edwards - genetical theory of natural selection

7/30/2019 Edwards - Genetical Theory of Natural Selection

1/8

Copyright 2000 by the Genetics Society of America

Perspectives

Anecdotal, Historical and Critical Commentaries on Genetics

Edited by James F. Crow and William F. Dove

TheGenetical Theoryof Natural Selection

A. W. F. Edwards

Gonvi ll e and Caius Coll ege, Uni versity of Cambri dge, Cambri dge CB2 1T A, Uni ted Ki ngdom

So melancholy a neglect o f D arwins work suggests reflections upon the use of those rare and precious

possessions of mangreat books. It was, we believe, the custom of the late Professor Freeman to warn

his students that mastery of on e great book was worth a ny amoun t of knowledge of many lesser on es. The

tendency of modern scientific teaching is to n eglect the great bo oks, to lay far too much stress upon

relatively unimportant modern work, and to present masses of detail of doubtful truth and questionable

weight in such a way as to obscure principles. . . . How many biological students of today have read TheOrigin? The majority know it only from extracts, a singularly ineffective means, for a work of genius doesno t easily lend itself to the scissors; its unity is too m arked. Not hing can really take the p lace of a first-

hand study of the work itself.

R. A. Fish er an d C. S. St ock (1915)

[Fisher was 25 years old and his student friend Stock 27 in 1915. Professor E. A. Freeman was Regius

Pro fessor o f Moder n H istory at O xford , 1884 1892.]

marked a turn ing point in the development of evolution-

O N May 14, 1929, R. A. Fisher wrote t o O xford U ni- ary thought, contributing fundamentally to the renais-versity Press, I should call the book somethingsance of Darwinism following a long period of neglect.

like The Geneti cal Theory of N atu ral Selecti on. Within theIt is a work remarkable for what i t reveals of Fishers

year Fisher s (1930) book had been published, and in creative genius and his insight into many of the pro blems

May 2000 we can celebrate its seventieth birthday. Last facing evolutionary biologists today. Benn et t (1983)year Oxford republished in facsimile the 1930 edition

as a Variorum Edition, edited with a foreword, notes,In this Perspecti vesseries, Cr ow (1990a) summarized

and appendices by Henry Bennett . Once again, thethe main themes of The Geneti cal T heory(arguably the

book, which J. B. S. Hal dane (1932) described a s bril-deepest and most influential book on evolution since

liant and Sewall Wr igh t (1930) as certain to takeDarwin), and in another celebration of the Fisher cen-

rank as one of the major contributions to the theory oftenary (Cr ow 1990b), he gave this succinct account:

evolution, is in print.

In the present panegyric I assume some familiarity [In addition to enunciating the fundamental theoremof natural selection, Fisher] developed a totally novelwith The Geneti cal Theory. The indefatigable scholarshipway (now standard) for determining the probability ofof Professor Bennett has left us with two detailed de-survival of a mutant gene. He worked out the partial

scriptions of the book, his Variorum Edition foreworddifferential eq uation for gene-frequency change, using

and an earlier essay in his volume of Fishers biological a trigonometric transformation that made the variance

correspond ence ( Benn et t 1971 74, 1983), b ut , like The independent of the allele frequencies. He generalizedHaldanes formula for the probability of survival of aOrigin of Species, The Geneti cal Theory cannot be appreci-mutant gene, making it possible to treat deleterious asated second hand. The book is not long; the main textwell as favorable mutan ts. The form ula, further impro ved

of the 1930 edition has 265 pages of 13-point lines.by Malecot an d Kimura, plays a crucial role in the a nalyti-

(Fairly large print is a real antidote to stiff reading cal treatment of the neutral theory of molecular evolu-wrote Fisher to his Oxford editorall quotations from tion. He worked out the stationary distribution of allele

freq uencies. H e gave us the first qua ntitative theories ofFishers correspond ence are from Benn et t 1983.) Ben-sexual selection, mimicry, polymorphism, evolution ofnett opens his foreword with the following paragraph:recombination rates, and supergenes. H e explained whythe sex ratio is nearly 1:1 even in highly polygamousThe genetical theory of n atu ral selecti on is celebrated as the

first major work to provide a synthesis of Darwinian selec- speciesa problem that baffled Darwin and that providesone of the best illustrations that natural selection doestion and Mendelian genetics. Its publication in 1930

Genetics 154: 1419 1426 (April 2000)


2/8

1420 A. W. F. Ed ward s

Laplace (1749 1827) an d G auss (1777 1855). Hal d

(1998), the foremost historian of mathematical statis-

tics, writes, Fisher was a genius who almo st single-ha nd -

edly created the foundations for modern statistical sci-

ence. . . .

Like Laplace an d G auss, Fisher a lso ma de out standin g

con tribution s to science outside the fields of proba bility

and statistics. While Laplace a nd Ga uss contributed to

the physical sciences such as astronomy a nd geodesy(not to mention pure mathematics itself), Fisher used

his mathematical gifts to till the fertile ground of biol-

ogy. C. G. Dar win (1930)Sir Charles Darwin, physi-

cist, grandson of Charles Darwinopened his review of

The Geneti cal Theoryplaintively with the words, When aman reveals himself as a master of two very different

subjects it is hard to find a critic of his work. How

did Fisher, th e eq ual of Laplace and Ga uss in statistics,

become the greatest of Darwins successors ( Dawkins

1986) as well?

Joan Box in her superb biography of her father (Box

1978) tells a story about Fisher s biology teacher a t H ar-

row School, Arthur Vassall, and the influence he had

on the boy. Vassall once told E. B. Ford, in answer to

his invitation to name the 10 or 12 cleverest boys he

had taught, that it would be d ifficult to do so, but in

terms of sheer brilliance h e could divide them into two

groups, Fisher in one and all the rest in the other.

Writing to Vassall in 1929, as The Geneti cal Theory ap-proached completion, Fisher said, The fact is that

nearly all my statistical work is based on biological mate-

rial and much of it has been underta ken merely to clearFigur e 1.R. A. Fisher, about 1932.up d ifficulties in exper imenta l techn ique.

Fishers early interest in natural history is reflected

not always do what is best for the species. In doing this, in the books chosen fo r schoo l prizes, leading up to, inFisher introduced the concept of parental expenditure, his last year, the complete works of Charles Darwin inthereby seeding a cloudburst of ecological literature. 13 volumes. He went up to Cambridge in possessionRarely have so many new and deep ideas been put into

of volumes he was to read and re-read with loving carea single book.throughout his life (Box 1978).Cr ow (1990b, p. 270)

There was some un certainty as to whether FisherI discuss the problem of the sex ratio below, as well as should try for a C ambridge scholarship in ma thematicsmention further themes in The Geneti cal Theory, such as or in biology, but in the end math ematics won. Whenkin selection. he arrived at Cambridge with a scholarship in 1909, the

Theauthor: Ronald Aylmer Fisher was born in Lon - Un iversity had just celebrated the D arwin centena ry anddon, England, in 1890, and died in Adelaide, Australia, the fiftieth a nn iversary of th e publication o f On theOriginin 1962. In 1928 1929, when h e was writing TheGenetical of Species(Dar win 1859). In connection with the cente-Theory, Fisher was head of the statistics department at nary, an anonymous benefactor had endowed a Profes-

Rothamsted Experimental Station, already taking over sorship of Biology at Cambridge to be devoted to thatfrom Karl Pearson th e man tle of Britains leading statis- branch of Biology now entitled Genetics (Heredity andtician. His Stat isti cal M ethods for Research Workers, which Variation), and William Bateson, who had coined thewas to revolutionize the application of statistics, had very word genetics only 3 years previously, had beenappeared in 1925 and was now in its second edition elected to it in 1908. Bat eson s (1909) book M endel s(Fisher 1925). By mid -centur y, when his collection o f Principl es of H eredity, containing an English translationpapers Contri but ions to M athematical Stati sti cs (Fisher of Mendels paper, and Francis Darwins edition of his1950) was published, it was clear that Fisher was the fathers unpublished essays of 1842 and 1844 (Dar win

leading world figure in statistics, and as the century 1909) were being printe d by Cam bridg e Un iversity Press

turns it is now generally accepted by historians of the at one end of King s Parad e just as the young math emati-

cian-biologist entered Go nville an d C aius College at th esubject that his contributions place him alongside


3/8

1421Perspectives

We usually think of stochastic diffusion theo ry as origi-other end. Fisher bought a copy of Bateson ( Fishernating in Kolmogorovs great paper of 1932. But hear1952), later remarking, The new school of geneticistswhat Feller said in a Princeton lecture: If Kolmogorov

using Mendels laws of inheritance was full of activity had never written, the whole theory would have developedand confidence, a nd the shops were full of boo ks good in much the same way starting from Fishers book The

Genetical Theory of Natural Selection, published in 1930.and ba d from which one could see how completely man ySome years ago at a small mathematical seminar in thewriters of th ism ovement believed tha t Darwins position

Bl ac k Forest , d e voted to the analyt ic al aspec ts ofhad been discredited ( Fisher 1947).branching processes, . . . Kolmogorov referred to das

Fisher th e mathema tician d eveloped and retained the wundervolle Buch von R. A. Fisher, an d I heard two

highest opinion of biologists work. In the Preface to American statisticians sitting near me whisper to one an-other: Well, it cant be the R. A. Fisher weknow. AndThe Geneti cal T heoryhe remarks:they were right: they had never before encountered R. A.

The types of mind which result from training in mathe- Fisher the gen eticist, much less R. A. Fisher the m ath ema-matics and in biology certainly differ profound ly; but the tician.difference does not seem to lie in the intellectual faculty. So in unveiling this plaque, a tribute to his memory,It would certainly be a mistake to say that the m anipula- let us for once recall the full sweep of his genius.tion of mathematical symbolsreq uiresmore intellect thanoriginal thought in biology; on the contrary, i t seems Leonard Darwin: In his second undergraduate year,much more comparable to the man ipulation of the micro- Fisher was instrumental in setting up the Cambridgescope an d its appu rten an ces of stains and fixatives; whilst

Un iversity Eugenics Society with the support of seniororiginal thought in both spheres represents very similarmembers of the Un iversity such as R. C. Punn ett (soonactivities of an identical faculty.

Fisher (1930, p. viii) to be the first Professor of Genetics), John Maynard

Keynes, and Horace Darwin (the youngest of Charless

Nevertheless, mathema tics was Fishers undergra du- surviving children). Galton, in his Presidential Addressate subject and he distinguished himself at it in the hard to the Eugenics Education Society of London in 1909,school of the Cambridge tripos, ob taining a first in had remarked, It cannot be too emphatically repeatedboth parts. Awarded a studentship in physics, he was that a grea t dea l of careful statistical work has yet to beable to return for a graduate fourth year a nd study, at accomplished before th e science of eugenics can makethe Cavendish Laboratory, the theory of errors under large advances (quoted in For r est 1974). AlthoughF. J. M. Stratton an d statistical mechanics and qua ntum this might have been Fishers motto for the work ahea d,theory under James Jeans. It is hard to imagine a more what cannot be seriously sustained, as it has been bycomplete biological and mathematical preparation for some historians of science, is that Fishers youthful en-the young man who was simultaneously to revolutionize thusiasm for eugenics was the engine for his advancesstatistics an d bring mathe matical and statistical thin king in statistical and evolutionary gen etics (see Bennet tinto evolutionary theo ry. Just as James Clerk Maxwell

1983). Nevertheless, his involvement was of vital influ-had conjured a theory of gases out of the motions of

ence in another way, for i t led him to meet Galtonsindividual mo lecules so Fisher, trained in the same successor as President of the London Society, Majorschool o f science, had conjured a theory of evolution

Leonard Darwin, Horaces immediately elder brother,out of the survival of individua l gen es (Edwar ds 1988).

surely the kindest and wisest man I ever knew (FisherPerhaps Fisher even dared to think that he bore to

in 1943; quoted by Benn et t 1983).Darwin the same relationship as Maxwell did to Faraday

Leonard Darwin was to the young Fisher what J. S.(see Fisher 1932).

Henslow, Professor of Botany at Cambridge, h ad beenBefore leaving the question of mathematics, we

to the young Cha rles Darwin. In 1836 Cha rles wrote onshould pause to record (with the authors permission)

his return from the Beaglevoyage, My dear Hen slow, ID. G. Kendalls remarks when he unveiled a plaque on

do long to see you; you have been the kindest friendthe building formerly occupied by Fishers Ca mbridge

to me that ever man possessed. In his AutobiographyDepartment of Genetics, on March 22, 1990 (see also

(Dar win 1893) Charles recorded his debt: I have notKend al l 1990):

yet mentioned a circumstance which influenced my

We have come here tod ay to pay hon our to Fisher th e whole career more than any other. This was my friend-great geneticist, rather than to Fisher the greatest of all ship with Professor Henslow. When Henslow died,statisticians. But I should also like to say a word about Cha rlescon cluded hiseulogy with a sentence that FisherFisher the great mathematiciana figure some of you

could so well have written of Leonard: Reflecting overmay not have met.his character with gratitude and reverence, his moralHe has many claims to th at third title. For example he

was one of the few who anticipated to some extent the attributes rise, as they should do in the highest charac-creation of th e Theory of Ga mes, which swept econ omists ter, in pre-eminence over his intellect (q uoted by Gar -off their feet an d which ha s more recently begun to play diner 1999).a role in biology.

Leonard was a powerful influence on Fisher in manyBut the chief claim rests on h iscon tributions to stochas-ways, an d th e direct link he provided with C harles Dar-tic diffusion theor y. It will suffice to present th is through

two a necdotes. win is of profound significance. The dedication of The


4/8


Genetical Theoryis TO MAJOR LEONARD DARWIN in description of the activities of the projectors in specula-tive learning in the Gran d Academy of Lagado in La-gratitude for th e encouragement, given to the author,

puta, a satire o n the Royal Society of Lond on.durin g th e last fifteen years, by d iscussing man y of th e

Against this background hostile to Darwinism, Fisherproblems dealt with in this book. From 1915 to the

set out to show that evolution by natural selection op-publication of The Geneti cal Theoryand beyond, the twoerating on Mendelian characters was a viable theory notmen were in regular contact, much of i t fortunatelyonly capable of explaining the observations that hadpreserved in their co rrespond ence, whe re we find Leo-led Darwin to his revolutionary views, but also rich innard urging Fisher, as early as 1922, to write his great

hidden consequences that o ffered explanations for evo-book: I hope to stir you up to write a great work onlutionary panoramas as yet unimagined. By applying thethe m athema tics of evolution. I n 1928 he wrote, I amtechniques of advanced, often stochastic, mathematicsglad you are at work on your evolution book. . . . Howto Mendels genetical theory, Fisher was able to give anabout your new statistical work? I hope they can go onaccount of the power of natural selection.simultaneously. Donthurry evolution, but dogo on with

Theimprobabilitygenerator: I once heard Fisher re-it. And simultaneo usly they did go on , with Fisher work-mark, Natural selection is a m echanism for genera tinging at Rothamsted during the day and dictating Thean exceedingly high degree of improbability. It wasGenetical Theory to his wife in the evenings.one of his favorite apho risms, first reported by JulianTheEclipseof Darwinism: It is difficult fo r u s, as th eHuxl ey in 1936 and often repeated in Huxleys workcentury turns, to imagine the scepticism that sur-(e.g., 1942, 1954) un til it fina lly pa ssed in to t he la ng ua gerounded Darwins theory of evolution by natural selec-unattributed throug h the writings of C. H. Wadd ington,tion 70 years ago and astonishing for us to recall thatGavin de Beer, Ernst Mayr, and Richard Dawkins. Al-Mendelism itself was regarded in some quarters as anti-though it does not occur in TheGeneti cal Theory, Fisher sthetical to it. In TheEclipseof Darwinism, Bow l er (1992)rebuttal of the objection which has been made, thatdescribes in detail th e anti-Darwinian arguments 100the p rinciple of Natura l Selection depend s on a succes-years agoLamarckism, orthogenesis, and the muta-sion of favourable chances embodies the thought intion theoryand Bennet t (1983) opens h is Introdu c-cha racteristic style:tion with a description of the doubts still prevailing in

1920 1930. (See also Mayr 1988; Ol by 1981; Tur nerThe objection is more in the nature of an innuendo

1985; and Fisher 1932, 1947, 1954.) than of a criticism, for it depends for its force upon theThe old criticism of Darwins theory, that chance ambiguity of the word chance, in its popular uses. The

income derived from a Casino by its proprietor may, inevents th emselves could never produce the marvelouson e sense, be said to depen d upo n a succession of favour-adaptations that nature exhibits, also refused to lieable chan ces, although the phrase contains a suggestion

down. It had originally appeared as an immediate re- of improbability more appropriate to the hopes of thesponse to The Ori gin of Specieswhen John Herschel had patrons of his establishment. It is easy without any very

profound logical analysis to perceive the difference be-commented that Darwins theory was the law of hig-tween a succession of fa voura ble deviation s from th e lawsgledy-piggledy. What exactly this means I do notof chance, and on the other hand, the continuous and

know, snorted Darwin, but it is evidentlyverycon temp-cumulative action of these laws. It is on the latter that

tuous. Herschel added a footn ote in the second ed ition the principle of Natural Selection relies.of his Physical Geography in 1861 in which he said: Fisher (1930, p. 37)

In 1954 Fisher contributed to Evolution as a Process,We can no more accept the principle of arbitrary andcasual variation an d n atura l selection a s a sufficient ac- edited by Julian Huxley, A. C. Hardy, and E. B. Ford,count, per se, of the past and present organic world, than an essay, Retrospect of the Criticisms of the Theorywe can receive the Laputan method of composing books of Natural Selection (Fisher 1954). In the words of(pushed a loutrance) a s a sufficient o ne o f Sha kespeare

Huxleys introductory essay, Fisher very appropriatelyand the Principia. Equally in either case, an intelligence,

deals with the main criticisms which have been levelledguided by a purpose, must be continually in action toagainst the general theory of natural selection, and thebias the directions of the steps of changeto regulate

their amountto limit their divergenceand to con- main difficulties which its opponen ts ha ve raised. Hetinue t hem in a definite co urse. points out that all of these h ave been in principle an-

Her sch el (1861, p. 12) swered or resolved by recent developments of mende-

lian or particulate genetics. It is uncertain when FisherLord Kelvin, then Sir William Thomson, told the 1871 wrote this essayprobably in the 1930s. When he sentmeeting of the British Association in his Presidential it to Ford in 1951 for Evolu ti on as a Process, he wrote, IAddress that Herschels objection to the doctrine of wrote it a long while ago when the possibility of mynatural selection, that it was too l ike the Laputan bringing out a second edition of the TheGeneti cal Theorymethod of making books, seemed to him to be a most was in my mind. It amplifies the rebuttal of the succes-valuable and instructive criticism. The references are, sion of favourable chances argument in two ways. First,

Fisher enlarges on his apho rism:of course, to Swifts Gullivers Travels (1726) an d his


5/8

1423Perspectives

. . . it was Darwins chief contribution, not on ly to Biology ing is beyond my power to express ( Leigh 1986). Thebut to the whole of natural science, to have brought to style belongs to Charles Darwins age, not ours, and itlight a process by which contingencies, a priori improba-

is illuminating to think of The Geneti cal Theoryas a kindble, are given, in the process of time, an increasing proba -of math ematical-Mendelian append ix to The Origin of bility, un til it is their no n-occurrence rather than their

occurrence which becomes highly improbable. Species. Fisher was probably the best-read Darwinian ofFisher (1954, p. 91) his generation, and it is not surprising that it affected

his style.

Another feature of The Geneti cal Theorythat is charac-Second, he emphasizes the importance of not being

carried away by improbab ilities viewed a fter the event teristic of its time is its cavalier way with references. TheOrigin of Species, it will be remembered, contains non eanyway:at al l , on the grounds that i t was but an abstract of

Consideration of the conditions prevailing in bisexualthe much larger work that Darwin had planned: Thisorganisms shows that . . . the chance of an organismAbstract, which I now publish, must necessarily be im-leaving at least one offspring of his own sex has a calcula-

ble value of about 5/8. Let the reader imagine that this perfect. I cannot here give references and authoritiessimple condition were true of his own species, and at- for my several statements; and I must trust to the readertempt to calculate the prior probability that a hundred

reposing some confidence in my accuracy ( Dar wingenerations of his ancestry in the direct male line should

1859). Perhaps the most astonishing of Fishers omis-each have left at least one son. The odds against such asionsishisown paper ( Fisher 1918) O n th e correlationcontingency as it would have appeared to his hundredth

ancestor (about the time of King Solomon) would require between relatives on the supposition of Mendelian in-for t heir expression fo rty-four figures of th e decimal n ota- heritance. It is true that th e 1918 paper is not primarilytion; yet this improbable event has certainly happened.

concerned with evolution, but what modern writerFisher (1954, p. 91) could resist citing such a pa th-brea king p aper wh en h e

came to write his magnu m opusin th e same gen era l field?These con sidera tions, flowing from TheGenetical The-ory in the 1930s, also provide a fascinating background Another notable omission from Fishers list of refer-

ences has led generations o f evolutionary biologists toto Fishers contempora ry discussions of the found ations

of statistical inference, for the logical weakness of any assume that perhaps the most famous single argument

in The Geneti cal Theory, on Natural Selection and thestatistical test of significan ce is tha t it argu es on the b asisof the probability of an event after the event has been sex-ratio, was original to Fisher. Although Fisher was

not aware tha t the argument was in th e first edition ofobserved. Such fundamental questions cannot be pur-

sued here, but the development o f Fishers own concept The Descent of M an (Dar win 1871), he was certainlyaware of some seconda ry sources for it (Edwar ds 1998)of likelihood is of course part of the response to the

dilemma (for likelihood and its history, see Edwar ds and , in particular, a paper by Cobb (1914) in TheEugen-ics Revi ew. Cobb did n ot refer to Darwin an y more than1992). We may note, however, that the theory of evolu-

tion by natural selection has a high likelihood precisely Fisher did to Cobb, and one must suppose th at theargument was circulating freely at the time among thebecause the ratio of the prob ability of the n atural world,

as we observe it, on the hypothesis of natural selection small number of people then interested in such things.

Neglectof The Genetical Theory: Bennet t (1983) hasto its probability on the hypothesis of pure chance isso enormous. d iscussed the reviews of The Geneti cal Theory. Those in

a position to judge it, especially Haldan e a nd Wright,ReadingThe Genetical Theory: The auth or, the editor,the reviewers, Fishers correspondents, and all more who both wrote long reviews, thought it a masterpiece.

In spite of Wr igh t s (1930) review, the book was notrecent commentators agree that The Geneti cal T heoryisnot an easy book. Seventy years after its publication the as widely appreciated in the Un ited States as it was in

Great Britain, and this was to remain the case untilcustom is now for such a bookif such a book were to

be writtento rely to a far greater extent on aids to quite recently. Leonard Darwin warned Fisher that its

influence would be slow to be felt: . . . my impressioneasy assimilation. Diagrams, pictures, and boxed text

relieve the burden of solid reading, and numerous meta- is. . . that it will be slowly recognized as a very important

contribution to the subject. But I am afraid it will bepho rs reflect reason ed argum ent like the curved mirrorsat a carnival. Fishers book, by contrast, is all argument. slow, because so few will really grasp all that it mean s.

. . . And Fisher himself observed, on hearing fromEach sentence must be considered, from the very firstone (Natural Selection is not Evolution). Each re- Oxford UniversityP ress in 1931 of better-than-expected

sales, It was so long before I heard from them that Ireading of this classic brings something new ( Cr ow

1990a). Leonard Darwin did his best as he read each had q uite made up my mind that it was one of those

books which everybody praised and nobody read, andchapter in draft: One idea one sentence is, I think, a

good rule, he wrote to Fisher in March 1929. One would have no influence on biological opinion.

In the United States, Dobzhansky (1937) clearly ap-reader has written, Not least is my debt to the ghost

of Ronald Fisher himself: in this wordy age, the pleasure preciated the significance of The Geneti cal Theory, but inhis influential Genetics and the Origin of Species he pre-of working with such closely and acutely reasoned writ-


6/8


ferred the mo re metaphorical line of thought estab- versy, and only now can we understand Fishers thinking

(see my review, Edwar ds 1994, and for a parallel be-lished by Wr igh t (1932). Simpson s (1944) Tempo andModein Evolution was rather more appreciative of Fish- tween the history of the fundamental theorem and that

of fiducial prob ability, see Edwar ds 1995).ers contributions. Mayr (1942), by contrast, never

seems to have come to terms with its importance, either RevivalofThe Genetical Theory:By the time of Fishersin Systemati cs and the Origin of Speciesor when writing on death in 1962, The Geneti cal Theory had thus had only athe h istory of biology. In TheEvolutionarySynthesis(Mayr limited direct influence. Like Mendels work, its fame1988; first edition 1980), he cont ributed a P rologu e: now rests more on the influence it should have had,

Some thoughts on the history of the evolutionary synthe- rather than the influence it did have. It lay fallow forsis without once mentioning TheGenetical Theory(unless just as long as did Mendels paper. The publication ofa remark about the supposedly evolutionary writings the second edition in 1958, however, mad e it moreof . . . R. A. Fisher counts as a mention). In his massive widelya vailable, an d th e next gen eration o f evolutionaryThe Growth of Bi ological T hought, Mayr (1982) expr esses biologists was to discover, or in some cases duplicate,opinions about Fishers work that can be explained on ly its riches. I make the point by considering the reprintby an inadequate study of The Geneti cal T heory, and his of G. C. Wil l iamss (1996) influential book AdaptationOne L ong Argument: Charles Darwin and theGenesis of M od- and Nat ural Selecti on. Williams, together with W. D. Ham-ern Evolutionary Thought (Mayr 1991) contains nearly ilton, has been held chiefly responsible for the modern200 references, but nothing of Fishers at all (or, indeed, view that natural selection operating within populationsof Wrights). is the primary mechan ism of evolution. Williams cites

In B ritain the advocacy of E. B. For d (e.g., 1938) and TheGenetical Theoryin the first para grap h of his In trod uc-Julian Hux l ey (e.g., 1942) in their numerous writings tion, but it is his 1996 account of what influenced himensured that The Geneti cal Theory was not entirely for- in the writing of his book to which I draw particulargotten in the early years. Ford wrote to Fisher in 1955: attention: I t ha s b een a n a m a zem ent t o m e t ha t t he o r i gi na l

Two publications that I read that year [195455] wereedition d id no t sell out long bef ore th e War [ 1939] but ,also of great importance to me, and mayhave deterred me

after all, a book is to be judged not by its sale but by from some alternative career. One was Shaw and Mohlersits effect upon science, and no book of the century has brief article on sex ratio (1953). The other was David

Lacks chapter in Huxley, Hardy, and Ford (1954). Shawhad a greater effect upon biology than has this one,and Mohlers superb work went largely unnoticed untilthe ideas spreading out from it through, apparently, ait wasd iscovered by Charno v (1982), who made th e Shaw-limited number o f readers of the original, but tha t kindMohler equation a fo cal concept for his broad treatment

of th ing is what both you a nd I are accustomed to findof sex allocation. . . . Right from its opening paragraph,

(people like to be given little summaries). Fisher, al- Lacks The Evolution of Reproductive Rates was a sub-though he was Arthur Balfour Professor of Genetics at l ime encouragement. I had found a biologist who be-

lieved, as decisively as I did, that natural selection is aCambridge from 1943 until his retirement in 1957, ran real scientific th eory. It logically predicts tha t th ere a rea very small department, which had little impact at thecertain sorts of properties that organisms must have, and

time (Edwar ds 1990), and the leading writer on evolu-oth ers, such as ada ptation s for the benefit of the species

tionary genetics in Britain has been Maynard Smith, a (Fisher 1958: 4950), tha t t hey could not possibly have.student of Haldanes. Mayn ar d Smit h s (1993) The

Theory of Evolution (first ed ition 1958) end s with a list

I may have been afflicted by a common form of megalo-of books for further reading, which starts with TheOrigin

mania in assuming tha t I had the un iquely right perspec-of Speciesand includes books by Mayr, D obzhan sky, a nd tive and was far ahead of my time. Awareness of the workSimpson, as well as P. M. Sheppard, R. C. Lewontin, of Shaw and Mohler should have made me doubtful about

any such claim. I was really forced to abandon it by theC. H . Wad ding ton , and G . C. Williams, but not TheGenet-discovery of other publications that anticipated some ofical Theory or, curiously enough, Haldanes The Causeswhat I was writing, especially W. D. Hamilton s 1964wor ksof Evolution (1932).on inc lusive fitn ess.

The Geneti cal Theory has not been translated into a ny Wil l iams (1996 p. x)foreign language, and outside the English-speaking

My point in reproducing this extract is, of course,world perhaps only France has given it serious attentionthat the cited works of Shaw and Mohler, Lack, and(LHer it ier 1934; Roger 1981).Ha milton all relate to topics Fisher treated in TheGeneti-The main influential topics from the early life o f Thecal Theory. Spe cifica lly, Shaw and Moh l er (1953) startGeneti cal Theory were Natural Selection and the sex-from The Geneti cal T heory having, like Fisher, consultedratio (mention ed above), which came to be seen as theonly the second edition of Darwins The Descent of M an.archetypal game-theoretic argumen t (Mayna r d Smit h[I had myself asked Fisher about natural selection and1982), and the fundamental theorem of natural selec-

the sex ratio in 1958, and he had told me to read Thetion. The latter, apart from having been the originalGenetical Theory, which I did, putting his argument intoinspiration for Wrights a daptive surface approach ,

has caused a great deal of discussion and indeed contro- mathematics with th e help of W. F. Bodmer (Bodmer


7/8

1425Perspectives

its supplement Of Man), and also rate as undoubtedlyand Edwar ds 1960), including the effect of parentalone of the greatest books of the [twentieth] century theexpenditure, which Shaw and Mohler had omitted]. Asappearance of a variorum edition is a major event. . . .

we now know, the argument was not original to Fisher, By the time of my ultimate graduation, will I have un-but it was The Geneti cal T heorythat made it famous. derstood all that is true in this book and will I get a First?

I dou bt it. In some ways some of us have overtaken Fisher;Lacks ( 1954) influential pa per The evolution ofin many, however, this brilliant, daring man is still far inreproductive rates appeared alongside Fisher s (1954)front.Retrospect in Evolution as a Process. In the words of

W. D. Ha mil t onHuxley in his introduction, Lack advances the view (fro m th e dust-jacket of th e Variorum Genetical Theory)

that basic clutch size is adaptive, genetically adjusted byIn th e 40 years since Fisher told me to read TheGeneti-s el ec t i o n. . . . I c a n d o no b et t er t ha n q uo t e fr o m

cal Theory, I have discussed it with more colleagues thanBur br idg e (1992):I can possibly record. I am grateful to all of them. But

This [explanation] has a long and tortuous history. R. A.no ne of these discussions could ha ve borne fruit witho utFisher, apparently unknown to Lack, discussed the gen-the scholarship of Henry Bennett and the encourage-eral prob lem in his classic boo k in 1930, har dly an obscure

work. Fishers chara cteristically incisive co mmen ts on fe- ment of James Crow.cundity may have been overlooked because they occur inthe chapters dealing with h uman evolution and eugenics, Note added in proof: W. D. H amilton d ied in Oxford on March 7,from which many readers avert their eyes. 2000, from malaria co ntracted during a field trip in Africa.

Fisher himself attributed the solution to Major Leon ardDarwin, a younger son of Charles Darwin. Leonard Dar-win did indeed cover the matter clearly. He thought histreatment of the subject was novel, but he was wrong. LI TERATU RE C ITED

Ironically, his father had discussed the problem in the Bat eson , W., 1909 M endel s Prin cipl es of H eredit y. Cambridge Univer-first editio n o f h is Descent of M an and given essentially thesity Press, Cambridge, UK.same solution. B ut in the second edition the relevant

Bennet t , J. H. (Editor), 1971 74 Collected Papers of R. A. Fisher, Vol.chapter was heavily revised and the passage on fecundityI V. Un iversity of Adelaide, Australia.

omitted. As later biologists seldom consulted the scarceBennet t , J. H. (Editor), 1983 Nat ur al Selecti on, H eredity, and Eugeni cs.

first edition, Dar wins contribution seems to have been Clarendon Press, O xford.forgotten, even by his own son. Bodmer , W. F., an d A. W. F. Edwar ds, 1960 Natural selection and

As a final complication, Cha rles Darwin ga ve the credit the sex ratio. Ann. Hum. Genet. 24: 239 244.for the crucial insight to Herbert Spencer . . . Bowl er , P. J., 1992 The Ecli pse of Darwini sm. Johns Hopkins Univer-

sity P ress, Baltimore.Bur br idg e (1992)Box, J. F., 1978 R. A. Fisher: The Li fe of a Scienti st. Wiley, New York.Bur br idg e, D., 1992 Lacking a solution? Nature 355: 118.As to W. D. H amilton s 1964 works on inclusive fit-Cobb, J. A., 1914 The problem of the sex ratio. Eugen. Rev. 6:ness, Ha mil t on (1964) ha d pointed out, with a quo ta-

157 163.tion , tha t Fisher ha d invoked th e idea o f inclusive fitn ess Cr ow, J. F., 1990a R. A. Fisher, a centennial view. G enetics 124:

207 211.in The Geneti cal Theory, Mimicry (Chap. 8), to explainCr ow, J. F., 1990b Fishers contributions to genetics and evolution.th e evolution of d istastefuln ess in in sects. The idea first

Theor. Popul. Biol. 38: 263 275.surfaces in a lecture Fisher g ave to the C ambridge U ni- Dar win, C., 1859 On t heOr igin of Species byM eans of Natu ral Selection.versity Eugen ics Society in 1912 (Fisher 1914; Edwar ds Murray, London .

Dar win, C., 1871 The Descent of M an, and Selection i n Relati on t o Sex.1993). It was later taken up by H aldane, and , tha nks toMurray, London .

the publication of Fishers correspondence, we n owDar win, C., 1893 In CharlesDarwin: HisLifeTold in an Autobiographi-

have a succinct statement of i t ( in a letter to A. G. cal Chapter and in a Selected Series of his Published Letters, edited byF. Dar win. D. Appleton, New York.Lowndes, June 1945): This does no t preclude adapta -

Dar win, C., 1909 In TheFoundati ons of the Ori gin of Species: Two Essaystions which are effectual through the survival of rela- written in 1842 and 1844 by C. Darwin, edited by F. Dar win.tives, for these share to a greater or lesser extent the Cambridge University Press, Cambridge, UK.

Dar win, C. G., 1930 Natural selection. Eugen. Rev. 22: 127 130.germ plasm of the ind ividual. So th e paren tal instincts,Dawkins, R., 1986 The Blin d Watchmaker. Longman, London.

though altruistic, are accessible to improvementDobzhansky, T., 1937 Genetics and the Ori gin of Species. Columbia

through natural selection, and in my book I do discuss University Press, New York.Edwar ds, A. W. F., 1988 Computers and genealogies. Biol. Soc. 5:

how far we may think of the development of nauseous 7381.flavours in insect larvae. . . .Edwar ds, A. W. F., 1990 R. A. Fisher: Twice professor of genetics:

I opened this discourse on The Geneti cal Theory with London and Cambridge orA fairly well-known gen eticist. Bio -metrics 46: 897 904.a quotation from Fisher an d St oc k (1915) on the im-

Edwar ds, A. W. F., 1992 Likelihood, expanded edition. Johns Hop-portance of a great book. I close i t with a quotationkins U niversity P ress, Baltimore.

from Hamilton: Edwar ds, A. W. F., 1993 Mendel, Ga lton, Fisher. Aust. J. Stat. 35:129 140.

This is a book which, as a student, I weighed as ofEdwa r d s, A. W. F., 1994 The fundamen tal theorem of natura l selec-

equal importance to th e entire rest of m y undergraduate tion. Biol. Rev. 69: 443 474.Cambridge BA course and, through the time I spent on Edwar ds, A. W. F., 1995 Fiducial inference and the fund amentalit, I think it notched down my degree. . . . theorem of natural selection. Biometrics 51: 799 809.

For a book that I rate only second in importance in Edwa r d s, A. W. F., 1998 Natural selection and the sex ratio: Fisherssources. Am. Nat. 151: 564 569.evolution theory to D arwins Origin (this as joined with


8/8


Fish er , R. A., 1914 Some hopes of a eugenist. Eugen. Rev. 5: 309 Lack, D., 1954 The evolution of repro ductive rates, pp. 143156 in315. Evoluti on as a Process, edited by J. Hu xl ey, A. C. Ha r dy an d E. B.

Fisher ,R. A., 1918 The correlation between relatives on the supposi- For d. Allen &Unwin, London.tion of Mend elian inheritance. Trans. R. Soc. Edinb. 52:399 433. Leigh , E. G., Jr ., 1986 Ronald Fisher and the development of evolu-

Fisher , R. A., 1925 Stat isti cal M ethods for Research Workers. Oliver & tionary theory. I. The role of selection, pp. 187223 in OxfordBoyd, Edinburgh. Surveys in Evolutionary Biology, Vol. 3, edited by R. Dawkins an d

Fisher ,R. A., 1930 TheGenetical T heory of Natural Selecti on. Clarendon M. Ridl ey. Oxford University Press, London.Press, Oxford [Variorum edition, Bennet t , J. H. (Editor), 1999, LHe r it ie r , P. , 1934 Genetique et Evolution. Hermann, Paris.Oxford University Press, Oxford]. Mayna r d Smit h, J., 1982 Evolu ti on and the Theory of Games. Cam-

Fish er , R. A., 1932 The bearing of genetics on theories of evolution. bridge University Press, Cambridge, UK.Sci. Prog. 27: 273 287. Mayn ar d Smit h , J., 1993 TheT heory of Evoluti on. Cambridge Univer-

Fisher ,R. A., 1947 The renaissance of Darwinism. Listener 37:1001. sity Press, Cambridge, UK.Fish er , R. A., 1950 Contr ibuti onst o M athematical Stati sti cs. Wiley, New Mayr , E., 1942 Systemati csand the Ori gin of Species. Columbia Univer-

York. sity Pre ss, New York.Fish er , R. A., 1952 Statistical metho ds in genetics. Hered ity 6:112. Mayr , E., 1982 The Growth of Bi ological Thought. Belknap, Cambridge,Fish er , R. A., 1954 Retrospect of criticisms of the theory of natural MA.

selection, pp . 8498 in Evoluti on as a Process, edited by J. Hu xl ey, Mayr , E., 1988 Prologue: some thoughts on the history of the evolu-A. C. Har dy an d E. B. For d. Allen &Unwin, London. tionary synthesis, pp. 148 in TheEvolutionary Synthesis, edited by

Fisher , R. A., 1958 The Geneti cal T heory of Natu ral Selection, Ed. 2. E. Mayr an d W. B. Pr ovine. Harvard Un iversity Press, C am-Dover, New York. bridge, MA.

Fisher , R. A., an d C. S. St ock, 1915 Cuenot on preadaptation: aMayr , E., 1991 One Long Argument: Charles Darwin and the Genesis of

criticism. Eugen. Rev. 7: 4661. M odern Evoluti onary Thought. Harvard University Press, Cam-For d, E. B., 1938 The Study of H eredity. Thornton Butterworth,

bridge, MA.London.

Ol by, R., 1981 La theorie gen etique de la selection naturelle vueFor r est , D. W., 1974 Francis Galton: T heL ife and Work of a Victorian

par un historien. Rev. Synth. 102: 251 289.Genius. Elek, London .

Roger , J. (Editor), 1981 R. A. Fisher et l histoire de la ge n etiqueGar diner , B. G., 1999 Picture quiz. Linnean 15: 411.

des populations. Rev. Synth. Ser. III 103104: 229 431.Ha l d, A., 1998 A H istory of Mat hematical Statistics from 1750 t o 1930.

Shaw, R. F., an d J. D. Moh l er , 1953 The selective significan ce ofWiley, New York. the sex ratio. Am. Nat. 87: 337 342.Hal dane, J. B. S., 1932 The Causes of Evolu ti on. Longman Green,

Simpson , G. G., 1944 Tempo and M odein Evolut ion. Columbia Univer-London.sity Pre ss, New York.Hamil t on, W. D., 1964 The genetical evolution of social behaviour,

Tur ner , J. R. G., 1985 Fishers evolutionary faith and th e challengeII. J. Theor. Biol. 7: 1751.of m imicry, pp. 159196 in Oxford Surveys in Evolutionary Biology,Her sch el , J., 1861 Physical Geography.Vol. 2, edited by R. Dawkins an d M. Ridl ey. Oxford UniversityHuxl ey, J. S., 1936 Natural selection an d evolutionary progress. Br.Press, London.Assoc. Annu al Meetin g Repo rt, 81 100.

Wil l iams, G. C., 1996 Adaptation and Natural Selection. PrincetonHu xl ey, J. S., 1942 Evolu ti on: T he M odern Synthesis. Allen &Unwin,Un iversity P ress, Princeton, NJ.London.

Wr igh t , S., 1930 The Genetical Theory of Natural Selection: aHu xl ey, J. S., 1954 The evolutionary process, pp. 1 23 in Evolutionreview. J. Hered. 21: 340 356.as a Process, edited by J. Huxl ey, A. C. Har dy an d E. B. For d.

Wr igh t , S., 1932 The roles of mutation, inbreeding, crossbreedingAllen & Unwin, London.and selection in evolution. Proc. Sixth Int. Con gr. G enet. 1:Kendal l , D. G., 1990 Obituary: Andrei Nikolaevich Kolmogorov

(1903 1987). Bull. Lon d. Math. Soc. 22: 31 100. 356 366.

edwards - genetical theory of natural selection

Documents