. Perhaps significantly, Syren and Luyckx (1977) point out that in some termites, the only non-haplodiploid group to have achieved full eusociality, “approximately half the genome is maintained as a linkage group with the sex chromosome” (Lacy 1980).
Wickler (1977), commenting on Whitney's (1976) rediscovery of Hamilton's X-chromosome idea, suggests that Y-chromosome effects are potentially even more powerful than X-chromosome effects, but Y chromosomes as a rule don't make up such a high proportion of the genome. In any case “sex-linked altruism” must be discriminating: individuals acting under the influence of their sex chromosomes should tend to show favouritism towards close relatives of the same sex rather than of the opposite sex. Genes for sex-blind sibling care would not be outlaws.
The value of the outlawed sex-chromosome thought experiment does not lie in its plausibility — which, like Hamilton, I do not rate highly — but in the fact that it focuses our attention on the importance of this discrimination. The sex of another individual is used as a label to identify it as a member of a class of individuals about whose genetics something is known. In the ordinary theory of kin selection, relatedness (or rather some proximate correlate of relatedness such as presence in one's own nest) is used as a label indicating higher than average probability of sharing a gene. From the point of view of a gene on a Y chromosome, the sex of a sibling is a label which signifies the difference between certainty of sharing the gene and certainty of not sharing it.
Note, by the way, the ineptness of notions of individual fitness, or even inclusive fitness as ordinarily understood, at dealing with situations like this. The normal calculation of inclusive fitness makes use of a coefficient of relationship which is some measure of the probability that a pair of relatives will share a particular gene, identical by descent. This is a good approximation provided the genes concerned have no better way of “recognizing” copies of themselves in other individuals. If a gene is on a sex chromosome and can use the sex of relatives as a label, however, its best “estimate” of the probability that a relative shares a copy of itself will be a better estimate than that provided by the coefficient of relationship. In its most general form, the principle of genes appearing to “recognize” copies of themselves in other individuals has been dubbed the “green-beard effect” (Dawkins 1976a, p. 96, following Hamilton 1964b, p. 25). Green-beard or {145} “recognition alleles” have been described in the literature as outlaws (Alexander & Borgia 1978; Alexander 1980), and they should therefore be discussed in this chapter, even though, as we shall see, their status as outlaws needs careful examination (Ridley & Grafen 1981).
The green-beard effect reduces the principle of “gene self-recognition” to its bare essentials in an unrealistically hypothetical, but nevertheless instructive, way. A gene is postulated which has two pleiotropic effects. One effect is to confer a conspicuous label, the “green beard”. The other effect is to confer a tendency to behave altruistically towards bearers of the label. Such a gene, if it ever arose, could easily be favoured by natural selection, although it would be vulnerable to a mutant arising which conferred the label without the altruism.
Genes are not conscious little devils, able to recognize copies of themselves in other individuals and to act accordingly. The only way for the green-beard effect to arise is by incidental pleiotropy. A mutation must arise which just happens to have two complementary effects: the label or “green beard”, and the tendency to behave altruistically towards labelled individuals. I have always thought such a fortuitous conjunction of pleiotropic effects too good to be true. Hamilton also noted the idea's inherent implausibility, but he went on “...exactly the same a priori objections might be made to the evolution of assortative mating which manifestly has evolved, probably many times independently and despite its obscure advantages” (Hamilton 1964b, p. 25). It is worth briefly examining this comparison with assortative mating, which for present purposes I shall take to mean the tendency of individuals to prefer to mate with individuals that genetically resemble them.
Why is it that the green-beard effect seems so much more far-fetched than assortative mating? It is not just that assortative mating is positively known to occur. I suggest another reason. This is that when we think of assortative mating we implicitly assume self-inspection as a means of facilitating the effect. If black individuals prefer to mate with black individuals, and white with white, we do not find this hard to believe because we tacitly assume that individuals perceive their own colour. Each individual, whatever his colour, is assumed to be obeying the same rule: inspect yourself (or members of your own family) and choose a mate of the same colour. This principle does not stretch our credulity by demanding that two specific effects — colour and behavioural preference — are controlled pleiotropically by the same gene. If there is a general advantage in mating with similar partners, natural selection will favour the self-inspection rule regardless of the exact nature of the recognition character used. It does not have to be skin colour. Any conspicuous and variable character would work with the identical behavioural rule. No far-fetched pleiotropism need be postulated.
Well then, will the same kind of mechanism work for the green-beard effect? Might animals obey a behavioural rule of the form: “Inspect yourself {146} and behave altruistically towards other individuals that resemble you”? The answer is that they might, but this would not be a true example of the green-beard effect. Instead, I call it the “armpit effect”. In the paradigm hypothetical example, the animal is supposed to smell its own armpits, and behave altruistically towards other individuals with a similar smell. (The reason for choosing an olfactory name is that police dog trials have shown that dogs presented with handkerchiefs held under human armpits can distinguish the sweats of any two individual humans except identical twins (Kalmus 1955). This suggests that there is an enormously variable richness of genetic labelling in molecules of sweat. In the light of the result with identical twins, one feels inclined to bet that police dogs could be trained to sniff out the coefficient of relationship between pairs of humans and that, for instance, they could be trained to track down a criminal if given a sniff of his brother. Be that as it may, “armpit effect” is here being used as a general name for any case of an animal inspecting himself, or a known close relative, and discriminating in favour of other individuals with a similar smell or with some other perceived similarity.)
The essential difference between the green-beard effect and the armpit self-inspection effect is as follows. The armpit self-inspection behavioural rule will lead to the detection of other individuals that are similar in some respect, perhaps in many respects, but it will not specifically lead to the detection of individuals that possess copies of the gene mediating the behavioural rule itself. The armpit rule might provide an admirable means of detecting true kin from non-kin, or of detecting whether a brother was a full brother or only a half brother. This could be very important, and it might provide the basis for selection in favour of self-inspecting behaviour, but the selection would be conventional, familiar kin selection. The self-inspection rule would be functioning simply as a kin-recognition device, analogous to a rule like: “Behave altruistically towards individuals that grew up in your own nest”.
The green-beard effect is quite different. Here the important thing is that a gene (or close linkage group) programs the recognition specifically of copies of itself. The green-beard effect is not a mechanism for the recognition of kin. Rather, kin recognition and “green-beard” recognition are alternative ways in which genes could behave as if discriminating in favour of copies of themselves.
To return to Hamilton's comparison with assortative mating, we can see that it does not really provide good grounds for optimism over the plausibility of the green-beard effect. Assortative mating is much more likely to involve self-inspection. If, for whatever reason, it is an advantage in general for like to mate with like, selection would favour an armpit type of behavioural rule: Inspect yourself, and choose a mate that resembles you. This will achieve the desired result — an optimal balance between outbreeding {147} and inbreeding (Bateson 1983) or whatever the advantage may be — regardless of the exact nature of the characteristics by which individuals differ.
Assortative mating is not the only analogy Hamilton might have chosen. Another one is the case of cryptic moths choosing to sit on a background that matches their own colour. Kettlewell (1955) gave dark-coloured carbonaria and light-coloured typical morphs of the peppered moth Biston betularia the opportunity to sit on dark- or light-coloured backgrounds. There was a statistically significant tendency for moths to choose a background matching their own body colour. This could be due to pleiotropism (or genes for colour being closely linked to genes for background choice). If this were so, as Sargent (1969a) believes, it might, by analogy, reduce our scepticism about the inherent plausibility of the green-beard effect. Kettlewell believed, however, that the moths achieved the matching by the simpler mechanism of “contrast conflict”. He suggested that a moth could see a small portion of its own body, and that it moved around until the observed contrast between its own body and the background reached a minimum. It is easy to believe that natural selection might favour the genetic basis for such a contrast-minimizing behavioural rule, because it would work automatically in conjunction with any colour, including a newly mutated colour. It is of course analogous to the “armpit self-inspection” effect, and is plausible for the same reason.
Sargent's (1969a) intuition differs from Kettlewell's. He doubts the self-inspection theory, and thinks that the two morphs of B. betularia differ genetically in their background preference. He has no evidence about B. betularia itself, but has done some ingenious experiments on other species. For example, he took members of a dark species and of a light species, and painted the circum-ocular hairs in an attempt to “fool” the moths into choosing a background that matched the painted hairs. They obstinately persisted in choosing a background that matched their genetically determined colour (Sargent 1968). Unfortunately, however, this interesting result was obtained with two different species, not with dark and light morphs of one species.
In another experiment which was done with a dimorphic species, Phigalia titea, Sargent (1969a) simply failed to get the result Kettlewell had with B. betularia. The P. titea individuals, whether the dark or the light morph, chose to sit on light backgrounds, presumably the appropriate background for the light ancestral form of the species. What is needed is for somebody to repeat Sargent's key experiment of painting the parts of a moth's body that it can see, but using a dimorphic species like B. betularia which is known to show morph-specific background choice. Kettlewell's theory would predict that moths painted black will select black backgrounds, and moths painted light will select light backgrounds, regardless of whether they are genetically {148} carbonaria or typica. A purely genetic theory would predict that carbonado, will choose dark backgrounds and typica light backgrounds, regardless of how they are painted.
If the latter theory turned out to be right, would this result give aid and comfort to the green-beard theory? A little, perhaps, since it would suggest that a morphological character and behavioural recognition of something resembling that morphological character can be, or can rapidly become, closely linked genetically. It must be remembered, here, that in the moth crypsis example there is no suggestion that we are dealing with an outlaw effect. If there are two genes, one controlling colour and the other background choice behaviour, both stand to gain from the presence of the appropriate other one, and neither of the two is in any sense an outlaw. If the two genes started out distantly linked, selection would favour closer and closer linkage. It is not clear that selection would similarly favour the close linkage of a “green-beard gene” and a gene for green-beard recognition. It seems that the association between the effects would have to be there by luck right from the start.
The green-beard effect is all about one selfish gene looking after copies of itself in other individuals, regardless of, indeed in spite of, the probability that those individuals will share genes in general. The green-beard gene “spots” copies of itself, and it thereby seems to work against the interests of the rest of the genome. It appears to be an outlaw in the sense that it makes individuals work and pay costs, for the benefit of other individuals who are not particularly likely to share genes other than the outlaw itself. This is the reason Alexander and Borgia (1978) call it an outlaw, and it is one of the reasons for their scepticism about the existence of green-beard genes.
But it is actually not obvious that green-beard genes would, if they ever arose, be outlaws. The following caution is urged by Ridley and Grafen (1981). Our definition of an outlaw makes reference to its provoking of modifiers at other loci which tend to suppress its phenotypic effects. At first sight it seems clear that green-beard genes would indeed provoke suppressing modifiers, because the modifiers would not tend to have copies in the bodies of the (unrelated) green-bearded individuals cared for. But we must not forget that a modifier, if it is to have any effect on the phenotypic expression of a green-beard gene, is likely itself to be in a green-bearded body, and is therefore in a position to benefit from receiving altruism from other green-bearded individuals. Moreover, since those other green-bearded donors are not particularly likely to be relatives, the costs of their altruism will not be felt by copies of the would-be modifiers. A case could therefore be made that would-be modifiers at other loci will gain rather than lose from sharing a body with a green-beard gene. To this it cannot be objected that the costs of paying out altruism to other green-bearded individuals may outweigh the benefits of receiving altruism from other green-bearded individuals: if this {149} were true there would be no question of the green-beard gene spreading in the first place. The essence of Ridley and Grafen's point is that if (which is unlikely) the green-beard gene has what it takes to spread through the population at all, the costs and benefits of the situation will be such as to favour modifiers that enhance rather than reduce the effect.
In evaluating this point, everything depends on the exact nature of what we are calling the green-beard phenotype. If the entire pleiotropic dual phenotype — green beard plus altruism towards green-bearded individuals — is regarded as a package deal, which modifiers can suppress or enhance only as a unit, then Ridley and Grafen are surely right that green-beard genes are not outlaws. But, of course, as they themselves stress, a modifier that could detach the two phenotypic effects from each other, suppressing the altruistic phenotype of the green-beard gene while not suppressing the green beard itself would certainly be favoured. A third possibility is the special case of a green-beard gene that caused parents to discriminate in favour of those of their children that happened to share the recognition character. Such a gene would be analogous to a meiotic-drive gene, and would be a true outlaw.
Whatever we may feel about Ridley and Grafen's point about the green-beard effect, it is clear that genes mediating altruism towards close kin, and favoured by conventional kin-selection pressure, are definitely not outlaws. All the genes in the genome have the same statistical odds of gaining from the kin altruism behaviour, for all have the same statistical odds of being possessed by the individual benefited. A “kin-selection gene” is, in a sense, working for itself alone, but it benefits the other genes in its genome as well. There will therefore not be selection in favour of modifiers that suppress it. Armpit self-inspection genes would be a special case of kin-recognition genes, and are likewise not outlaws.
I have been negative about the plausibility of the green-beard effect. The postulated favouritism based on sex chromosomes, which I have already mentioned, is a special case of the green-beard effect, and is perhaps the least implausible one. I discussed it in the context of within-family favouritism : elder siblings were supposed to discriminate among their younger siblings on the basis of probability of sharing sex chromosomes, sex itself being used as the label (“green beard”). This is not too wildly improbable because, if Y chromosomes do not cross over, instead of having to postulate a single pleiotropic green-beard “gene” we can postulate a whole “green-beard chromosome”. It is sufficient that the genetic basis for sexual favouritism should occur anywhere on the sex chromosome concerned. One might make a similar argument for any substantial portion of a chromosome which, say because of inversion, did not cross over. It is conceivable, therefore, that a true green-beard effect of some kind might one day be discovered.
I suspect that all present examples of what might appear to be green-beard effects are, in fact, versions of the armpit self-inspection effect. Thus Wu et al. {150} (1980) placed individual monkeys, Macaco, nemestrina, in a choice apparatus where they had to choose to sit next to one of two offered companions. In each case one of the two companions offered was a half sibling, related through the father but not the mother; the other was an unrelated control. The result was a statistically significant tendency for individuals to choose to sit next to half siblings rather than unrelated controls. Note that the half siblings concerned were not related in the maternal line: this means that there is no possibility of their recognizing an odour acquired from the mother, say. Whatever the monkeys are recognizing, it comes from the shared father, and this suggests, in some sense, recognition of shared genes. My bet is that the monkeys recognize resemblances of relatives to perceived features of themselves. Wu et al. are of the same opinion.
Greenberg (1979) Studied the primitively social sweat bee Lasioglossum zephyrum. (Seger, 1980, refers to this work under the picturesque heading, “Do bees have green setae?”.) Where Wu et al. used choice of sitting partner as a behavioural assay, Greenberg used the decision of a guard worker whether to admit or exclude another worker seeking entrance to the nest. He plotted the probability of a worker's being admitted against her coefficient of relationship with the sentinel. Not only was there an excellent positive correlation: the slope of the line was almost exactly one, so the probability of a sentinel admitting a stranger was approximately equal to the coefficient of relationship! Greenberg's evidence convinced him that “The genetic component is therefore in odor production and apparently not in the perceptual system” (p. 1096). In my terminology, Greenberg's words amount to the statement that he is dealing with the armpit effect, not with the green-beard effect. Of course the bees may, as Greenberg believes, have inspected relatives with whom they were already familiar rather than their own “armpits” (Holldobler & Michener 1980). It is still essentially an example of the armpit rather than the green-beard effect, in which case there is no question of the genes responsible being outlaws. A particularly elegant study which comes to a similar conclusion is that of Linsenmair (1972) on the family-specific chemical “badge” of the social desert woodlouse Hemilepistus reaumuri. Similarly, Bateson (1983) provides intriguing evidence that Japanese quail discriminate their first cousins from their siblings and from more distant relatives, using learned visual cues.
Waldman and Adler (1979) have investigated whether tadpoles preferentially associate with siblings. Colour-marked tadpoles taken from two clutches were allowed to swim freely around a tank; then a grid was lowered into, the tank, trapping each tadpole in one of sixteen compartments. There was a statistically significant tendency for tadpoles to end up closer to siblings than to non-siblings. Perhaps unfortunately, the experimental design does not rule out a possible confounding effect of genetically determined “habitat selection”. If there was a genetically determined tendency to, say, {151} hug the edges of the tank rather than the middle, genetic relatives might as a consequence be expected to end up in the same parts of the tank. The experiment does not, therefore, unequivocally demonstrate recognition of relatives or a preference for associating with relatives as such, but for many theoretical purposes this does not matter. The authors introduced their paper with reference to Fisher's (1930a) kin selection theory of the evolution of aposematism, and for purposes of that theory relatives simply have to end up together. It doesn't matter whether they are together because of shared habitat preference or because of true relative recognition. For our present purpose, however, it is worth noting that, if further experiments confirm the rule of “incidental habitat selection” for the tadpoles, this would rule out the “armpit” theory but would not rule out the green-beard theory.
Sherman (1979) invokes the idea of genetic favouritism in an ingenious theory about chromosome numbers in social insects. He presents evidence that eusocial insects tend to have higher chromosome numbers than their nearest non-social phylogenetic relatives. Seger (1980) independently discovered the effect, and has his own theory to account for it. The evidence for the effect is somewhat equivocal and would perhaps benefit from a demanding analysis using the statistical methods developed by modern students of the comparative method (e.g. Harvey & Mace 1982). But what I am concerned with here is not the truth of the effect itself but Sherman's theory to account for it. He correctly notes that high chromosome numbers tend to reduce the variance in fraction of shared genes among siblings. To take an extreme case, if a species has only one chromosome pair with no crossing-over, any pair of full siblings will share (identical by descent) either all, none, or half of their genes with a mean of 50 per cent. If there are hundreds of chromosomes, on the other hand, the number of genes shared (identical by descent) among siblings will be narrowly distributed about the same mean of 50 per cent. Crossing-over complicates the issue, but it remains true that a high chromosome number in a species tends to go with low genetic variance among siblings in that species.
It follows that if social, insect workers wished to discriminate in favour of those of their siblings with whom they happened to share the most genes, it would be easier for them to do so if the species's chromosome number were low than if it were high. Such preferential discrimination by workers would be detrimental to the fitness of the queen, who would “prefer” a more even-handed treatment of her offspring. Sherman therefore suggests, in effect, that the high chromosome count in eusocial insects is an adaptation to cause “offsprings' reproductive interests to more closely coincide with those of their mother”. We should not forget, by the way, that workers will not be unanimous. Each worker might show favouritism towards younger siblings that resemble her, but other workers will tend to resist her favouritism, for the same reason as the queen will resist it. The workers cannot be treated as {152} a monolithic party in opposition to the queen, in the same way as Trivers and Hare (1976) were able to treat them in their theory of conflict over the sex ratio.
Sherman very fairly lists three weaknesses of his hypothesis, but there are two more serious problems with it, Firstly, unless we are careful to qualify it further, the hypothesis appears to come dangerously close to a fallacy which I have dubbed Misunderstanding Number 11” (Dawkins 1979a) or the “Ace of Spades Fallacy” (Chapter 10). Sherman assumes that the extent of cooperation among coespecifics is related to the “mean fraction of alleles they share” (my emphasis) whereas he should think in terms of the probability that a gene “for” cooperation is shared (see also Partridge & Nunney 1977). On the latter assumption his hypothesis, as at present expressed, would not work (Seger 1980). Sherman could rescue his hypothesis from this particular criticism by invoking the “armpit self-recognition” effect. I shall not spell the argument out in detail, because I suspect that Sherman accepts it. (The essential point is that the armpit effect can use weak linkage within a family, while the green-beard effect requires pleiotropy or linkage disequilibrium. If workers inspect themselves, and show favouritism to those of their reproductive siblings that share features that they perceive themselves to possess, ordinary linkage effects become all-important and Sherman's hypothesis could escape the “Ace of Spades Fallacy”. It would also, incidentally, escape the first of Sherman's own objections, namely that the hypothesis “depends on the existence of alleles that enable their bearers to recognize their alleles”. “Such recognition alleles have never been discovered...”, and, by implication, are pretty implausible. Sherman could make things easier for himself by tying his hypothesis to the armpit rather than the green-beard effect.)
My second difficulty with Sherman's hypothesis was brought to my attention by J. Maynard Smith (personal communication). Taking the “armpit” version of the theory, it is indeed conceivable that workers might be selected to inspect themselves, and show favouritism towards those of their reproductive siblings that share their own individual characteristics. It is also true that queens would then be selected to suppress this favouritism if they could, for instance by pheromonal manipulation. But, in order to be selected, any such move on the part of a queen would have to have an effect as soon as it arose as a mutation. Would this be true of a mutation that increased a queen's chromosome number? No it would not. An increase in chromosome number would change the selection pressures bearing on worker favouritism, and many generations later it might produce an evolutionary change which would be to the advantage of queens in general. But this would not help the original mutant queen, whose workers would follow their own genetic, programming and would be oblivious to changes in selection pressures. Changes in selection pressures exert their effects over the longer time-scale of {153} generations. A queen cannot be expected to initiate a program of artificial selection for the long-term benefit of future queens! The hypothesis may be rescued from this objection by suggesting that high chromosome numbers are not an adaptation to facilitate queen manipulation of workers, but rather are a preadaptation. Those groups that happened, for other reasons, to have high chromosome numbers, were the most likely to evolve eusociality. Sherman mentions this version of his hypothesis, but sees no reason to favour it over the more positive, maternal-manipulation version. In conclusion, Sherman's hypothesis can be made theoretically sound, if it is phrased in terms of preadaptation rather than adaptation, and in terms of the armpit effect rather than the green-beard effect.
The green-beard effect may be implausible, but it is instructive. The student of kin selection who first understands the hypothetical green-beard effect, and then approaches kin selection theory in terms of its similarities to and differences from “green-beard theory”, is unlikely to fall prey to the many tempting opportunities for error that kin selection theory offers (Dawkins 1979a). Mastery of the green-beard model will convince him that altruism towards kin is not an end in itself, something that animals are mysteriously expected to practise in accordance with some clever mathematics that field workers don't understand. Rather, kinship provides just one way in which genes can behave as if they recognized and favoured copies of themselves in other individuals. Hamilton himself is emphatic on this point: “...kinship should be considered just one way of getting positive regression of genotype in the recipient, and ... it is this positive regression that is vitally necessary for altruism. Thus the inclusive fitness concept is more general than "kin selection"” (Hamilton 1975a, p. 140–141).
Hamilton is here using what he earlier described as “the extended meaning of inclusive fitness” (Hamilton 1964b, p. 25). The conventional meaning of inclusive fitness, the meaning upon which Hamilton himself based his detailed mathematics, is incapable of handling the green-beard effect and, indeed, outlaws such as meiotic-drive genes. This is because it is firmly tied to the idea of the individual organism as “vehicle” or “maximizing entity”. Outlaw genes demand to be treated as selfish maximizing entities in their own right, and they constitute a strong weapon in the case against the “selfish organism” paradigm. Nowhere is this better exemplified than in Hamilton's own ingenious extensions of Fisher's sex ratio theory (Hamilton 1967).
The green-beard thought experiment is instructive in further ways. Anyone who thinks about genes literally as molecular entities is in danger of being misled by passages like “What is the selfish gene? It is not just one single physical bit of DNA... It is all replicas of a particular bit of DNA, distributed throughout the world ... a distributed agency, existing in many different individuals at once ... a gene might be able to assist replicas of itself which are sitting in other bodies”. The whole of kie selection theory rests on this general {154} premise, yet it would be mystical and wrong to think that genes assist copies of each other because those copies are identical molecules to themselves. A green-beard thought experiment helps to explain this. Chimpanzees and gorillas are so similar that a gene in one species might be physically identical in its molecular details to a gene in the other. Is this molecular identity a sufficient reason to expect selection to favour genes in one species “recognizing” copies of themselves in the other species, and extending them a helping hand? The answer is no, although a naive application of “selfish gene” reasoning at the molecular level might lead us to think otherwise.
Natural selection at the gene lev