DEAN KEITH SIMONTON

ORIGINS OF GENIUS

This page intentionally left blank

~ ORIGINS OF GENIUS ~

Darwinian Perspectives on Creativity

Dean Keith Simonton

New York Oxford

OXFORD UNIVERSITY PRESS

Oxford University Press

Oxford New York

Athens Auckland Bangkok Bogota Buenos Aires Calcutta

Cape Town Chennai Dar es Salaam Delhi Florence Hong Kong Istanbul Karachi Kuala Lumpur Madrid Melbourne Mexico City Mumbai Nairobi Paris Sao Paulo Singapore Taipei Tokyo Toronto Warsaw

and associated companies in

Berlin Ibadan

Copyright © 1999 by Oxford University Press

Published by Oxford University Press, Inc.

198 Madison Avenue, New York, New York rooi6 http://www.oup-usa.org

Oxford is a registered trademark of Oxford University Press

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, electronic, mechanical, photocopying, recording, or otherwise, without the prior permission of Oxford University Press.

Library of Congress Cataloging-in-Publication Data Simonton, Dean Keith.

Origins of genius:

Darwinian perspectives on creativity/

Dean Keith Simonton.

p. cm. Includes bibliographical references and index.

ISBN 0-19-512879-6

1. Genius. 2. Creative ability.

3. Darwin, Charles, 1809-1882. I. Title.

BF412.S58 1999 153.9’8 — dc2i 98-45044

987654321

Printed in the United States of America

on acid-free paper

To all Darwinists

This page intentionally left blank

Preface

IX

• 1. GENIUS AND DARWIN

The Surprising Connections

1

• 2. COGNITION

How Does the Brain Create?

25

• 3. VARIATION

Is Genius Brilliant—or Mad?

75

• 4. DEVELOPMENT

Are Geniuses Born—or Made?

109

• 5. PRODUCTS

By What Works Shall We Know Them?

145

• 6. GROUPS

The Right Place and the Right Time?

199

• 7. DARWINIAN GENIUS

The Future of an Idea

243

Notes

249

References

261

Index

297

This page intentionally left blank

For nearly a quarter of a century, I have been conducting scientific inquiries into the nature and origins of creative genius. The subjects of these investigations have included thousands of eminent figures from most of the world’s civilizations and key domains of creative activity: scientists and artists, philosophers and composers, poets and psychologists. During the course of my research, I have come to admire many outstanding exemplars of creativity. From European civilization, for instance, come such personal idols as Beethoven, Shakespeare, and Leonardo da Vinci. Nonetheless, because I am a behavioral scientist, my private list of heroes tends to be heavily weighted toward those who have made signal contributions to scientific knowledge. At the top of the roster are such notables as Galileo, Pascal, Newton, Faraday, Pasteur, and Einstein. But my all-time favorite is Charles Darwin. There are probably three main reasons for this choice.

First, Darwin had an unusually attractive and approachable personality, at least as revealed in his autobiography and correspondence. He had a rare combination of frankness, modesty, and persuasive self-confidence. Seriously dedicated to the hard work of science, Darwin also could enjoy the everyday world of family and community. He was arguably the most human of all scientific luminaries. He is certainly more accessible than Newton.

Second, all of his landmark contributions are accessible to educated lay readers—including me. One does not need special training, mathematical or otherwise, to read the Origin of Species, for example. Indeed, I can think of no scientific masterpiece that can boast such universal appeal. Newton’s Principia, in contrast, is very tough going, even for physicists (who are now unschooled in its archaic mathematics). At the same time, Darwin did not compromise his scientific integrity or effectiveness in producing such a popular work. It is rich in logic and fact—features that render the work all the more thought provoking and convincing.

Third, and probably most important, no scientist, living or deceased, has more influenced my own thinking than has Charles Darwin. I may admire Einstein or Newton or Galileo, but their epoch-making ideas belong to someone else’s discipline. In comparison, Darwin’s powerful contributions have left their indelible imprints throughout the biological and behavioral sciences, psychology not excluded. Great psychologists as diverse as William James, Sigmund Freud, and B. F. Skinner have all acknowledged Darwin’s penetrating influence. In fact, among the first behavioral scientists to feel the impact of Darwin’s ideas was Francis Galton, whose 1869 Hereditary Genius is considered one of the landmarks in the scientific study of creative genius.

Which brings me to my purpose in writing this book: Not only do I admire Darwin as a creative genius par excellence, but in addition I believe that Darwin provided the secure foundations for a comprehensive theory of extraordinary creativity. For more than a decade now, I have been grappling with whether the diverse features of creative phenomena are best understood from a Darwinian perspective. This potential application is by no means new with me. I have many predecessors going as far back as Galton and James. Nevertheless, I believe the time has come to consolidate all of the diverse efforts at constructing a Darwinian theory of creative genius. Besides advancing our understanding of creativity, such a consolidation would also serve as a kind of homage to that scientist who most profoundly shaped my own thinking.

I would like to thank Joan Bossert and Philip Laughlin, my editors at Oxford, whose initial reactions to an early draft of this book were most encouraging. I am particularly grateful to Philip for his having obtained a balanced set of external reviewers. Two of these evaluators provided detailed critiques along with a willingness to be identified: Colin Martindale and Howard Gardner. In addition, Frank Sulloway provided numerous, detailed, and useful comments on what I had—quite mistakenly!— thought to be the final draft. Although none of these readers can be said to be sympathetic with everything in the final product, I hope that all will appreciate my attempts to accommodate their most significant criticisms. The book certainly has much fewer faults as a result of their input.

Charles Darwin did all of his research at home. He was therefore fortunate to have a family willing to provide him with that special environment he needed to read and write, contemplate and experiment. Because I also do virtually all of my own research at home, I must feel myself equally blessed. Specifically, I must explicitly thank my wife, Melody, and my daughter, Sabrina. Melody’s role was especially crucial, and in all aspects of my life and work. I can think of no better way of honoring her contribution than to quote what Darwin said of his own wife:

She has been my greatest blessing, and I can declare in my whole life I have never heard her utter one word which I had rather have been unsaid. She has never failed in the kindest sympathy towards me, and has borne with the utmost patience my frequent complaints.... I do not believe she has ever missed an opportunity of doing a kind action to anyone near her. I marvel at my good fortune that she, so infinitely my superior in every single moral quality, consented to be my wife. She has been my wise adviser and cheerful comforter throughout life, which without her would have been ... a miserable one.... She has earned the love and admiration of every soul near her.

ORIGINS OF GENIUS

This page intentionally left blank

c

V_>ivilizations are often defined by the lives and works of their creative geniuses. The glory that was ancient Greece was built on the achievements of Homer, Pythagoras, Herodotus, Sophocles, Plato, Hippocrates, Phidias, and hundreds of other great creators. Modern European civilization was illuminated by the likes of Galileo, Descartes, Tolstoy, Rembrandt, and Mozart. The same story may be told of the world’s other civilizations. The histories of Persia, India, China, and other high cultures are to a very large extent chronologically arranged biographies of notable creative minds. As Thomas Carlyle proclaimed in his famous 1841 essay On Heroes, “Universal History, the history of what man has accomplished in this world, is at bottom the History of the Great Men who have worked here.” Among those he discussed at length as illustrations were Dante and Shakespeare, two of the greatest writers in any language.

So obvious is the debt civilizations owe these exceptional individuals that the appearance of such geniuses is often considered an indication of the creative health of a civilized culture. For example, this linking is evident in the 1944 Configurations of Culture Growth by Alfred Kroeber, an eminent cultural anthropologist. Wishing to gauge the emergence, growth, and stagnation of civilizations throughout the world, Kroeber could conceive of no better method than to assess the appearance of famous creators across a culture’s history. A civilization enjoyed a golden age when it overflowed with first-rate creative minds, experienced a silver age when the creative activity descended to a less notable level, and suffered a dark age when creators became few and far between. Indeed, this association goes beyond the abstract operational definitions of behavioral scientists. What nation does not take pride when one of its writers or scientists is awarded the Nobel Prize?

Yet who are these creative luminaries? Where do they come from? When do they appear? What are they like? Can any one of us become a creative genius? Or is high-caliber creativity limited to the one in a million?

These questions are important. In fact, they provide the impetus behind my writing this book. But before I can even begin to address these issues, I must first define what we will take to mean “creative genius.”

Creative Genius

Actually, I need to define two distinct even if overlapping terms: genius and creativity. I will begin with a look at the multiple ways that genius can be defined, identifying the definition that will prove most useful in this book. I will then turn to the task of defining creativity, another concept that can be defined in more than one way. By merging these definitions, we end up with the book’s subject matter.

Genius

The word genius has a curious etymology. It dates from Roman times, when the word signified the guardian spirit of a particular individual or location. Because this spirit provided for the distinctiveness or uniqueness of the entity with which it was identified, it came to represent that which was special about the person, place, period, or other entity. Thus, we could speak of the “genius” of a culture or era or even people, such as the genius of Native American culture, the genius of the Elizabethan age, or the genius of the Arabic language. With respect to individuals, the term genius became descriptive of some natural talent, ability, or disposition, especially when it goes well beyond the norm. Hence, a person might have a genius for making the most apt remarks in socially difficult situations. In the extreme case, an individual might achieve fame for the realization of this special talent. As a consequence, famous composers, artists, writers, and scientists began to be called geniuses. Beethoven was a musical genius, Shakespeare a dramatic genius, Michelangelo an artistic genius, Newton a scientific genius, and so forth. Because the successful exercise of these special aptitudes seemed to imply an extraordinary degree of intellect or talent, the word genius also began to be used in the more generic fashion favored today—as someone who exhibits exceptional intellectual or creative power. The specific application of this power was less important than the possession of the capacity. It is for this reason that Samuel Johnson, the author of the first English dictionary, could claim that “the true Genius is a mind of large general powers, accidentally determined to some particular direction.”

Behavioral scientists have recognized the significance of the term genius by attempting to provide it with a more exact meaning. If the concept has any scientific content, it should be possible to devise quantitative measures to evaluate it. These measures may then be used to gauge the magnitude of genius an individual can claim, or at least to identify those who most deserve the appellation. Such measures are of two kinds, namely, those who attempt to assess intelligence and those who try to estimate eminence.

Intelligence. At the beginning of this century, psychologists tried to build a more precise definition on the more generic meaning suggested by Samuel Johnson. This new conception was founded on the intelligence test. Alfred Binet had already devised a measure of intellectual ability, which others transformed into a measure of IQ, the intelligence quotient. People with average powers received scores of 100, and scores lower or higher than this baseline figure indicated whether an individual was below or above average in intellectual ability. Eventually, a number of psychologists, such as Lewis Terman at Stanford University and Leta Hollingworth at Columbia University, began identifying children as “geniuses” who obtained exceptionally high scores on these tests. The exact cutoff would vary from researcher to researcher, although it seems that the threshold IQ ranged somewhere between 130 and 140. The basic assumption was that someone with an IQ in this range or higher could boast the “mind of large general powers” that might be channeled successfully into almost any activity.

This high-IQ definition of genius became very popular not only among psychologists but among the general public besides. Parents of high-IQ children would learn to call their offspring “geniuses.” There even exists a society named Mensa that consists solely of “geniuses” who have scored a bit better than 130 on some standard IQ test. Even so, this psychometric definition leaves much to be desired. As will become apparent later in this book, a high IQ by no means ensures that an individual will display any special talent or achievement beyond the rather restricted ability to score high on standardized tests. For instance, the Guinness Book of Records notes that Marilyn vos Savant holds the record with an IQ of 228, and yet she does not have the kind of accomplishments to her credit that we might predict from such an exceptional intellect. Rather than discovering the cure for cancer or even making a better mousetrap, her achievements thus far have been limited to writing the weekly column “Ask Marilyn,” in which she responds to readers’ questions that supposedly only a real brain can answer.

Even worse, those with accomplishments worthy of the designation “genius” do not always make the IQ cut. When Terman first used the IQ test to select a sample of child geniuses, he unknowingly excluded a special child whose IQ did not make the grade. Yet a few decades later that overlooked talent received the Nobel Prize in physics: William Shockley, the cocreator of the transistor. Ironically, not one of the more than 1,500 children who qualified according to his IQ criterion received so high an honor as adults. Clearly, a Nobel laureate has much greater claim to the term genius than those whose achievements did not win them such applause.

Eminence. The last statement brings us to an alternative definition—the one preferred for use throughout this book. This particular definition’s origins go back over a century, to Francis Galton, whose 1869 classic, Hereditary Genius, defined genius in terms of enduring reputation. By this Galton meant “the opinion of contemporaries, revised by posterity... the reputation of a leader of opinion, of an originator, of a man to whom the world deliberately acknowledges itself largely indebted.” Certainly, those men and women who receive Nobel Prizes are honored for this very reason. When Niels Bohr received the Nobel Prize for physics, when Marie Curie received one for chemistry, when Ivan Pavlov received one for medicine and physiology, and when Toni Morrison received one for literature, they were being acknowledged for their notable contributions to their respective domains, and hence to human culture as a whole.

The only qualification that we must impose on this attribution is that, as Galton noted, the “opinion of contemporaries” must be “revised by posterity.” There do exist occasions when contemporary judgment errs according to the retrospective assessment of subsequent generations. Banting and Macleod, for example, shared a 1923 Nobel Prize for figuring out a way to isolate insulin, yet now the credit is given to Banting and his laboratory assistant Best, the original role of Macleod now being considered minimal at best. Nonetheless, empirical studies have amply demonstrated that there exists a strong consensus linking the judgments of contemporaries with the evaluations of posterity. Those who were most famous in their own times tend to be the most eminent decades, even centuries, later. Hence, Dante was being overly pessimistic when he claimed that “worldly renown is naught but a breath of wind, which now comes this way and now comes that, and changes name because it changes quarter.” In fact, eminence relies not only on transhistorical stability but on cross-cultural consensus besides. Individual differences in distinction cut across national boundaries and transcend subcultures within nations. For example, differential fame of African Americans within that minority culture correlates very highly with the differential fame of those same luminaries within the majority European American culture.

This eminence definition of genius has four major advantages. First, it automatically avoids the problem of the so-called unrecognized genius. If an individual commands no reputation, and thus is unrecognized, then it is not possible for him or her to claim status as a genius. Indeed, the phrase unrecognized genius becomes an oxymoron. Second, the eminence definition comes closer to what genius means in everyday language. The word is commonly used to refer to those individuals whose impact on history is most widely recognized as broad and enduring. Third, this definition captures something of the notion of uniqueness that is present in the word’s etymology. People do not claim fame because they do what everyone else does. On the contrary, they attain distinction because they accomplish that which sets them apart from the crowd. Every first-rate genius is necessarily sui generis. Fourth, because eminence varies immensely from person to person, the current definition allows us to speak of degrees of genius. For instance, the place of Beethoven in the history of European music amply exceeds that of his contemporary Anton Reicha. On that basis we can style Beethoven the superior genius. Only Bach and Mozart have reputations that rival Beethoven’s in the world of classical music.

Of course, no definition is perfect. Fame is at times capricious. Eminence is sometimes bestowed without complete regard to actual achievement. As Shakespeare once put it, “some are born great, some achieve greatness, and some have greatness thrust upon ’em.” Even so, the connection between overt accomplishments and ultimate distinction is not so whimsical as to render this operational definition invalid. It will be sufficient for our purposes to concur with Thomas Carlyle when he noted that “fame, we may understand, is no sure test of merit, but only a probability of such.” That probability is especially high when we deal with those who attain distinction for creativity.

Creativity

Another contemporary of Beethoven, Napoleon, also earned the title of genius, in his case military genius. Indeed, by the eminence definition, Napoleon’s genius may have far exceeded that of Beethoven. Winning decisive battles probably has far more impact than writing popular symphonies. However, in this book our interest will be in genius associated with creativity rather than that associated with leadership. This imposes the need to define what counts as creativity. Fortunately, this is a somewhat easier task. Psychologists have reached the conclusion that creativity must entail the following two separate components.

First, a creative idea or product must be original. Producing exact copies of someone else’s paintings, or reproducing verbatim quotes from other people’s poetry, or repeating scientific theories that others have already presented before the world—none of this can be considered original. Hence, not one of these activities is deemed creative. However, to provide a meaningful criterion, originality must be defined with respect to a particular sociocultural group. What may be original with respect to one culture may be old news to the members of some other culture. Thus, Galileo’s discovery of sunspots counts as an original contribution to European civilization even though the Chinese had noted their existence for well over a thousand years.

Second, the original idea or product must prove adaptive in some sense. The exact nature of this criterion depends on the type of creativity being displayed. In terms of technology, for example, an invention must not only be new, but it must also work. A rocket made of cinder blocks may be original, but if it cannot get off the ground, the conception cannot be considered creative. A scientific theory, in contrast, must be logically coherent and factually correct to count as adaptive. A theory that is self-contradictory or that conflicts with the best established empirical findings may be original, but it cannot be considered creative. In the arts, finally, adaptiveness often entails the capacity to maintain interest through novel expression as well as through powerful emotional appeal. For instance, a symphony that lacks beautiful or exciting themes and that fails to make a deeper emotional connection with the audience will fail by the criterion of adaptiveness. Clearly, an original idea or product is judged as adaptive not by the originator but rather by the recipients. Accordingly, we have another reason for maintaining that creativity entails an interpersonal or sociocultural evaluation. Not only must others decide whether something seems original, but they are also the ultimate judges of whether that something appears workable.

Given the foregoing two-criterion definition of creativity, we are now ready to define more precisely what we mean by creative genius. Essentially, these are individuals credited with creative ideas or products that have left a large impression on a particular domain of intellectual or aesthetic activity. In other words, the creative genius attains eminence by leaving for posterity an impressive body of contributions that are both original and adaptive. In fact, empirical studies have repeatedly shown that the single most powerful predictor of eminence within any creative domain is the sheer number of influential products an individual has given the world. Mozart is considered a greater musical genius than Tartini in part because the former accounts for 30 times as much music in the classical repertoire as does the latter. Indeed, almost a fifth of all classical music performed in modern times was written by just three composers: Bach, Mozart, and Beethoven. Parallel points can be made with regard to scientific creativity. The most potent predictor of contemporary and posthumous fame is the number of citations a scientist receives from other scientists who publish in professional journals. In fact, those scientists who receive the most journal citations also have the highest odds of earning the Nobel Prize for contributions to their scientific discipline.

Notice that this definition of creative genius agrees with Galton’s eminence criterion. These individuals are originators whose contributions are acknowledged by contemporary and future generations. These are people to whom others feel indebted. And this indebtedness can be illustrated in countless ways. It is shown when we walk through an art museum, attend a concert or opera, see a classic play, or read a great book. Homage is paid when the discoveries and inventions of the past are used to construct the miracles of today, whether they be drugs, telephones, computers, automobiles, bridges, jet airliners, or rockets. Of course, some of us display our indebtedness in more explicit fashion when we attempt to build on the achievements of a notable predecessor. As Newton advised, we can see farther than the rest if we stand on the shoulders of giants. If we wish to see farther into the origins of genius, the giant on whose shoulders we must stand may just be Galton’s cousin, Charles Darwin.

Charles Darwin

Was Darwin a genius? If we had to give an answer based on the intelligence definition, the answer might be negative. Darwin admitted in his autobiography that he was considered “much slower in learning than my younger sister” and that his teachers and father viewed him “as a very ordinary boy, rather below the common standard in intellect.” At school he found himself “singularly incapable of mastering any language.” Nor did he mature into a particularly brilliant adult, as he revealed in the following self-assessment:

I have no great quickness of apprehension or wit which is so remarkable in some clever men, for instance, Huxley. I am therefore a poor critic: a paper or book, when first read, generally excites my admiration, and it is only after considerable reflection that I perceive the weak points. My power to follow a long and purely abstract train of thought is very limited; and therefore I could never have succeeded with metaphysics or mathematics. My memory is extensive, yet hazy: it suffices to make me cautious by vaguely telling me that I have observed or read something opposed to the conclusion which I am drawing, or on the other hand in favour of it; and after a time I can generally recollect where to search for my authority.

Naturally, such modest remarks are hard to translate into a specific IQ score necessary for the application of the intelligence criterion. Nonetheless, one of Terman’s graduate students at Stanford provided an estimate of Darwin’s IQ score based on his childhood and adolescent activities and achievements, including any clear signs of intellectual precocity. He weighed in with an IQ of only 135. Although this would be high enough to earn Darwin membership in Mensa, it would not have been sufficiently high to be included in Terman’s primary sample of intellectually gifted children. Because such membership nominally demanded an IQ score of at least 140, Darwin might have joined a Nobel laureate among the rejects. Hence, by this definition Darwin would at best be considered a borderline genius.

Yet there can be no doubt that Darwin fits the Galtonian definition of creative genius. The magnitude of his influence on both the biological sciences and the world at large is unquestionable, made evident in a multitude of ways. For example, one ranking of “the too most influential persons in history” placed Darwin 17th, immediately after Moses! Darwin’s 1859 volume On the Origin of Species has been identified as one of the “books that changed the world.” And this same book, coupled with his Descent of Man, constituted a whole volume of the anthology Great Books of the Western World. Even a century after his death, hundreds of scientific journal articles appear each year that pay explicit homage to his powerful ideas. And thousands of books have been written on Darwin, his theories, and his empirical discoveries.

Perhaps the most striking evidence for Darwin’s status is the fact that his name has become an eponym. An eponym is a word that originated in someone’s proper name, such as is so often seen in place-names. Many creative geniuses have had their names expropriated in this fashion—another form of acknowledging our indebtedness. In the sciences, for example, there are laws attached to the names of Coulomb, Dalton, Fechner, Joule, Ohm, and Weber; temperature scales named after Celsius, Fahrenheit, and Kelvin; measurement units named after Ampere, Angstrom, Bell, Faraday, Henry, Orsted, and Watt; as well as miscellaneous elements, effects, concepts, theorems, processes, instruments, and even whole fields named after Boole, Curie, Descartes, Einstein, Fermi, Fourier, Galileo, Galvani, Gauss, Lamarck, Linnaeus, Lister, Mach, Maxwell, Mendel, Newton, Pasteur, Pavlov, Ptolemy, and Pythagoras. In Darwin’s case, of course, the relevant additions to the list of eponyms are the adjective Darwinian and the noun Darwinism.

But what is Darwinism? Why is it so important? What did Darwin contribute that earns him eponymic status? To address this question, we must recognize that there actually exist two kinds of Darwinism, the primary and the secondary. The primary form comprises Darwin’s theory of biological evolution, along with the many scientific developments extending from this theory to explain the diverse features of living organisms. The secondary form of Darwinism, in contrast, has to do with the explanation of other phenomena not directly related to biological evolution. Darwinian theory provides the basis for describing analogous processes that operate outside the sphere of biological evolution proper. This distinction between primary and secondary Darwinism is extremely important, because their scientific standings are not identical and the phenomena they treat are usually very different. Therefore, I must devote some time to outlining what these two forms entail.

Primary Darwinism

In essence, Darwin’s Origin of Species was devoted to establishing a phenomenon and a process. The phenomenon was the evolution of life-forms. Darwin spent many pages trying to show how all biological species, both living and extinct, descended from common ancestral forms. Earlier species would not only change over time but they would also frequently undergo speciation, producing two or more species where previously there had been only one. One of the more effective illustrations is that of the extremely varied species of finches Darwin observed on the Galapagos Islands during his famed Beagle voyage. Although at first Darwin made no explicit effort to include Homo sapiens in this documentation, a dozen years later, in Descent of Man, he made it clear that he believed the human species had also evolved. Humans shared ancestors with the apes, who in turn had common origins in the early mammals—and so on down to the very beginnings of life on this planet.

The very idea of biological evolution was extremely controversial at the time. Most Europeans, and even most naturalists, believed that the biblical accounts of creation, as described in the first chapters of Genesis, were literally true. All species were created in one grand miracle by God, and for the most part species had been fixed in form since the creation, an event that had transpired as late as around 4000 b.c. To be sure, some naturalists would allow for catastrophic events, such as Noah’s flood, that might have caused the extinction of some early life-forms and thus account for the previously unknown, fossilized animals that were being excavated. Even so, there was no reason to doubt the basic veracity of the creationist theory. Yet not only did Darwin struggle to undermine this prevailing view, his eventual inclusion of the human species in the evolutionary model struck at the very heart of the biblical story. God did not especially create Adam out of clay or create Eve out of Adam’s rib, but rather all of us were descended from lowly apes and monkeys.

Of course, not everyone was adamantly opposed to the doctrine of evolution. Many were willing to weigh the evidence and arrive at their own conclusions. Gradually, Darwin was able to win over many of his most important contemporaries, such as T. H. Huxley, who joined him in propagating the revolutionary idea. Moreover, in some respects evolution had the Zeitgeist on its side. Victorian England was obsessed with the idea of incessant progress, and evolution seemed to imply that the very origins of life involved an unrelenting progression from the less adaptive to the more adaptive and from the simple to the complex. Thus, human beings could be placed at the pinnacle of the evolutionary progression. And, taking things a step farther, some Darwinians would descry a similar progression taking place within our species, placing some races above others in some hypothetical hierarchy of humanity. For example, just such a racial ranking was advocated by Galton in his Hereditary Genius, in which he placed the Athenians above the British, and the British above the Africans.

Although the phenomenon of evolution was the most controversial feature of Darwin’s contribution in his own day, evolution is now considered by almost all biological scientists to be an established fact rather than a speculative theory. Even the majority of the Christian faiths have come around to accepting the phenomenon, interpreting Genesis in a more figurative sense. Where there still remains room for scientific controversy, however, is in the Darwinian process by which evolution was said to take place. Put in a nutshell, the process maybe characterized by the following six assertions.

• 1. There occurs spontaneous variation in life-forms within any given species. For example, the beaks of finches may be big or small, long or short. Of course, finches may vary on a host of other dimensions as well, such as coloration, overall size, wing design, digestive enzymes, and various instinctive behaviors involving courtship, nest building, and so forth.

• 2. These variable characteristics are to some extent subject to biological inheritance. That is, parents with certain traits tend to produce offspring with very similar traits. Like breeds like.

• 3. Some trait-variant life-forms are better adapted to the environment than others. For instance, some types of beaks may work best for small seeds, others for large seeds, and still others for insects or other classes of food. If small seeds are the predominant food source, a particular type of beak will become more adaptive.

• 4. The capacity of any species to reproduce their kind well outstrips the capacity of the environment to feed and shelter the potential progeny, which necessitates a “struggle for existence.” The inspiration for this proposition came from the 1798 Essay on the Principle of Population by Thomas Malthus, who first discussed the negative consequences of humanity’s aptitude for reproducing itself faster than it can feed itself.

• 5. Those variants that are more fit are more likely to survive and reproduce their kind. Those finches with beaks that best permit the exploitation of available food resources will have higher odds of surviving to maturity and of procreating through several breeding seasons. This is the critical Darwinian mechanism of natural selection.

• 6. With each successive generation, the more fit variants will gradually replace the less fit variants in a given population. The outcome is a population that maximally adapted to the resources of a particular environment. Over time what will emerge is an entirely new species.

Not only could this process explain the evolution of a species over geological time, but it could also account for the divergence of single species into two or more species—the phenomenon of speciation. Any complex environment contains a large variety of resources, such as different types of food or forms of shelter. Although an organism might endeavor to exploit all available resources, pursuing the strategy of a generalist, such an organism would find it difficult to compete with other organisms that have specialized in the exploitation of a subset of those resources—those that belong to what is now termed an “ecological niche.” The result is the phenomenon of adaptive radiation. Naturally, one example of such speciation is the way that “Darwin’s finches”—another eponym!—filled up the ecological niches created by the volcanic emergence of the Galapagos Islands. A more dramatic illustration is the manner in which the dinosaurs differentiated into their extremely varied forms: carnivores and herbivores, big and small, in a myriad of structural configurations, walking over the surface of the earth, swimming the planet’s oceans, and winging their way through its skies.

Hence, although this Darwinian process of natural selection could explain many aspects of the phenomenon of evolution, it was a far cry from a complete system. Darwin himself was well aware of many of the difficulties in his theory, and he spent much of his life trying to make improvements in his theoretical system. Moreover, after Darwin’s death others developed various aspects of his basic theory, expanding the explanatory power and sophistication of primary Darwinism. Three developments since the Origin of Species are worth special mention here: sexual selection, the “modern synthesis,” and sociobiology. Besides documenting the tremendous explanatory power of Darwinian theory, we will come across ideas that may have potential for helping us understand Darwinian perspectives on the origins of creative genius.

Sexual selection. Darwin realized very early that not all traits organisms possess can be easily explained according to the principle of natural selection. Indeed, sometimes a species would exhibit certain structures or behaviors that seem to be outright maladaptive. The classic example is the long and brilliant plumage of the peacock. If this grand display of feathers were so adaptive, one would think the peahen would boast the same characteristic. Instead, her tail has an appearance presumably more conducive to survival in the wild. Darwin recognized that such exaggerated traits were probably the upshot of sexual selection. Once the peahens, for whatever reason, began to prefer the peacocks with more impressive displays, those that beat out the competition sired more offspring. Each generation of males would possess even more impressive tail feathers. To be sure, the process cannot continue indefinitely. Eventually, the selection pressure exerted by the need for survival—the finding of food and the avoidance of predation—will outweigh any gains acquired in the sexual prowess conferred by these feathery advertisements. An equilibrium will be reached where the push of sexual selection toward larger and more brilliant tails will be counterbalanced by the pull of natural selection toward tails less cumbersome and conspicuous. The current peacock epitomizes the point of the even trade-off.

It is important to note that these two counteracting forces are really two means to the same end. Those traits that enhance the individual’s ability to reproduce its kind will be the ones that will become more prominent in the next generation. Natural selection will determine which variant will survive to sexual maturity; but under circumstances in which females can choose among males who must compete for mates, such survival may not suffice to ensure reproductive success. To obtain that goal, males may have to incur some costs to win the sexual competition. But, on the average, the benefits compensate for the costs.

The “modern synthesis.” Darwin’s initial theory encountered two difficulties that were even more fundamental. First, his theory did not specify precisely how the spontaneous variations came about. Second, his theory did not provide a mechanism for the inheritance of adaptive traits. These two problems are closely related. Before the advent of modern genetics, there existed a common belief in “blending” inheritance: the offspring would possess traits that would be some average of those of its parents. One consequence of such a form of inheritance would be that variation in any trait would diminish with each successive generation until all members of a species would exhibit the same inherent value. To be sure, there might survive variation due to environmental circumstances, but such variation would not be inheritable and thus not subject to natural selection. Another consequence of blending inheritance is that if by some fluke an individual happened to possess a highly adaptive trait that departed from the average, that characteristic would quickly become diluted once that individual mated with the rest of the population. Under such conditions, it is difficult to imagine how evolution could take place.

Although Darwin did not realize it, the solution to this problem had appeared within the decade that followed the Origin of Species. An obscure Austrian monk named Gregor Mendel had published the results of a series of breeding experiments with garden peas. These results established the basis of modern genetics. Mendel showed that various traits that characterize different varieties of peas, such as the color of the flowers, were carried by discrete units of inheritance that we now call genes. Moreover, during sexual reproduction, the diverse units freely recombine to produce an incredible variety of possible variants. This helped solve the problem of where the spontaneous variations originate. Just as important, the traits do not blend, but rather they maintain their integrity from generation to generation. Indeed, according to Mendel’s discovery of the distinction between dominant and recessive traits, an individual who received a dominant gene from one parent and a recessive gene from another would only exhibit the trait carried by the dominant gene. In any case, it soon became evident that variation in a population would be maintained across consecutive generations, unless selective forces operated to the contrary. Although Mendel’s original laws of heredity were to undergo many transformations over the years, it soon became clear that genetics helped resolve the difficulties in Darwin’s theory.

So primary Darwinism merged with Mendelism to produce a new “ism” known as neo-Darwinism, or the “modern synthesis.” Besides the laws of heredity, this updated version of Darwin’s theory incorporated the phenomenon of mutation into the mechanism of spontaneous variations. Not all variants that appear in a population represent straightforward recombinations of genes received from parents. Sometimes totally new genes will appear as well. Such haphazard additions are most often deleterious, and many mutant variations do not survive or reproduce their novel kind. Even so, mutations can provide a powerful resource for evolution on those rare occasions when the new trait is highly adaptive.

Neo-Darwinism gave primary Darwinism a new lease on intellectual life. It thereby became a rich theory that could be expanded in many directions. For example, with the discovery of DNA (deoxyribonucleic acid) and the emergence of molecular biology, both inheritance and mutation could be understood in more fundamental terms. The evolution of species could be investigated from the perspective of transformation in the makeup of the DNA molecules that contain the genetic information. An earlier, but just as potent, development was the emergence of the notion of population genetics, which holds that a species can be conceived as a population with a set of genes that have a particular frequency distribution that transfers from generation to generation according to genetic laws. Evolution takes place when the frequencies of various genes change over time in a manner that departs from the laws of heredity. Such changes in gene frequencies may reflect the operation of selection pressures. Or they could reflect the influence of such phenomena as genetic drift, which occurs in small populations when each generation obtains a random sample of the genes in the preceding generation. The discipline of population genetics provides a highly rigorous way of treating these diverse evolutionary changes by specifying the relevant processes in mathematical terms. For instance, these methods could be used to determine the precise circumstances under which sexual selection would produce a “runaway process,” in which traits become increasingly exaggerated and maladaptive. Hence, with the advent of population genetics, primary Darwinism advanced from a qualitative explanatory theory to a quantitative predictive theory. Darwinism thus enabled biology to join the ranks of other mathematical sciences, such as physics, chemistry, and economics.

Sociobiology. Some gaps in Darwinism’s explanatory scope still remained, however. One of the more irksome problems had to do with altruism. The Darwinian process appeared to favor only selfish behavior. In the struggle for existence—in the competition for food, shelter, and mates—there seemed no room for true self-sacrifice. To be sure, there might be occasions in which animals might engage in behavior with the superficial appearance of altruism. For example, animals might cooperate in the search for food or defense of their territory. And they might engage in acts of reciprocity, where favors might be exchanged according to the principle of “I will scratch your back if you will scratch mine.” But insofar as this reciprocal altruism directly benefits the individuals engaged, such behaviors cannot be considered bona fide altruism. At bottom, cooperation and reciprocation can still be seen as manifestations of underlying “selfish genes.”

Nevertheless, apparent acts of genuine altruism do seem to exist in nature. The most conspicuous examples, perhaps, come from the complex worlds of the social insects, such as honeybees and ants. The workers in these societies forfeit the opportunity to reproduce themselves in order to raise offspring on behalf of their queen. Besides putting their lives at risk when foraging for food, these workers will often sacrifice their lives in defense of the colony. This altruistic heroism is perhaps most dramatic in bees, whose barbed stingers often oblige their owners to commit suicide by disembowelment when attacking would-be invaders. Although perhaps not so sensational, similar acts of authentic altruism are sometimes prominent in the world of other social animals, such as the primates—including, of course, the human species.

One early solution to this enigma was to introduce the doctrine of “group selection.” Although altruistic behavior might be costly to the wellbeing of individual organisms, it was argued that those species with more altruists may be more apt to survive and reproduce when that species competes with others in the exploitation of a particular ecological niche. In other words, the unit of selection was neither the gene nor the organism but rather the entire group making up a species. Although this account was not without attractions or adherents, it fell short of a convincing explanation. Population genetics could easily demonstrate that group selection would favor the increased frequency of altruistic genes only under the rarest of circumstances. Indeed, within any given population there will always exist selection pressures in favor of selfish individuals willing to take advantage of their altruistic conspecifi.es. Eventually, the genes carried by these “cheaters” will overwhelm those of the altruists. Hence, the very existence of altruistic proclivities would seem to be most perilous.

The real breakthrough came with the recognition that it was not the individuals that were being selected, but rather it was the genes. Not only could those genes reside in more than one individual, but also related individuals shared more genes than did unrelated individuals, and the higher the degree of relationship the larger the expected proportion of shared genes. Therefore, a gene that promoted altruistic behavior on behalf of close relatives could indeed survive and proliferate in a population. Individuals might even sacrifice their own reproductive success for the sake of related individuals with whom they shared enough genes. This possibility is most apparent in ants and bees. Owing to a peculiarity in the reproductive systems of these hymenopterans, a daughter is so highly related to her mother that considerable adaptive advantage can accrue to those genes that encourage the subordination of the daughters’ reproductive interests to those of their queen. So the daughters become sterile workers who labor on behalf of the queen mother, helping the latter maximize the production of more sterile daughters. Thus, the apparently altruistic behavior remains selfish from the standpoint of the genes that support this subordination of individual interests to that of the colony. It is inclusive fitness, not individual fitness per se, that is being maintained—maintained through the process of kin selection.

Although the specific details differ, this same concept of inclusive fitness proved useful in explaining the appearance of altruistic behavior in other species, including mammals. For instance, the first member of a prairie dog colony to spot a predator will bark a quick warning to others rather than quietly disappear within the security of the colony’s underground tunnels system. This seemingly unselfish behavior makes more sense when we learn that the individual is acting to maximize inclusive fitness. After all, the members of any given colony are likely to be very closely related. As a consequence, the modest risk incurred by the individual in possibly attracting the predator’s attention is adequately compensated by the higher odds that other members of the colony who carry the same genes will be able to seek safety in time.

Once it became obvious that paradoxical behaviors like altruism could be explicated in terms of standard evolutionary theory, researchers endeavored to explain a great many social behaviors in strict Darwinian terms. The upshot was the emergence of the discipline of sociobiology, a research program most strongly associated with the zoologist Edward O. Wilson. Of course, the theorizing of sociobiologists was not always welcomed by other scientists, especially in the social sciences. It seemed like a classic instance of intellectual imperialism in which the theories of one discipline are imposed on the phenomena of quite another discipline. Moreover, some sociobio-logical explanations seemed not only reductionistic and deterministic but potentially sexist and racist besides. Sometimes sociobiologists seemed to defend the status quo, however inequitable, with assertions that such social injustices ensued from the evolutionary foundations of human nature.

Nonetheless, I think it remains fair to say that once we have time to weed out the good from the bad applications of sociobiological theory, the achievements of this form of Darwinism will be better appreciated. At the very least, sociobiology illustrates how very powerful were Darwin’s original insights. More than a century after his death, his theory of evolution by natural selection can still inspire new attempts to expand its explanatory scope and accuracy.

Secondary Darwinism

After the advent of the “modern synthesis,” primary Darwinism emerged triumphant in the scientific community. As its explanatory and predictive power became increasingly obvious, the same basic ideas soon began to be applied successfully to other phenomena besides organic evolution. These applications represent the emergence of secondary Darwinism. In these derivative extensions there was seldom a need to demonstrate the existence of a phenomenon analogous to evolution. On the contrary, these secondary applications usually begin with some occurrence of growth, change, or development assumed as empirical fact. As a consequence, the key goal of these secondary forms was to show that these evolutionlike phenomena represent environmental adaptations that are acquired through a Darwinian process of variation and selection.

To give a clearer idea of the nature of secondary Darwinism, I will offer brief examples of the most prominent applications. These illustrations fall into three categories according to the nature of the phenomena explained: the biological, the behavioral, and the cultural.

Biological phenomena. One of the most successful applications of secondary Darwinism is the immunological theory of antibody production. According to this theory, special cells (B lymphocytes) are constantly engaged in the random recombination of the various components of the antibodies that may successfully combat an antigen, such as a virus. When an antigen actually invades, a selection process takes place whereby those antibody combinations that best attack the intruder are selected at the expense of those that offer no defense. In a short time, the useful antibodies can be mass-produced until the immune system emerges victorious. This Darwinian procedure means that the system can guard against the evolution of new disease agents. This capacity is most critical, for the rate at which such agents can evolve is far more rapid than that of larger organisms such as the human being. Therefore, if humans and other large organisms with long life cycles had to acquire all the antibodies through organic evolution, they would probably face extinction. In a sense, every time we come down with a common cold, an eternal battle is being staged between a primary Darwinian process that yields ever more virulent antigens and a secondary Darwinian process that tries to counter with the most effective antibodies.

Another biological illustration involves the growth and development of complex nervous systems, such as the human brain. In central nervous systems of any complexity, there may be billions of neurons, each with thousands of synapses that support untold numbers of neuronal connections. It is inconceivable that the intricate neuronal circuitry can be completely specified in the genetic code. Therefore, much of the detailed structure of the brain must emerge developmentally, by some kind of interaction between genetic specification and environmental stimulation. Many neuroscientists have described this developmental process as entailing an essentially variation-selection procedure. In fact, these theorists have explicitly identified the hypothesized process as entailing “Darwinism of the synapses” or “neural Darwinism.” Although the particulars of the supposed mechanism depend on the theorist, the basic principle is the same. There first occurs an undisciplined “blooming” of neuronal connections—far in excess of what is required for efficient information processing. Next, this profusion of potential neural networks is followed by a “pruning” of connections that fail to prove useful according to subsequent sensory experience. As one advocate put it, the brain is a “Darwin machine” that endeavors to “make lots of random variants by brute bashing about, then select [ing] the good ones.” Such a flexible, Darwinian process permits the development of a nervous system maximally adapted to its experiential world.

Behavioral phenomena. It is one of the miracles of organic evolution that the primary Darwinian mechanism should discover two secondary Darwinian mechanisms for improving an individual’s immunological and neurological systems. But these are not the only illustrations of secondary processes evolving out of the primary process. Another example comes from the realm of an organism’s behavior. Obviously, one of the central means of adapting to the environment is to respond to it—to search for food, locate shelter, flee from predators, find mates, and so forth. Each organism possesses a whole repertoire of such behaviors dedicated to maximizing its adaptive fitness. For many species, the bulk of this behavioral repertoire may be provided by the primary Darwinian process of organic evolution. That is, the behaviors represent instinctive motor patterns that were the genetic gifts of many generations of evolutionary variation and selection. So essential is the evolution of these adaptive instincts that Darwin allotted an entire chapter of the Origin of Species to explaining their emergence.

Yet for those species with sufficiently complex nervous systems, the largest portion of the behavioral repertoire is acquired through direct experience with the world. Moreover, this learning process itself may be largely Darwinian. The most prominent proponent of this view was the psychologist B. F. Skinner, who devoted his life to outlining the principles of what he called “operant conditioning.” Furthermore, he explicitly identified this learning process as functioning in the same way as Darwin’s theory of evolution by natural selection. Each organism begins with a diverse assortment of rudimentary motor patterns that can be combined and recombined in an unlimited number of ways to produce more complex behaviors. The organism is thus capable of generating behavioral variations that provide the basis for environmental selection. Behavioral variants that produce positive consequences are retained, whereas those producing negative consequences are extinguished. In Skinnerian parlance, the organism emits “operants” that are then subjected to “reinforcement contingencies.” To study this phenomenon scientifically, behaviorists would typically place a food-deprived rat or pigeon in what came to be known as a “Skinner box.” To these conditions the creature would produce a wide variety of responses, most of which would prove futile; but eventually, by trial and error, the rat might press down a lever or the pigeon peck at a disk, and suddenly a food pellet would appear. With continued conditioning, the irrelevant actions vanish, and solely the appropriate operants remain. The animal is now optimally adapted to its environment, its behavioral repertoire shaped by the circumstances in which the organism must thrive and procreate.

Using this learning paradigm, Skinner and his followers were able to demonstrate that extremely complex behaviors could be established by this Darwinian procedure. Pigeons have been trained to use mirrors, play tunes on the piano, compete against each other at Ping-Pong, and engage in symbolic communication. Moreover, behaviorists have taken operant conditioning outside the laboratory in an attempt to show its utility in solving applied problems (through such applications as learning machines, token economies, and behavioral therapy). Skinner has used operant conditioning as a general explanatory framework for understanding the acquisition and use of language. Even more interesting, given the subject matter of this book, Skinner has attempted to explain creative behavior as another example of a conditioned operant. Although he only speculated on the possibility, some empirical research suggests that creativity may respond positively to reinforcement, at least under the proper circumstances. Hence, this instance of secondary Darwinian theory can claim tremendous explanatory scope.

Cultural phenomena. The adaptations acquired through organic evolution are retained in the population gene pool. There they can be readily passed down from generation to generation as a form of “received wisdom.” In contrast, the adaptations acquired through operant conditioning are retained in the individual’s memory. That means that the individual’s death signifies a loss of that accumulated fitness. Therefore, it would seem more efficient to devise some means by which the skills and knowledge learned by one generation can be passed down to the next. The latter could then skip the trial-and-error process of their progenitors. Indeed, each generation could advance further than the previous generation, building on prior achievements to attain adaptive accomplishments ever more impressive.

The most obvious way of effecting this end is for the members of each successive generation to imitate the adaptive behaviors of their parents or other elders. The appearance of such “monkey see, monkey do” transferal of acquired expertise illustrates on a single level culture as an alternative means of retaining selected variants. If members of a species form social groups, with sophisticated means of communication—especially language—then the dissemination of adaptations can also occur by direct instruction, whether by parents, teachers, peers, or experts. Indeed, in literate civilizations that have produced books, the transfer of expert ise can even skip generations. The European Renaissance was partly stimulated by the rediscovery of the classical Greek creators whose works had been lost to Western civilization since the decline of the Roman Empire.

Many researchers have suggested that cultures evolve just as species do, and that the mechanism underlying cultural evolution is closely analogous to that seen in organic evolution. Where the evolution of life-forms entails the selection of genetic variations, the evolution of cultures involves the selection of variations in acquired behaviors or beliefs. The units of selection in the former instance are genes, whereas the units in the latter case have been called “memes.” This word derives from the Greek root for imitation; it thus represents the unit of imitation just as the gene stands for the unit of inheritance. According to Richard Dawkins, memes may include “tunes, ideas, catchphrases, clothes fashions, ways of making pots or building arches.” Moreover, “just as genes propagate themselves in the gene pool

by leaping from body via sperms or eggs, so memes propagate themselves in leaping from body to body via a process which, in the broad sense, can be called imitation.” In addition, cultural evolution may exhibit many features found in organic evolution, such as analogs of random variation, genetic drift, and sexual selection.

Admittedly, there also exist striking contrasts between these two forms of evolution. For example, the transfer of adaptations in organic evolution usually occurs from parents to offspring, whereas the transfer in cultural evolution may take place between any two individuals. Indeed, the generational transmission may even be reversed, as when children find themselves teaching their embarrassed parents how to use the family computer. Furthermore, for organic evolution the genes are passed from genotype to genotype, ignoring whatever adaptations individuals may have learned during their lifetimes. For cultural evolution, in contrast, the memes go from phenotype to phenotype, bypassing the genotype altogether—but thereby speeding up the rate by which a population can adapt to environmental conditions.

These and many other differences notwithstanding, the similarities between the two evolutionary phenomena are close enough to inspire valuable theoretical and empirical analyses. For instance, the populationgenetics methods that have proven so useful in the study of organic evolution have been fruitfully applied to the study of cultural evolution. In fact, these mathematical analyses have permitted the creation of “dual-inheritance” theories that treat organic and cultural evolution simultaneously. This is an important intellectual advance, for all human beings are ultimately the product of the interplay of two long evolutionary histories, one determining a person’s genetic constitution, the other governing that person’s cultural heritage. These dual-inheritance theories also illustrate how the distinction I introduced between primary and secondary Darwinism can sometimes break down. In such systems, the variation-selection process operates at two interacting levels, so that genes and memes undergo coevolution. Yet the very possibility of such a synthesis of the two types dramatically demonstrates how very powerful were Darwin’s initial insights.

Darwinian Genius

What makes Darwinism so attractive to so many scientists is its parsimony. Extremely complex, varied, and even unusual phenomena can be explicated in terms of a simple variation-selection mechanism. The richly diverse forms of life on this planet, the adaptive flexibility of the immunological and neurological systems, the unlimited elaborations of learned behaviors, and the remarkable variability of human cultures and societies— all these and more have been granted Darwinian interpretations. To be sure, not everyone is sympathetic to Darwinian theories. From the very onset critics felt that an intelligent universe filled with purpose had been replaced by a chaotic world without a set direction. The infinite wisdom of an allpowerful but caring Creator gave way to billions of hapless little creatures, each struggling to survive and reproduce by seemingly senseless trial and error. If primary Darwinism evokes so much complaint, secon dar y Darwinism cannot be expected to do much better. Critics of B. E Skinner maintained that the behaviors that most distinguish humans from the brutes were precisely those that could not be explained by operant conditioning.

Even so, when a Darwinian explanation is found not to prove valuable in understanding a certain phenomenon, it often happens that the fault lies in which Darwinian mechanism was selected, not in Darwinism per se. Variation-selection processes occur at multiple levels, and so the investigator must be ever careful to adopt the proper level of operation. For instance, there exist many human social behaviors that may not comply with sociobi-ological interpretations because those behaviors have a cultural et iology. Yet insofar as cultural evolution is itself shaped by a Darwinian process, a valid explanation may finally emerge when the analysis is switched to populations of memes rather than genes. Similarly, although the Skinnerian conditioning paradigm stressed the variation and selection of behaviors, it is manifest that for complex cognitive systems, such as possessed by human beings, the Darwinian process more often operates internally. Before an individual acts in a novel situation, cognitive representations of alternative responses can be first generated and then tested against mental models of the external world. Finally, the researcher must always be wary of assuming that all secondary processes operate in precisely the same way as the primary process. As already noted, numerous functional differences divide organic and cultural evolution, and these differences must be accommodated for any scientifically useful account.

Once these precautions are taken, however, Darwinian theories can provide extremely potent explanations for a great many natural phenomena. Among those phenomena is human creative genius. Indeed, there is something quite natural about such a theoretical application. The very definitions of creativity and genius seem almost to beg for a Darwinian perspective. If creativity is defined as the output of ideas that are both original and adaptive, then the creative act may approximate a variation-selection process. The creator must generate many different novelties from which are selected those that satisfy some intellectual or aesthetic criteria. Those creative individuals who have produced an unusual number of original and adaptive ideas will then attain eminence, and thus be counted among the geniuses. An eminent creator, or genius, is someone who has contributed to subsequent generations an impressive body of achievements. Stated in terms of cultural evolution, the creative genius is a person who has produced a large assortment of memes to posterity. Creative geniuses exhibit extraordinary “reproductive success” through their “productive success.”

Why, for example, do we call William Shakespeare a literary genius? The reason is obvious when we consider the Bard’s legacy. Probably only the Bible is more likely to be found in English-speaking homes than is a volume containing the complete works of Shakespeare. His sonnets and other poems are still included in anthologies of great literature and are even recited as marriage oaths. His plays continue to be produced throughout the world, and in all the world’s major languages. Indeed, his dramatic creations can be found on record, CD, videotape, and in full-length feature films. They have also been adapted or transformed in an astonishing variety of ways, such as in Akira Kurosawa’s take on King Lear in his movie Ran, the updating of Romeo and Juliet in Leonard Bernstein’s musical West Side Story, or the remake of Hamlet found in Tom Stoppard’s Rosencrantz and Guildenstern Are Dead. Various other renditions of his dramas can be found in cartoons, comic books, and children’s storybooks as well as in tone poems, songs, modern dance and ballet compositions, puppet shows, and kabuki theater. More operas are based on Shakespeare than on any other author; a partial list of composers who have written Shakespearean operas includes such names as Adam, Barber, Bellini, Berlioz, Britten, Bruch, Goldmark, Gounod, Halevy, Holst, Nicolai, Purcell, Rossini, Smetana, Vaughan Williams, Verdi, Wagner, and Wolf-Ferrari. Tourists still visit Hamlet’s castle in Denmark and the (fictitious) tomb of Romeo and Juliet in Italy—not to mention all of the pilgrimages made to Shakespeare’s home in Stratford-upon-Avon.

Moreover, it is not just the poems and plays that define his legacy to human culture. English is the second most widely spoken language in the world, and no other individual has contributed more to the richness of that language than Shakespeare. Hundreds of words and expressions that have now become part of everyday speech first appeared on his writing desk. Assassination, birthplace, critical, droplet, equivocal, fashionable, go-between, hostile, invitation, lament, majestic, ode, pious, quarrelsome, retirement, shooting star, transcendence, useless, vulnerable, watchdog, and zany—these are just a smattering of the many words first coined by his genius. And such expressions as “All the world’s a stage,” “caviar to the general,” “the dogs of war,” “eaten me out of house and home,” “household words,” “a lean and hungry look,” “the milk of human kindness,” “one fell swoop,” “the primrose path,” “strange bedfellows,” “wild-goose chase,” and “the world’s mine oyster” have become so commonplace that some grammar checkers in word-processing software will tag many of these as cliches.

Furthermore, these gems of expression have inspired their counterparts in the main languages of the world, enriching each tongue with new terms, images, and metaphors. No wonder, then, that near the beginning of this century the Chinese poet Liu Boduan could write an “Ode to Shakespeare” that claimed “Three hundred years have passed ‘twixt then and now, Yet all the world looks to that mountain’s brow!” Liu’s praise confirms what Ben Jonson said of his illustrious contemporary: “He was not of an age but for all time.” If Jonson’s prediction holds, then there still should live memes descended from the Bard’s pen when the last lights of civilization expire.

Naturally, it was the same enduring nature of his legacy that obliged us to call Charles Darwin a creative genius. This legacy is acknowledged by the eponym that has attached his name to so much theoretical and empirical research in the biological and behavioral sciences. In the remainder of this book, I endeavor to carry this legacy forward into the very phenomenon that Darwin himself represents. For I believe that Darwinian theories can help us better appreciate the nature of creative genius. Indeed, I will argue during this book’s course that such theoretical models enhance our understanding of Charles Darwin himself, whom we may take as an incontrovertible exemplar of genius-grade creativity. In a certain sense, I will sketch out Darwin’s psychobiography, one conceptually Darwinist rather than psychoanalytic.

To make my case, each chapter will examine a different aspect of the phenomenon. Chapter 2 concentrates on the thought processes responsible for creative ideas, thus setting the stage for all subsequent discussion. Chapter 3 then covers the characteristics of creative individuals that enable them to generate those ideas. Chapter 4 continues the focus on the creator but this time scrutinizes the basis for the development of creative talent. Chapter 5 switches the unit of analysis from the creative genius to the products on which his or her reputation is based. In contrast, chapter 6 shifts the analytical unit in the opposite direction, to the groups, cultures, or civilizations in which the creative geniuses emerge. The final chapter reviews and evaluates what I consider the landmarks of our Darwinian tour of creativity.

I must stress that this panoramic view will not be totally uncritical, despite my professed enthusiasm for Darwinian principles. Theoretical interpretations can be bad as well as good. One of my goals is to examine the strengths and weaknesses of Darwinian accounts of exceptional creativity. The famous evolutionary biologist Theodosius Dobzhansky once made the often-quoted claim that “nothing in biology makes sense except in the light of evolution.” We seek to know the extent to which a similar remark may be applied to the origins of genius.

This page intentionally left blank

J—Xr win’s seminal ideas had an impact on epistemology, the philosophical enterprise devoted to comprehending the nature and origins of knowledge. The Darwinian theory of knowledge is called “evolutionary epistemology.” As the name implies, this discipline attempts to comprehend the foundations and status of knowledge in evolutionary terms. Actually, as there are two kinds of Darwinism, so are there two types of evolutionary epistemology.

On the one hand, the theory of organic evolution can provide a basis for explaining the specific nature of the human mind. Presumably, the brain exists in its present state as the result of adaptation, like any other adaptation, and its structure and function may have been shaped by generations of natural and sexual selection. Indeed, the very nature of what constitutes “common sense” may have roots in our ancestors’ needs to survive and reproduce. At the same time, any quirks and flaws in the human information-processing apparatus may be ascribed to the fact that certain intellectual skills may have contributed nothing to the reproductive fitness of ancestral Homo sapiens. In either case, primary Darwinism is being used to understand the human brain as a knowledge-making organ. Because the philosopher Immanuel Kant argued that human knowledge is shaped by innate categories of thought, Konrad Lorenz, the Nobel Prize-winning ethologist, has identified this first form of evolutionary epistemology as the “biologizing of Kant.” This intellectual movement has become especially conspicuous in recent years, as is evident in Steven Pinker’s 1997 book How the Mind Works.

On the other hand, evolutionary epistemology may assume the guise of secondary Darwinism. In this case the theory of evolution by natural selection provides a generic metaphor for understanding how organisms acquire knowledge. Although such analogical reasoning can prove hazardous, there is a certain aptness about this particular application. After all, to some extent inherited adaptations can be viewed as implicit forms of knowledge about the world. For instance, digestive enzymes reflect a knowledge about certain chemical processes, just as the wings of birds reflect a knowledge about inescapable aerodynamical laws. Similarly, learned behaviors that are acquired through operant conditioning constitute adaptations that are shaped by environmental consequences through a variation-selection procedure. And these operants, too, reflect implicit knowledge about causeeffect relationships in the world, such as the link between pecking at a disk and receiving a food pellet in a Skinner box. Hence, it seems easy to go the additional step and claim that all knowledge has its origins in some variety of variation-selection process. In the sciences, for example, the philosopher Karl Popper has observed: “The growth of our knowledge is the result of a process closely resembling what Darwin called ‘natural selection’; that is, the natural selection of hypotheses: our knowledge consists, at every moment, of those hypotheses which have shown their (comparative) fitness by surviving so far in their struggle for existence; a competitive struggle which eliminates those hypotheses which are unfit.” One of the most outstanding proponents of this secondary type of evolutionary epistemology is the psychologist Donald T. Campbell. More important for our current purposes, Campbell argued that the creative process itself can best be construed in Darwinian terms. In particular, he maintained that creativity involves the following three conditions:

• 1. There exists some process that generates variations. Just as biological evolution must begin with numerous genetic recombinations and mutations, so must creativity begin with the production of many diverse ideational variants.

• 2. These variations are subjected to some consistent selection mechanism. For biological evolution the fitness of variants is decided by natural or sexual selection. In the case of human creativity, the selectors are more likely to be cognitive or cultural in nature.

• 3. There is some retention procedure that preserves and reproduces the variations so selected. Where natural evolution retains and propagates the best genes through biological inheritance, the mental evolution that produces creative ideas requires a memory system, plus an ability to communicate the stored ideas to others.

In line with Darwinian thought, Campbell claimed that this variational procedure at some point becomes essentially blind. By this qualifier Campbell did not insist that the variations be absolutely random, although they may be. He held only that the mind eventually reaches the point where it has no a priori basis for knowing which ideational variations will prove most effective. Neither prior experiences nor current environmental circumstances will provide sufficient clues about how to restrict the range of choices, nor does there exist any rationale for assigning useful priorities to the various alternatives. As a consequence, the process is reduced to a basically trial-and-error procedure, whether through cognitive rumination or behavioral experimentation. It is for this reason that Campbell chose to identify creativity as a process of “blind-variation and selective-retention.”

The rest of this chapter will be devoted to assessing the plausibility of this Darwinian model of the creative process. I begin by examining what creators themselves have said about creativity. From there I will review the experimental research that bears on this question. Then I discuss what we might infer from computer programs designed to emulate the creative process. The chapter concludes with an overall evaluation of the variationselection model.

Biographical Illustrations

Many notable creators have explicitly claimed that their creativity appears best described by a process akin to a variation-selection procedure. For example, Paul Valery, the French poet and essayist, claimed that “it takes two to invent anything. The one makes up combinations; the other chooses, recognizes what he wishes and what is important to him in the mass of the things which the former has imparted to him.” John Dryden, the English playwright and poet, conveyed a similar idea with more vivid imagery when he described the composition of a play as beginning “when it was only a confused mass of thoughts, tumbling over one another in the dark; when the fancy was yet in its first work, moving the sleeping images of things towards the light, there to be distinguished, and then either chosen or rejected by the judgment.”

Nor are such claims confined to artistic creators, for similar reports come from scientific creators, suggesting that the creative process is the same for both artists and scientists. For instance, Michael Faraday, the great chemist and physicist, noted that “the world little knows how any thoughts and theories which have passed through the mind of a scientific investigator have been crushed in silence and secrecy by his own severe criticism and adverse examinations; that in the most successful instances not a tenth of the suggestions, the hopes, the wishes, the preliminatory conclusions have been realised.” Likewise, the Nobel Prize-winning chemist Linus Pauling advised that the successful scientist must “have lots of ideas and throw away the bad ones.... [Y] ou aren’t going to have good ideas unless you have lots of ideas and some sort of principle of selection.” Charles Darwin himself confessed in his autobiography that “I cannot remember a single first-formed hypothesis which had not after a time to be given up or greatly modified.” Hence, behind the clean logic and solid documentation of a scientific publication may hide a far more tentative history of trial-and-error.

Below I take advantage of the biographical reports of illustrious creators to fathom this process further. In particular, I will examine the role of associative richness, mental imagery, intuitive cognition, and serendipitous discovery.

Associative Richness

In organic evolution, variations are constantly created by the recombination of genes. In creativity, on the other hand, variations must be created by the recombination of ideas. This ability to engage in ideational permutations requires more than simply the capacity to retrieve concepts and images from memory. Also required is the ability to link these ideas by rather unusual pathways. This component was noted by the Austrian physicist Ernst Mach in a paper published more than a century ago. Although Mach allowed the advantages of possessing “a powerfully developed mechanical memory, which recalls vividly and faithfully old situations,” he insisted that “more is required for the development of inventions.” In particular, “more extensive chains of images are necessary here, the excitation by mutual contact of widely different trains of ideas, a more powerful, more manifold, and richer connection of the contents of memory.” If the memory cannot support this associative chaos, there can be no creative imagination at all.

Mach was discussing scientific creativity, but the same may be said of any form of creative genius. For example, the eminent psychologist William James described “the highest order of minds” in this way: “Instead of thoughts of concrete things patiently following one another in a beaten track of habitual suggestion, we have the most abrupt cross-cuts and transitions from one idea to another, the most rarefied abstractions and discriminations, the most unheard of combination of elements, the subtlest associations of analogy; in a word, we seem suddenly introduced into a seething cauldron of ideas, where everything is fizzling and bobbling about in a state of bewildering activity, where partnerships can be joined or loosened in an instant, treadmill routine is unknown, and the unexpected seems only law.” Henri Poincare described this process in a more picturesque manner when he recounted an episode in which he suffered insomnia as a consequence of an evening cup of coffee. “Ideas rose in crowds; I felt them collide until pairs interlocked, so to speak, making a stable combination. By the next morning I had established the existence of a class of Fuchsian functions.” Poincare compared these colliding images to “the hooked atoms of Epicurus” that jiggle and bump “like the molecules of gas in the kinematic theory of gases” so that “their mutual impacts may produce new combinations.” This choice of words is most provocative, for it implies a high level of randomness in the way that concepts and images are interconnected. The creative mind here seems capable of generating truly blind variations, as required by Campbell’s Darwinian model.

Mental Imagery

Poincare’s description illustrates another common feature of creative thought: the frequent use of vivid or unusual imagery. Max Planck, the physics Nobel laureate, explicitly affirmed that creative scientists “must have a vivid intuitive imagination, for new ideas are not generated by deduction, but by an artistically creative imagination.” And Albert Einstein held that “imagination is more important than knowledge” in the creation of new scientific ideas. This imaginative capacity can take many forms.

Einstein himself reported that “combinatory play seems to be the essential feature in productive thought.” Furthermore, “the psychical entities which seem to serve as elements in [this] thought are certain signs and more or less clear images which can be ‘voluntarily’ reproduced and combined.” These cognitive “elements are ... of visual and some of muscular type.” Although “the desire to arrive finally at logically connected concepts is the emotional basis of this rather vague play with the above mentioned elements,” the combinatory play takes place “before there is any connection with logical construction in words or other kinds of signs which can be communicated to others.” “The words or the language, as they are written or spoken, do not seem to play any role in my mechanism of thought.” Indeed, “conventional words or other signs have to be sought for laboriously only in a secondary stage, when the mentioned associative play is sufficiently established and can be reproduced at will.” The French mathematician Jacques Hadamard concurred when he described his own creative imagination, adding that “as to words, they remain absolutely absent from my mind until I come to the moment of communicating the results in written or oral form.”

Because words and logic may come into play late in the mental process, the creative genius often finds it difficult to translate the images into a form accessible to others. Darwin repeatedly complained to his colleagues about the extreme difficulties he encountered trying to put his ideas down on paper, however unambiguous and self-evident they were to him. Galton observed his similar frustration:

It is a serious drawback to me in writing, and still more in explaining myself, that I do not so easily think in words as otherwise. It often happens that after being hard at work, and having arrived at results that are perfectly clear and satisfactory to myself, when I try to express them in language I feel that I must begin by putting myself upon quite another intellectual plane. I have to translate my thoughts into a language that does not run very evenly with them. I therefore waste a vast deal of time in seeking for appropriate words and phrases, and am conscious, when required to speak on a sudden, of being often very obscure through mere verbal maladroitness, and not through want of clearness of perception. That is one of the small annoyances of my life.

Curiously, Galton admitted that words would sometimes appear in his creative thought, but when this auditory imagery took place it would not conform to standard linguistic expression. On the contrary, Gabon’s mind would experience nonsense words running by “as the notes of a song might accompany thought.” Other creators have also reported unusual verbal associations, such as puns. Because these capricious associations are governed by superfluous sound patterns rather than logical linkages, they boast a superior capacity to originate new and inherently blind connections between otherwise isolated concepts.

When the average person on the street thinks about vivid and unusual imagery, dreams are what most often come to mind. There do exist biographical accounts of creative ideas occurring during this altered state of consciousness. The most often cited illustration is the insight that led Friedrich August Kekule to decipher the ring structure of benzene. As someone who once aspired to become an architect, Kekule was a great visu-alizer and dreamer. Earlier he had already dreamt of atoms in playful recombinations, but on this evening, “I turned my chair to the fire and dozed. Again the atoms were gambolling before my eyes. This time the smaller groups kept modestly to the background. My mental eye, rendered more acute by repeated visions of the kind, could now distinguish larger structures, of manifold conformation; long rows, sometimes more closely fitted together; all twining and twisting in snake-like motion. But look! What was that? One of the snakes had seized hold of its own tail, and the form whirled mockingly before my eyes. As if by a flash of lightning I awoke.” Although there has been some doubt cast on the veracity of this story, the notion that creativity can emerge from dreamlike imagery has independent support from another quarter.

Psychiatrist Albert Rothenberg has systematically probed the relationship between creative thought and dream phenomena. His investigations were based partly on more than 1,670 hours of structured interviews with 57 highly creative artists and scientists, including recipients of the Nobel and Pulitzer Prizes, the National Book Award, and other prestigious honors. These interviews often took place while the subject was right in the middle of some creative project. By combining his interview data with retrospective studies of the products and introspections of other creators, Rothenberg identified two processes that bear a close relationship to dreams.

The first dreamlike process is called “homospatial thinking.” This involves “actively conceiving two or more discrete entities occupying the same space, a conception leading to the articulation of new identities.” In other words, two or more images are superimposed in the mind’s eye, and by a cognitive fusion a novel image results. Although homospatial thinking applies to images of any modality, its operation is most obvious in visual imagery. Rothenberg presents an ample number of examples drawn from the visual arts. Leonardo da Vinci, Paul Klee, Oskar Kokoschka, Henry Moore, Claes Oldenberg, and others all provide illustrations. For instance, in Klee’s 1927 Physiognomic Lightning the chief features of a man’s face are delineated by a bolt of lightning, an integrated image ensuing from two heterogeneous elements. Later, Klee even documented the steps by which he conceived this painting, and the procedure was patently homospatial.

The second dreamlike process is called “Janusian thinking.” This entails “actively conceiving two or more opposite or antithetical ideas, images, or concepts simultaneously.” The term comes from the Roman god Janus, who had two faces simultaneously looking in opposite directions. Rothenberg found that his distinguished creators resorted to this paradoxical maneuver quite often in the act of producing original insights. In addition, he finds numerous illustrations of this trick in the historical record. For example, Rothenberg showed how Janusian thinking helped Einstein to arrive at the general theory of relativity. Einstein realized that an observer who jumped off a rooftop would not, in his immediate vicinity, find any evidence of a gravitational field. This apparent absence arises even though gravitation causes the observer’s accelerating plunge. Rothenberg found another illustration in Niels Bohr’s conception of the principle of complementarity. The very claim that light is both a particle and a wave is inextricably Janusian. Moreover, according to one of Bohr’s sons, this Nobel laureate had elevated Janusian thinking to a basic technique of scientific discovery: “One of the favorite maxims of my father was the distinction between the two sorts of truths, profound truths recognized by the fact that the opposite is also a profound truth, in contrast to trivialities where opposites are obviously absurd.”

Homospatial and Janusian thinking, when combined with other forms of imagination, provide the creative genius with the ability to generate exceptionally original ideational variants. Indeed, some of these imagery processes appear capable of producing the blind variations suggested by a Darwinian model of creativity. It is hard to conceive of a process more blind than one that deliberately thrusts together opposites in the hope of discovering some meaning far more novel and profound.

Intuitive Cognition

I must hasten to emphasize that not all ideational variants arise through such dramatic means. Indeed, viable recombinations of ideas may emerge even when there appears to be no imagery at all! For instance, when 64 eminent scientists were interviewed about their mental processes, a large proportion reported the frequent occurrence of “imageless thought,” especially just before a breakthrough idea. Representative statements include: “I just seem to vegetate; something is going on, I don’t know what it is”; “I often know intuitively what the answer is, and then I have to work it out to show it”; “You feel it in your guts.”

Hadamard, the mathematician, even argued that much of the process by which new ideas emerge must take place at unconscious levels. In particular, he claimed that mathematical creativity requires the discovery of unusual but fruitful combinations of ideas. To find such novel variations, it is “necessary to construct the very numerous possible combinations.” But “it cannot be avoided that this first operation take place, to a certain extent, at random, so that the role of chance is hardly doubtful in this first step of the mental process.” Even so, “we see that the intervention of chance occurs inside the unconscious: for most of these combinations—more exactly, all those which are useless—remain unknown to us.” Thus, Hadamard was asserting that the variation process was blind in two different senses. First, to a certain degree, the production of ideational recombinations occurs at least in part by some largely random or chance process. Second, the conscious mind may not be aware of the combinations produced until some variant arises that passes some selection criterion. So much of the variation procedure transpires unseen.

Poincare, also a French mathematician, has described this process in more detail. In general, creative individuals begin with a problem, for which they attempt to find a solution. This is the phase of conscious preparation. When creators discover that a solution is not immediately forthcoming— when they find they are “hitting their heads against a brick wall”—they eventually give up, turning to other more fruitful activities, whether to ere-ating other products or to the everyday chores of work and life. The creators thereby enter an incubation period with respect to the initial problem. Then without warning, creative persons may experience a sudden illumination. For instance, Poincare offered the following observation: “I turned my attention to the study of some arithmetical questions apparently without much success and without a suspicion of any connection with my preceding researches. Disgusted with my failure, I went to spend a few days at the seaside, and thought of something else. One morning, walking on the bluff, the idea came to me, with just the same characteristics of brevity, suddenness and immediate certainty, that the arithmetic transformations of indeterminate ternary quadratic forms were identical with those of non-Euclidean geometry.” Poincare took this unexpected inspiration as a “a manifest sign of long, unconscious prior work.” He concurs with Hadamard in holding that much of the ideational recombinations take place below the threshold of consciousness. In fact, “the sterile combinations do not even present themselves to the mind of the inventor. Never in the field of his consciousness do combinations appear that are not really useful, except some that he rejects but which have to some extent the characteristics of useful combinations.”

Poincare even explicitly identifies the intuitive process as fundamentally blind: “Among the great numbers of combinations blindly formed by the subliminal self, almost all are without interest and without utility; but just for that reason they are also without effect upon the esthetic sensibility. Consciousness will never know them; only certain ones are harmonious, and, consequently, at once useful and beautiful.” To be sure, Poincare admitted that sometimes “a sudden illumination seizes upon the mind of the mathematician ... that... does not stand the test of verification; well, we almost always notice that this false idea, had it been true, would have gratified our natural feeling for mathematical elegance.”

Finally, Poincare uses the atomistic imagery quoted earlier to venture the following conjecture about the progression from preparation to incubation to illumination.

The role of the preliminary conscious work... is evidently to mobilize certain of these atoms, to unhook them from the wall and put them in swing. We think we have done no good, because we have moved these elements a thousand different ways in seeking to assemble them, and have found no satisfactory aggregate. But, after this shaking up imposed upon them by our will, these atoms do not return to their primitive rest. They freely continue to dance.... The mobilized atoms are ... not any atoms whatsoever; they are those from which we might reasonably expect the desired solution. Then the mobilized atoms undergo impacts which make them enter into combinations among themselves or with other atoms at rest which they struck against in their course.... However it may be, the only combinations that have a chance of forming are those where at least one of the elements is one of those atoms freely chosen by our will. Now, it is evidently among these that is found what I called the good combination.

Hence, in the preparation period creators establish some of the material that is going to enter into the variational procedure. In the incubation period these and related ideas are freely varied at unconscious levels of mental processing. When an acceptable combination appears, the outcome is thrust into consciousness during the illumination period. Of course, this is just the beginning of another variation-selection stage. Not all original ideas that appear in consciousness will survive verification, and of those that do survive, only a few will prove so fruitful to be worthy of extensive elaboration and development.

In any case, Hadamard and Poincare viewed the incubation as an operation in which unconscious mental processes are spontaneously generating ideational recombinations. However, the actual mechanism may be less active than this. For instance, Darwin’s younger contemporary Herbert Spencer—the evolutionist who introduced the term survival of the fittest— once offered the following speculation: “While thought continues to be forced down certain wrong turnings which had originally been taken, the search is in vain; but with the cessation of strain the true association of ideas has an opportunity of asserting itself. . . . [Qjuiet contemplation of the problem from time to time, allows those proclivities of thought which have probably been caused unawares by experiences, to make themselves felt, and to guide the mind to the right conclusion.” In other words, during the incubation phase, the mind is incessantly bombarded by a diversity of stimuli. Indeed, the longer the incubation, the more different circumstances the creators will find themselves in—whether at work, at home, or on vacation—and hence the greater the variety of such essentially random stimulation. By chance, some of these external inputs may set the chain of associations off in a direction more likely to lead to a solution. Because this sensory bombardment will often take place at relatively subliminal levels of awareness, the subjective experience under Spencer’s mechanism would likely be indistinguishable from that produced by the Hadamard-Poincare process. Moreover, this alternative mechanism would still entail an intuitive variation-selection procedure. Instead of internal associations being selected for their utility, haphazard sensory influx is undergoing selection, the most fecund exciting the associative response that leads to an adaptive intellectual variant.

I return later to this issue when I examine the experimental literature. But first I need to turn to the last form of biographical evidence.

Serendipitous Discovery

Let us go back to Poincare’s remark about the “brevity, suddenness and immediate certainty” with which the solution to a problem appeared to him. Many other creators have reported similarly unexpected and dramatic episodes of illumination. The first recorded historical example is the famous episode in which Archimedes suddenly discovered the solution to a problem upon settling himself into a bathtub. Needing to find a way to measure the volume of an irregular solid—a recently crafted gold crown for the king—Archimedes realized that the water displaced by his body was exactly equal to the volume of his body that was submerged. Reportedly, he became so immediately excited by the realization that he ran outside into the streets screaming “Eureka,” the Greek for “I have found it!” These eureka experiences are often so striking that the episodes become permanently engraved in the creator’s memory. Charles Darwin recorded in his autobiography the moment when he arrived at his epochal solution to the problem of the origin of species: “1 can remember the very spot in the road, whilst in my carriage, when to my joy the solution occurred to me.”

Now there is one type of illumination event that occupies a special place in the Darwinian theory of creativity—the phenomenon of serendipity. This term originated in a letter by the author Horace Walpole, who derived the word from a fairy tale called “The Three Princes of Serendip,” whose protagonists “were always making discoveries, by accident or sagacity, of things which they were not in quest of.” This happy invention was revived by physiologist Walter Cannon in an essay entitled “The Role of Chance in Discovery.” There he documented the numerous occasions when major contributions seem more attributable to accident than to genius. A similar point was made by physicist Ernst Mach in his article “On the Part Played by Accident in Invention and Discovery,” albeit without providing such a convenient term for this phenomenon. Since these classic contributions, many books have been written on the subject, and the list of examples has become very long. Some of the most frequently cited cases include: European discovery of the New World (Columbus, in 1492), discovery of the interference of light (Grimaldi, in 1663), discovery of animal electricity (Galvani, in 1781), creation of laughing gas anesthesia (Davy, in 1798), discovery of electromagnetism (0rsted, in 1820), discovery of ozone (Schonbein, in 1839), invention of synthetic coal-tar dyes (Perkin, in 1856), invention of dynamite (Nobel, in 1866), invention of the phonograph (Edison, in 1877), development of vaccination (Pasteur, in 1878), discovery of X rays (Rontgen, in 1895), discovery of natural radioactivity (Becquerel, in 1896), innovation of classical conditioning (Pavlov, in 1902), discovery of penicillin (Fleming, in 1928), invention of Teflon (Plunkett, in 1938), and invention of Velcro (de Maestral, in 1948). Although we often think of serendipity as a phenomenon that occurs in the sciences, instances do in fact take place in the arts as well. For example, Henry James, the fiction writer, revealed in his preface to The Spoils ofPoynton how his story grew from “mere floating particles in the stream of talk” during a casual dinner conversation. Hence, the phenomenon of serendipity is far more universal than is generally thought.

Upon closer examination, it becomes clear that serendipity may take a number of different forms. These forms vary according to the relative amount of “chance” involved as well as to the degree of intentionality underlying the discovery. To appreciate the potential contrasts, consider the following five cases.

Case 1: A person attempts to solve a problem and repeatedly fails. After incubating the question for some time, the person chances upon a fairly common circumstance that immediately inspires the solution. In this case, it is simply a matter of time before the required event will happen, and hence, to a certain degree the “accidental” input is inevitable. For example, when Gutenberg tried to solve the problem of how to mass-produce Bibles, he had worked out the notion of reusable type but had not yet figured out a workable device by which to make an imprint. It was during the wine harvest that he suddenly realized that the wine press provided the mechanical prototype for the printing press. He had found what he was looking for once the sequence of events had put him in the right place at the right time. There can be no doubt that all the other pieces of the puzzle were assembled, and the final piece would soon appear in the normal course of events.

Case 2: An individual attempts to solve a particular problem, but the solution requires a truly fortuitous event or sequence of events to be realized. The discovery in this case can be considered truly lucky, for it might not have happened at all. Charles Goodyear’s discovery of vulcanization is a prime example. India rubber had the unfortunate characteristic of changing flexibility with the ambient temperature. It would become hard and brittle in cold weather, pliant but sticky in hot weather—which severely limited its utility as a production material. Because it was already known that sulfur would remove the adhesiveness, Goodyear began to experiment with this mixture. One day he accidentally dropped some of this mixture on a hot stove, and thereby discovered the secret. Presumably, without this highly fortunate mistake Goodyear might never have found the solution.

Case 3: Someone works on a problem and manages to find a solution. But in the process of making the intended discovery, the individual stumbles upon a totally unanticipated finding. For instance, when James Clerk Maxwell developed his electromagnetic theory, his original goal was not to explain light. But when he used his equations to calculate the speed of electromagnetic waves, he was surprised to discover the figure equaled the known velocity of light. Similarly, Paul Dirac’s initial purpose behind creating his quantum-mechanical model of the electron was to make provision for relativity theory. Yet the existence of a hitherto unknown particle, the positron, just popped out of his equations. Physics was thus led to the entirely novel realm of antimatter. Unlike the cases of Gutenberg and Goodyear, Maxwell and Dirac discovered more than they had intended.

Case 4: A person is attempting to solve a given problem and has not yet succeeded (and may never do so). But along the way, the person chances across an unexpected discovery, such as a solution to an entirely different problem. The goal of Columbus was to find an alternative route to the Far East, but he ended up discovering the New World, albeit it was not at once obvious that he had done so. Observe that case 4 takes the lack of intentionality one step beyond case 3. The intended discovery is never made, while an unintended discovery is made in its stead. Serendipitous events like these play an important role in the theory of scientific revolutions proposed by Thomas Kuhn. In particular, Kuhn argues that the appearance of anomalies is what gradually undermines the confidence of a scientific community in a current theoretical paradigm. These findings may involve disconfirmed predictions or the discovery of unexpected phenomena. After a sufficient number of such anomalous results accumulate, the community enters a crisis stage that paves the way for the emergence of a new paradigm that displaces the old.

Case 5: Sometimes an individual is not looking for anything in particular, or at best is operating under only the vaguest of hunches. Perhaps the person is just “monkeying around” or “seeing what happens if”—that is, engaged in curious exploration or playful tinkering. By a stroke of blind luck, the individual stumbles upon an original but useful idea. This last case involves the highest degree of chance and the least degree of intentionality. It was an act of blind curiosity that motivated Antoni van Leeuwenhoek to direct his high-powered microscope at diverse substances and thus to discover a whole world of bacteria, protozoa, spermatozoa, and other fantastic organisms. His curiosity could tell him only that the findings might be interesting, and they fulfilled his expectations. Likewise, when Galileo pointed his newly improved telescope toward the heavens, he had no idea of what he would find. But what he saw was beyond what anyone at that time could possibly imagine—sunspots, lunar craters, Venusian phases, Jovian moons, Saturnian rings, and a host of other curiosities. Suddenly there appeared a vast array of anomalous discoveries that would destroy forever the Aristotelian and Ptolemaic cosmology that had dominated science since antiquity.

Some scholars would like to separate the first two cases from the last three. When individuals solve a problem they set out to solve, it is called “pseudoserendipity,” no matter how much luck may have participated in the favorable outcome. On the other hand, whenever people come across an unexpected discovery, as in cases 3 to 5, the result is said to involve true serendipity. This distinction fits the folktale from which the word was born, and it seems more in accord with a Darwinian theory of creativity. Indeed, a number of scholars have suggested that true serendipity more closely approximates the genetic mutations of organic evolution. In the first two cases, the creator meant to make a particular discovery and eventually did so. The only surprise was the specific means by which the desired end was attained. But in the last three cases, ideational variants emerged that were totally unanticipated, even less intended. These useful and adaptive variations are thus truly blind in the sense implied by Campbell’s theory. Moreover, like a chance mutation in organic evolution, these serendipitous ideas may radically alter the course of cultural evolution, obliging it to pursue unforeseen paths.

Although this Darwinian distinction between the pseudo- and true serendipity makes sense, too much can be made of the contrast as well. In some respects, the underlying thought processes may be very similar. After all, the same individuals who make unintended discoveries are often those who make intended discoveries, and the role of chance sensory input may be much the same. Some people have the capacity to take advantage of the opportunities provided by their experiential world, while others fail to do so. As Ernst Mach pointed out, many lucky discoveries “were seen numbers of times before they were noticed.” Fleming was certainly not the first to see a bacterial culture ruined by mold, but he was evidently the first to notice the significance of this observation. This suggests that creative geniuses may enjoy a special openness to the potential implications of the innumerable stimuli that impinge daily upon their brains. Charles Darwin himself provides a good example of what may be lacking in those of us who seem not to have Fortune on our side. When Darwin had to specify what he considered his single best mental attribute, he claimed, “I think that I am superior to the common run of men in noticing things which easily escape attention.” This exceptional capacity was confirmed by Darwin’s son Francis, who was often his father’s scientific collaborator. Francis took special note of his father’s

instinct for arresting exceptions: it was as though he were charged with theorizing power ready to flow into any channel on the slightest disturbance, so that no fact, however small, could avoid releasing a stream of theory, and thus the fact became magnified into importance. In this way it naturally happened that many untenable theories occurred to him; but fortunately his richness of imagination was equalled by his power of judging and condemning the thoughts that occurred to him. He was just to his theories, and did not condemn them unheard; and so it happened that he was willing to test what would seem to most people not at all worth testing. These rather wild trials he called “fool’s experiments,” and enjoyed extremely.

Francis provided a curious example of one of these experiments: After noticing that the leaves of a certain plant seemed sensitive to table vibrations, Darwin asked his son to play his bassoon close to the plant, to determine if it might perceive sound! Francis’s father was here displaying “his wish to test the most improbable ideas.” Yet by conducting such experiments, Charles Darwin was sending open invitations to Fortune in the hope that something serendipitous would be sent his way. Sometimes she complied, and when she did, Darwin always took note.

Experimental Investigations

If the foregoing biographical reports represented the sole evidence for a Darwinian theory of creativity, the latter would be in a very sorry state indeed. After all, introspections into higher mental processes are fraught with all sorts of inherent difficulties, and memories about past events can undergo considerable distortion with the passage of time. Indeed, I must confess that it is quite possible to find notable creators offering descriptions of the creative process that run quite counter to what the preceding accounts seemed to imply. A case in point is “The Philosophy of Composition,” an essay in which Edgar Allan Poe recounted the creation of his classic poem “The Raven.” Because the essay appeared only one year after the poem’s own publication, Poe’s description certainly enjoys the advantage of being close to the events described. Yet in this account Poe said, “It was my design to render it manifest that no one point in its composition is referable either to accident or intuition—that the work proceeded, step by step, to its completion with the precision and rigid consequence of a mathematical problem.” He then went on to show how logic dictated every single choice, from the optimal number of words right down to single words and images. Poe makes the whole creative process so straightforward, even mundane, that there seems little latitude for the kind of chaotic and unpredictable thinking discussed earlier.

Of course, Poe’s report could be dismissed as just a piece of propaganda. He was engaged in attacking the prevailing Romantic notions of his day. As he put them, “Most writers—poets in especial—prefer having it understood that they compose by a species of fine frenzy—an ecstatic intuition.” Moreover, Poe somewhat prided himself on his analytical skills, a trait that shows itself quite vividly in his detective stories, such in as “The Purloined Letter.” Nevertheless, if his testimony is dismissed as biased reporting, then why should not the same judgment be passed on all the biographical evidence that has been handed down to us? Certainly there is reason to believe that others have distorted the stories behind their creation to better fit the Romantic view of the creative process. Samuel Coleridge, for instance, reported how his poem “Kubla Khan” originated in an opium-induced stupor—a nice touch given the great vogue of this drug at that time. But scrutiny of his drafts shows that Coleridge probably made up this episode. Hence, rejection of any given bit of biographical evidence can cut both ways.

Rather than debate this matter further, it may behoove us to look elsewhere to see if there exist other, more scientific sources of evidence that endorse the Darwinian view. Here I review a diverse assortment of laboratory experiments that deal with pertinent aspects of the creative process. Although the participants in these experiments are more likely to be college students than established creators, the studies at least avoid the problems associated with using introspections and anecdotes. The specific experiments concern insight, imagery, and intuition.

Insight

Creativity is often identified with the process of insight. In fact, the term insight is sometimes used interchangeably with the moment of illumination in the creative process. Insight has also attracted a considerable amount of experimental research. The earlier Gestalt psychologists were especially interested in the phenomenon, and much of the subsequent research is founded on those pioneering inquiries. Although the bulk of the studies use human beings as experimental subjects, some studies have examined insight in animals. Let us review the latter inquiries before turning our attention to the former.

Animal studies. Because insight seems to be such a supreme mental ability, it may appear strange to look for insight in nonhuman organisms. Indeed, if not every human being enjoys the capacity for insight, it may seem absurd to study the phenomenon in animals. However, because Darwinism implies that there exists some degree of evolutionary continuity between humans and other animals, there should appear mental processes analogous to insight in our closest phylogenetic relatives. One likely prospect is the chimpanzee, an organism whose cognitive skills are remarkably close to those of Homo sapiens. One of the classic Gestalt studies of insight, in fact, was conducted by Wolfgang Kohler, using chimps. Kohler wished to show that the solution of problems required a perceptual process, one that entailed a sudden and novel reorganization of experience—the sudden insight. To make his case, he would present a chimp with an urgent but difficult problem: how to reach a banana that was set deliberately beyond arm’s reach. Successful solutions required the chimp to put together a long pole from two short poles or to pile boxes on which the chimp could stand. In the typical experiment, the animal would try the most obvious tactics until it became apparent that nothing in its standard repertoire of behaviors would allow it to attain the goal. The chimp would then enter an incubation period, after which it would exhibit an apparent eureka experience. The necessary behaviors would then quickly follow until the banana was at hand. Because finding a solution required the organization of a novel sequence of behaviors, and because the solution came so suddenly, the chimpanzee can be said to have displayed insight. Outwardly, at least, there does not appear to be much of a difference between the manner in which the animals and Archimedes solved their problems.

The Gestaltists believed that such insightful behavior allows the organism to bypass trial-and-error learning. But this conclusion goes well beyond the evidence. Critics were quick to point out that the various behaviors that the animals had to assemble to solve the problem were already part of their repertoire. The chimps had previously learned to put the sticks together and to pile boxes. Moreover, the chimps surely could have used this basic input as part of an internal process of trial and error. Remember what creative individuals have said about the importance of generating lots of ideational variants—what Einstein called “combinatory play.” Certainly, chimpanzees have the brains to engage in the same process, even if at a more limited level. If so, the mechanism underlying insight would still be Darwinian. Creative chimps would be manipulating representations of their behavioral world until they found a combination that satisfied the constraints of the problem situation, such as the height and distance of the hanging banana.

Actually, there already exists a highly successful explanation of Kohler’s results completely within the framework of Skinner’s operant conditioning paradigm. And, as pointed out in chapter 1, this learning paradigm is explicitly Darwinian in nature. The proponent of this theory is Robert Epstein, a close collaborator of B. F. Skinner’s. Together they had launched a simulation project in which they attempted “to get pigeons to do some of the complex and mysterious things people do.” At first, these behavioral simulations did not depart far from usual learning studies. The pigeons would simply undergo reinforcement for certain behaviors until the desired operants were established. But a serendipitous finding emerged that opened a whole new avenue of inquiry. Often when a pigeon was placed in a new situation in which previously learned behaviors did not apply, it would spontaneously generate new behaviors that were intriguing combinations of acquired behaviors. This led Epstein to examine the possibility that the insightful solutions displayed by Kohler’s chimps might be exhibited by pigeons as well.

The experimental simulation required that the pigeon learn three behaviors. First, the pigeon learned it would receive food when it successfully pushed a box toward a green spot located at random positions on the floor of the Skinner box. Second, the pigeon learned it could obtain food by pecking at an object—made to look like a little banana—suspended from the top of the chamber. However, it also learned that the food would be delivered only if the pigeon’s feet were standing on something. Jumping and flying at the toy banana accomplished nothing. Third, the pigeon learned how stepping on a box to peck at the banana was permissible, and the position of the banana-box combination was freely varied about the cage. After acquiring this behavioral repertoire, the pigeon was confronted with a novel situation: A banana was suspended out of reach, with the box located in another part of the chamber. None of the pigeon’s learned behaviors will solve the problem, although the right combination will do the trick. So what did the pigeon do? It behaved in a manner virtually indistinguishable from Kohler’s chimps. “At first the pigeon looked confused. It stretched toward the banana, looked back and forth from the banana to the box, and so on. Then, quite suddenly, it began to push the box toward the banana, sighting the banana as it pushed. Each pigeon stopped pushing when the box was beneath the banana and then immediately climbed and pecked.” Significantly, pigeons that did not acquire the full behavioral repertoire either failed to solve the problem or solved it only through trial and error (depending on which operants had been learned). This replicates what has been found in chimpanzee studies as well. Chimps will show insight only when they have learned all the behavioral components required of the solution.

A critic might consider Epstein’s simulation of insight more cute than convincing. But, on the contrary, Epstein has established the theoretical plausibility of this model in three important ways. First, he has carefully studied the behavioral processes that underlie the insight phenomenon in order to comprehend precisely how it works. For example, he has advanced the “principle of resurgence,” which states, “When, in a given situation, a behavior that was previously successful is no longer successful, behavior that was previously successful in similar situations tends to recur.” Thus, the organism begins to generate behavioral variants based on all those past circumstances that might bear some relation to the current problem. Second, Epstein has cast his explanation in more formal terms, producing equations that provide the basis for computer simulations. These equations describe how the probability of the appearance of various behaviors changes over the course of the organism’s interactions with the environment. In other words, the model describes the manner in which the organism achieves increased adaptive fit by generating and testing behavioral variants. Third and last, Epstein did not rest content with applying his “generativity theory” to pigeons. He has applied the same paradigm to human subjects trying to solve a classic insight problem. The computer simulation predicts the behavior of humans just as well as it does for animals. As a consequence, insofar as creativity requires insightful behavior, Epstein’s theory can be considered a viable Darwinian model of the creative process. Certainly the model incorporates a mechanism by which animals can generate behaviors that are both original and adaptive.

Human studies. The insight problem that Epstein successfully modeled using his generativity theory was the famous two-string problem. The participant in the experiment enters the laboratory and sees two strings hanging from the ceiling. The subject is told that the object is to tie the two strings together. In the vicinity are a number of objects that may be used to solve the problem, including a pair of pliers. The subject soon learns the task is more difficult than first meets the eye. The relationship between the distance between the strings and their length is such that one cannot simply grab the end of one string, walk over to where the second string hangs, grab that second end, and complete the task. Try as he or she might, the other string always remains frustratingly beyond reach. The solution requires the subject to use the pliers in an unusual fashion—to create a pendulum. Once the pliers are tied to the end of one string, and that string is set in oscillation, it is an easy matter for the subject to walk over to the second string, pull its end toward the first string, and grab the pliers as they swing toward the subject. The rest is simple.

Many subjects experience severe difficulty solving this problem, and research has unearthed some of the factors that enhance the odds of successful solution. One factor concerns the nature of the objects available to tackle the task. To use a pair of pliers as a weight requires the subject to perceive that tool in a different manner, as merely a mass suitable for making a pendulum. But if the available objects include a plumb bob, the probability of a solution increases dramatically. Another factor is even more provocative: The subject may respond to “hints.” In one experimental condition the experimenter would casually walk by one of the strings, brush it “accidentally” with his shoulder, and thus set it in motion. Shortly after this clue the subject was much more likely to solve the problem. It is interesting to note that when subjects were asked how they arrived at the necessary insight, they would often fail to report the hint as the precipitating factor—the impact of the cue would be subliminal or unconscious.

This is but one example of the kind of problem studied in the rich literature on insightful problem solving. Although the details vary, the findings are generally comparable. When an individual first confronts a novel problem, the most obvious responses may utterly fail. To succeed, the various features of the problem must be reformulated in a different and often highly unusual fashion. As a consequence, the problem solver may enter an incubation period, in which the mind must open up to various possibilities. During this interval the individual is exposed to all sorts of extraneous input. Some of this input is external (everyday events as well as work on other projects), and other input is internal (retrieved memories, chains of associative thought). But whatever the specific source, this bombardment is constantly “priming,” or exciting, different aspects of the mnemonic and semantic networks surrounding a given problem. This largely random influx of priming stimuli produces a series of alternative formulations, some more fruitful than others, but with only one leading the individual down the correct path to solution. In other words, during incubation the mind is engaged in an inadvertent blind-variation process. The process is “blind” because the order in which the new conceptions appear is determined by factors pretty much irrelevant to the problem. Indeed, it is the very extraneous nature of this input that is so essential to the solution. Subjects get stumped on insight problems because the most obvious ways of thinking about them prove abortive.

Because much of this haphazard input takes place at unconscious levels of information processing, the problem solver might not be cognizant of the stimulus that led to the successful solution. Like for the subjects in the two-strings problem, the solution may be partly intuitive. As noted earlier, this lack of awareness fits with the observations of notable creators themselves. For instance, Carl Friedrich Gauss, the reknowned mathematician, once recorded that, after years of failing to solve a problem, “finally, two days ago, I succeeded, not on account of my painful efforts, but by the grace of God. Like a sudden flash of lightning, the riddle happened to be solved. I myself cannot say what was the conducting thread which connected what I previously knew with what made my success possible.” Presumably, it was some arbitrary event that inspired following the appropriate pathways to the requisite insight. In the long interim between the posing of the problem and the solution, Gauss’s mind could do no more than generate unconsciously and blindly ideas that led nowhere.

One last point must be made with regard to incubation: This period may facilitate solution finding by more than one means. Usually when people fail to solve a problem, their level of arousal increases—they experience excitement and frustration. Such heightened emotion tends to constrict the width of attention. In addition, higher arousal tends to make high-probability associations even more probable and low-probability associations less so. Given that the solution requires the ability to look at the problem in an original way, the individual must attain a more relaxed state to allow the low-probability associations a reasonable chance to emerge. Hence, during the incubation period arousal may be lowered enough to make the person more able to take advantage of the sometimes subtle cues offered by the surrounding environment. In other words, the low-arousal state is more conducive to the Darwinian process needed to arrive at an insightful solution.

Imagery

Although I believe the experimental research on insight lends support to a Darwinian view of the creative process, the case is somewhat weakened by the nature of the problems normally found in such research. The typical insight problem has a well-defined answer. In fact, the connection between the problem and the solution is quite logical once the “trick” is known. Real creativity, however, does not begin with the knowledge that there even exists a true answer, and thus the phenomenon is far more open-ended. Recent attempts to develop the “creative cognition approach” have increasingly recognized this contrast, leading many investigators to ask that subjects generate truly original ideas. From the perspective of this book, the most interesting of these studies is the experimental research on the role of imagery in the production of creative ideas.

An excellent example is the work associated with the “Geneplore” model recently advanced by Ron Finke, Tom Ward, and Steve Smith. Although not explicitly formulated in Darwinian terms, the connection with a variationselection framework is quite apparent. The very term Geneplore stands for “generate and explore.” The authors see the creative process as consisting of the generation of combinations, followed by the exploration of their possibilities. In other words, the first stage entails a variation process in which new arrangements of mental imagery are freely generated; the second stage involves a selection process in which certain novelties are chosen for further development.

In the experimental research on the geneplore model, the connection with a variation-selection process becomes even more obvious. Subjects were given shapes or forms—lines, circles, triangles, letters, spheres, cubes, rings, hooks, etc.—from which they had to create objects with recognizable functions (e.g., furniture, appliances, tools and utensils, weapons, or toys). The products of this inventive activity were then evaluated by judges for their creativity. Some of the inventions arrived at were truly ingenious, including a hip exerciser, a shoestring unlacer, and a hamburger maker. More interesting still was how the creativity of the inventions would depend on the experimental conditions. In some conditions subjects themselves could select the shapes for the imaginative constructions, whereas in other conditions the subjects were simply given a random selection of forms. In yet another manipulation subjects themselves could choose the category of object they had to invent or the category would be selected randomly by the experimenter. The researchers showed that subjects arrived at the most innovative solutions when both the object parts that they had to work with and the category of object they had to invent were randomly selected from the larger pool of possibilities. The best creativity tends to be serendipitous rather than deliberate. One could hardly obtain a more blind basis for launching the combinatory process. By beginning with the totally unexpected, the participants in these experiments were forced to stretch their creativity to the highest degree.

This conclusion receives endorsement from experiments that have an entirely different theoretical foundation. Earlier I mentioned the work of Rothenberg on homospatial and Janusian thinking. Although the bulk of his research on these forms of imagery depended on interviews and other nonexperimental data, Rothenberg has also examined these processes in the laboratory. Of special interest are his investigations of homospatial thought. He and a colleague began by making up a set of visual stimuli that involved the superimposition of visual images. The superimpositions included all sorts of unrelated subject matter. For example, one contained a photograph of an empty French four-poster bed placed in a period room superimposed over a group of soldiers in combat who were taking cover behind a tank. These highly incongruous homospatial images were then shown to writers and to artists, the latter including individuals selected in a national competition by faculty at the Yale School of Art. The writers were instructed to create new metaphors inspired by the stimuli, whereas the artists were instructed to make pastel drawings. In comparison with the control group (e.g., subjects who saw the images only separately), individuals exposed to these visual juxtapositions of unrelated images generated more creative products, as judged by independent raters.

These are provocative results, for they suggest that creative imagery is excited by sensory input that is random or at least unusual. In fact, artists often report engaging in activities that might stimulate their visual imagination after a similar fashion. For instance, the surrealist painter and sculptor Max Ernst described how he so used the artistic technique of frottage: “I was struck by the obsession that showed to my excited gaze the floor-boards upon which a thousand scrubbings had deepened the grooves.... [I]n order to aid my meditative and hallucinatory faculties, I made from the boards a series of drawings by placing on them, at random, sheets of paper which I undertook to rub with black lead. In gazing attentively at the drawings thus obtained ... I was surprised by the sudden intensification of my visionary capacities and by the hallucinatory succession of contradictory images superimposed, one upon the other.” In this passage Ernst even makes explicit reference to homospatial imagery.

Together the above experiments underline the creative utility of mental imagery, while at the same time showing how the creative imagination seems to operate according to Darwinian principles. Not only must the mind be capable of engaging in wild combinatory play, but in addition this play often appears to be encouraged by the random juxtapositions of visual stimuli. Such stimuli optimally elicit cognitive mutations.

Intuition

The notion that much of the business of creativity can be ascribed to intuition is very old. We have already quoted the opinions of Poincare and Hadamard on this subject. Many theorists have also placed considerable emphasis on the role the unconscious plays in the creative process. The most notable example is the attempts of Sigmund Freud and subsequent psychoanalysts to explain creativity in terms of primary-process thought. This mode of thinking is largely subconscious, highly primitive, even infantile, and rich in irrational associations. Indeed, the primary process was held responsible by psychoanalysts for the highly creative imagery and symbolism experienced in dreams. Although experimental research on intuitive processes has been going on for some time, it is perhaps only since the advent of modern cognitive psychology that genuine headway has been made. Certain of these findings suggest that unconscious mental processes may be a good source of ideational variants that would have a low probability of being produced by conscious mental processes. The following five conclusions deserve special attention:

• 1. Current research on implicit learning and memory suggests that the human mind can acquire a vast set of expectations in the absence of any awareness of the basis for those expectations. For example, experiments on artificial grammars show that individuals can judge the difference between permissible and impermissible symbol strings without being able to specify the corresponding grammatical rule. Psychologists have known for some time that classical and operant conditioning can sometimes occur in the absence of awareness, but results such as these indicate that unconscious expectations can be even more sophisticated.

• 2. Unconscious associations can greatly influence the course of thought in the absence of any conscious intervention. Indeed, the activation of a given part of a memory network can ramify to other portions of the network, activating associations that are sometimes only very remotely connected with the initial impetus. This unconscious process of “spreading activation” can play a major part in getting the mind to isolate components of a problem’s solution that might otherwise be missed by directed, conscious thought.

• 3. The unconscious material in the mind is often stored in multiple ways, including in manners that are unusual, if not illogical. For example, words are associated not only according to denotative meanings but also according to sound qualities and emotional connotations. Hence, in the tip-of-the-tongue phenomenon, individuals trying to retrieve a particular word will often extract from memory words that sound similar with respect to specific phonemes and the number of syllables. This permits the activation of associative material to follow several different paths simultaneously in a variety of “parallel processing.”

• 4. Experimental research indicates that unconscious mental operations may be particularly useful to the solution of problems that require creative insight. For instance, cognitive psychologists sometimes study problem-solving processes using “protocol analysis,” in which subjects are asked “to think aloud” while working on a given problem. Although this experimental instruction may not affect the performance of subjects working on more everyday problems, the imposition of such a condition does noticeably undermine performance when the subjects are trying to solve insight problems. Thinking aloud by necessity obliges the subject to rely exclusively on conscious mental processes in circumstances when unconscious processes are more likely to overcome the misperceptions and implicit constraints that obstruct finding the solution.

• 5. Unconscious mental processes can even support “feeling of knowing” states that are comparable to the unjustified “hunches” that are so often reported by creative individuals. Although these feelings are by no means infallible, and may often err, they can provide support for pursuing avenues that might be bypassed otherwise. In fact, experimental research has shown that even when hunches are incorrect, they often bear an associative link to the correct answer. These bad guesses show that a subject is “getting warmer” long before reaching the right solution to the problem. These results also indicate that the unconscious mind can generate ideational variants, thrusting into consciousness those that exhibit superior prospects of success.

The foregoing characteristics of intuition suggest a curious parallel between the emergence of creative ideas in the mind and the evolution of new life-forms in the organic world. When Darwin contemplated some of the possible objections to evolutionary theory, he realized that among the most crucial was the lack of well-documented lineages in the paleontological record. Instead, there were countless “missing links,” or what Darwin called “transitional varieties.” Even today, with the fossil evidence tremendously expanded, these gaps in the evolutionary record are the rule rather than the exception. In Origin of Species, Darwin dealt with this problem by arguing that these transitional species would probably be represented by very small populations. Moreover, such populations may not endure very long before facing extinction. After all, these transitional varieties would most likely be less well adapted than the forms that evolved from them. For example, species that can actually fly may quickly win the struggle for existence against species that can only glide or flutter. Hence, the paleontological record is dominated primarily by the successful varieties, only an occasional transitional form surviving to the present. This process is quite similar to that of the creative mind. Intuition is spontaneously generating all sorts of ideational combinations, but most of these are transitional ideas that fail to enter consciousness. Only a few ideational variants—either the solution itself or ideas that have some close associative connection—will emerge from the unconscious mind. As a consequence, the creator’s memory retains only the highlights of the sequence of mental events that led to the solution. Solely the consciously deliberated ideas will become a part of the mind’s “fossil record.” The result is numerous cognitive missing links. The creative insight then so often emerges de novo, like Minerva from the head of Zeus.

Finally, it is important to recognize that the unconscious mind is not a single unit but rather an often loosely connected collection of sometimes rather distinct processes. For instance, some strongly practiced skills have become so automatic that there is no longer any cognitive need to attend to their execution. Other unconscious processes are assigned to the recognition of faces, the mastery of language, and the acquisition of emotional meanings. Some of these processes are potentially accessible to consciousness, and others not. Moreover, some of these processes are carried out in the more advanced portions of the brain, whereas others are apparently located in more primitive parts of the nervous system. The significant point is that the workings of the conscious mind represent only a small fragment of the total activity carried on by the mind at a particular time. Consequently, the unconscious mind is probably more capable of generating the blind variations required by a Darwinian theory of creativity. That is not to say that all ideational variants have intuitive origins. I have already noted how unusual imagery may also provide a valuable source of intellectual recombinations. Yet intuition just might provide the single most potent resource for the creative genius.

Computer Simulations

Modern cognitive science has relied ever more on computers as a means of developing a theory of the mind. In fact, to a very large degree computer models of mental processes have replaced mathematical models as the preferred approach to building comprehensive and precise theory. For the most part, unfortunately, these computational models deal with rather basic cognitive processes that appear far removed from what are required to achieve creative thoughts. Nonetheless, there have been a number of attempts to model the operations that can produce original ideas. These can be grouped into two distinct categories. The first set of models are founded on principles that seem totally antithetical to a Darwinian position. The second set, in contrast, consists of models that are explicitly Darwinian.

Discovery Programs

Earlier I mentioned Poe’s claim that the composition of “The Raven” involved nothing more than the step-by-step application of a conscious, rigorous, and inevitable logic. Many cognitive scientists who study human problem-solving behavior would concur. Foremost among these advocates of the conscious and deliberate application of logic is Herbert Simon, a Nobel laureate. Simon has attempted to demystify the creative process by arguing that even the renowned achievements of so-called geniuses can easily be ascribed to far more basic mechanisms. For example, he has claimed that Mendeleyev’s origination of the Periodic Law of the Elements required nothing more advanced than what is “required to handle patterned letter sequences.” In fact, the bulk of Simon’s work has concentrated on scientific discovery, which Simon believes follows logical principles that may be applied by anyone. To help make his case, Simon has conducted some informal experiments in which naive subjects were presented with a problem whose solution won renown for some past scientist. For instance, Simon reported, “On eight occasions I have sat down at lunch with colleagues who are good applied mathematicians and said to them: ‘I have a problem that you can perhaps help me with. I have some very nice data that can be fitted very accurately for large values of the independent variable by an exponential function, but for small values they fit a linear function accurately. Can you suggest a smooth function that will give me a good fit through the whole range?’” Of the eight lunch companions, five found an answer within a couple of minutes or less. None suspected what Simon was up to, nor did any realize the historic nature of the problem given them. Still, those five anonymous individuals had independently arrived at Planck’s formula for blackbody radiation. In another mini-experiment a mere graduate student in chemical engineering was able to derive the Balmer formula for the hydrogen spectrum. Moreover, this subject was asked to think aloud while solving the problem, and therefore protocol analysis could be applied to learn the search processes that led to the discovery. The thought processes were comparable to those revealed in Balmer’s surviving documents. And

~ Cognition ~                             51

those processes seemed to involve nothing more than straightforward logical reasoning.

These informal results have been confirmed in a more formal laboratory experiment. Subjects drawn mostly from a student population were presented raw data for five cases on two variables, s and q. The specific scores were: 36 and 88, 67.25 and 224.7,93 and 365.3,141.75 and 687, and 483.8 and 4,332.1, respectively. The subjects were told that the experimenters were “interested in how a human being discovers a scientific law.” Accordingly, each subject was to identify a precise functional relationship between s and q. Four out of 14 subjects tested found the correct relationship. Yet what these successful problem solvers accomplished was a rediscovery of the third law of planetary motion first formulated by Johannes Kepler! In fact, the subjects wrestled with data for all practical purposes the same as those used by Kepler. The five sets of observations regard the planets Mercury, Venus, Earth, Mars, and Jupiter. Here s is the distance from the sun in millions of miles and q is the period of revolution in earth days. According to Kepler’s third law, the distance cubed is proportional to the period squared, or s³ = kq~, where k is a constant. Furthermore, the protocols indicated that the successful subjects used the same methods found in other problemsolving experiments. For instance, to find the relationship between two variables one needs to find functions for each of the variables that produce a constant ratio. In this case, one must find a function jj of the distance and another function/^ of the period such that/j ($)lf₂(q) returns about the same quotient k across all observations.

The computer models. The supremely logical nature of these thought processes implies that a computer might be programmed to duplicate the discoveries of illustrious scientists. Herbert Simon and his colleagues have attempted to do just that, writing a large number of “discovery programs.” These programs are often christened with names of famous creative individuals, such as OCCAM, BACON, GALILEO, GLAUBER, STAHL, FAHRENHEIT, BLACK, and DALTON. And these tags are not wholly incidental. For example, BACON specializes in the inductive method, yielding data-driven discoveries as advocated in Francis Bacon’s Novum Organum; GLAUBER, of Glauber’s salt fame, concentrates on discoveries in chemistry.

But even more important, these programs have shown the ability to rediscover scientific laws or principles that have made human scientists famous. GLAUBER can learn to distinguish acid and alkali. STAHL can identify elements as components of substances. DALTON can take the output from STAHL to generate structural formulas consistent with atomic theory. Yet it is BACON that is the powerhouse of discovery programs. BACON has discovered Kepler’s third law of planetary motion, Black’s law of temperature equilibrium, Ohm’s law of current and resistance, Prout’s

hypothesis of atomic structure, the Gay-Lussac law of gaseous reaction, the Dulong-Petit law of atomic heats, and the derivation of atomic weights by Avogadro and Cannizzaro. Some programs even duplicate some of the fine details of the process by which a scientist made a given discovery. For instance, KEKADA models the heuristics that Hans Krebs used to get the urea cycle. The programmers fixed this concordance by comparing the computer’s output with both the notebooks and the living testimony of Krebs himself.

The criticisms. If these discovery programs accurately represent the creative process, then the Darwinian model advanced by Donald Campbell would seem most implausible. Nonetheless, several critics have pointed out limitations to these programs that render them less convincing as models. To begin with, the programs do not always re-create the cognitive processes of creative scientists. For example, one critic compared the logical operations presumed in the programs with the processes revealed in the extensive notebooks left by the physicist Michael Faraday. Faraday’s dynamic use of visual imagery had no counterpart in these computer models. Although some progress has been made in incorporating imagery into discovery programs, these innovations have a long way to go before they can accurately simulate the imaginative experiences reported by creative individuals. In addition, the laboratory notebooks of Faraday, like those of many other notable creators, exhibit considerable cross-talk between separate projects. Ideas or themes that originally appeared to help solve one problem have a way of stimulating developments with other problems. An instance cited earlier is Poincare’s seaside discovery of the link between indeterminate ternary quadratic forms and non-Euclidean geometry. Yet such ideational interchanges have no counterpart in any of the discovery programs, all of which are designed to work on only one project from start to finish.

Furthermore, unlike scientists in the real world, discovery programs must always be given the problem to solve rather than discovering the problems for themselves. And some have argued that problem finding may be a skill every bit as important as problem solving. As Einstein noted, “The formulation of a problem is often more essential than its solution, which may be merely a matter of mathematical or experimental skill. To raise new questions, new problems, to regard old problems from a new angle, requires creative imagination and marks real advances in science.” Einstein practiced what he preached. His contemporary physicists were not really bothered about inconsistencies between Newtonian mechanics and Maxwell’s equations for electromagnetism. Not only was Einstein perplexed by the discrepancies, but he discovered that to solve the problem he himself had invented required him to devise a revolutionary new physics, the theory of special relativity. Hence, so long as these discovery programs only tackle problems given to them by human beings, we can only suspect their capacity to provide a comprehensive explanation for scientific creativity.

Even worse, these discovery programs can only work by extensively simplifying the problems they are given. This simplification occurs at both input and output ends. At the input end, the programs are fed the data in a highly abstract, even sanitized form. Real scientists, in contrast, often operate with a much more complex array of information that can vary tremendously in precision and relevance. As a consequence, scientific discovery normally requires a more haphazard examination of various possibilities, and only far later will it become more obvious what data are essential and what are superfluous. At the output end, moreover, the discovery programs begin with a well-defined set of criteria of what counts as a successful solution. Scientific creativity in the actual world, in contrast, may have to contend with considerably more ambiguity on this score. Not only must the scientist assess the solution’s adequacy on the basis of personal criteria— including more nebulous aesthetic factors—but in addition the scientist will have to convince other members of the scientific community who won’t be applying the same criteria to the solution. A good illustration concerns the discovery that the elements exhibit periodicities in their chemical and physical properties. Although Herbert Simon claimed that this was an easy discovery to make, that claim can be made only if we ignore the context in which that discovery had to be evaluated. Shortly before Mendeleyev offered his solution, John Newlands presented his comparable law of octaves at a professional meeting. A distinguished scientist chairing the session waxed sarcastic, asking whether Newlands had also tried placing the elements in alphabetical order. Clearly, these two scientists disagreed dramatically on what qualified as a scientific discovery. So far, discovery problems cannot implement a complex search through the data space, and even less can they display flexibility in the application of criteria regarding valid output.

Besides a certain rigidity in data input and theory output, discovery programs are rather fixed in their “throughput.” These rule-driven programs successfully solve given problems by applying a predetermined set of heuristic principles to the data provided. Trial-and-error procedures are minimized by assigning an a priori ordering to the operations that will be applied. When BACON rediscovered Kepler’s third law, for example, it began with the linear relationship between the two variables. Since that failed the criterion of a proportional relationship, one of the variables was squared and the empirical test again applied. This iterative process continued until the cube of one variable was found to be proportional to the square of the other. Although in a certain sense this sequence of transformations and tests involved some trial and error, there was too much wisdom underlying the procedure to make it count as truly blind. Rather than randomly trying out all possible mathematical relationships, the hypothesized functions proceeded from the simple to the complex by successive powers.

The rigidity of this information-processing procedure raises two related issues. First, the hierarchical ordering of the data manipulations was obviously installed after the fact, and therefore one must wonder whether the programs were specifically tailored to fit the problems in a post hoc manner. For example, Hermann von Helmholtz, the illustrious physiologist and physicist, once admitted

that I had only succeeded in solving such problems after many devious ways, by the gradually increasing generalisation of favourable examples, and by a series of fortunate guesses. I had to compare myself with an Alpine climber, who, not knowing the way, ascends slowly and with toil, and is often compelled to retrace his steps because his progress is stopped; sometimes by reasoning, and sometimes by accident, he hits upon traces of a fresh path, which again leads him a little further; and finally, when he has reached the goal, he finds to his annoyance a royal road on which he might have ridden up if he had been clever enough to find the right starting-point at the outset. In my memoirs I have, of course, not given the reader an account of my wanderings, but I have described the beaten path on which he can now reach the summit without trouble.

To what extent do these programs merely reflect the advantage of hindsight, taking the preordained “royal path” so as to bypass the trial and error inherent in the genuine creative process?

Second, the kinds of heuristic searches that work well for one problem will not work well for a completely different problem. It is for this reason that there does not exist a single discovery program, but rather many discovery programs, each containing a subset of heuristic principles designed to solve a specific class of problems. Because real scientists are confronted with multiple problems, they must possess a diversity of problem-solving techniques, without knowing which particular trick or set of tricks will get them where they want to go. Indeed, not only must scientists select among alternative heuristic principles, they also must choose among alternative problems, trying to guess which problems are most likely to be solved with the tools at hand.

It should not be surprising, therefore, that these programs have so far been limited to making mere rediscoveries. Hence, they should be called “rediscovery programs.” So let us ask: What would it take to devise discovery programs that are truly worthy of the name? These programs would have to increase the diversity of mental modalities and operations available, including adding various kinds of imagery and free association to the logical operators already incorporated in the computer models. The programs would have to be able to work on several projects simultaneously, permitting the free exchange of ideas across problems. The programs would have to feature the capacity not only to select among problems at any moment but even to discover new problems. The programs would have to deal with much more messy data sets in which it is not always clear what is useful and what not. At the same time, the programs would need the ability to consider a wider diversity of criteria by which potential discoveries are judged, and allow these criteria to change over time according to feedback from other projects or the environment. Moreover, the programs would contain a huge set of heuristic principles, with only the vaguest information on which principles will best work for the various problems in queue.

I would argue that such a bona fide discovery program, with so many choices at its disposal at any particular moment, would have to depend very heavily on trial and error in maneuvering through the intricately interconnected problem spaces. Moreover, the greater the novelty and complexity of the problems under consideration, the stronger the role of that dependence. To a large degree, the highly creative scientist may have little option other than to sample blindly modalities, processes, operations, problems, criteria, heuristics, and the other components of the rich phenomenon of scientific discovery. In short, I believe that when genuine discovery programs appear, they will in all likelihood operate according to a Darwinian view of the creative mind.

Genetic Algorithms

Perhaps the discovery programs were doomed to fail because they operate under a computer model of the brain that is fundamentally at odds with the reality. These programs work by a sequential, step-by-step logic defined by a priori decision trees that specify the if-then statements to which there can be provided discrete yes-no answers. Of course, this is the modus operandi of the conventional digital computer, founded on Boolean algebra, which has made computers the only machines comparable to the human brain in terms of information-processing power. But that success does not mean that these traditional simulations offer an adequate representation of the central nervous system. When a computer system (BIG BLUE) finally beat the world chess champion (Gary Kasparov), it did not do so by emulating the operations of the chess genius. On the contrary, the human brain was simply overwhelmed by the “brute force” capabilities enjoyed by the phenomenally fast machine. The mind was not outthought but rather outcalculated.

Fortunately, recent advances in computer models of the mind have taken an entirely different approach. In particular, these new “connectionist” and “neural network” models operate in a far more diffuse and probabilistic manner. Such modern approaches, moreover, are more likely to function by Darwinian principles. From a multitude of potential interconnections are selected that subset that best satisfy some external criterion of performance. Although at present such models are designed to simulate relatively simple mental processes, some psychologists have already suggested that these models hold far more promise in the eventual simulation of the complex process of creativity. In particular, Colin Martindale has delineated the intimate conceptual relationships among (a) Campbell’s blind-variation-and-selective-retention model of creativity, (b) connectionist, neural-network, and spin-glass models (in solid-state physics), (c) such creativity mechanisms as remote associations and primary process, and (d) individual differences in associative gradients, defocused attention, and arousal level. In essence, Martindale has sketched the route by which connectionist models may eventually be able to simulate creativity in the fashion that actually takes place in the human mind.

At present, however, the more fruitful approach has been to design explicitly Darwinian computer programs. These can then be used to test whether blind-variation-and-selective-retention models can indeed support the evolution of complex, organized structures from simple elements. For example, researchers have constructed computer models of organic evolution by natural selection. In these simulations the characteristics of a hypothetical organism are specified by strings of computer code. These genetic strands are placed in an electronic “soup” and given the opportunity to reproduce their kind, often with the addition of sporadic random mutations. Quickly a struggle for existence ensues that permits the influence of selection pressures. Accordingly, over time the creatures evolve—yielding progeny with new genetic instructions. Even more fascinating, novel evolutionary phenomena will often appear. For instance, certain mutants may develop into parasitic organisms, which will then display population oscillations with their hosts that are indistinguishable from those observed in the natural world. In addition, eventually there may emerge more complex processes, such as sexual reproduction, sexual selection, and kinship selection. These computer simulations of organic evolution provide proof that the Darwinian process can indeed provide an explanation for complex adaptation.

It is a short step from these simulations of primary Darwinism to applications of secondary Darwinism in action. The pioneer in this development was John Holland, who received the first American Ph.D. in computer science. Holland was inspired by R. A. Fisher’s 1930 book The Genetical Theory of Natural Selection, the landmark contribution to the development of a mathematical theory of evolution. The result was the problem-solving strategy of the computer-driven genetic algorithm. Genetic-algorithm programs start with a population of randomly generated strings of ones and zeroes, such as 011010. These binary strings function the same way as a genetic code. That is, the strings define the traits of some entity or system, in this case a potential solution to a given problem. The trial solutions represented by this random collection of binary strings can then be tested to determine which come closest to solving the problem. The least successful strings are then deleted, while the most successful are allowed to reproduce. This reproduction occurs sexually. Specifically, each strand pairs off with another strand, and then each exchanges a portion of its strand with its mate (the exact point of crossover being itself randomly determined). For example, suppose that 001001 and 110100 represent two strands that survived the first round of selection. They may mate by splitting after the fourth bit, the first part of one strand joining with the second part of the other strand. The outcome would be 001000 and 110101. Furthermore, genetic mutations can be introduced by randomly changing one or two bits on a subset of strands. Once this new generation is produced, the corresponding trial solutions can again be tested against the criterion. The whole Darwinian process may repeat, cycle after cycle, until the criterion of success is fully attained.

Despite the utterly blind procedures for producing variations, genetic algorithms have proven to be quite effective problem solvers. They can now already solve real-world problems, such as planning fiber-optic telecommunication networks, designing gas and steam turbines, enhancing the efficiency of jet engines, making forecasts in currency trading, and improving oil exploration and mining operations. Moreover, the basic Darwinian approach could be taken a step further to generate much more complex problem-solving systems. One of Holland’s students, John Koza, extended genetic algorithms to a higher level of creativity by conceiving, the procedure of genetic programming. In this technique, whole components of programs are subjected to blind variation, and thus the Darwinian process operates at a higher level of structure. This approach has worked well with such problems as designing electrical circuits, solving algebraic equations, determining animal foraging behavior, and finding optimal game-playing strategies.

It is interesting to note that these Darwinian programs can even make rediscoveries, just as claimed by the discovery programs discussed earlier. For example, genetic programming also managed to arrive at Kepler’s third law of planetary motion. In fact, during the course of the evolution of a solution process, the program first came across a less accurate statement of the relationship between planetary distance and time of revolution—the same solution Kepler himself had published a decade before arriving at the more precise law! Genetic algorithms have even re-created lines from William Shakespeare in only a few dozen variation-selection cycles. For instance, Hamlet’s “Methinks it is like a weasel” will appear in 10 to 50 generations. In contrast, the probability of the line being produced randomly— by a monkey typing madly away at a typewriter—is 28'²⁷, a prohibitively minuscule likelihood. The Darwinian process may be blind, but its selectionist feature renders it cumulative, making it far superior to pure chance.

Nonetheless, these acts of creative duplication are less important than the fact that these techniques have generated creative ideas of their own. Besides solving difficult problems in science, engineering, mathematics, and economics, Darwinian methods have even created original art and composed bebop jazz melodies. Although evolutionary programs are a relatively recent development, they hold great future promise for allowing machines to come close to matching the creativity of humans. Moreover, unlike discovery programs, these programs have very little intelligence built into them. They are merely endowed with a method of blindly generating variations and a criterion for selecting the best variations to add to the next evolutionary cycle. The discovery programs can probably know no more than their programmers, whereas genetic algorithms can acquire an expertise that exceeds that of their programmers.

Admittedly, unlike the discovery programs, genetic algorithms and their descendants were not specifically designed to simulate the creative process of the human mind. Their roots were directly in primary Darwinism, not in the secondary Darwinism represented by Campbell’s blind-variation-and-selective-retention model. Even so, it seems that these programs inadvertently come very close to reflecting the actual processes reported by creative individuals. For example, earlier I quoted William James as observing how creative minds do not have “thoughts of concrete things patiently following one another in a beaten track of habitual suggestion.” On the contrary, their intellects exhibit “the most abrupt cross-cuts and transitions from one idea to another” as well as “the most unheard of combination of elements, the subtlest associations of analogy.” In short, the creative mind displays “a seething cauldron of ideas, where everything is fizzling and bobbling about in a state of bewildering activity, where partnerships can be joined or loosened in an instant.” This graphic description comes remarkably close to what takes place during the sexual reproduction of binary strings that occurs in genetic algorithms. These strings break apart at random places and recombine with the fragments of other strings, generating thereby novel genetic instructions. By the same token, the creative mind may contain a chaotic soup consisting of strings of vague thoughts and images.

These ideational strings, too, may split up at unexpected places and recombine until a new, more adaptive configuration appears. A great part of this combinatory play is involuntary and unguided, and hence, to a large degree, blind.

The operation of these genetic algorithms can also be compared with the research done on the creative imagination. Particularly provocative are comparisons with Rothenberg’s work on Janusian and homospatial processes. For instance, in the experiments on homospatial thinking, the production of creative drawings was enhanced by viewing two unrelated images superimposed. The resulting drawings tended to combine and integrate forms that emerged from the juxtaposition of the shapes. The outcome is artistic creations that look like visual analogs of the crossover occuring both in the binary strands of genetic algorithms and in the chromosomes of organic evolution.

Hence, despite the fact that genetic algorithms were designed to manifest Darwinian creativity rather than simulate human creativity, they may actually provide a crude model of the latter process. To the extent that the creative mind engages in the spontaneous splitting and recombining of chains of thought, we can say that the intellect is operating according to an analogous mechanism. Indeed, it may even be fruitful to apply some of the findings from these computer programs to help us better appreciate some of the finer details of the creative process. For example, the research on genetic algorithms has found that recombination plays a far more important role than mutation. That is, randomly flipping bits on strings contributes relatively little to reaching a solution in comparison to the crossover of two “mating” binary strings to produce two new “offspring” strings. The advantage of the recombinatorial procedure is especially important, moreover, when selection has had the opportunity to operate for several generations. Even so, sometimes a particular combination of bits that disappeared early in the evolution of the strands will need to reappear if the optimal solution is to emerge. Yet the extinct binary sequence may not easily appear through recombination of the surviving sequences. In this case, mutation provides a vehicle for reintroducing into the population the required combinations. It is conceivable that creativity in human beings functions in an analogous fashion. Most creative ideas may emerge from simple recombinations of material of previously established utility. Only when the creator is stumped on a truly difficult problem will it be necessary to depend on some ideational mutation that can propel the mind toward the solution. This “shot out of the blue” may require a totally serendipitous event.

Before advancing to a general assessment of Campbell’s model of creativity, I wish to mention one recent development that has introduced a fascinating twist to Darwinian computation strategies. Genetic algorithms simulate the variation processes of primary Darwinism, but the simulation proceeds electronically rather than biochemically. That is, the program operates on binary computer codes rather than DNA strands. But now that we have the biochemical wherewithal to manipulate DNA, it would seem feasible to use DNA to generate the blind variations. The result would be a molecular computation technique that could be used to solve conceptual and mathematical problems. Indeed, because a test tube full of DNA can encode huge amounts of information and simultaneously process that information, molecular computation should originate rapid answers to the most difficult questions. This alternative Darwinian strategy was actually implemented in the mid-1990s by Leonard Adleman, a computer scientist at the University of Southern California. By substituting biochemistry (DNA) for electronics (silicon), this approach can quickly provide solutions to otherwise computationally intractable puzzles (e.g., the classic “traveling salesman problem”). This novel development in computer science represents another way that the distinction between primary and secondary Darwinism sometimes becomes blurred. Here the very molecular substance that provides the variational material for all organic evolution has been coopted as the basis for creative problem solving by an equally blind varia-tion-and-selection procedure.

Evaluations

Judging from the foregoing discussion, Campbell’s blind-variation-and-selective-retention model seems to have considerable merit. The model appears consistent with what geniuses themselves tell us about the creative process, and it seems compatible with experimental research on insight, imagery, and intuition. Although this model is inconsistent with the underlying assumptions of discovery programs, these programs may have to acquire a Darwinian basis if they are ever to replicate creative behavior in the real world. More important, current research on genetic algorithms and genetic programming shows that computer systems based on Darwinian principles have the expected capacity for creativity. Of course, none of this can be taken as conclusive proof that Campbell’s model is correct. At this point, it is probably safer to offer the more modest conclusion that at least the model has yet to be proven utterly implausible. Additional experiments and simulations are probably necessary before more can be claimed.

Nonetheless, I should end this evaluation of the Darwinian model by addressing two related issues. The first concerns a matter of secondary Darwinism: Do the blind variations that figure so prominently in organic evolution really have their counterpart in the emergence of creative ideas? Or is this just another one of those disanalogies that so often plague secondary theories? The second question involves a matter of primary Darwinism: Assuming that Campbell’s model is correct, how would organic evolution produce a brain capable of creativity?

Variational Blindness

Ever since the advent of Darwin’s theory of natural selection, the concept of “blindness” has provoked controversy. Many critics could not tolerate the notion that the spontaneous variation that fed the evolutionary process was unguided by some intelligence. This criticism was especially prominent among those who believed that an all-wise and all-powerful Creator had provided a guiding hand behind the cornucopia of life-forms that populate the planet. Yet even many notable scientists had difficulty accepting the idea that the sophisticated adaptations seen in the world could have been produced by a chance mechanism. This disbelief did not vanish with the advent of the “modern synthesis,” despite the provision of specific processes— genetic recombination and mutation—that would support the production of blind variations. Even a century after the publication of the Origin of Species, a significant number of scientists argue that the variational procedure displays more insight. The most recent example is the research on whether certain bacteria exhibit “directed mutations,” in which the resulting mutants are biased toward successful adaptations. Although the consensus view is that this phenomenon probably does not occur, a minority still maintains that organic evolution need not be as blind as stated by Darwinian theory.

It goes without saying that if primary Darwinism must contend with this issue, secondary Darwinism is even more vulnerable to the same objection. Few psychologists would deny that creativity involves some type of variation process. Even the discovery programs function by a generate-and-test procedure. The question of debate, rather, is about the degree of blindness underlying the production of the ideational variations. The very concept of blindness seems to run counter to the fundamental nature of the human being as a goal-directed organism. And as humans go, the lives of creative geniuses overflow with plans, purposes, and aspirations. They are individuals obsessed with a mission, with a destiny, devoting their whole careers to the creation of marvelous works in their chosen domains. Moreover, the commonplace conception of genius seems to be antithetical to the notion of someone who works by trial and error. Genius holds the connotation of intelligence, and haphazard searches seem the very opposite of intelligent behavior. Indeed, someone who insisted on generating utterly random thoughts might even be thought to suffer from a maladaptive psychopathology.

Yet objections such as these only exhibit a misunderstanding of the Darwinian theory of creativity. To see how, let us begin with the question of the creator’s reason, and then turn to the matter of the creator’s will.

Creative rationality. It cannot be overly stressed that logic and reason must play a critical part in any Darwinian model. Often the criteria by which ideational variations are selected are logical and rational, especially in the sciences. Even in the arts, logic and reason may impose constraints on what variants are chosen for further development. A plot in a novel, a character in a play, a figure in a painting, or an architectural plan are often governed by rational rules about what is plausible or workable. Thus, the selection portion of the variation-selection process is usually far from blind. The real question, then, is the blindness of the variational procedure.

It would seem that a Darwinian model requires that ideational variations would be strictly blind in the sense of being completely random or unpredictable. Yet such an extreme claim would not even apply to organic evolution. Of the various sources of genetic variability in the external world, only mutation can be considered totally blind in the strictest sense. Indeed, it is for this very reason that mutations are far more likely to be deleterious than adaptive. Genetic recombination, on the other hand, is far less blind. After all, the genes that enter into the combinatory process are mostly those that have proven their adaptive utility in previous generations. The only major exception to this rule is recessive genes that carry certain maladaptive traits, and even these will be common in the gene pool only when they confer some advantage under certain circumstances (e.g., sickle-cell anemia in malaria-plagued environments). In addition, the genes do not always undergo completely independent assortment in a classic Mendelian fashion, for those contained on the same chromosome will be subject to the phenomenon of linkage. The closer the two genes are on a given chromosome, the higher the probability that they will be inherited together. Moreover, because there may exist selection pressure in favor of maintaining certain combinations of genes intact, some of these constraints on pure random recombination may reflect the acquired wisdom of selection having operated in previous generations. In population genetics, this constraining outcome is known as “linkage disequilibrium.”

Thus, blindness should not be viewed in all-or-nothing terms; but rather as a characteristic that admits degrees ranging from total randomness (mutation) to complete a priori constraint (linkage), with many gradations between (the probability of crossover). The same continuity applies to the term chance, which is often taken as being equivalent to blindness in many Darwinian models. There exist degrees of “chanciness.” If you stand right in front of a dart board, you can push the darts into the bull’s-eye with a 100% chance of success. If you step far enough back that you must actually throw the darts, the probability may shrink to 98%, only an occasional slip leading a projectile astray. But each successive step leads to a reduced odds of a hit. There will be a distance at which the probability will be only 50%, another at which it will reduce to 10%, and so on, approaching the point where a bull’s-eye would be a matter of pure luck, and yet another point where you would say that a hit is absolutely impossible, or 0% with certainty (namely, the target lies outside your maximum throwing range). The blindness of creative variations has the same underlying continuity. This continuity will become more evident when I scrutinize three radically different theories of creativity—the behavioral, the psychoanalytic, and the cognitive.

• 1. Let us return to Epstein’s generativity theory. When a hungry pigeon first enters the experimental chamber, it scans the environment, seeking the stimuli that inform it about the operants most likely to activate the release of a food pellet. If it sees a box beneath a toy banana, it goes to the box, climbs it, pecks at the banana, and gets the reward; if it sees a box over to the side and a target somewhere on the floor, it pushes the box toward the target to receive the expected reinforcement. The odds of a bull’s-eye in either case is 100%. However, if we alter the conditions ever so slightly—start modifying the shape, color, locations, or other features of the box, the target, the banana, or other discriminative stimuli—the pigeon will experience more difficulties. Not only will the likelihood of the appropriate operant be a function of some generalization gradient for the discriminative stimuli, but in addition alternative operants learned under comparable circumstances will begin to have higher probabilities of activation. Indeed, if the operant with the highest probability does not produce the anticipated effect, the pigeon has no other choice but to start descending down a hierarchy of diverse operants. Moreover, if none of these pass the test, the pigeon must begin trying out operants in various combinations in a more or less haphazard fashion. Eventually, one combination does the trick—pushing the box under the banana as if the latter were the floor target—and the pigeon experiences a moment of creative insight. The important point here is that no abrupt switch takes place from the rational application of old behaviors to the blind search for some new behaviors. Rather, there occurs a roughly continuous descent into ever more blind behavioral variations.

• 2. Freud introduced the distinction between primary and secondary thought processes. Secondary process is that of the conscious mind. It operates according to the “reality principle,” and it is therefore rational, pragmatic, logical. Consequently, according to psychoanalytic theory, the creative imagination must have its locus elsewhere in the mind. That locus, as noted earlier, is in the primary process, which is far more primitive, even infantile. Ruled by fantasy, and by unconscious impulses (the “pleasure principle”), this part of the mind lacks the social and intellectual constraints of secondary process. Being more irrational, the associative material produced is much more blind—wild, unpredictable, juxtapositional, incongruous.

Therefore, it is not surprising that psychoanalytic theory has often discussed creativity in terms of “regression in service of the ego.” The creative individual must leave secondary-process thought to descend into primaryprocess thought in order to arrive at truly original ideational variants. Nonetheless, we must take care not to see this regressive process as a discrete all-or-none affair. There are degrees of regression into primary process, a point adequately demonstrated by Colin Martindale in a long series of experimental and content-analytical investigations (see chapter 5). Dreams provide a good example of the range. Sometimes we have dreams so realistic that when we wake up we find it hard to believe that the events did not really happen. Other times our dreams consist of such bizarre ideas, such strange juxtapositions of people and circumstances, that they seem to make no sense whatsoever (to the secondary-process mind).

The fact that regression exhibits degrees is important. To a great extent, successful creativity entails finding the right level of primary-process imagery. Too much, and the outcome is highly original but useless variants, such as the crazy thoughts of psychotics or people under the influence of hallucinogenic drugs. Too little, and the result is adaptive, but lacking in originality, as in everyday thinking. Somewhere in the middle is the happy medium in which the product of originality and adaptiveness is maximized, thereby optimizing the magnitude of creativity. Thus, the regressive process must descend to the proper level of associative blindness.

Epstein has pointed out an interesting connection between psychoanalytic regression and the principle of resurgence in his generativity theory. The resurgence of older, once-adaptive operants will essentially represent a process of behavioral regression. As Epstein put it, “Freud’s concept of regression could be considered a special case of resurgence in which the behavior that recurs is infantile.” “If you are turning a doorknob that has always turned easily, for example, and it fails to turn, any and perhaps all of the behaviors that have ever gotten you through doors are likely to appear: You may turn harder, pull up on the knob, kick the door, shout for help, and so on.” Depending on how desperate you are, this regression may descend yet more. You might begin to swear, utter “open sesame,” whimper, fantasize about being rescued, throw a tantrum, or engage in some other more childish activity. I think this conceptual link is significant, for according to the behavioral interpretation, the creator descends to a deeper level of primary process only after the more shallow levels of primary process have failed to solve the problem. In other words, so long as a solution is not found, the variants must become ever more blind with time. At the ultimate end point are the cognitions of psychotics, whose consistent failure to resolve life’s problems has backed them into a corner of absurd beliefs and behaviors.

• 3. The research on human problem solving, such as that represented by Herbert Simon and his associates, also supports the conclusion that blindness has degrees rather than yes-no attributes. If asked to multiply 216 by 37 (without a calculator), you apply the arithmetic method learned in elementary school to get the answer. The solution is thus acquired by a straightforward algorithm that requires the intrusion of no blind guesswork. More difficult problems, in contrast, may require a heuristic search through an array of possible solutions. For instance, one of the difficulties faced by students learning integral calculus is that the antiderivative of a function cannot always be found by an algorithmic solution. There exist a great diversity of integration tools, not all of which will succeed. So the student must acquire some “rules of thumb” that will restrict the range of possibilities to that smaller subset that have the best chance at helping to arrive at a solution. These rules provide hints about which equations are best integrated by substitution, which by parts, and so forth. Even then, a certain amount of trial and error may be necessary before the function is successfully integrated. Matters get ever more complicated when the student begins to study differential equations. Not only is the range of integration techniques vastly enlarged, but in addition the student learns that some differential equations may not even feature definite solutions. The heuristics become far richer, trial and error more prominent, and the blindness of the search far more conspicuous. Indeed, in the mathematical sciences a researcher will often devote years of effort in trying to solve a single differential equation, and frequently never succeed in doing so. Newton, for example, was never able to solve the differential equations necessary to describe the orbit of the moon around the earth (i.e., the “three-body problem”)—despite his being one of the greatest mathematical geniuses of all time.

The above illustration comes from mathematics, where the role of rationality would seem paramount. Needless to say, searching for solutions becomes ever more blind when dealing with forms of creativity that lack precise criteria for evaluating successful solutions. Contrary to Poe’s claims, writing a poem cannot proceed with the logic of a mathematical proof. The poet must acknowledge that the factors contributing to a poem’s aesthetic success are far too complex, vague, and transient to permit composition to be reduced to heuristic principles, even less algorithms. The same holds for other forms of creativity, such as music and art. Even the writing of a successful scientific journal article demands the satisfaction of a myriad independent and constantly varying requirements. Moreover, the more original, daring, revolutionary, or breakthrough the creative endeavor, the less guidance can be expected from the application of logical principles. The number of potential heuristics increases dramatically, and new rules of thumb may have to be tried and tested. Even the creator’s conception of the problem must undergo a series of unanticipated transformations before settling on the representation that supports a viable solution. In brief, as the amount of creativity required increases, the blindness of the search procedures will proportionally increase.

Note that this continuity can also be cast in the language of Epstein’s gen-erativity theory. When no algorithm exists that will immediately solve a given problem, the creator must fall back on one or more heuristic principles that seem to have worked well for similar problems in the past. If these fail, the criteria must be relaxed to allow the resurgence of heuristics ever more remotely related to the intransigent problem. Furthermore, eventually the individual will have to resort to the unguided recombination of the diverse techniques and approaches acquired over a considerable range of situations. This progression, again, entails a descent into an increasingly Darwinian form of creativity. Because the blindness can vary continuously as a function of problem difficulty, we should not really speak of creativity as Darwinian or not Darwinian. Instead, it is more accurate to speak in terms of the relative importance of Darwinian processes in the origination of a creative solution to a specific problem.

This notion that Darwinian blindness represents a continuous dimension rather than a discrete quality is absolutely essential to our understanding of creative genius. As will become apparent in later chapters, different creative domains, as well as distinct forms of creativity within a given domain, contrast greatly in the extent to which cognitive variations must be blind. Sometimes blindness plays a major role, and other times its involvement is minimal. Nevertheless, according to a Darwinian perspective, no supreme genius can operate without at least some variational blindness during critical points in his or her career.

Creative volition. Long before the advent of Darwin’s Origin of Species, his grandfather Erasmus Darwin had offered a theory of evolution in which purpose played a major role. Each life-form on this planet is driven by a “lust, hunger, and a desire for security,” he argued. As a consequence, by some mysterious process each creature would willfully modify its organs and behaviors to enhance its adaptation to the ever-changing surroundings. A bit later, but still a half century before Origin of Species, Jean-Baptiste Lamarck elaborated a somewhat similar theory, one that has earned its own eponym. Strictly speaking, Lamarckism makes two rather restricted claims. First, the organs that make up any individual organism can undergo changes in size, shape, or other features according to the amount of use or disuse received during that individual’s life. Second, according to the doctrine of acquired characters, the adaptations thus obtained during the life of the parents would be passed down to the offspring. However, through some misunderstanding, Lamarckism is often incorrectly taken to encompass a purposive component as well. According to this misconception, the giraffe got a long neck because it wished for just such an improvement to reach the leaves growing higher in the trees.

On the basis of this misunderstanding, the term Lamarckian is sometimes applied to creativity as well. Creative people are actively engaged in creating new ideas, and creativity is a purposive activity that assigns meaning to their lives. For example, it was certainly Charles Darwin’s lifetime goal to reach a deeper understanding of the biological world. He did not flutter randomly from topic to topic; instead his entire career following the Beagle voyage exhibited a developmental progression. Darwin’s various projects are interwoven as part of his grand quest. And this ever-evolving tapestry is embroidered throughout with certain persistent metaphors that Darwin used to make sense of the natural world (e.g., the evolution of species as a branching tree). In short, the evolution of ideas in Darwin’s brain exudes purpose in a manner seemingly “Lamarckian” (in the loose sense) rather than Darwinian (in the secondary sense). However, such an assertion, besides misapplying Lamarck’s name, fails to comprehend the potent but circumscribed function of volition in a Darwinian theory.

On the one hand, it is doubtful that anyone can be outstandingly creative without really wanting to be so. Volition permeates many aspects of the phenomenon. For instance, seldom does a person crack a difficult problem without first spending considerable time preparing the groundwork. Not only must the individual be willing to devote years of effort in the acquisition of the necessary knowledge and skills, but also the creator must have struggled with the particular problem, or at least with a closely related riddle. Even serendipitous discoveries are not bestowed on creators indiscriminately. On the contrary, the foundation usually must first be laid by a person who has made a considerable intellectual and emotional commitment to a relevant domain. In the often-quoted words of Pasteur, “chance favours only the prepared mind.”

More important, perhaps, is the fact that creative individuals are obsessed with certain problems and will not rest until those problems find solutions. Even when they give up on a problem, creators essentially enter an incubation period in which the question always lurks in the back of the mind. Creators thus demonstrate a certain, large-scale, form of the “Zeigar-nik effect,” which is what happens when experimental subjects who are not allowed to finish some tasks are more likely to recall those than other tasks they successfully completed. Because creators’ lives are replete with such incomplete tasks, they are incessantly on the lookout for cues about how those tasks might be completed. Moreover, there will intrude many false alarms and blind alleys. Accordingly, creators cannot give up easily and may spend years before coming across a successful resolution. Indeed, Howard Gardner has suggested that a creator’s major breakthroughs are usually separated by a full decade. Einstein had contemplated a paradox in physics for nearly ten years before he finally arrived at the solution in the form of his special theory of relativity; the more complete solution, the general theory of relativity, had to wait another decade to emerge. In the restrictive sense of maintaining the necessary long-term commitment, then, Einstein can be said to have willed his theory into existence. He persisted when most others would have quit.

Finally, we must not forget another necessity of creative genius. It is not sufficient to resolve an obsession. The answer must undergo testing, elaboration, development. Revision and amendment may be required, and disappointing setbacks overcome. The creator may have to run a gauntlet of criticism and collegial advice. All this hard work that connects the initial creative insight to a finished creative product must be executed. There is no room for the weak of will or faint of heart in this arduous business. So, again, volition plays a crucial role even in a Darwinian theory of creativity.

On the other hand, a Darwinian theory makes a distinction between willfully seeking solutions to problems and willing the appearance of those sought-for solutions. It is one matter to ask a question, quite another to find an answer. And it is the latter process that is far more likely to be Darwinian in nature. Recall how Helmholtz described his own trial-and-error problem-solving process. The economist and logician William S. Jevons generalized Helmholtz’s realistic modesty to all scientists when he wrote, in his book The Principles of Science, that

it would be an error to suppose that the great discoverer seizes at once upon the truth, or has any unerring method of divining it. In all probability the errors of the great mind exceed in number those of the less vigorous one. Fertility of imagination and abundance of guesses at truth are among the first requisites of discovery; but the erroneous guesses must be many times as numerous as those that prove well founded. The weakest analogies, the most whimsical notions, the most apparently absurd theories, may pass through the teeming brain, and no record remain of more than the hundredth part.... The truest theories involve suppositions which are inconceivable, and no limit can really be placed to the freedom of hypotheses.

If the creator could so easily will into existence novel and workable solutions to troublesome problems, we would be hard-pressed to explain the reports of historic creators. Very often they describe the sudden appearance of a solution without the slightest participation of the will. I have already mentioned Archimedes’ moment of insight in the bathtub. In a similar class is Hadamard’s affirmation that “one phenomenon is certain and I can vouch for its absolute certainty: the sudden and immediate appearance of a solution at the very moment of sudden awakening.” In the same category, finally, I may assign this episode recorded by Poincare: “Just at this time I left Caen, where I was then living, to go on a geologic excursion under the auspices of the school of mines. The changes of travel made me forget my mathematical work. Having reached Coutances, we entered an omnibus to go some place or other. At the moment when I put my foot on the step the idea came to me, without anything in my former thoughts seeming to have paved the way for it, that the transformations I had used to define the Fuch-sian functions were identical with those of non-Euclidean geometry.” So typical are reports like these that Margaret Boden concluded, “The bath, the bed and the bus: this trio summarizes what creative people have told us about how they came by their ideas.” Indeed, so involuntary are these events that creators sometimes insist illuminations took place largely in opposition to volitional control. The poet William Blake, for example, once noted that “I have written this poem from immediate dictation, twelve or sometimes twenty or thirty lines at a time without premeditation, and even against my will.” And the philosopher Friedrich Nietzsche described the moment of volition-free inspiration in these terms: “One can hardly reject completely the idea that one is the mere incarnation, or mouthpiece, or medium of some almighty power. The notion of revelation describes the condition quite simply; by which I mean that something profoundly convulsive and disturbing suddenly becomes visible and audible with indescribable definiteness and exactness. One hears—one does not seek ... a thought flashes out like lightning, inevitably without hesitation—I have never had any choice about it.... Everything occurs quite without volition, as if in an eruption of freedom, independence, power and divinity.” The highly romantic, overwrought nature of this expression should not cause us to dismiss the veracity of the underlying proposition. Acts of creativity are not necessarily acts of will. To the degree that volition is absent, the creative process must be viewed as a form of secondary Darwinism, not pseudoLamarckism.

Before proceeding to the question of primary Darwinism, I would like to touch on a matter raised by some of the preceding introspective reports. According to the testimony of creative geniuses, novel insights sometimes arise through an illumination that is as sudden and complete as it is involuntary. The moment of inspiration often seems to embody a “quantum leap” or “breakthrough” that departs significantly from all that has gone before. This unexpected discontinuity would seem to run counter to the historical continuity often thought to be characteristic of organic evolution. Species were supposed to evolve by almost imperceptible steps, not massive jumps. In fact, these revelations appear more comparable to the “punctuated-equilibrium” theory that Eldredge and Gould first advanced in 1972 as a direct challenge to Darwinian gradualism. According to this alternative, the most frequent condition of most species is stasis, in which generation follows generation without noticeable change. Then, in a manner not unlike the sudden illumination of the creative genius, the species rapidly transforms into one or more new species. After speciation, stasis takes over again, as the new species attain new equilibria with their environments. To understand better how punctuated equilibrium connects with a Darwinian theory of creativity, I should note two facts.

First, Darwin himself was not of the mind that continuity necessarily implied constancy of evolutionary change. Nor did Darwin maintain that change was an inevitable feature. In particular, he observed: “Many species once formed never undergo any further change ..., and the periods during which species have undergone modification, though long as measured in years, have probably been short in comparison with the periods during which they retained the same form.” Thus, the pace of evolution may vary from stagnation to almost explosive change. Darwin required only that even the most rapid changes would take many generations to transpire. Especially complex adaptations could not appear overnight; rather there needed to be time for the variation-selection process to operate.

Second, evolutionary biologists recognize that phyletic gradualism and punctuated equilibria may represent end points of a continuum. Some evolutionary lineages may show one extreme, other lineages the other, but most are probably distributed somewhere between. In addition, different traits within the same lineage may exhibit distinct patterns. For example, the size of an organ might change more or less gradually, while the structure of another organ might follow the course of punctuated equilibria. The reasons for these contrasts are varied. Catastrophic environmental or ecological transitions—such as sudden changes in sea level or climate—can certainly affect abrupt alterations in corresponding adaptations. But there are also occasions when a species may chance upon a revolutionary adaptation that enables the quick exploitation of an ecological niche previously unoccupied.

I believe that this same dimension underlies what we know of the creative process. At one extreme are the sudden illuminations in which a grand idea appears only after a long incubation period of subconscious exploration—the cognitive analog of evolutionary stasis. At the other extreme are ideas that develop very gradually from crude beginnings, eventually reaching the status of a work of genius. Judging from notebooks kept by great creators, including Darwin, the creative career usually consists of a chaotic mix of sudden inspirations and step-by-step refinements of initial, uncultivated ideas. The illuminations are more dramatic, and thus most often remembered (at least if they have later proven correct). But a creative career cannot subsist on revelations alone. On the contrary, the bulk of the ideas on which creators stake their fame is probably the consequence of a more gradual ideational evolution.

The very capriciousness of this cognitive mixture of the quick and the slow serves only to accentuate the Darwinian claim: The creative genius has no real volitional control over the emergence of ideas.

Evolutionary Origins

Thus far, the focus has been on secondary Darwinism. Does the creative process in human beings really operate in a fashion analogous to what supposedly happens in organic evolution? But the next issue takes us back to primary Darwinism. How can a Darwinian theory of organic evolution account for the emergence of the creativity in the first place?

The most obvious answer is to claim that creative behavior enhances a creature’s adaptive fitness. In other words, the evolution of creative behavior may have been favored by natural selection. This possibility is readily apparent in Epstein’s research on generativity theory. The pigeons in his experiments seem to be operating according to highly adaptive principles. If what to do in a given situation is self-evident, by all means do it. If not, try out all behaviors that have worked under comparable circumstances. If none of those do the trick, then generate various combinations of behaviors that have solved similar problems, until a behavioral combination is found that receives reinforcement. Thus, trial and error is a last resort, but it is a resource that must be available if the pigeon is to attain full mastery of its environmental niche.

Naturally, trial and error is an inefficient process. Much time and energy can be expended to little or no avail. For example, the kittens in Edward Thorndike’s classic puzzle box thrashed about chaotically until they chanced upon the latch that allowed escape. Such behaviors appear less than intelligent. However, for organisms with a highly sophisticated cognitive apparatus, this Darwinian procedure can be rendered far more efficient. Higher intellects possess internal representations (or models) of their external world. These representations are abstract, but they contain the core features of the real world and thus can be used to test potential solutions with a freedom and speed often impossible in the physical environment. Moreover, the more complex minds are able to insert themselves as actors in their mental worlds, as when intelligent organisms display the capacity for planning behaviors in advance and then execute those behaviors precisely according to plan. Once Kohler’s apes had figured out how to get the out-of-reach banana, they could often do so smoothly and quickly, just as if they had performed that trick before. But, of course, the chimps had already carried out the appropriate behaviors—in their heads. What appears to be behavioral foresight is actually cognitive hindsight.

Human beings take the chimpanzees’ cognitive capacity even further. Besides being able to engage in behavioral variations using internal representations of the self, others, and the world, humans can transform the problem into media and modalities unavailable to even the most intelligent of the beasts. The visual givens of the problem might be translated into verbal representations, for example. Those with sufficient training, moreover, may transform the problem into mathematical or logical representations. The trial-and-error procedures can then be applied to these translations, and often with superior success. In fact, with the advent of modern computers, programs can be written that will perform the necessary manipulations automatically, just as we saw in the case of genetic algorithms and programming. Darwinian creativity then for the first time can occur outside an organic system, whether cell nucleus or complex intellect.

Thus, human creativity seems a highly adaptive behavior that should be favored by natural selection. Given that Homo sapiens appears to be the most successful single species on this planet, maybe this argument is too obvious. Humans have emerged as monarchs of the ecological mountain because they have exhibited a creativity unrivaled by even the most impressive intellectual competitors, such as chimpanzees and dolphins. This creativity has enabled our species to fill niches far more diverse than those occupied by any other organism.

The foregoing adaptationist account is certainly plausible. Yet its very plausibility may in fact undermine its credibility. The adaptiveness of creative behavior seems so transparently true that this account may represent nothing more than another of those “just so” stories that plague so many Darwinian explanations. Indeed, this explanation maybe so glib as to raise more questions than it answers. Consider the following three issues:

• 1. Although the adaptive value of creative behavior is apparent when it comes to technological domains—activities having to do with obtaining food, obtaining shelter, and self-protection—it is less apparent that creativity has evolutionary utility in fields that make no direct contribution to reproductive fitness. Michelangelo may have fed, clothed, and sheltered himself by painting frescoes in chapels and carving chunks of marble, but he also died childless. The average sixteenth-century Tuscan farmer, not showing an ounce of originality but instead relying on the tried-and-true agricultural techniques passed down for generations, would certainly have claimed more fitness than this exalted genius. In fact, Michelangelo’s creative genius may have actually detracted from his reproductive fitness. A similar conclusion may be drawn with respect to Newton, Descartes, Beethoven, and innumerable other creators who apparently left no biological progeny. Indeed, as we will see in chapter 3, creative individuals often feature many characteristics that would seem to militate against attaining fitness in the sense of primary Darwinian selection.

• 2. A related issue is the existence of pronounced individual differences in the amount of creativity displayed. Creativity seems to be a trait most people wish they had more of. Yet only a relatively small percentage of the population appears to display creative behavior with any conspicuous regularity. And creativity of the highest order is more rare still. Moreover, as will be seen in chapter 5, the distribution of creativity may differ from that of other characteristics presumably selected for their adaptive value. Physical traits like height as well as mental traits like intelligence tend to be normally distributed. That is, the distributions are described by a symmetric, “bellshaped” curve. The distribution of creative behavior, in contrast, exhibits an extremely asymmetrical curve, with a small proportion of the population exhibiting a disproportionate amount of the creativity. It is not clear how this elitist distribution might be reconciled with the assumption that creativity has adaptive value.

• 3. Besides these individual differences within a particular society at a specific point in time, creativity can vary appreciably from culture to culture and period to period. This fundamental fact was pointed out at the beginning of chapter 1 when I spoke of golden ages and dark ages. Does this mean that creative behavior has more adaptive value in some societies and eras in comparison with others? If so, perhaps creativity is indeed an operant that can be reinforced under some circumstances and extinguished under others. Alternatively, of course, someone might adopt Gabon’s position that such contrasts across space and time are the inevitable consequence of contrasts in the innate capacity of various biological populations to produce genius. Then the more “advanced races” in some supposed evolutionary progression would boast the superior creative capacity—a potential interpretation that I shall consider in chapter 6.

In light of the above enigmas, it is not immediately clear whether the primary Darwinian process of biological evolution can account for the appearance of the secondary Darwinian process of creative thought. Compare this problem with attempts to explain the origins of, say, the secondary mechanism of the immune system. In the latter case, the adaptive value is quite evident, the individual differences much less pronounced, and the variation across cultures and periods probably minimal. All of this is guaranteed, courtesy of the straightforward fact that organisms incapable of defending themselves against pathogens are rather easily (and often dramatically) removed from the gene pool. Natural selection in favor of a Darwinian immune system can seem cruel, but ultimately the results are kind.

There are a number of potential resolutions of this enigma. For example, creative genius might be the product of some form of group selection something along the lines of altruistic behavior. Alternatively, creativity might result by some coevolutionary process involving the complex interaction of natural and cultural selection. Some of these possibilities will be discussed in later chapters. But for now let it suffice to say that it is not an easy task to explain how the primary Darwinian process could give rise to the appearance of creative genius.

Although the evolutionary origins of creative genius present such a puzzle, that mystery should not lead us to reject the idea that creative thought is fundamentally Darwinian. In fact, creativity probably constitutes the most successful of all Darwinian processes, whether primary or secondary. That is, if the success of a Darwinian mechanism is judged by the diversity of creative forms it generates, the creative mind just may come out on top. For instance, one scholar inferred from patent statistics that human inventions are as diverse as all the species that currently inhabit the earth. Yet if you add to this list of technological accomplishments all scientific journal articles, all musical compositions, all artistic masterworks, and all literary creations, it may be argued that the creative mind represents the single most potent Darwinian force on this planet, if not the universe. In future generations, for good or ill, its supremacy may emerge unchallenged, as the products of human genius continue to expand their presence over every inch of this globe, and sometimes reach to the farthest corners of the solar system.

1-^

A                                        .            .

Ikn essential component of all Darwinian theories is variation. Neither primary nor secondary theories can function without assuming that the units of selection display some degree of inheritable or transferable variation. The variants are what feed the selection process, whether natural or sexual, immunological or sociocultural. In the absence of such variability, Darwinian evolution and development would screech to a halt. It should not surprise us, therefore, that the behavioral scientists most influenced by Charles Darwin’s ideas have displayed a fascination with how human beings can differ in so many of their attributes. Francis Galton counts among the earliest and most prominent examples. Profoundly inspired by reading Origin of Species, Galton became convinced that individual differences are both substantial and consequential. For example, Gabon’s 1869 Hereditary Genius argues emphatically that individuals vary immensely in what he styled “natural ability,” a character that includes “those qualities of intellect and disposition, which urge and qualify a man to perform acts that lead reputation. I do not mean capacity without zeal, nor zeal without capacity, nor even a combination of both of them, without an adequate power of doing a great deal of very laborious work. But I mean a nature which, when left to itself, will, urged by an inherent stimulus, climb the path that leads to eminence, and has strength to reach the summit—one which, if hindered or thwarted, will fret and strive until the hindrance is overcome.” Galton added that “it is almost a contradiction in terms, to doubt that such men will generally become eminent,” for “the men who achieve eminence, and those who are naturally capable, are, to a large extent, identical.” On the other hand, individuals who lack natural ability will be incapable of rising above the vast crowd of nonentities.

This hypothesized association between individual differences in natural ability and a person’s eventual societal distinction is central to the fundamental question of this book—what are the origins of genius? As discussed in chapter 1, Galton defined genius in terms of reputation. Geniuses are those who have earned a name in the annals of human civilization. Moreover, in the case of creative genius, we have already claimed that this broad influence can be best gauged in terms of the creation of concrete products. Just as adaptive fitness in organic evolution is assessed by reproductive success, fitness in sociocultural evolution can be judged by productive fertility. In support for this Darwinian linkage, I should make two empirical observations.

First, creative personalities exhibit tremendous variety in the number of products they offer the world. At one extreme are those “one-idea” creators who have gone down in history for a single contribution. Gregor Mendel provides perhaps the most conspicuous illustration, for his entire posthumous reputation is founded on his experiments on trait inheritance in garden peas. At the other extreme are those “multiple-idea” creators whose brains seem to overflow with brilliant contributions. Charles Darwin himself provides an exquisite illustration. Despite sometimes severe health problems, Darwin was able to generate 119 publications on a tremendous diversity of topics in geology, zoology, botany, ecology, and psychology. Even when Darwin’s theory of evolution suffered some neglect before its rebirth under the auspices of the “modern synthesis,” he was still widely recognized as one of the most brilliant scientists of his era. Notwithstanding the protests of the religious establishment, Darwin was buried in Westminster Abbey, not far from Newton, Faraday, and Lyell.

Second, individual differences in total lifetime output are indeed associated with the degree of eminence achieved. In fact, research has consistently shown that the most powerful single predictor of reputation among both contemporaries and future generations is the person’s sum total of contributions. Furthermore, almost all other variables that may correlate with the difference in fame between individuals do so only because they affect the output of creative products. To be sure, the relationship between productivity and eminence is by no means perfect. For example, John Edward Gray, an English naturalist and approximate contemporary of Darwin, could boast 883 publications. This figure is almost eight times what Darwin produced, and more than a hundred times more prolific than Mendel’s measly seven publications! Yet Gray’s reputation matches neither Darwin’s nor Mendel’s. Later I will discuss the nature of these imperfections in the productivity-eminence association. The main point here is that exceptions such as these do not suffice to overthrow the fundamental conclusion: Total lifetime productivity provides the best behavioral indicator of creative genius.

Given this productivity-eminence criterion for genius, the obvious question is whether creative productivity is linked to other individual-difference variables. After all, human beings may vary in a wide range of cognitive, motivational, and personality dimensions. Gabon’s definition of natural ability includes three distinct dimensions—intellect, motivation, and persistence. Psychologists who investigate human variation in intellectual and dispositional variables could certainly list a host of additional possibilities. Moreover, there already exists a huge literature on the individual-difference variables that correlate most highly and most consistently with creative behavior. Although the findings are quite rich, the psychometric instruments diverse, and the theoretical underpinnings quite varied, I argue that a consistent pattern emerges. Specifically, the pattern underlying all research findings is what would be expected from a Darwinian concept of creative genius. In other words, creative individuals may be said to have a “Darwinian personality.” By this I mean that creators tend to have the attributes necessary to generate the numerous and diverse ideational variants required by Campbell’s Darwinian model of the creative process. Some character traits may enhance this capacity for generating variations, whereas other traits may inhibit this capacity. The Darwinian personality features more of the former, fewer of the latter, yielding a distinctive trait profile.

Yet it is critical to recognize that different domains of creativity require varying amounts of variation-generation capacity. In some endeavors, ideational variations are highly constrained, whereas in others the variations must be free. For example, scientific creativity, on the average, operates under greater constraint than does artistic creativity. The concepts and techniques that the scientist employs in combinatorial play are rather abstract, specialized, and divorced from everyday experience. Accordingly, the variation process will be rather compartmentalized for the scientific genius. The artist, in contrast, must allow the imagination more freedom, encompassing the ideas and feelings of everyday life and using a vocabulary of expression more intelligible to a wider audience. Thus, in searching for the profile of the Darwinian personality, we should make allowance for the greater prominence of that profile among artists than among scientists.

However, even this complication must be rendered more complex. From a Darwinian perspective, not all forms of scientific creativity are the same, nor are all types of artistic creativity identical. Within the sciences, for example, Thomas Kuhn advanced the distinction between normal and revolutionary scientists. The former conduct research within the confines of a received body of theory and method—the old paradigm—whereas the latter offer novel theoretical and methodological approaches—the new paradigm. Because the revolutionaries allow greater scope to their search for an original paradigmatic system, one might infer that they are more likely to display Darwinian characteristics than the practitioners of normal science. Likewise, the art world does not represent a homogeneous set of creators either. Certainly, one of the most critical contrasts apparent in literature, music, and the visual arts is the distinction between classical and romantic styles, the former exhibiting more orderly restraint and logical structure, the latter more freedom and impulsive emotionality. Presumably the classicists would be less Darwinian in disposition than the romanticists, at least on the average. It may even be that the personality profiles of artistic creators overlaps those of scientific creators. A revolutionary scientist may have a similarly Darwinian disposition to a classicist artist to gain optimal success as a generator of original ideas. In any case, in this chapter we must take care not to assume that all creators display equivalent personal attributes.

These precautions in mind, I now review the research findings relevant to two broad domains of individual-difference variables, namely, intellect and character. In each domain I shall scrutinize a tremendous diversity of variables associated with creative genius that are consistent with the hypothesized Darwinian personality. To the extent that this demonstration is successful, this chapter should reinforce the argument advanced in the previous chapter that creative thought is at root Darwinian.

Intellect

As pointed out in chapter 1, although psychologists have often defined genius in terms of exceptional intelligence, the actual empirical relationship between the two phenomena is more ambiguous. No doubt a certain minimal level of intelligence seems necessary to demonstrate significant levels of creative behavior. That is, there exists some threshold level of IQ below which it is virtually impossible to claim any distinctive level of creativity. Although the exact location of this threshold is not known, there is no well-documented case of an eminent creator having a below-average intelligence. Even so, a high level of intelligence by itself cannot guarantee that a person will display an impressive degree of creativity. There are plenty of people with high-IQs, for example, who do not seem any more creative than individuals with average or even low IQ scores. On the other hand, the possession of an extremely elevated intelligence does not automatically mean that a person is doomed not to be creative. Indeed, an exceptionally bright individual can potentially boast more creative brainpower than can a less brilliant person. Consequently, as intelligence increases beyond whatever the threshold level may be, the height of eminence possible to achieve increases as well, producing what is sometimes called a “triangular distribution.” At the lower levels of intelligence, the maximum level of achieved eminence is not very high, whereas at higher levels of intelligence the maximum is higher, albeit there will also exist a considerable number of creators who rise just barely above the nonentities in the annals of history.

From a Darwinian perspective, this complex linkage should not be all that surprising. Intelligence involves the capacity for acquiring and applying knowledge. To generate ideational variations, a person must have a sufficient repertoire of ideas that can be subjected to some combinatorial procedure. The more powerful the intellect, the larger the potential size of that repertoire. Even so, having an impressive collection of information and skills does not by itself suffice to support the production of ideational recombinations. The information must be organized in the appropriate fashion. Remember the earlier discussion of genetic recombination. If all genes are contained in a very small number of chromosomes, linkage is going to severely restrict the independent assortment of the genetic material. In contrast, if the genes are distributed over a large number of chromosomes, the genetic variations will much more readily proliferate.

Thus, it is the structure of intelligence rather than intelligence per se that should be of the greatest importance in a Darwinian model of the creative genius. Below this point is elaborated by looking at two illustrations, namely, remote association and divergent thinking.

Remote Association

In 1962, Sarnoff Mednick proposed an influential associative theory of the creative process. This theory was based on the premise that creativity requires the ability to make rather remote associations between separate ideas. Highly creative individuals were said to have a flat hierarchy of associations in comparison with the steep hierarchy of associations of those deemed less creative. Figure 3.1 provides a hypothetical illustration of the two opposing types of associative hierarchies. As this figure shows, a flat associative hierarchy means that for any given stimulus, the creative person has a great many associations available, all with roughly equal probabilities of retrieval. Because such an individual can generate many associative variations, the odds are increased that he or she will find that one association that will make the necessary remote connection.

A concrete example can be devised that uses actual word frequencies calculated from a large sample of people. Let us first imagine a person with steep associative hierarchies who is given the two words foot and command.

Probable Order of Association

Figure 3.1 Steep and flat associative hierarchies according to Mednick’s theory of creativity.

The first word might quickly elicit such associates as “shoe, hand, toe, and leg,” while the second word might rapidly evoke such associates as “order, army, obey, and officer.” The individual would then fall silent, waiting for the next verbal stimulus. Because the two chains of association share not a single word in common, they fail to converge. Thus, “foot” and “command” remain unconnected in the person’s mind. In contrast, an individual with a flat associative hierarchy might begin with the same initial associates, albeit with considerable alteration in order. More important, this second person might then append to “leg” the additional associates “soldier, ball, walks, amble, arm, sore, inch, rat, snow, person, physics, dog, mule, wall, shin, wash, hat, and end,” plus trail “officer” with “performance, do, tell, shout, halt, voice, soldier, hat, polite, plea, book, salute, fulfill, obedience, war, and stern.” In the second case, the two strings of thought actually contain not one but two associates in common, namely “soldier” and “hat.” Hence, a person with a flat associative hierarchy will be far more likely to find an associative nexus between the two remote ideas of “foot” and “command.” Actually, this illustration understates this ability, for it has relied on relatively commonplace associates according to standardized association tests. If we allow for the high probability that creative people will include rather idiosyncratic associations in their semantic networks, the odds of successful remote association becomes greater still.

Mednick’s associative theory is quite provocative from the standpoint of Campbell’s model of the creative process. Because flat associative hierarchies imply that the numerous alternative chains of associations feature comparable odds of activation, the search through the inventory of associative linkages will be largely unpredictable, perhaps even random. The flatter the associative hierarchy, the greater is the unpredictability or randomness of the exploration. Indeed, flat hierarchies would be quite susceptible to even the most subliminal of priming from external stimuli—effects that could haphazardly and transiently alter the comparative associative strengths of the available avenues of thought. Hence, highly flat hierarchies would tend to produce blind associative variations. Such mental explorations would be, in a word, Darwinian.

To give this associative theory some empirical grounding, Mednick devised the Remote Associates Test, or RAT. This instrument attempted to assess individual differences in the capacity to generate numerous associations. Each item on the RAT would list words to which the subject taking the test would have to pick another word that represented an associat e of all the given words. To offer an easy example, “rat,” “blue,” and “cottage” would all share the associate “cheese.” Because the common associate would not be a close associate of all the given stimulus words (e.g., “cat,” “red,” and “house”), the person would have to have access to a large number of associations to do well on the RAT. Although the test is not without its imperfections—most notably its use of a multiple-choice format that requires a single right answer—the RAT has done fairly well in validation studies. High scores on this instrument have been associated with higher levels of creativity. Just as important, scores on the RAT are associated with other attributes that should facilitate the variation process, such as the capacity for primary-process thought and synesthesia. The latter correlate is especially fascinating, because the ability for cross-modal associations would certainly yield associates far more remote than those confined to a single modality. One could hardly conceive an ideational junction more blind than one synesthetic!

Yet it remains fair to say that Mednick’s associationist theory has more plausibility than the psychometric test used to provide an operational definition of the core concept of that theory. In fact, subsequent researchers have elaborated various aspects of Mednick’s associationist theory to grant it greater explanatory power. For instance, many years ago I argued that this theory can help account for the importance of intuitive thought in creative individuals. As discussed in the preceding chapter, the autobiographical accounts of creators are replete with episodes of unconscious mental processing. According to this theory, there exists a correspondence between the strength of an association and cognitive and behavioral consequences. In

82                           - Origins of Genius ~

particular, I proposed that the associative content of the mind may be stratified into four levels.

• 1. Habitual associations are those so automatic that they are elicited without conscious effort. For example, in most people the word dog will immediately evoke the response cat. These habitual responses are the first to emerge when a person generates associations to a particular stimulus.

• 2. Cognitive associations are those sufficiently well established to enter core awareness. Because such associates are conscious, they can be articulated and symbolized, and thus successfully communicated to others.

• 3. Behavioral associations are sufficiently strong to have repercussions for behavior but not strong enough to enter consciousness. As noted in chapter 2, there is ample evidence in studies of implicit knowledge that individuals can acquire expertise long before they become aware of what has been mastered. This information is therefore intuitive in nature.

• 4. Attention associations are sensory expectations too weak to support behavioral reactions, and yet strong enough to alert the mind to potential regularities in the environment. Associative material at this lowest level serves simply to orient attention toward sources of prospective knowledge.

This associative stratification was then integrated with Mednick’s concept of associative hierarchies. At one end of an intellectual continuum are those individuals with steep hierarchies, that is, those whose minds hold largely a restricted supply of habitual and cognitive associations and an even more restricted reservoir of behavioral and attention associations. Because these people have so few total associations, they will have relatively few ways to connect between various concepts, and hence they will only rarely successfully link ideas by remote associates. At the other end of the continuum are those with flat hierarchies. Such individuals will have fewer habitual associations but will have more cognitive and, especially, behavioral and attention associations. Not only will their store of knowledge be more richly interconnected with various associative links, but in addition a large proportion of these associations will be intuitive in nature. As a consequence, they will probably more often report instances of intuitive hunches and imageless thought. There thus should be more instances of creative ideas emerging without the slightest hint about the chain of reasoning behind the discovery. In addition, minds with flat associative hierarchies are likely to more freely scan the environment for potential cues, and thereby

~ Variation ~                               83

they exhibit a greater propensity for making serendipitous discoveries. These individuals are likely to be more attentive to their surroundings.

It deserves emphasis that individuals with steep associative hierarchies can be every bit as intelligent as those with flat associative hierarchies. In the foregoing model, intelligence is determined by the total number of concepts forming associative networks. Therefore, two people can boast the same stock of conceptual material but differ immensely in the richness of associations linking their concepts. Individuals may be equal in intelligence while differing in how that intelligence is structured. In the above model, highly intelligent minds with steep associative hierarchies are styled “analytical geniuses,” whereas those with flat hierarchies are styled “intuitive geniuses.” Both may display genius-grade intellects, but only the intuitives are capable of thinking in a fully Darwinian fashion.

Figure 3.2 illustrates the theoretical contrast between the analytical and intuitive minds. Here the letters A through U represent concepts while the lines represent the associative linkages between those concepts. Both intellects hold the same number of concepts, but they differ greatly in the number of associations and in the strength of those associations. This contrast means that the intuitive genius will have more possible associative pathways connecting any two given concepts, with many of these pathways operating at unconscious levels of information processing. Given the large number of available routes and the relative associative weakness of any single path, the search for a connection between any two ideas will be more likely to be unpredictable, even random. In terms of Campbell’s Darwinian model, ideational variations will be necessarily more blind. Note, too, that for the analytical genius there will exist a number of concepts that cannot be linked by any associative pathway whatsoever, no matter how long or tenuous. Individuals with this mental makeup may thus overlook ideational combinations that will be readily available to those with a richer associative network.

Naturally, analytical and intuitive geniuses are not types. Instead, they represent opposite ends of a continuum connecting those with the steepest associative hierarchies to those with the flattest hierarchies. Hence, certain kinds of creators will tend to fall at distinctive locations on this cognitive line. Artistic creators will be more prone to be intuitive, whereas scientific creators will tend to reside closer to the analytical end of the spectrum. Again, variable placement is likely to appear within each field. In particular, revolutionary scientists will be more intuitive than normal scientists, romantic artists more intuitive than the classical artists. Nonetheless, even the normal scientists should tend to be more intuitive than their noncreative colleagues.

Figure3.2 Hypothesized associative networks for analytical and intuitive geniuses according to the Simonton model.

Fortunately, there does exist empirical support for various predictions generated by this extension of the Mednick associationist theory. Particularly provocative is the evidence that successful creativity requires optimal placement along the intuitive-analytical dimension. In one study of scientific creativity using a free-association test, the generation of moderately atypical responses was more highly correlated with creativity than was either the production of commonplace associations or the production of extremely remote associations. The creative scientist must conceive original connections between ideas as long as those connections are not too strange. In contrast, as will be shown in chapter 5, the associative processes of artistic geniuses often must reach for the more bizarre associative links. Even so, even in artistic creativity there exists an optimum level of originality beyond which artists begin to lose their audiences.

Divergent Thinking

J. P. Guilford, one of the pioneers in the psychometric study of creativity, proposed a critical distinction between two kinds of thinking—convergent versus divergent. Convergent thought involves the convergence on a single correct response, such as is characteristic of most aptitude tests, like those that assess IQ. Divergent thought, in contrast, entails the capacity to generate many alternative responses, including alternatives of considerable variety and originality. Guilford and others have devised a large number of tests that purport to measure the capacity for divergent thinking. Typical is the Alternate Uses test, in which the subject must come up with many different ways of using a common object, such as a paper clip, toothpick, or brick.

The responses of the person taking the test can be scored on several dimensions, but three have received the most attention:

• 1. Fluency is the capacity to generate a large number of responses. In the Alternate Uses test, for example, an individual who thinks of a great many ways of using a paper clip would be scored as highly fluent on this task.

• 2. Flexibility is the capacity to generate responses that can be assigned to many different conceptual categories. In other words, this measure is a gauge of the ability to “change set” rather than always exploit the same function or property. For instance, a person who realizes that a toothpick can serve as a construction material or as something flammable is going to be more flexible than someone who thinks that a toothpick can only be used for picking at something.

• 3. Originality is the capacity to generate unusual but appropriate responses. Using a brick as a doorstop or paperweight is less novel than using broken bricks of different colors to create a wall mosaic.

Theoretically speaking, these three dimensions could be relatively independent of each other. For instance, a person who thought of dozens of uses for a brick would get a high score on fluency yet would get a low score on flexibility if all the uses entailed taking advantage of a single property, such as the brick’s weight. Likewise, an individual might conceive of only a few uses for a paper clip, but all uses might be highly original. Nevertheless, scores on fluency, flexibility, and originality all display strong positive correlations, suggesting that they all measure the same underlying capacity for generating ideational variations. In particular, the more ideas a mind can produce, the higher the odds that those ideas will be original and varied. Flexibility and originality are both to a very large extent mere consequences of fluency. This is precisely what would be anticipated from the standpoint of a Darwinian theory of creativity.

Many investigators have tried to validate the various divergent-thinking tests against other criteria of creative performance. Although these validation studies have had some modicum of success, two main complications have emerged.

The first complication is that different forms of creativity tend to require contrasting proportions of divergent and convergent thinking. Just as was noted in the case of remote associations, scientific creativity needs a stronger dose of convergent thought, whereas artistic creativity is far more dependent on divergent thought. As before, these contrasts should reflect the degree to which a particular form of creativity depends on the production of blind variations. On the average, creative ideas in science have more a priori guidance than do those in the arts. The less divergent thinking required in a given domain of creativity, the less Darwinian it can be said to be.

The second complication is that generalized tests do not have as much predictive validity as tests more specifically tailored to a particular domain of creativity. It is difficult to predict creativity in music, for example, on the basis of how many uses the subject can imagine for a toothpick. A similar domain specificity has been found for word-association tests as well. Even so, from a Darwinian perspective we would not expect it to be otherwise. The variation process must operate on those concepts that belong to a specific discipline. Hence, tests of divergent thinking must be tailored to each domain. This constraint is somewhat analogous to what is seen in biological evolution: Short of a highly fortuitous coincidence in protein synthesis, abundant spontaneous variation in wing pigmentation will probably not aid a herbivorous insect species that needs to generate variations in digestive enzymes in order to counteract a new chemical defense adopted by its main food source.

From the Darwinian perspective, the research on divergent thinking has one major drawback: Because the precise cognitive mechanism for generating divergent ideas is not specified by any theory, it is not clear whether divergent thought entails any blind variation. This disadvantage stands in stark contrast to Mednick’s theory, where it is easily seen how chains of association can quickly descend to the realm of nearly equiprobable associates (i.e„ where the hierarchies become almost perfectly flat). Nonetheless, the associative theory itself may help resolve this question. The fluency, flexibility, and originality of divergent production may ultimately depend on an intellect in which concepts are richly interconnected by diverse associative linkages. As a consequence, divergent thinking of the highest caliber—especially as seen in artistic genius—would be contingent on a certain amount of decidedly blind variation. The empirical research using divergent-thinking measures seems to endorse this position. It typically happens that once the most obvious answers are given, the sequence descends to a series of highly unusual and unpredictable responses. Moreover, these latter responses are quite variable in quality, highly creative responses being mixed randomly with worthless responses. The emission of ideational variants thus becomes increasingly chaotic, as the mind regresses into increasingly Darwinian thought.

Character

Not all creativity tests concentrate on individual differences in the structure of a person’s intelligence. Some gauge the individual’s personality instead. The underlying assumption behind such assessments is that creativity is as much a dispositional characteristic as it is an intellectual trait. Each human being is a cluster of motives, traits, interests, and values, some of which are conducive to creativity, and some not. If the Darwinian viewpoint is correct, the characteristic personality profile of creators should not be arbitrary, but rather it should consolidate those personal attributes that are most conducive to the generation of ideational variations. Below I offer some evidence on behalf of this theoretical conjecture. I first examine the general profile and then turn to one additional trait that has the most profound implications for both primary and secondary Darwinian models of human creativity.

Personality

What is the typical creative genius like? According to the accumulated literature, creative geniuses are open to diverse experiences, display exceptional tolerance of ambiguity, seek out complexity and novelty, and can engage in defocused attention. They display a wide range of interests, including interests that extend beyond their immediate domain of creative activity. They are far more likely to be introverted than extroverted, and they may sometimes appear remote, withdrawn, and perhaps even antisocial. They also exhibit tremendous independence and autonomy, often refusing to conform to conventional norms—at times exhibiting a pronounced rebellious streak. They deeply love what they do, showing uncommon enthusiasm, energy, and commitment, usually appearing to friends and family as “workaholics.” They are persistent in the face of obstacles and disappointments, but at the same time they are flexible enough to alter strategies and tactics when repeated failure so dictates.

Although no single creator will match this profile perfectly—for reasons discussed below—the foregoing personality sketch matches Darwin’s own personality fairly closely. For example, in his autobiography he reports that “I had, as a very young boy, a strong taste for long solitary walks,” in which he “often became quite absorbed.” Later as an adult, he continued to engage in solitary activities, and he spent most of his career retired at his country home in Down rather than taking full advantage of all the social and professional stimulation available in London. Darwin also reported that early on he “had strong and diversified tastes, much zeal for whatever interested [him], and a keen pleasure in understanding any complex subject or thing.” In fact, “independently of science, [he] was fond of reading books, and [he] used to sit for hours reading the historical plays of Shakespeare” as well as much other poetry. One book entitled Wonders of the World Darwin read often, and he claimed that it inspired in him “a wish to travel in remote countries, which was ultimately fulfilled by the voyage of the Beagle” Darwin always remained an omnivorous reader with a tremendous curiosity about diverse subjects. “His wide interest in branches of science that were not specially his own was remarkable,” said his son Francis. Yet Darwin was not a mere cloistered bookworm. As a youth he loved horseback riding, hunting, fishing, and collecting all kinds of natural samples—enthusiasms that would prove most useful when he served as the Beagle naturalist. Thus, Darwin had the appearance of being rather well-rounded. Yet at the same time, he was persistent at any task he undertook. For instance, he devoted eight years to a monumental work on the cirripedes (barnacles), continuing his labors long after he found the work tiresome. His Origin of Species consumed even more time: He began accumulating notes on the subject in 1837, the theory assumed abstract form in 1842, a more complete essay was finished in 1844, and the final product emerged in 1859—22 years later. Darwin expressed his tenacity of purpose with the motto “It’s dogged as does it.” Finally, I might mention Darwin’s unconventional attitudes about a wide range of matters, including his strong antipathy to slavery and his more generous attitudes toward the “inferior races” and the “lower classes.” Darwin also entertained less than conforming attitudes toward the prevailing religious views of his time and station. Of course, when he finally ventured to replace the Genesis view of creation with his theory of evolution by natural selection, Darwin’s religious iconoclasm became a cause celebre.

But rather than show that the personality profile fits Darwin, it is more important to show how the character of the creative genius fits a Darwinian theory of creativity. To accomplish this task, we must recognize that these traits actually form two separate clusters. One set of characteristics probably belongs most properly to primary Darwinism, whereas only the second set pertains to secondary Darwinism.

Primary Darwinian traits. When Galton first introduced “natural ability” as a construct to explain individual differences in eminence, he perceived it in rather inclusive terms. The concept was inclusive, first of all, because it encompassed a set of traits rather than just one—intellect, motivation, and persistence. The concept was inclusive in a second sense as well, for natural ability was hypothesized to underlie success in many different domains of human achievement. It is for this reason that the book Hereditary Genius treats not just famous creators and leaders but eminent divines and athletes besides. Indeed, Galton might also have included celebrities in the performing arts, business, and maybe even organized crime. Behind the attainment of greatness in any domain is a common cluster of attributes essential for success. To be successful at any demanding profession requires intelligence, enthusiasm, drive, persistence, commitment, and plain hard work. Moreover, these same traits may be considered extremely adaptive in a more general sense. It is hard to imagine the survival value of stupidity, apathy, passivity, inconstancy, irresolution, and laziness! Such a personality profile would be as dysfunctional for a prehistoric hunter and gatherer as for a contemporary investment banker. In short, certain characteristics in the profile of the creative genius are part of a generic recipe for success in life. Given this universality, they are probably best considered among the human features favored by primary Darwinism. These traits are surely those that help ensure the reproductive success of any individual who possesses them.

Secondary Darwinian traits. In contrast, other traits seem to provide exactly what is necessary to fulfill the requirements of a variation-selection model of the creative process, such as that put forward by Donald Campbell. As pointed out at the close of the preceding chapter, we have not yet determined how to account for creative genius from the perspective of organic evolution. Therefore, it is best to attach these characteristics to the secondary class of Darwinian processes. This maneuver is especially necessary because it is not clear how some of these traits would be adaptive to a particular organism. For example, wide interests might not be of much service if the pursuit of those interests consumes time that might be better devoted to the development of some special expertise. Even worse, some of these attributes might actually be maladaptive. What possible biological utility can introversion have for a social animal such as Homo sapiens? Is a lone wolf better off than the leader of the pack?

Once we assign these residual traits to secondary Darwinism, in contrast, their functional value becomes clear. If the aim is to generate and retain numerous and diverse ideational variants, we should predict that creativity

90                           - Origins of Genius ~

is more likely to appear in a person who has the following six characteristics:

• 1. Highly creative people harbor an impressive array of intellectual, cultural, and aesthetic interests. Individuals with broad interests, after all, are more likely to encounter concepts in multiple contexts, thereby enriching the associative linkages between various ideas in their heads. The result would be the flat associative gradients required by Mednick’s theory, and hence the capacity for divergent thought demanded by Guilford’s theory. The history of great creative ideas is replete with examples of people finding a solution to a major problem in one domain while engaged in “recreational reading” in an entirely different domain. Darwin reported that “fifteen months after I had begun my systematic inquiry [into the origins of species], I happened to read for amusement Malthus on Population, and being well prepared to appreciate the struggle for existence which everywhere goes on from long-continued observation of the habits of animals and plants, it at once struck me that under these circumstances favourable variations would tend to be preserved and unfavourable ones to be destroyed.”

• 2. Highly creative individuals are widely open to novel, complex, and ambiguous stimuli in their surroundings. People receptive to a wide range of stimuli in the external world will necessarily be exposed to a broad spectrum of potential priming stimuli during the incubation phase of the creative process. The creative individual must remain always open to subtle cues in the environment that may prime just that missing chain of images or associations that eventually leads to the long-delayed solution of a particular problem. Indeed, as the occurrences of serendipity amply demonstrate, an unexpected input from the world may stimulate a train of thought that leads to totally unforeseen places.

• 3. Highly creative people are capable of defocused attention. As one creativity researcher put it, “the greater the attentional capacity, the more likely the combinatorial leap which is generally described as the hallmark of creativity.” Defocused attention permits the mind to attend loosely to more than one idea or stimulus at the same time, even when these cognitions and perceptions bear no obvious relationship to each other. Yet certain combinations of these ideas may provide the basis for a creative insight. Most commonly, while creators are working on one problem, or even engaged in some irrelevant activity, they will be carrying around in the back of the mind or have “on the back burner” ideas belonging to seemingly different problems. The chance concurrence of these unconnected fragments then puts the missing pieces of the puzzle together. Such fortuitous coincidences, though rare, are

supremely blind, as demanded by Campbell’s theory of creativity.

• 4. Highly creative individuals are unusually flexible both cognitively and behaviorally. For example, when a person cannot solve a problem immediately, it may not be an efficient strategy to doggedly insist on working on the problem until a solution is found. Rather, it may be more effective to incubate the puzzle for some time, taking advantage of any accidental stimulation provided in the interim. It is for this reason, as noted earlier, that eminent creators tend to work on several interrelated projects at once, jumping back and forth according to whatever happens to appear most promising at the moment. Darwin’s own work habits are typical of what Howard Gruber aptly styled a “network of enterprises.” As Darwin put it, “I have always had several quite distinct subjects in hand at the same time, [and] I keep from thirty to forty large portfolios, in cabinets with labelled shelves, into which I can at once put a detached reference or memorandum.” Furthermore, sometimes a serendipitous event will hand-deliver an entirely new project. The gifted creators are those who are flexible enough to be opportunistic when Lady Luck thus knocks on the door. As B. F. Skinner expressed it, “a first principle not formally recognized by scientific methodologists: when you run onto something interesting, drop everything else and study it.”

• 5. Highly creative people are introverted. Because the variation process is blind, and because so much of the variation-generation procedure must take place through long sequences of free association, creativity requires long hours of solitary contemplation. What William Wordsworth said of Newton would apply to any creative genius, namely that he was “a mind forever / Voyaging through strange seas of thought, alone.” Hence, smoking a pipe in an armchair, taking a walk in the woods, engaging absentmindedly in some routine activity—these are the circumstances most supportive of Darwinian incubation. Experimental research has actually demonstrated rather conclusively that group problem-solving using more egalitarian “brainstorming” techniques usually yields dismal results in comparison with more solitary forms of problem solving. Individuals working alone will generate more and better ideas than will the same individuals working in a group. To be sure, sometimes interaction with other people, and especially with fellow creators, can sometimes stimulate the mind in new directions, promoting the discoveries that might otherwise remain out of reach. Nevertheless, for the most original geniuses, this social contact is always subordinate to the internal ruminations of their eternally preoccupied minds. Einstein once admitted: “I am a horse for a single harness, not cut out for tandem or teamwork; for well I know that in order to attain any definite goal, it is imperative that one person do the thinking and the commanding.”

• 6. Highly creative individuals are independent, autonomous, unconventional, and perhaps even iconoclastic. As a consequence, they will impose fewer a priori constraints on the scope of what they are willing to consider. Such people will be less prone to dismiss an idea as preposterous without first giving it a fair hearing. Many of the most original contributions to human civilization seemed absurd on first glance. Once when a world-class orchestra first rehearsed a new symphony, the musicians all laughed after playing the opening motif, and put down their instruments, believing that it was some practical joke. Had Beethoven had the same reaction when composing the piece, he would not have produced the Fifth Symphony, which is now believed to boast one of the most effective openings in the classical repertoire. Another lesson on this point is found in an episode in the early history of quantum theory. Once Wolfgang Pauli, the discoverer of electron spin, was presenting a new theory of elementary particles before a professional audience. Extended discussion followed, which Niels Bohr summarized to Pauli as follows. “We are all agreed that your theory is crazy. The question which divides us is whether it is crazy enough to have a chance of being correct. My own feeling is that it is not crazy enough.” Because Bohr rejected conventional criteria of what counts as a good scientific theory, he was able to originate the complementarity principle and other radical ideas that revolutionized modern physics.

When I used Darwin to exemplify the creative genius, I admitted it is most unlikely that any given creator will fit the general profile perfectly, the theoretical foundation be what it may. There are three main reasons why these discrepancies may appear.

First, because there are so many personality traits associated with the capacity for creative thought, weaknesses in one trait may be compensated by exceptional strengths in one or more different traits. For example, someone who did not have very wide interests might still be exceptionally open to whatever novel experiences might come his or her way. Or, turning to the primary Darwinian traits, a relative weakness in intellectual power might be compensated by a superlative drive and determination. As observed in chapter 1, Darwin himself admitted that he was not the most brilliant naturalist of his day, but nonetheless he claimed compensatory virtues that would put him in the front ranks of world scientists. To be sure, these tradeoffs can only be carried so far. Utterly lazy people who cannot force themselves to master the basic principles and skills of a creative domain are never going to rise above the status of amateurs or dilettantes.

Second, I have already emphasized the critical relevance of differences across domains of creative activity. Because scientific creativity is less Darwinian than artistic creativity, scientists will less likely fit the secondary profile than will artists. In the case of anticonformity, for instance, Kuhn has spoken of the “essential tension” that underlies all scientific creativity, for “very often the successful scientist must simultaneously display the characteristics of the traditionalist and of the iconoclast.” And once more, within the sciences differences in traits can be expected between the normal and revolutionary scientists, the latter displaying more iconoclasm than the former. Similar qualifications apply to the other attributes that compose the general personality profile of creative genius.

Third, if individuals vary immensely in their creative power, they will also differ to the extent to which they will epitomize the Darwinian personality. Only the most prolific and profoundly original of creative geniuses would be expected to exhibit the full panoply of traits. In contrast, a creator who can claim but one, relatively minor contribution need not fit the profile at all. A one-idea person may not have engaged in any variationselection process, and thus did not need the matrix of personality traits that support such Darwinian thought. For example, sometimes a person will introduce into a given discipline a concept or technique that is already well-established in another discipline. Such individuals may receive credit for the dissemination of knowledge, but their accomplishments may not have required great creative insight. Indeed, Mednick long ago recognized the possibility of a third type of association gradient besides those discussed earlier: Some individuals may have steep association gradients as seen in figure 3.1, but rather than generating the most obvious associates, all of their associations may be bizarre or unusual. According to Mednick’s theory, although such individuals would not produce many ideas, the few they would generate would be highly original. Among the novelties produced may be one excellent idea on which to stake a claim to fame, diminutive though it be in comparison to the output of truly prolific creators.

The two factors of domain contrasts and individual differences are more or less unrelated to each other. You can be a great scientist or a mediocre one, an illustrious artist or an obscure one, and so forth. Accordingly, these two dimensions may jointly determine the odds of exhibiting a Darwinian personality in the manner depicted in figure 3.3. This shows the interrelation among the creative domain (e.g., artistic versus scientific), the magnitude of genius (e.g., level of productivity), and the degree to which a creator should display the Darwinian profile. On the average, scientific creators are

likely to exhibit less of a Darwinian disposition than artistic creators. In addition, there should appear contrasts among the scientists (normal versus revolutionary) and the artists (classical versus romantic). Nevertheless, within a particular creative domain, such as classical artists, the greater the manifestation of genius, the more we would predict the appearance of the secondary attributes of the Darwinian personality. Because of this within-domain variation, there may appear some overlap across domains. The greatest revolutionary in science may have a personality profile not all that distinguishable from some lesser artist of a classicist inclination.

The foregoing complications should not obscure the central conclusion, however. The general profile of the creative genius seems explicable in Darwinian terms, at least if we combine both primary and secondary processes. Even so, it now comes time to admit that the above broad portrait of the creative genius remains incomplete. A critical feature was deliberately omitted because it requires rather special treatment.

Psychopathology

Ever since the times of the ancient Greeks, genius has often been linked with mental disturbance. Aristotle is said to have claimed, “Those who have become eminent in philosophy, politics, poetry, and the arts have all had tendencies toward melancholia.” Similarly, Seneca held that “no great genius has ever existed without some touch of madness.” In more recent times, the same link was still proclaimed. Thus, Shakespeare wrote, “The lunatic, the lover, and the poet I Are of imagination all compact.” Nor has this viewpoint been confined to philosophers and dramatists. By the end of the nineteenth century, many behavioral scientists were espousing similar views. A notable example was the Italian criminologist Cesare Lombroso. In his 1891 classic The Man of Genius, Lombroso confidently asserted that genius was associated with “degenerative psychosis,” especially that of the “epileptic group.” Indeed, reputable psychiatrists would claim that genius was among the symptoms of an inferior genetic endowment. For instance, an 1895 article published in the Journal of Nervous and Mental Disease listed the four possible repercussions of a single congenital defect: “First, and most prominent in the order of frequency is an early death. Second, he may help swell the criminal ranks. Third, he may become mentally deranged and ultimately find his way into a hospital for the insane. Fourth, and least frequently, he startles the world by an invention or discovery in science or by an original composition of great merit in art, music or literature. He is then styled a genius.” Psychologists as distinguished as Sigmund Freud and William James espoused comparable opinions. They might disagree on the details, such as the specific syndromes entailed, but they would concur that creativity and psychopathology bear an intimate relationship. The only dissonant voices in this overwhelming chorus are a handful of humanistic psychologists, such as Abraham Maslow, Carl Rogers, and Rollo May, who were so bold as to claim that creative individuals enjoyed better mental health than the rest of us. Yet however loud these dissidents might shout, they could hardly be heard through the grand chord of consensus.

A champion of Darwinist theories cannot help but feel a little discomfort with the received tradition. To begin with, the mad-genius association seems to fly in the face of primary Darwinian theory. Mental illness and emotional disturbance seem so antithetical to adaptive fitness that it appears impossible that either natural or sexual selection would favor the survival of pathological genes. If psychopathology fed creativity, how could there be any creativity at all? Furthermore, the notion of the mad genius appears inconsistent with some aspects of creativity, and thus would seem to undermine a secondary Darwinian theory of creativity. Specifically, creativity demands the origination of ideas that are both novel and adaptive. While madness might certainly generate novelty, it is not at all apparent that those novel notions would satisfy essential scientific or aesthetic criteria. The inhabitants of mental institutions are often the source of original ideas, but these ideas are not considered sane—rather, quite the contrary. Indeed, it is far from clear whether people with severe psychological disorders can

Table3.1 alleged psychopathology among eminent creators

Schizophrenic disorders (and other cognitive psychoses)

Scientists: Tycho Brahe, Cantor, Copernicus, Descartes, Faraday, W. R. Hamilton, Kepler, Lagrange, Linnaeus, Newton, Pascal, Semmelweiss, Weierstrass, Horace Wells; Thinkers: Kant, Nietzsche, Swedenborg; Writers: Baudelaire, Lewis Carroll, Hawthorne, Holderlin, S. Johnson, Pound, Rimbaud, Strindberg, Swift; Artists: Bosch, Cellini, Diirer, Goya, El Greco, Kandinsky, Leonardo da Vinci, Rembrandt, Toulouse-Lautrec; Composers: Donizetti, MacDowell, Felix Mendelssohn, Rimsky-Korsakov, Saint-Saens.

Affective Disorders (depression, mania, or bipolar)

Scientists: Boltwood, Boltzmann, Carothers, C. Darwin, L. De Forest, J. F. W. Herschel, Julian Huxley, T. H. Huxley, Jung, Kammerer, J. R. von Mayer, V. Meyer, H. J. Muller, J. P. Muller, B. V. Schmidt, J. B. Watson; Thinkers: W. James, J. S. Mill, Rousseau, Schopenhauer; Writers: Balzac, Barrie, Berryman, Blake, Boswell, Van Wyck Brooks, Byron, Chatterton, Coleridge, William Collins, Conrad, Cowper, H. Crane, Dickens, T. Dreiser, R. Fergusson, F. Scott Fitzgerald, Frost, Goethe, G. Greene, Hemingway, Jarrell, Kafka, Charles Lamb, Jack London, Robert Lowell, Maupassant, O’Neill, Plath, Poe, Quiroga, Roethke, D. G. Rossetti, Saroyan, Schiller, Sexton, Shelley, C. Smart, T. Tasso, V. Woolf; Artists: Michelangelo, Modigliani, Pollock, Raphael, Rothko, R. Soyer, Van Gogh; Composers: Berlioz, Chopin, Elgar, Gershwin, Handel, Mahler, Rachmaninoff, Rossini, Schumann, Scriabin, Smetana, Tchaikovsky, Wolf.

Personality Disorders (including severe neuroses)

Scientists: Ampere, Cavendish, A. S. Couper, Diesel, Einstein, Frege, Freud, Galton, Heaviside, Huygens, Marconi, Mendel; Thinkers: J. Austin, Beccaria, Comte, Descartes, Hegel, Hobbes, Hume, Kierkegaard, B. Russell, Spencer, Voltaire, Wittgenstein; Writers: H. C. Andersen, E. B. Browning, R. Browning, Bunyan, Carlyle, Dickinson, Dostoyevsky, T. S. Eliot, Emerson, Flaubert, Garcia Lorca, Gide, Allen Ginsberg, Gogol, Heine, G. M. Hopkins, A. Huxley, W. M. Inge, Melville, Pavese, Proust, S. Richardson, Rimbaud, Ruskin, Tennyson, Tocqueville, Tolstoy, Verlaine, Tennessee Williams, Zola; Artists: Borromini, Bramante, Caravaggio, Cezanne, Coco Chanel, Munch, Romney; Composers: Beethoven, Bruckner, Orlando de Lasso, Schubert, Wagner.

even make satisfactory judgments about what is right and wrong, true and false, beautiful and ugly.

Given these difficulties, I have two goals in this section. First, I will provide a brief review of the research results suggesting that some kind of relationship exists between psychopathology and genius. Second, I will give an overview of a recent theory proposed by Hans Eysenck, who has offered a foundation for that relationship from the standpoint of secondary Darwinism.

Empirical findings. It is easy to conjure up long lists of creative geniuses who have exhibited some guise of mental illness. Table 3.1 lists some of the more frequently cited examples. I must stress that few of these cases are founded on modern clinical diagnosis, and many are extremely conjectural. On the other hand, many other examples are quite secure. Some creators were actually placed in mental institutions, like Van Gogh and Schumann, while others either committed suicide, like Hemingway and Virginia Woolf, or became severe alcoholics, like Modigliani and Mussorgsky. Still others displayed symptoms of abnormality too conspicuous to be ignored, including the paranoia of Newton and the hysteria of Tchaikovsky.

Darwin’s own pathology may fall in this category. He would often be prostrated by mysterious attacks of dizziness, heart palpitations, violent shivering, nausea, vomiting, and hysterical crying. These episodes were usually brought on by the mere act of conversing with people—revealing their psychosomatic origins. To avoid a recurrence, Darwin did everything possible to isolate himself from all but the briefest social interactions, even with his nearest relations. It is possible that Darwin suffered from the hyperventilation syndrome, which ensues from excessive arousal of the sympathetic nervous system. Whatever the correct diagnosis, there is no doubt that Darwin was emotionally and mentally unwell. His attacks would often incapacitate him to the point that he could not work at all. Even when the consequences were not so drastic, he avoided society to an unhealthy degree. Darwin himself confessed that he was “confined to a living grave.”

At the very least, the names in table 3.1 suggest that people with mental disturbances can still attain distinction as creators. But do such lists imply any inference stronger than that? Probably not. There are thousands upon thousands of individuals who have gone down in history as creative geniuses. Hence, even if the incidence rate of psychopathology were far lower among such creators than in the general population, we should still be able to compile impressive lists of supposed mad geniuses. Thus, we cannot draw any conclusions without having the proper baselines or control groups. This methodological essential has been achieved in the following three sets of empirical investigations:

• 1. For nearly a century, behavioral scientists have conducted historio-metric research on the likelihood that notable creators will exhibit some form of psychopathology. The bulk of these inquiries have claimed incidence rates higher than that found in the general population. The most recent historiometric inquiry on this subject is that reported by Arnold Ludwig in his 1995 book The Price of Greatness. After studying the lives of over a thousand eminent personalities, he showed that (a) the rate of mental disorder exceeded expectation and (b) a positive association exists between the presence of pathological symptoms and the magnitude of achievement. Just as critical, he found considerable variation across different domains. In general, notable leaders displayed less evidence of disorder than did famous creators—something found by other historiometric inquiries as well. Moreover, among the various types of creative activities, illustrious scientists appeared more stable than did distinguished artists and authors. To offer specifics, whereas only 28% of the notables in the natural sciences experienced some sort of mental disorder, psychological problems plagued 60% of the composers, 73% of the visual artists, 74% of the playwrights, 77% of the fiction writers, and 87% of the poets. The last figure seems to endorse Thomas Macaulay’s comment: “Perhaps no person can be a poet, or can even enjoy poetry, without a certain unsoundness of mind.” It is interesting to note that Ludwig has more recently shown how parallel differentials hold within specific creative domains, depending on whether the emphasis is on logic, objectivity, and formalism or on intuition, subjectivity, and emotionalism. In painting, for example, only 22% of painters practicing a formal style exhibited some form of mental disorder, whereas this incidence rate increased to 52% for those painting in a symbolic style and to 75% for painters who expressed themselves in an emotive style. According to Ludwig, the data display the fractal characteristic of “self-similarity.” The same pattern that separates leaders from creators, and then divides scientific creators from artistic creators, also permeates down various levels of resolution, culminating in the stylistic contrasts within a single creative domain, such as in painting.

• 2. Psychiatric research lends additional support to the hypothesis that outstanding creativity bears some relationship to psychopathology. Unlike the historiometric studies, these investigations examine samples of eminent contemporaries, using diagnostic criteria to establish the appearance of certain syndromes. These psychiatric inquiries have tended to focus on successful artists and writers, and within this group the inclination toward affective disorders (including bipolar or manic-depressive) is conspicuous, along with corresponding tendencies toward alcoholism, drug abuse, and suicide. Among the most prominent contributors to this research literature is Kay Jamison, who has persuasively argued for a firm linkage between affective disorders and the artistic temperament. Although some have criticized this research on methodological grounds, the findings are not at all discrepant with what has been found using more rigorous methods.

• 3. Like the psychiatric work, psychometric investigations apply their techniques to contemporary creators of note. An exquisite example is the research conducted at the Institute for Personality Assessment and Research (IPAR) at the University of California at Berkeley. Eminent architects, writers, mathematicians, and other creators were invited to IPAR for intensive psychometric evaluation. Besides submitting to interviews and other assessments, the participants took a large inventory of standard psychometric tests, such as the Minnesota Multipha-sic Personality Inventory (MMPI). Because these tests had norms based on general populations, standards of comparison were available. In addition, the IPAR researchers would invite less successful colleagues in the same creative domains to serve as matched controls. The IPAR studies again demonstrated that some degree of psychopathology is associated with creative achievement. The creative writers, for example, scored higher than normal on all the clinical subscales of the MMPI (namely, depression, hypomania, schizophrenia, paranoia, psychopathic deviation, hysteria, hypochondriasis, and psychaesthenia). Comparable findings have been reported by other researchers. Highly creative people show elevated scores on dimensions that indicate the presence of psychopathological symptoms. And elevated scores are especially common among those active as artistic creators. “What garlic is to salad, insanity is to art,” said sculptor Augustus Saint-Gaudens.

To sum up, the historiometric, psychiatric, and psychometric literature converge on the same conclusion: The genius-madness link may be more than myth. Furthermore, the association between creativity and psychopathology is particularly prominent in those domains where the creative process must be the most Darwinian, as in poetry, fiction, and the visual arts—and especially in the most intuitive, subjective, and emotional styles of artistic expression. However, to comprehend better the relationship between psychopathology and creativity, two qualifications must be recognized.

First, although creative geniuses seem to exhibit above-average levels of psychopathological symptoms, these levels are seldom so high as to translate into total mental and emotional deterioration. Indeed, if they do suffer from such extreme degrees of disturbances, their creative careers terminate, whether by suicide or by complete intellectual or emotional incapacitation. Consequently, creative individuals tend to exhibit symptoms midway between those of the normal and abnormal personality. Dryden captured the fine distinction when he composed the often-quoted lines “Great Wits are sure to Madness near ally’d, I And thin Partitions do their Bounds divide.”

Second, creative geniuses tend to possess other cognitive and emotional resources that help to channel and contain any potential psychopathology. Besides superior intelligence, eminent creators will possess considerable “ego-strength” and other traits of personal fortitude and self-discipline. These moderating attributes enable creators to exploit the strange ideas that fill their heads without allowing those ideas to take over the organization of their personality. Such traits also allow a person to engage in the full range of behaviors required for creative accomplishment. Albert Rothenberg has provided excellent examples in his psychiatric studies of the creative process in both schizophrenic and authentic poets. Unlike true poets, schizophrenic poets refuse to revise their initial drafts, revealing an inability to adopt a more objective perspective on their work. They are all inspiration without verification, variation without selection.

Unfortunately, although the link between psychopathology and creativity has been empirically determined, the precise causal foundation of the mad-genius syndrome has yet to be established. In fact, the number of potential causal explanations is actually rather large. For example, it could be that the achievement of eminence is profoundly stressful, and that the stress can sometimes provoke a mental breakdown. Under such a causal scenario, genius is only incidentally mad. Nonetheless, a distinguished researcher recently advanced a theory arguing that psychopathology may actually make a positive and direct contribution to the manifestation of creative genius.

Theoretical interpretation. Hans Eysenck was a pioneer in the empirical study of personality. Among his many contributions was the development of the Eysenck Personality Questionnaire, or EPQ. One subscale of this instrument has special pertinence for the question at hand, namely that which measures individual differences in psychoticism. At the low end of this dimension are those people who are socialized, conventional, and conformist but also empathic and even altruistic. Such low scorers are not only normal, but perfectly well adjusted for life in human society. At the high end of this dimension are those who are impulsive, egocentric, antisocial, impersonal, hostile, and aggressive, at times to a criminal degree, and who, at the higher extremities, display tendencies toward psychopathic, affective, and schizophrenic disorders. According to Eysenck, although high scores are associated with the appearance of pronounced psychopathology, moderately high scores are also linked with the manifestation of exceptional creativity. A considerable amount of psychometric research, in fact, supports this assertion.

Eysenck did not believe that the correlation between psychoticism and creativity is a mere empirical curiosity. On the contrary, Eysenck constructed a Darwinian theory that builds directly on the previous variationselection models—he styled it the Campbell-Simonton theory. Eysenck began by recognizing the creator’s need for an exceptional reservoir of remote associations that will provide a rich source of divergent responses. He then reviewed an impressive number of empirical studies showing that psychoticism positively correlates with performance on word-association tests as well as on various measures of divergent thinking. Individuals scoring higher on psychoticism do indeed seem more capable of producing a large quantity of unusual ideas.

Furthermore, such intellects exhibit several cognitive quirks that make their thought processes depart from the norm—sometimes in a fashion that approximates the pathological. For instance, high scorers on psychoticism are more likely to engage in “allusive” or “overinclusive” thinking, in which the sharp distinctions between separate ideas are loosened, yielding overgeneralized concepts. They seem to lack the strong “filter mechanism” that keeps ideas within their usual conceptual boundaries. This cognitive deficiency is related to other peculiarities about the ways that high scorers process information. For example, higher psychoticism is associated with a weakened ability to inhibit environmental stimulation (i.e., reduced “negative priming” or “latent inhibition”). This means that their intellects are more subjected to the random influx of extraneous stimuli, which would in turn elicit incongruous associations. Of course, taken to the extreme, these same cognitive attributes will produce the kinds of symptoms that make life so difficult for the psychotic. But at less conspicuous levels, these same proclivities would permit the production of numerous ideas that are highly unexpected. People with just the right amount of psychoticism would be prone to have all sorts of seemingly irrelevant ideas pop into their heads almost randomly, and without control. Even if not meaningful in themselves, this chaotic influx of ideas can prime new chains of associations that lead to insights otherwise missed.

Indeed, some of these ideational variants will be bizarre enough to count as cognitive analogs of the genetic mutations in organic evolution. Consequently, the intellectual variations would satisfy the most controversial claim of Campbell’s Darwinian theory of creativity—that the process operates with a certain degree of blindness. Some valuable proportion of the ideas that pass through the mind of a creative genius would be like “bolts out of the blue.”

Eysenck’s theoretical discussion concentrated on the cognitive implications of a certain degree of psychoticism. Nonetheless, the very nature of this dimension suggests two other ways that psychoticism may be associated with a high order of creativity:

1. The creative genius must impose few constraints on the range and freedom of thought. Conformity to conventional norms and attitudes will unduly restrict the number and diversity of ideational variations. Furthermore, during the selection phase of the variationselection process, many original variants will be summarily dismissed as inconsistent with some arbitrary but traditional standard and will thus never see the light of day. Hence, insofar as psychoticism is associated with independence, unconventionality, anticonformity, and iconoclasm, high scores on this dimension should enhance a person’s ability to engage in Darwinian creativity. Like Henrik Ibsen’s protagonist in An Enemy of the People—or the real-life example of Ignaz Sem-melweiss confronting the medical establishment over the contagious nature of puerperal fever—the great innovator must be willing to take issue with the received tradition, even when it requires an epic battle against all the powers that be. It is no accident, for instance, that the greatest thinkers in the Western intellectual tradition tend to be those who refuse to conform to the contemporary philosophical Zeitgeist, who advocate extreme positions, and who insist on combining those positions in highly unconventional ways.

2. The creative genius must often be willing to work hard and long on ambitious projects that verge on the unrealistic—a scientific magnum opus, an epic novel, a grandiose opera. To the extent that psychoticism is correlated with the appearance of manic affective disorders, individuals with this disposition will more likely be motivated to take on such lofty tasks. After all, during manic episodes creators would enjoy heightened levels of energy combined with intensified optimism about what they can accomplish. Indeed, the biographies of eminent creators often report periods in which they work on projects feverishly, and with little or no attention to life’s pragmatics. Beethoven would often compose his masterpieces in an elevated state of “rapture.” Once while Beethoven was working on the Missa Solemnis, his friend Schindler reported:

I arrived at the master’s home in Modling. It was 4 o’clock in the afternoon. As soon as we entered we learned that in the morning both servants had gone away, and that there had been a quarrel after midnight which had disturbed all the neighbors, because as a consequence of a long vigil both had gone to sleep and the food that had been prepared had become unpalatable. In the livingroom, behind a locked door, we heard the master singing parts of the fugue in the Credo—singing, howling, stamping. After we had been listening a long time to this almost awful scene, and were about to go away, the door opened and Beethoven stood before us with distorted features, calculated to excite fear. He looked as if he had been in mortal combat with the whole host of contrapuntists, his everlasting enemies. His first utterances were confused, as if he had been disagreeably surprised at our having overheard him. Then he reached the day’s happenings and with obvious restraint he remarked “Pretty doings, these, everybody has run away and I haven’t had anything to eat since yesternoon!”

Moving beyond the realm of mere anecdote, a detailed quantitative analysis in fact identified a conspicuous relationship between manic episodes and prolific output in the career of Robert Schumann, whose bouts with manic depression eventually led to his attempted suicide and institutionalization. The composer’s creative effusions of 1840 and 1849, in particular, took place during manic highs.

After extensively illustrating the consequences of high psychoticism, I believe it is instructive to offer an example of a historical personality who would probably have scored extremely low on this dimension had the EPQ been available 150 years ago. The person is John Stearns Henslow, Darwin’s botany professor at Cambridge, and the individual who nominated Darwin to serve as naturalist aboard the Beagle. Darwin recalled that Henslow’s “knowledge was great in botany, entomology, chemistry, mineralogy, and geology,” and “his judgment was excellent, and his whole mind well-balanced.” Moreover, according to Darwin, “his moral qualities were in every way admirable. He was free from every tinge of vanity or other petty feeling.” Darwin “never saw a man who thought so little about himself or his own concerns. His temper was imperturbably good, with the most winning and courteous manners . . . [and his] benevolence was unbounded.” Yet Henslow was also “deeply religious, and so orthodox, that... he: should be grieved if a single word of the Thirty-nine Articles were altered.” Darwin also admitted that he did “not suppose that any one would say that he possessed much original genius.” Indeed, it was actually Henslow who was first invited to serve aboard the Beagle, but he declined the opportunity to take so venturesome an action. One can only speculate how the history of evolution would have differed had Henslow been the naturalist to scramble over the Galapagos Islands. In fact, because Henslow could not advance beyond the restraints of his conservative spirit, he never was able to accept the bold evolutionary theory put forward by his most distinguished student.

Because Eysenck has provided extensive discussion of his theory, I cannot do full justice to his ideas here. Nor do I have the space to examine all the criticisms that might be inveighed against his theory. I will only direct the reader to his 1995 book Genius: The Natural History of Creativity and to a target article and ensuing commentary that appeared in a 1993 issue of Psychological Inquiry. Such additional readings will document the profound manner in which Eysenck connects his Darwinian theory with research on (a) individual differences in personality, (b) the cognitive processes underlying creativity, and (c) the genetic and environmental factors behind the emergence of creative genius. Eysenck himself admitted that his attempted integration may sometimes amply illustrate the phenomena of remote association and overinclusive thought. Nonetheless, the resulting synthesis is expansive, bold, creative—and explicitly Darwinian, in the secondary sense.

Evolution

As noted earlier, the personality profile of the creative genius contains certain features that seem inconsistent with what would be predicted according to the principles of organic evolution. Of all those features, it is creativity’s affiliation with psychopathology that causes the most theoretical discomfort. Hence, the big question that comes to the fore is how to make the mad-genius relationship square with primary Darwinism. Whatever the assets of creativity, overt psychopathology does not appear to constitute adaptive behavior. So why would evolution favor its emergence?

One facile solution would be simply to deny that either madness or genius represent genetically transferable traits. The natural selection that drives organic evolution can only operate on inheritable variation among individuals. If both psychopathology and creativity are exclusively products of environmental influences, then there would be no necessity to look for circumstances that render these attributes adaptive. But this response will not do. There is already ample empirical evidence that both syndromes may exhibit a strong genetic foundation. Indeed, according to family pedigree studies, mental illness and exceptional achievement tend to appear in higher than normal frequencies in the same lineages.

Take, for example, a comprehensive Icelandic survey. Because the number of Icelanders does not get much beyond 200,000, and because immigration and emigration are relatively small, it is possible to trace genealogies for the bulk of its population. In addition, the mental hospital in Reykjavik can keep comprehensive records of mental disturbances among Icelandic citizens. These records were compared against luminaries who had made names for themselves as novelists, poets, painters, composers, and performers—according to Who Is Who in Iceland. It was thus possible to trace the genealogies of both achievement and insanity with a completeness unimaginable in gigantic and unstable populations like that of the United States. The family pedigrees were revealing. The relatives of schizophrenics were two to six times more likely to earn a place of creative distinction in Icelandic society. The incidence of insane relatives definitely exceeded what occurred in families whose members had not left their marks. Because compatible results have emerged in the survey of family pedigrees in other cultures, this co-occurrence cannot be a peculiarity of Icelandic culture. Illustrious creators do seem to originate from families in which mental disorders are rampant—at least in comparison to pedigrees that produce less outstanding minds.

There are many dramatic cases of these convergent pedigrees in the annals of human history. One especially fascinating example is the Huxley family. Thomas Henry—“T. H.”—Huxley often suffered from depressive states, and his father had died in an asylum. Moreover, of T. H.’s seven siblings, only one sister could be considered relatively normal. T. H.’s own daughter, Marian, became extremely melancholic, lost her sanity, and died young. One of T. H.’s grandsons, Trevenen was also melancholic, and committed suicide. A second grandson, Julian, attempted suicide and suffered from depression. And a third, Aldous, experimented with hallucinogenic drugs and the occult. Several other temperamental disorders permeate the family history. Yet all of these disturbances notwithstanding, the Huxley family is distinguished for its output of first-rate creative minds. Sir Andrew Fielding Huxley, another grandson of T. H., shared the 1963 Nobel Prize for physiology or medicine. Sir Julian Huxley was a noted biologist who became secretary of the Zoological Society of London and the first director general of UNESCO. Aldous Huxley, though not knighted, became a famous author, most notably of the novel Brave New World. And, of interest to our discussion, two members with Huxley pedigree made critical contributions to the development and dissemination of Darwinian theory. In 1942, Sir Julian wrote Evolution: The Modern Synthesis, which helped establish the current version of the theory. In this he was following in his grandfather’s footsteps, for T. H. was an early champion of the original theory of evolution by natural selection. So prominent and successful was his advocacy that T. H. earned the nickname “Darwin’s bulldog.”

The frequent appearance of such pedigrees implies that creativity and psychopathology feature a common genetic component. Hans Eysenck has actually proposed a sequential model that links these two individualdifference variables to underlying intellectual faculties (involving weakened cognitive inhibition), which is itself a consequence of a particular psychoneurological constitution (entailing hippocampal functioning and the neurotransmitters dopamine and serotonin). The latter, finally, is a function of genes that provide the basis for individual differences in the organization and operation of the nervous system. Eysenck’s model is little more than a highly speculative sketch, but it at least provides a theoretical framework for understanding how there might exist a shared genetic foundation for the creative hits and crazy misses seen in such family pedigrees, as the Huxley line.

Whatever the fate of Eysenck’s genetic model, the fact remains that individual differences in the capacity for genius and madness seem to share a hereditary component. This means that we cannot ignore the evolutionary consequence of having such genes in the population. If psychopathology is maladaptive, and if both psychopathology and creativity come with the genotype, it would seem that both behavioral phenomena would became increasingly rare. Depending on the selection pressures, creative genius might eventually disappear from the gene pool.

Curiously, one potential solution to this problem was offered in a 1964 Nature paper published by two distinguished evolutionists, Julian Huxley and Ernst Mayr, in collaboration with two psychiatrists who had special expertise in schizophrenia. After noting the strong evidence for the her-itability of this common mental disorder, the authors observed that the gene for schizophrenia appears too frequently to be maintained by mutation alone. They accordingly examined the benefits and costs of possessing a genetic inclination toward schizophrenic disorder. For example, they discussed data showing that schizophrenics may be more physiologically robust than normal members of the population. The authors also suggested that the low fertility of schizophrenic males may be more than compensated by the high fertility of schizophrenic females. Therefore, Huxley, Mayr, and their coauthors concluded that the high incidence rate of schizophrenia could be “the result of a balance between its selectively favourable and unfavourable properties.”

Unfortunately, these authors focused on the biological repercussions of mental disorder. It could just as well be argued that psychopathological symptoms may have certain social consequences that contribute to the survival value of the corresponding genes. This very possibility was put forward by Hammer and Zubin in 1968. They looked at psychopathology as a manifestation of a more general syndrome of what they styled “the cultural unpredictability of behavior.” Some individuals in a population inherit a certain tendency to do the unanticipated according to societal norms and role expectations. Although psychopathology is one manifestation of this genetic inclination, it is not by any means the only one. The innate proclivity for unpredictability may take the form of creative genius, which can prove adaptive both to the individuals and to the culture that produces them. Whether this genetic endowment is positive or negative depends on certain cultural circumstances that channel the tendency in different directions. Hammer and Zubin directly compared this phenomenon with sicklecell anemia. Although the gene for this ailment is disadvantageous in temperate environments, it acquires a selective advantage in tropical climates where people heterozygous on this trait can gain increased resistance to malaria. In a sense, the mental disorder of some humans is the price that society pays for the benefits of creative genius.

More recently, Geoffrey Miller has argued for the selective advantage of a genetic trait quite similar to cultural unpredictability. But his explanation was founded on the notion of “Machiavellian intelligence.” According to the latter concept, human beings have had to evolve extremely complex cognitive and behavioral skills to survive the interpersonal politics of primate social systems. Such intricate systems require considerable acumen and dexterity to negotiate the elaborate web of cooperative and competitive activities that define an individual’s place in the dominance hierarchies. Because a premium is placed on being able to “outsmart” rivals, the social primates have evolved a number of strategies that help prevent the disclosure of intentions. Among those strategies may be “proteanism”—the capacity to be unpredictable when necessary in a given social situation. Moreover, social proteanism would be useful in a variety of circumstances besides domestic politics. A warrior locked in mortal combat, for example, would certainly benefit if an enemy were unable to anticipate his next move.

One striking feature of Miller’s theory of protean behavior is that it provides an evolutionary explanation for the emergence of a mechanism that can generate true randomness, the prerequisite for the production of genuinely unpredictable behavior. Many psychologists have claimed that human beings are incapable of producing unpredictable behaviors, but this only holds in highly artificial situations (e.g., when individuals are asked to write a sequence of random numbers). In contrast, this skill appears to be highly developed in social interactions in which there are painful costs to being successfully anticipated (e.g., business, sports, and games). Miller even made a specific connection between this intellectual capacity and Donald Campbell’s blind-variation-and-selective-retention model. In doing so, Miller offered a primary Darwinian mechanism that would support the evolutionary emergence of the secondary Darwinian process of creativity.

The only feature missing from Miller’s theory is an explicit treatment of psychopathology. However, it is not too far-fetched to conjecture that mental disorder may be the unfortunate consequence of inheriting too much proclivity for proteanism. Presumably, the optimal amount of protean behavior in the population would represent some equilibrium point between two maladaptive extremes on this trait—between psychosis and its opposite. At some point on the continuum between the optimally protean and the outright psychotic may then emerge the creative genius. This interpretation suggests that creativity need not have any special adaptive advantage from the standpoint of natural selection. Rather the creative genius may reside at the delicate neutral point between the highly fit protean intellect and the sadly unfit pathological mind. Perhaps only the greatest creators can stand at the very pivot between success and failure.

This page intentionally left blank

Aou care for nothing but shooting, dogs, and rat-catching, and you will be a disgrace to yourself and all your family.” Such was the prophesy of Darwin’s father, who had lost all patience with his young son’s inability to show direction and aptitude. Matters did not improve when Darwin entered his college years. He first enrolled at Edinburgh with the halfhearted intention of becoming a physician, like his father. Finding medical studies quite unattractive, Darwin proceeded to Cambridge at his father’s urging, with the plan of becoming a minister. But that career goal, too, soon ran aground. After he earned an undistinguished bachelor’s degree at age 22, he had no firm idea of what to do with himself. As is not surprising, when Darwin was invited to serve as the Beagle naturalist, his father was not inclined to give his consent. Rather than pursue a true profession, his son was apparently going to spend the next five years with only room and board as pay, and with no prospect of further advancement. His father made just one fateful concession, “If you can find any man of common -sense who advises you to go I will give my consent.” Fortunately, Darwin’s uncle was considered “one of the most sensible men in the world,” and it was he who persuaded Darwin’s father to relent. At this point, in the year 1831, Darwin’s biography becomes history.

In modern terminology, I guess we would say that the young Darwin did not strike his father as particularly gifted or talented—except at wasting time and money. Had the senior Darwin wanted to, he certainly could have compared his son to Charles’s young cousin, Francis Galton. Although more than a dozen years younger, Galton had already proven himself to be far more precocious. For example, several years earlier he had written the following letter to his sister:

MY DEAR ADELE,

I am 4 years old and I can read any English book. I can say all the Latin Substantives and Adjectives and active verbs besides 52 lines of Latin poetry. I can cast up any sum in addition and can multiply by 2,3,4,5,6,7,8, [9], 10, [11].

I can also say the pence table. I read French a little and I know the clock.

FRANCIS GALTON, Febuary 15,1827.

The numbers in brackets were those that Galton, in a display of second thoughts, erased from the letter; he used a knife to scratch out one number and, evidently finding this unsatisfactory, glued paper on top of the other number. Only one misspelling appears, the month that this letter was written (an error that some adults still make). There may even be a little brilliance hidden in the dating of this letter, which was written the day before his fifth birthday—as if young Francis were trying to extract the most possible credit for the precocity of a mere four year old.

In chapter 1,1 mentioned how one of Lewis Terman’s students had used biographical data about childhood and adolescent accomplishments to calculate Darwin’s IQ. He scored 135. Using the same methodology, Terman himself estimated Gabon’s IQ as close to 200—or four standard deviations higher. Thus, in today’s language, Galton would certainly be called a gifted or talented child, if not an unequivocal case of a child prodigy. Even so, Gal-ton’s ultimate achievements pale in comparison to those of his older cousin. By Gabon’s own criterion of genius, Darwin was the greater genius of the two. Only Charles Darwin belongs in that exclusive pantheon that includes Copernicus, Galileo, Newton, and Einstein. Wordsworth once said, “The Child is father of the Man.” Yet that observation seems invalid here. The discrepant childhoods of Darwin and Galton do not predict the equally discrepant but inverted adulthoods.

This curious inconsistency in outcomes certainly raises the larger issue of talent development. What is the connection between the events of youth and the attainments of maturity? What are the factors responsible for the conversion of giftedness into genius? And, most critical, do Darwinian principles participate in any significant manner in the growth process? To

~ Development ~                         111

address these questions properly requires that I first examine development from two contrary perspectives.

Nature and Nurture

Perhaps the most troublesome single debate in developmental psychology is the so-called nature-nurture issue. What are the relative contributions of genetic endowment versus early experience to the development of the human organism? This recurrent controversy has its natural counterpart in the phenomenon of genius. On the one side are those who would second John Dryden’s assertion that “genius must be born, and never can be taught.” Galton himself was initially in this camp. The reason Galton named his 1869 book Hereditary Genius was because he thought genius was precisely that—hereditary. If supreme natural ability drove the inevitable attainment of distinction, and if individual differences in natural ability are inherited from parents, then genius should run in families. A major portion of Gabon’s book, in fact, is expressly devoted to presenting long lists of eminent personalities who emerged from distinguished family lines. Indeed, his chapter headings read like a list of the ways one can attain a durable reputation: Statesmen, Judges, Commanders, Divines, Men of Science, Literary Men, Poets, Musicians, and Painters.

Naturally, in the “Men of Science” chapter, Galton gave the Darwin family its due. He noted that Charles Darwin (born 1809) was the grandson of Erasmus Darwin (born 1731), the poet and scientist who was an early advocate of evolutionary ideas. Galton also mentioned that Charles had a number of promising sons, although at the time the book was written Galton could not anticipate the scope of their achievement. Charles Darwin’s immediate progeny included: Sir George Howard (born 1845), a distinguished mathematician and astronomer; Sir Francis (born 1848), a famous botanist; Leonard (born 1850), a noted engineer, economist, and eugenicist; and Sir Horace, (born 1851) an eminent civil engineer. Another illustrious Darwin necessarily omitted from Hereditary Genius was Charles’s grandson through George, George Galton Darwin, who was a notable physicist. Less explicable is the exclusion of Charles Darwin’s maternal grandfather, Josiah Wedgwood (born 1730), the potter who founded the Etruria factory that produced the celebrated “Wedgwood ware.” Another descendent of Wedgwood’s was Thomas Wedgwood (born 1771), an illustrious physicist and inventor. On the other hand, it was probably Victorian modesty that obliged Galton not to cite his own name in this pedigree, because he avoided doing so even in the 1892 edition that came out when he was quite famous himself. The most he would do was to say that he “could add the names of others of the family who, in a lesser but yet decided degree, have shown a taste for subjects of natural history.” Galton and Charles Darwin were both grandsons of Erasmus Darwin, albeit through a different grandmother.

At first glance, the extensive lists of such lineages make a persuasive case on behalf of the innate genius thesis. Charles Darwin was himself quite convinced upon reading Gabon’s book, informing his cousin that “I do not think I ever in all my life read anything more interesting and original.” In his autobiography Darwin further endorsed this genetic determinism when he admitted that he was “inclined to agree with Francis Galton in believing that education and environment produce only a small effect on the mind of anyone and that most of our qualities are innate.” Yet although other researchers have replicated Gabon’s findings, Galton also quickly learned that other scientists would challenge his hereditarian stance. In his next book, English Men of Science, Galton yielded a little ground when he looked at some of the environmental factors that might contribute to the emergence of genius. In fact, the subtitle to this 1874 book is telling: “Their Nature and Nurture.” Although these words are first associated together in Shakespeare’s Tempest, it was Galton who introduced these terms into the language of behavioral science. As Galton explained, “the phrase ‘nature and nurture’ is a convenient jingle of words, for it separates under two distinct heads the innumerable elements of which personality is composed. Nature is all that a man brings with himself into the world; nurture is every influence from without that affects him after his birth.”

Unfortunately for Galton, although he tried to integrate these two perspectives, later researchers increasingly tended to emphasize nurture over nature. Only recently has hereditarianism witnessed some revival in the face of the hegemony exercised by environmentalism. So let us first examine the nurture position before returning to the nature position.

Environmentalism

To comprehend the origins of genius from the perspective of a Darwinian environmentalism, I must begin by asking the same type of question examined in the preceding chapter. There the goal was to isolate those individual-difference variables that might facilitate a person’s ability to engage in the creative process, as defined by Donald Campbell’s variationselection model. Similarly, certain kinds of environmental experiences may be more likely to enhance an individual’s capacity for Darwinian creativity. Those experiences would presumably increase the number and diversity of ideational variations that could emanate from a creator’s mind. These developmental experiences may facilitate this potential in two ways, the direct and the indirect. Direct developmental effects are those that expand the intellectual capacity for remote association and divergent thinking—the very cognitive processes that produce ideational variations. Indirect effects, on the other hand, are those that may encourage the development of the Darwinian personality that optimally supports engagement in the variational process.

Nevertheless, the search for both direct and indirect environmental factors must also take into consideration the domain of creative activity (see figure 3.3). If domains vary tremendously in the amount of Darwinian creativity required—from the minimal level of normal science to the maximal level of romantic art—then the developmental circumstances must differ in the same manner. The value of considering domain contrasts will become apparent when I scrutinize some of the factors that contribute to the development of creative talent. Below I examine the potentially nurturing consequences of enriched home environments, adversity and trauma, education, and marginality.

Enrichment. In chapter 1,1 mentioned Lewis Terman’s classic longitudinal study of over a thousand intellectually gifted children. Among the mounds of data he collected about these high-IQ children was information on the quality of their home environments. Their parents tended to have higher than average levels of formal education, and at least one parent tended to work at an intellectual profession, such as a doctor or lawyer. The homes of these bright children were also likely to contain private libraries well stocked with books of all kinds. Other studies of the gifted have found similar results. The parents highly value learning and supply their homes with intellectually and culturally stimulating magazines, games, and similar materials. Family outings will often include visits to museums, exhibits, galleries, libraries, and other places that stimulate intellectual development. Moreover, studies of eminent creators report similar findings of geniuses originating from such enriched family environments.

All early stimulation no doubt makes some contribution to intellectual development. Insofar as intelligence constitutes a primary Darwinian trait required for general success in any domain, such home environments can be said to contribute to the development of creative genius as well. Yet I have already pointed out that intelligence alone does not support creative genius. The knowledge in the brain must be organized in a structure that permits remote association and divergent thinking. As noted previously, few of Terman’s children attained the highest levels of eminence, and those who did tended to achieve distinction in conventional occupations—professions in which creative ideas are not at a premium. The few who became notable for their creativity, moreover, were more likely to be scientists than artists. Therefore, the development of creative talent requires that the home feature enriching experiences that encourage the diversification of the intellect. The most diverse environments will be those of artistic creators, whereas the homes of scientific creators will fall somewhere between those of the artists

114                           ~ Origins of Genius ~

and those of individuals who fail to display any pronounced levels of creativity.

There is already evidence for the existence of such a ranking of home environments. For example, one study examined U.S. male high school students who demonstrated special talent in either scientific or artistic creativity. Compared with a control group of students who demonstrated no particular talent, both groups of talented teenagers came from homes that were academically and intellectually superior. The parents were likely to be college educated, tended to engage in reading as a leisure-time activity, and were prone to offer models of creativity and interest in the student’s chosen field. The parents were more likely than those of the control group to play a musical instrument, to engage in creative hobbies, and to visit art museums or galleries. Thus, the general pattern was a home environment that would provide rich stimulation for scientific and artistic talents alike.

However, in terms of environmental diversity, the homes of the young creative artists were more distinctive than those of the young creative scientists, the latter coming closer to what is normal for academically superior students, such as those studied by Terman. At the one end, a larger proportion of the parents of the artistic talents were born in a city, state, or country different from the one in which the family currently resided. The artistically talented teenagers were themselves more prone to have traveled to various parts of the country, as well as to have visited more distant locales. In addition, these same teenagers were more likely to have lived in more than one state during their childhood and adolescence. At the other end, a smaller proportion of the parents of the scientific talents were born in a foreign country, and a larger proportion were actually born in the city in which they were currently living. Moreover, the parents tended to have more conventional occupations and interests. For example, the young scientific creators were more likely “to have fathers who majored in business subjects in college, who play bridge as a hobby, and from whom the students had learned about sports.” All in all, the data make it clear that the creative artists come from homes that have a greater likelihood of supporting the development of the mental capacities required for Darwinian creativity.

Adversity. Ironically, one of the reasons so few of Terman’s gifted children grew up to become geniuses may be that they had it too good in childhood. The “Termites” were not only physically robust, but in addition most grew up in ideal homes of stable marriages and financial security. The English poet Dylan Thomas once warned: “There’s only one thing that’s worse than having an unhappy childhood, and that’s having a too-happy childhood.” Whatever the intellectual talents of Terman’s children, their potential for genius may have been destroyed by a superfluity of happiness.

There is, in fact, empirical reason for believing that the development of

genius may sometimes be enhanced by traumatic or adverse experiences in childhood and adolescence. For instance, eminent creators may display a disproportionate number of physical or sensory disabilities. Handicaps afflicted such diverse individuals as Thomas Edison, Homer, Aldous Huxley, Karl Jaspers, Frida Kahlo, Rudyard Kipling, Sean O’Casey, Joaquin Rodrigo, Charles Proteus Steinmetz, Henri Toulouse-Lautrec, Konstantin Tsiolkovsky, Carl Maria von Weber, and Stevie Wonder. Or, the disability will take the form of chronic illness in childhood or adolescence. Other times the adversity pertains more to the family circumstances. An example is the tendency for eminent personalities to come from homes that have experienced economic reversals or changes of fortune, even to the point of bankruptcy or impoverishment.

But the type of adversity that has attracted the most scientific research is early parental loss or orphanhood. This literature has found a tendency for geniuses of all kinds to have experienced the death of one or both parents at an early age. Table 4.1 provides some examples. Of course, without an adequate baseline for comparison, lists such as those in table 4.1 prove nothing. Nonetheless, several investigators have found the incidence rates of parental losses to be noticeably higher than what holds for the general population. Thus, one ambitious study of 699 eminent figures of world history discovered that 61% lost a parent before age 31,52% before 26, and 45% before 21. Another study of 301 geniuses found that over one-fifth were plagued by orphanhood. A follow-up investigation, also based on famous people from all areas of accomplishment, discovered that nearly one-third had lost their fathers early on.

The orphanhood effect has been documented on more narrowly defined samples, too. Studies of large samples of scientists, for example, suggest that the incidence of orphanhood is indeed high; in one sample of 64 great scientists, 15% had lost a parent before age 10. Among eminent mathematicians, the percentages may be higher still: one-quarter had lost a parent before age 10 and nearly one-third before age 14. The results for creative writers are even more dramatic, for 55% were found to have lost a parent before age 15. Finally, an inquiry into British prime ministers found rates of parental loss of nearly 63%, a percentage far higher than any reasonable comparison group, such as those contemporaries who were English peers.

The effects of parental loss seems verifiable enough to have inspired investigators to propose explanations for how such traumatic events might contribute to the development of genius. Three environmentalist hypotheses are perhaps the most prominent. First, the trauma of parental loss produces a so-called bereavement syndrome, in which acts of achievement serve as emotional compensation. Second, such adverse events nurture the development of a personality robust enough to overcome the many obsta-

Table 4.1 instances of parental loss in

THE LIVES OF CREATIVE GENIUSES

Lost One or Both Parents in First Decade

Scientists: d’Alembert, I. Barrow, Berzelius, Boerhaave, Boyle, W. Bragg, Buffon, Carver, Cavendish, Copernicus, C. Darwin, Eddington, Flamsteed, A. Fleming, Fourier, Fulton, Haber, Haller, Helmont, Humboldt, J. Hutton, Huygens, Jenner, Kelvin, Kolmogorov, Kummer, Laennec, Lavoisier, Lobachevski, Maxwell, Newton, Paracelsus, Pascal, Priestley, Quetelet, J. Rennie, Count Rumford, W. Smith, Steinmetz, Steno, Telford, Volta, C. T. R. Wilson; Thinkers: W. Blackstone, Confucius, Descartes, Hobbes, Hume, Leibniz, G. Marcel, Mencius, Montesquieu, Nietzsche, Rousseau, B. Russell, Shankara, P. Sarpi, Sartre, Shinran, Spinoza, Swedenborg, Voltaire; Writers: Baudelaire, Bronte sisters, Byron, Camus, Coleridge, Conrad, A. Cowley, W. Cowper, Dante, DeQuincey, Donne, Dumas pere, Emerson, E. M. Forster, Gibbon, M. Gorky, brothers Grimm, Hawthorne, Holderlin, Hu Shih, Ben Jonson, Keats, Lermontov, E. B. Lytton, Mallarme, Maugham, G. Meredith, Moliere, Montagu, Montaigne, Neruda, Poe, Propertius, Racine, Sainte-Beuve, Solzhenitsyn, Steele, Stendhal, Sterne, Swift, Thackeray, Tolstoy, Villon, Wordsworth, Zola; Artists: Canova, Delacroix, Diaghilev, D. W. Griffith, Fra Lippi, Masaccio, Michelangelo, Munch, Murillo, Raphael, S. Ray, Rubens; Composers: J. S. Bach, Corelli, Puccini, Scriabin, Sibelius, Wagner.

Lost One or Both Parents in Second Decade

Scientists: N. Abel, Ampere, J. Bruner, M. Curie, H. Davy, Durkheim, Galois, J. Gibbs, W. R. Hamilton, J. Henry, Hooke, Humboldt, Joule, A. Keith, Lamarck, Leeuwenhoek, Malinowski, Mendeleuev, Newcomb, B. Silliman, J. J. Thomson, Tsi-olkovsky, An Wang, J. Watt, Weierstrass, E. Whitney, W. Wundt; Thinkers: Thomas Aquinas, Aristotle, Augustine, F. Bacon, Comenius, Croce, Erasmus, Frege, Hegel, Ibn Khaldun, Kant, Melanchthon, R. Niebuhr, Origin, Santayana, Schleiermacher, Schopenhauer, Zhu Xi; Writers: H. C. Andersen, Ariosto, Bellow, Bunyan, Calderon, Chateaubriand, J. F. Cooper, Dostoyevsky, Dreiser, G. Eliot, H. Fielding, Frost, Gide, Goldsmith, L. Hearn, Hugo, Malamud, Mann, Melville, Petrarch, Plutarch, D. Richardson, R. Sheridan, T. Tasso, Turgenev, Mark Twain, Vega Carpio, H. Walpole; Artists: Caravaggio, Claude Lorraine, J. L. David, Degas, Hiroshige, Spike Lee, Magritte, Whistler; Composers: Beethoven, Bruckner, F. Couperin, Handel, Liszt, Mussorgsky, Schoenberg, Schubert, Schumann, Tchaikovsky, von Weber.

cles and frustrations standing in the path of achievement. Third, parental loss and other forms of extreme adversity may set a young talent along a developmental trajectory that diverges from the conventional.

Each of these three interpretations has certain explanatory advantages and disadvantages. The bereavement-syndrome hypothesis has the asset of explaining the motivation for exceptional achievement in the first place. The robust-personality hypothesis, in contrast, presumes the existence of such a drive and concentrates instead on the essential role that persistence and determination play in the final realization of talent. The divergentdevelopment hypothesis, finally, focuses on the notion that a highly traditional upbringing may encourage the growth of a conformist disposition, and thus ultimately thwart true innovation. All three accounts allow for creative genius as well as exceptional achievement in other domains, such as political leadership. All three can also incorporate various forms of adversity besides parental loss as part of the developmental process. And last, all three hypotheses allow for the possibility that the effects of early adversity might be too extreme, nipping the talent in the bud. This provision is crucial. Because juvenile delinquents and depressive or suicidal psychiatric patients may exhibit orphanhood rates similar to those of the eminent, it is clear that adversity can backfire. As one investigator noted, parental loss “can be an impetus for creative effort, a force for good, or it can have the effect of stunting personality growth and producing the concomitant antisocial acts, destruction of social relationships, and even the taking of one’s own life.”

From the standpoint of a Darwinian creativity theory, the divergentdevelopment hypothesis may have the most explanatory power. One problem with the other two hypotheses is that they assume that the young talent is dramatically affected by the supposedly traumatic event. Yet in many cases this seems not to be the case. For example, Sir Isaac Newton’s father died three months before his birth. Hence, young Newton could suffer no bereavement nor feel that he had some life challenge to overcome. Even when the timing seems more likely to produce effects, the expected trauma may not appear. For instance, although Darwin’s autobiographical recollections go back to when he was four years old, his memory about his own encounter with parental loss was quite thin and cold. “My mother died ... when I was a little over eight years old, and it is odd that I can remember hardly anything about her except her deathbed, her black velvet gown, and her curiously constructed work-table.”

In comparison, the divergent-development interpretation does not presume that the effects of adversity are so sensational. Indeed, rather than one big trauma, a large number of rather small events might accomplish the same end, which is to set the talent on a developmental path that sets him or her apart from the rest of the crowd. So neither Newton nor Darwin needed to have felt the deaths of their parents in order to have experienced the consequences of being thrust along a divergent growth trajectory. In addition, this third hypothesis can more easily account for the fact that the incidence rates for trauma and adversity tend to vary according to the domain of creative achievement. Because various domains differ widely regarding the amount of Darwinian creativity they require, this variation should correspond to the extent to which talent development diverges from the norm. If artistic creativity is more Darwinian than scientific creativity, then development should be more divergent for the artists than for the scientists. This theoretical expectation is vindicated by research. For instance, the incidence of orphanhood for recipients of the Nobel Prize for literature is over eight times higher than that for winners of the Nobel Prize for physics.

The above points do not prove that the bereavement-syndrome and robust-personality hypotheses are completely wrong. It may even be that all three developmental processes work together in the nurturance of creative talent, the specific role of each varying from person to person according to the circumstances. Nonetheless, of the three explanations, the divergentdevelopment hypothesis may have the greatest consistency with a Darwinian conception of the emergence of creative genius.

Education. Terman’s intellectually gifted children were, if nothing else, highly accomplished students. Already in elementary school they were well in advance of their peers in academic achievement, and they continued to demonstrate scholastic prowess throughout their student years. Besides getting excellent grades and earning special honors, they attained high levels of formal education, including more than their share of doctorates and professional degrees. Terman’s children not only did very well in school, but genuinely enjoyed school besides. It was the one place in their life where their talents could really shine. But if they were so proficient in school, college, and university, why didn’t they achieve ever more successes once they launched their adulthood careers? Why doesn’t graduating summa cum laude necessarily predict the later receipt of a Nobel Prize?

One Nobel laureate may have provided the beginning of the answer to this mystery. Albert Einstein once complained that “it is, in fact, nothing short of a miracle that the modern methods of instruction have not yet entirely strangled the holy curiosity of inquiry; for this delicate little plant, aside from stimulation, stands mostly in the need of freedom; without this it goes to wreck and ruin without fail. It is a very grave mistake to think that the enjoyment of seeing and searching can be promoted by means of coercion and a sense of duty.” The tests that plague the school year were singled out for special condemnation. “One had to cram all this stuff into one’s mind for the examinations, whether one liked it or not. This coercion had such a deterring effect on me that, after I passed the final examination, I found the consideration of any scientific problems distasteful to me for an entire year.” Given Einstein’s negative attitudes, his teachers were astonished when he became famous for his creative contributions to mathematical physics. For example, one university professor, Hermann Minkowski, admitted that his former pupil’s achievements “came as a tremendous sur-

- Development ~                         119

prise ... for in his student days Einstein had been a lazy dog. He never bothered about mathematics at all.”

Nor would Darwin have disagreed with Einstein’s negative opinions about school, as is evident from reading his autobiography. To provide but one illustration, Darwin complained that “during my second year at Edinburgh I attended Jameson’s lectures on Geology and Zoology, but they were incredibly dull. The sole effect they produced on me was the determination never as long as I lived to read a book on Geology, or in any way to study the science.” Fortunately, Darwin eventually overcame his repugnance: Had he not taken Lyell’s Principles of Geology along with him on the Beagle, it is questionable whether he would have reached his insights about the origin of species.

But we have to be careful about condemning formal education on the basis of these personal testimonials. After all, many eminent creators were quite excellent students. Marie Sklodowska, later to assume the surname Curie, was two years ahead of her elementary school classmates in all subjects. She received a special gold medal when she graduated from the Russian lycee at age 16. Sigmund Freud was at the head of his class at the gymnasium, and graduated summa cum laude. J. Robert Oppenheimer graduated summa cum laude from Harvard with the highest honors ever awarded. So rather than play anecdote against anecdote, let us consider the following three empirical findings:

• 1. While the quality of scholastic performance may be modestly correlated with adulthood success in some domains, it bears no relation with achievement in other domains, especially in those areas requiring creativity. For example, one investigator scrutinized the undergraduate records of those elected as Fellows of the Royal Society. The academic performance of an FRS was usually undistinguished, and decidedly no better than a comparison group of scientists who were not so honored. Inquiries that examine a broader range of domains allow us to generalize this conclusion. Singular creativity may not correspond to earning high marks in school and college. This lack of correspondence between scholastic performance and creative achievement is particularly conspicuous in artistic creativity. Although D. H. Lawrence would eventually become a lauded British writer, in his secondary school composition class he ranked only 13th out of the 21 who took the course. Artistic creators are also more likely to have much more negative attitudes about their educational experiences in comparison with scientific creators.

• 2. Creative genius is not necessarily associated with attaining high levels of formal education. Certainly many great scientists—Newton and Darwin among them—managed to attain fame with no more than a bachelor’s

degree, and many others—most notably Faraday—never even went to college. The irrelevance of advanced degrees in artistic endeavors is even more conspicuous. In the domain of creative writing, for example, Vera Brittain warned: “The idea that it is necessary to go to a university in order to become a successful writer, or even a man or woman of letters (which is by no means the same thing), is one of those phantasies that surround authorship.” Indeed, empirical research has often found that achieved eminence as a creator is a curvilinear, inverted-U function of the level of formal education. That is, formal education first increases the probability of attaining creative success, but after an optimum point, additional formal education may actually lower the odds. The location of this peak varies according to the specific type of creativity. In particular, for creators in the arts and humanities, the optimum is reached in the last two years of undergraduate instruction, whereas for scientific creators the optimum may be delayed until the first couple of years of graduate school.

• 3. Although formal education thus seems to bear an ambivalent relationship to the development of creative talent, it must be stressed that those who later attained status as creators almost invariably engaged in the arduous process of self-education. Darwin disliked school and was quite content to be a mediocre student at the university; yet he was also deeply committed to self-education through extensive reading, scientific explorations of the English countryside, and conversations with established scientists. “I consider that all that I have learned of any value has been self-taught,” Darwin once told his cousin Galton. In fact, one of the reasons creative talents often dislike school is that it can seem to interfere with their learning what they really want to know. When faced with the choice of reading a good book or studying for an exam, the extracurricular but still instructive diversion may win out. “I have never let my schooling interfere with my education,” said Mark Twain. As should be of no surprise, one of the prime predictors of success in adulthood is being an omnivorous reader.

The foregoing empirical findings seem quite in line with the requirements of Darwinian creativity. To become a creative genius requires the acquisition of a certain minimum amount of knowledge and technique in the chosen domain. Therefore, some amount of formal training will certainly be helpful in talent development. Yet when taken to extremes, formal education can undermine the capacity for generating ideational variations. To obtain high marks in school often requires a high degree of conformity to conventional ways of looking at the world and people. Moreover, the increased specialization that becomes so conspicuous at higher levels of formal education can excessively narrow the range of perspectives that a person is able to entertain. Essentially, as the student progresses through the sequence of ever-higher degrees, the range of subjects, issues, theories, and techniques becomes ever more restricted. Such restriction will tend to confine the number and diversity of ideational variations that the individual can conceive. Hence, self-education may have the critical function of maintaining the necessary breadth in the face of the narrowing tendencies of formal education.

From a Darwinian perspective, moreover, it is telling that the development of scientific talent seems to be more intimately connected with academic training than is the development of artistic talent. As already observed many times, artistic creativity tends to lean more heavily on the blind-variation process than does scientific creativity, and thus the ill effects of formal education are more serious for artists than for scientists. However, the same theoretical argument would oblige us to make distinctions among artists and among scientists according to the degree to which the creative process is Darwinian. Hence, revolutionary scientists should tend to exhibit less academic success and accomplishment than normal scientists. Although no empirical study has looked at this possibility directly, there certainly seems to be ample anecdotal evidence on behalf of this conjecture.

Einstein, for example, earned his Ph.D. while working full-time at the Swiss patent office, and thus he could be said to have had no graduate education at all. Curiously, he earned his doctorate by submitting the least important of the papers he had finished in 1905. In contrast, the other papers—on Brownian motion, the photoelectric effect, and the special theory of relativity—he used to ensure his status as a scientific genius. Obviously, the pursuit of a Ph.D. was a mere sideshow in the development of his creative talent. Indeed, when Einstein first encountered difficulties getting the University of Zurich to approve the original version of his dissertation, his response was, “I shall not become a Ph.D. . . . The whole comedy has become a bore to me.” Happily, the complaint was that the paper was too short. So he played the cynical minimalist, added but one sentence, resubmitted the thesis, and became Dr. Einstein.

Marginality. There is one more peculiarity about Terman’s sample of 2,500 gifted children that may help explain why so few attained the status of creative genius. The children came predominantly from majority-culture backgrounds. That is, they were overwhelmingly native-born, Englishspeaking, white, Anglo-Saxon, and Protestant. Yet this hegemony of the majority seems to run counter to the often-expressed view that creativity may be nurtured by the experience of being a “marginal” person. The sociologist Robert Park, for example, emphasized the significant place immigrants may have in cultural innovation: “One of the consequences of migration is to create a situation in which the same individual... finds himself striving to live in two diverse cultural groups.” Consequently, “the ‘cake of custom’ is broken and the individual is freed for new enterprises and new associations.” In a compatible vein, the historian Arnold Toynbee spoke of the “creative minority” who further human progress by their “withdrawal and return” relative to the majority culture.

The relevance of ethnic marginality in the development of talent to a Darwinian theory of creativity was made explicit by Donald Campbell. In his classic paper presenting his variation-selection model, he claimed that “persons who have been uprooted from traditional culture, or who have been thoroughly exposed to two or more cultures, seem to have an advantage in the range of hypotheses they are apt to consider, and through this means, in the frequency of creative innovation.” The marginal person should display more associative richness, divergent thinking, and other cognitive processes that provide the foundation of ideational variations.

There exists some evidence, in fact, that ethnic marginality may make some contribution to the development of talent. For example, one study of eminent contemporaries found that around 19% were either first- or second-generation immigrants. In the United States alone, newcomers accounted for 44% of the eminent Americans. Indeed, their representation was seven times higher than of those Americans descended from families that had resided in the United States since shortly after the Revolution if not before.

But perhaps the history of the Jews in Western civilization provides the best-known illustration. One study in the 1970s found that although Jews made up only between 1% and 3% of the worl d’s population, their presence in the lists of eminent creators exceeds statistical expectation by a factor of 10 or more. This distinction holds especially for such areas of creativity as mathematics, physics, chemistry, medicine, and economics. In fact, almost one out of five Nobel Prize recipients had come from Jewish backgrounds. This phenomenon probably has many sources, including the superior family environment often found in Jewish homes. Nevertheless, not the least of the developmental factors is the marginal position Jews have in Western culture. They are a Near Eastern people transplanted to a European culture, enabling them to bring the novel insights of the ethnic outsider. Moreover, Western Jews often exhibit high rates of bilingualism, frequently combining some European tongue with either Hebrew or Yiddish or both. Research has shown that intensive exposure to two or more different languages helps build the cognitive basis for creativity. After all, concepts will be coded in multiple ways, enriching the associative interconnections among various ideas. The process operating here is not unlike the possible role of hybridization in the generation of new biological species.

Although the emphasis of much of the empirical research has been on ethnic marginality, other forms of marginalization may also have a beneficial impact on talent development. Another possibility is religious marginality. In Great Britain, for example, many famous scientists and inventors did not belong to the established Church of England. Rather, they subscribed to one of the dissenting faiths. Joseph Priestley was a Nonconformist minister, John Dalton a Quaker, Michael Faraday a Sandemanian. Likewise in the United States, the per capita representation of Unitarians among notable scientists is well over 100 times greater than that of Methodists, Baptists, and Roman Catholics—the churches with the biggest memberships.

Finally, also of relevance is professional marginality. Arthur Koestler claimed in his Act of Creation that “all decisive advances in the history of scientific thought can be described in terms of mental cross-fertilization between different disciplines.” Similarly, F. C. Bartlett, the cognitive psychologist, held that “it has often happened that critical stages for advance are reached when what has been called one body of knowledge can be brought into close and effective relationship with what has been treated as a different, and largely or wholly independent, scientific discipline.” Such crossfertilization of scientific domains is more likely to emerge from those who have switched fields. In fact, Thomas Kuhn, the historian of science, maintained that “almost always the men who achieve... fundamental inventions of a new paradigm have been either very young or very new to the field whose paradigm they change.... [TJhese are the men who, being little committed by prior practice to the traditional rules of normal science, are particularly likely to see that these rules no longer define a playable game and to conceive another set that can replace them.” In support of these claims, one empirical study showed that the innovators in X-ray astronomy were likely to have been from fields marginal to the astronomical profession. A second investigation found a similar marginality effect concerning Alfred Wegener’s theory of continental drift. In particular, the theory’s opponents were usually well published in mainstream geology, whereas the main advocates often hailed from scientific disciplines outside the geosciences.

Because Darwin himself could be considered a professional outsider, these findings may help us appreciate the basis for his own scientific breakthroughs. Even so, from a Darwinian perspective, professional marginality has two features that distinguish its operation from ethnic marginality.

First, most of the cases of innovations by professional outsiders seem to involve the scientific enterprise. To be sure, examples can be given in the arts—such as the invention of mobiles by Calder, who had some training in engineering. Nonetheless, marginality of this kind seems to play a relatively minor role in the arts. The reason may be that scientists, unlike artists, tend to impose rather severe constraints on the production of ideational variants. Sometimes these a priori restrictions exclude the consideration of the very ideas that might have a reasonable chance of generating a solution to some enigma. Hence, it may require a total outsider to introduce a novel set of considerations. For example, paleontologists had grappled with the extinction of the dinosaurs ever since these behemoths first became apparent in the fossil record. However varied the speculations, the hypotheses stayed bound to terrestrial processes, such as climate changes or competition from early mammals. It took a nuclear physicist, Luis Alvarez, to propose that a 10-kilometer-wide asteroid, striking the earth at the end of the Cretaceous period, precipitated the catastrophic events that wiped out the dinosaurs.

Second, although ethnic marginality may be associated with highly prolific creators, professional marginality often seems linked to creators who had only one noteworthy idea. This difference may reflect a fundamental contrast in the developmental implications of these two forms of marginality. On the one hand, it is easy to conceive how early exposure to different cultures and languages might enhance the capacity for remote association and divergent thinking. On the other hand, the professional outsider may be doing nothing more than transplanting some established concept or technique from one discipline to another. If this is the case, then, as mentioned in the previous chapter, it is not essential that the outsider be engaged in generating ideational variations.

Even so, in this latter case we may still speak of creativity as being Darwinian, albeit at a different level of analysis. Instead of thinking in terms of a single mind producing blind variations of ideas, we can interpret the phenomenon in terms of a whole society producing blind variations of individuals. In other words, in any culture with sufficient occupational mobility, a certain number of people will be engaged in career changes. These changes may combine the expertise of different professions in a number of ways. A small proportion of these two-career individuals will be fortunate to have entered a field in which knowledge from their first career will contribute to their creative success in their second career.

Many others will not be so lucky. I say “lucky” because to a great extent the career switches may be blind, thus making the multiple-career combinations truly Darwinian at the individual level of analysis. Although most people do not change professions without the expectation of success in their next career, for some creators the choice may be wrong. Rather than a positive transfer of knowledge and expertise from one domain to another, the transfer may prove negative. The story of the German chemist Justus Liebig provides a needed antidote to enthusiasm over the potential fruitfulness of career changes. Liebig’s creative achievements in pure organic chemistry earned him such acclaim that he was created a Baron. Yet when he later moved into applied agricultural chemistry, the results were extremely disappointing. Because his chemical fertilizers were based on false conceptions about plant nutrition, they failed miserably. Of course, Liebig would not have changed fields had he been able to anticipate the outcome. But his career switch was blind to its true prospects.

On the other hand, the story of another career change proves that the blindness can occur in the inverse direction. A person can change fields with no expectation of success and be proven wrong. When in the 1860s the French silk industry was being destroyed by some mysterious ailment, the chemist Dumas urged his former pupil Pasteur to work on the problem. The latter protested, “But I never worked with silkworms”; Dumas returned with the strange reassurance, “So much the better.” Although the practical problem that Pasteur took on was much further removed from his field of expertise than was the case for Liebig, and despite the fact that Pasteur had much lower initial expectations of succeeding, it was Pasteur, not Liebig, who provided an example of successful professional marginality. Thus, Pasteur’s career change was just as blind as Liebig’s, but the Pasteur variant proved to have more adaptive fitness from the standpoint of the society at large.

Hereditarianism

The picture looks pretty bleak for the doctrine that genius is born and not made. A large number of environmental factors seem to nurture the cognitive and dispositional attributes required for Darwinian creativity. In particular, early experiences at home and school may facilitate the development of the capacity for generating ideational variations, and even instill the motivation and persistence necessary to exploit that capacity. Even Gal-ton’s family pedigrees might be explained away as evidence of nurture rather than nature. As will be discussed in chapter 6, creative potential also seems to be fostered by exposure to role models and mentors within one’s chosen domain of creative activity. Therefore, achievement could run along certain family lines simply because of the greater availability of such environmental influences. Is it really necessary to explain the origins of the creative achievements of George, Francis, Leonard, and Horace Darwin when they all were exposed daily to a father who could serve as their model scientist and even scientific mentor?

Although Gabon’s hereditarianism appears defeated, Galton himself set the stage for a revival of a genetic theory of genius. He pioneered several techniques that would grow to become critical methods in the field of behavior genetics, including the use of twins to gauge the impact of heredity on individual differences. Let us begin by reviewing some of the key findings of this research, and then we can examine the implications for understanding the origins of genius.

Modern behavioral genetics. A great deal can be learned about human heredity by looking at twins. As is well known, twins come in two basic types, dizygotic (fraternal) and monozygotic (identical). The first share virtually identical environments but are no more similar genetically than other full siblings. The latter, in contrast, are genetically indistinguishable. To the extent that certain individual differences are genetically transferred, then, monozygotic twins should be much more similar than dizygotic twins. In fact, for a large inventory of cognitive and personality traits, monozygotic twins are extremely alike, whereas dizygotic twins are no more similar than other siblings. Particularly provocative are the results of studies examining monozygotic twins who were separated at infancy and raised in separate homes. These reared-apart identical twins share virtually no characteristics with their foster parents and siblings yet retain a supreme level of similarity to each other. Indeed, as these twins get older, and especially after they leave their foster homes, the resemblances increase, even when they have had no contact whatsoever with one another. Moreover, the genetically endowed similarities persist until old age. These solid findings prove that a substantial portion of a person’s psychological constitution is determined by the genes received at the moment of conception. To be sure, the specific heredity coefficients vary from trait to trait. For cognitive characteristics, the degree of inheritance can be quite high, whereas for complex personality traits, such as religiosity, the genetic influence is lower. Yet somewhere between one-third and two-thirds of who we are seems attributable to our genetic endowment.

This remarkable but secure conclusion is joined by a second finding that is no less striking. Behavior geneticists have scrutinized the contributions made by environmental factors, taking special care to distinguish two kinds of such influences—the shared versus the nonshared. The shared environment is what all siblings experienced together because they all grew up in the same home with the same parents. The nonshared environment, on the other hand, is what is unique to each sibling. An excellent example is birth order: Only the firstborn has solely younger siblings, only the last-born has solely older siblings, and only the middle children have siblings both younger and older. The empirical research has convincingly demonstrated that for a great variety of individual-difference variables, it is the nonshared environment, not the shared, that is responsible for whatever cannot be attributed to heredity. It is for this reason that siblings turn out so different even when parents make every effort to raise all their children the same way. Cognitive and personality development is usually more contingent on what siblings do not have in common than on what siblings do share. The ineffectiveness of the shared environment also explains why identical twins reared apart are so similar to each other even when brought up in foster homes that feature quite contrasting life experiences and parenting styles.

In fact, some behavior geneticists have argued that parents play a very minimal role in the development of their children. Various socialization practices will simply have no consistent effect on how their children turn out. Because of the impact of the genes and the nonshared environment, a parenting technique that works well on one child might backfire on another. Only if the parents carefully tailor their interactions to meet the unique needs of each child can they hope to exert any developmental influence at all. Indeed, often it may be more appropriate to speak of children raising their parents than the other way around. For example, children may request toys and games that are consistent with their innate interests and talents and ignore gifts that they find boring despite encouragement given by well-meaning parents. Children try to shape their environment to comply with their inborn needs.

One final empirical lesson of behavior genetics deserves mention: On certain characteristics monozygotic twins are too similar relative to dizygotic twins. That is, a trait that shows practically no heritability in fraternal twins and other siblings might exhibit extremely high heritability for identical twins. This will happen when the trait in question requires the simultaneous presence of several separate traits, each of which is inherited independently. If one monozygotic twin inherits the entire collection, its twin will obviously do so as well. But it is extremely unlikely that two siblings originating from different zygotes would inherit all the requisite traits together.

David Lykken, a leading behavior geneticist, gave the example of “social potency.” This trait involves “the self-perceived ability to influence, lead, or dominate others.” Monozygotic twins, whether reared apart or together, closely match on this trait. Yet dizygotic twins are no more alike than any two randomly selected people from the population. Lykken speculated that this characteristic is not only polygenetic but in addition may require all the genes to participate if the trait is to appear at all. In particular, social potency “probably depends on some configuration of attractiveness, selfconfidence, assertiveness, dominance—whatever the ingredients are of ‘charisma.’” If one component is lacking, social potency cannot emerge as a character trait.

So distinctive is this form of inheritance that it has been given a special name: emergenesis. An individual-difference variable is emergenic when it consists of multiple genetic components, all of which must be inherited for the main character to appear at all. The confluence of these diverse but well-defined traits is what underlies a new, more comprehensive, emergent trait. The genotypic whole is greater than the mere sum of its genetic parts.

Genetic origins of genius. The preceding general conclusions, although derived from studies of everyday populations, have profound implications for our understanding of talent development. Most obviously, behavior genetics provides a firm basis for arguing that at least a portion of creative talent may indeed be innate. Presumably, talent in a particular domain depends on the possession of a particular set of cognitive abilities, interests, motives, and dispositions. To the degree that these are genetically endowed, it can be affirmed that the resulting talent is correspondingly inherited. Admittedly, some components of the required profile may have very low heritability, and thus these facets of genius may have to be nurtured by the environment. Nonetheless, it is likely that environmental nurturance can only go so far. If a person fails to inherit certain traits crucial to creative behavior—such as a sufficiently high intelligence—it is most improbable that even the most advantageous environment will succeed in making up the difference. Thus, genetic endowment may provide the necessary even if not sufficient conditions for the emergence of genius.

The behavior genetics research undermines the environmentalist position in yet another manner. A large hunk of the research on the childhood and adolescent antecedents of adulthood creativity assumes that shared environment has a powerful developmental impact. But if shared circumstances have minimal impact for most of the traits that define creative talent, all of this research may provide no support whatsoever for the nurture position. On the contrary, this environmentalist research may actually endorse the hereditarian view. After all, many in behavior genetics argue that several so-called environmental factors are more correctly placed in the column of genetic factors. This inversion of attribution can occur in two major ways:

First, because parents and children share genes in common, the parental genotype will result in a phenotype that will seem to influence the child’s phenotype even though the latter is more accurately said to be a function of its own genotype, which is itself a direct function of the parental genotype. A concrete example will make this more clear. It may be that people who inherit a high degree of information-processing capacity—or native intelligence—will have a tremendous need for intellectual and cultural stimulation. They will accordingly buy books, subscribe to magazines, visit museums and galleries, engage in diverse hobbies, and do whatever else they need to maintain a mentally active life. Of course, they will also tend to choose mates who will closely match them in intellectual ability as well. Consequently, when individuals enjoying high natural ability decide to have children, they will more likely have offspring who will display a comparable need for intellectual and cultural stimulation. These progeny will take advantage of the opportunities available in the home. The environmentalist researcher then looks at this pattern and falsely concludes that it was the enriched environment that the parents offered their children which stimulated the latter’s intellectual growth. This conclusion is false because it assumes that it was the home environment that produced their children’s phenotype. Instead, the intellectual aptitude acquired by the progeny may reflect more strongly the underlying genotype, which just so happens to correlate strongly with the parental genotype with respect to native intelligence. Indeed, if the parents had not provided opportunities in the home for the desired stimulation, the children would probably seek out stimulation elsewhere—in the homes of friends, at school, at the public library, and so forth.

Second, a child with certain inborn talents may soon put pressure on the environment to make it conform more closely to feed those talents. Parents with sufficient resources will often respond accordingly, thereby producing an environment that conforms ever more closely to the child’s gifts. To the outside observer it may appear as if the environment is influencing the child’s development, but instead it is the child’s genetic disposition that is influencing the home circumstances. The research literature on child prodigies is replete with examples of future geniuses who insist on pursuing specific enthusiasms even in the face of parental discouragement. A classic example is Blaise Pascal. Because his father wanted his son to learn Latin and Greek before undertaking mathematics, he had all books on the latter subject removed and even refrained from mentioning the subject as much as possible. However, with what little mathematical knowledge Blaise possessed, he was able to devise his own version of Euclidean geometry, inventing his own terms and theorems. He was not yet 12 years old. When his father discovered his son’s secret activities, he realized that Blaise had a rare gift, and he adopted a more receptive attitude toward its further development. His son’s inherent talent had won the day. By the time Blaise was 16 he had produced an original work on conic sections that provoked the envy of Rene Descartes, 27 years his senior. And by 19 Blaise Pascal had invented a calculating machine. Even greater scientific and mathematical achievements lay before him.

These two illustrations concerned the intellectual and cultural opportunities available in the home environment. An advocate of environmentalism might argue that the hereditarian viewpoint would have a hard time accounting for other developmental factors that seem obviously to involve nurture rather than nature. What about the developmental impact of early trauma and adversity, education, or marginality? How can these events possibly be ascribed to genetic endowment? Actually, a hypothetical connection can be conceived in every case. Take parental loss, for example. Parents who are more intelligent tend to delay reproduction until later in life, after their professional careers are established. In fact, data show that the parents of eminent personalities were older than is the norm when their illustrious progeny were born. Because the parents are much older, the odds will be greater that they might die when their offspring are in childhood or adolescence. The latter, of course, already inherited their parents’ intellectual superiority, and so when they become more successful in life, their success may be falsely attributed to the traumatic experiences. I might add that although Darwin suffered parental loss at age eight, his mother was already in her 50s when she died, his parents having married in their mid-3os. Darwin himself put off family life until he had entered his 30s.

Similar arguments can be made to convert education and marginality into indicators of biological inheritance rather than determinants of development. In the first case, individuals with certain innate gifts may provide for those talents by purposively selecting the level and type of education most suitable for their cultivation. Hence, artistic talents may drop out of school sooner than scientific talents simply because the former discover much more quickly that fulfilling scholastic requirements has become irrelevant to the development of their gifts. In the case of ethnic marginality, a different process may be operating. For instance, it may be that those individuals who take the trouble to immigrate to a foreign land will tend to possess personal qualities that lead to success, such as intelligence, energy, motivation, and initiative. These advantageous traits are then passed down to their offspring. The causal basis is nature, but the superficial appearance is nurture. This was actually Gabon’s own explanation for any ethnic marginality effect.

I am not claiming that every putative environmental influence represents an exclusively genetic influence operating incognito. The claim is simply that it is not easy to tease out the relative contributions of nature and nurture. Developmental events and circumstances may not always be what they seem; what look like causes may actually be effects. An undetermined proportion of the so-called environmental determinants of talent development may be the mere consequences of underlying genetic processes. Doubts about the impact of nurture must be especially potent when the factors involve the shared rather than the nonshared environment. Although behavior genetics has identified some traits that are nourished by shared environmental influences, these are relatively rare, and none so far involve characteristics proven essential for creative genius. Consequently, we must be ever wary about claiming that certain home conditions encourage the emergence of creative talent.

However, we must be equally careful not to take the results of behavior genetics as endorsements of Galton’s original hereditarianism. In some respects, contemporary research in behavior genetics has weakened that very theory as well. In particular, the phenomenon of emergenesis can undermine Galton’s thesis that creative genius must run in family lineages.

If creativity consists of several critical components, all of which must be inherited before creativity can emerge, then it will be rather unlikely that creative genius can be passed down from parent to child. Nor will siblings exhibit any family resemblances in innate creative capacity—with the sole exception of monozygotic twins. Hence, creative giants may not necessarily arise from Galtonian pedigrees.

The concept of emergenesis is a relatively new idea in behavior genetics. Even so, there exists some empirical research suggesting that creativity may indeed constitute an emergenic trait. In addition, Lykken offered an interesting historical example—that of Carl Friedrich Gauss, the great mathematician. Gauss’s father was a bricklayer, his mother a peasant, and Gauss’s own son came nowhere close to matching his father’s mathematical skills. Apparently, to become a Gauss requires a distinctive convergence of many abilities, interests, and values. If a single attribute is missing, you no longer have a Gauss. Curiously, Galton failed to discuss this discrepancy, and he even neglects to mention Gauss in Hereditary Genius. This omission is surprising, because Gauss had a telling influence on Galton’s own work.

In fact, Galton’s classic treatise contains other mysterious hints that emergenesis may account for the appearance of the most monumental genius. Many of the top-notch creative geniuses appear to have no family pedigree whatsoever. Consider the cases of Newton, Shakespeare, Beethoven, and Michelangelo. These four individuals epitomize the highest levels of creativity in the fields of science, literature, music, and art. Yet what sorts of pedigrees was Galton able to devise for such exalted geniuses? In the case of Newton, Galton devotes nearly two pages to discussing a highly speculative link with other relatives of far lesser distinction. Likewise, the best Galton could do for Shakespeare’s pedigree is to dig up a lineage for Francis Bacon, a contemporary who some have unsuccessfully claimed to be the real author of the Bard’s works. But Shakespeare himself has no pedigree. Beethoven created yet another anomaly for Galton’s theory. Galton’s only recourse was to repeat an old rumor that Beethoven was the illegitimate son of Frederick the Great of Prussia. Even if the rumor had foundation, it is hard to fathom how a merely competent flute player could provide the genetic foundation for Beethoven’s immense musical powers. And with respect to Michelangelo’s artistic lineage, Galton could find absolutely nothing at all. If height were inherited in the same fashion as these anomalous cases indicate, the tallest professional basketball players could have the shortest parents.

Emergenesis does more than merely create difficulties for Galton’s simple genetic theory of talent development. It also introduces complications for understanding the evolution of creative genius from the perspective of primary Darwinism. In previous chapters I have tried to discern the selection pressures that might favor the appearance of genes for exceptional ere-ativity. But if creativity is an emergenic trait consisting of a large number of independent traits that must be simultaneously present to produce a creative genius, then creativity can only undergo consistent selection when all the component genes exhibit a high degree of chromosomal linkage. If the traits are not easily inherited as a package—the most likely situation—then each of the separate genes making up the genotype of the creative genius must undergo its own individual evolution in the population. The greater the number of requisite genes, and the more varied the basis for reproductive fitness provided by those genes, the less the lucky conjunction of these genius-generating genes will be under the control of natural selection. In the extreme case, it might make little difference whether creative geniuses are at an advantage or a disadvantage in the struggle for existence. Darwin may have had seven children, and Galton zero, but this differential success may have had little impact on the relative frequencies of the various component traits in the gene pool.

This interpretation is consistent with Gabon’s contention that true genius is very rare. He estimated that only 1 out of every 4,000 individuals would qualify for this distinction—a very small percentage, roughly comparable to the proportion of people who would score 150 or higher on an IQ test. With only 0.025% having what it takes to be a genius, it hardly matters whether great creators are fertile or sterile. In comparison with the huge reservoir of genes available in the human population, the reproductive success of the creative genius would count for almost nothing. Hence, to the extent that exceptional creativity is highly emergenic, we need not even worry whether individual differences on this polygenetic characteristic are consistent with Darwin’s theory of organic evolution. In a sense, Newton, Shakespeare, Beethoven, and Michelangelo were all freaks of nature. Such fortuitous freaks will always appear against heavy odds, regardless of their reproductive fitness.

Evolutionism

Earlier in this chapter I noted that Galton himself tried to reach an accommodation between the positions of nature and nurture. In the survey of eminent scientists presented in his 1874 English Men of Science, Galton went beyond the strictly genetic stance of his 1869 Hereditary Genius. In particular, he distributed a questionnaire among members of the Royal Society that asked them about various environmental circumstances that might have contributed to their scientific success. Among the conditions examined was birth order. Galton was, in fact, the first behavioral scientist to introduce this variable into research on talent development. Moreover, Galton reported the striking empirical finding that firstborns are overrepresented among illustrious scientists. This result is remarkable because the result seems contradicted by Gabon’s own family history, as well as by that of Charles Darwin. Galton was the last of nine children, Darwin the fifth of six. Yet Gabon’s own data indicate that he and his cousin had to rub shoulders primarily with firstborns.

Gabon’s finding has been replicated many times since 1874. For example, one study of 64 notable scientists found that fully 39 were firstborn, and another 13 were second-born. Furthermore, of the 25 who missed out on primacy of birth, 5 were oldest sons and 2 had an older sibling who had died in infancy or early childhood. Many of the rest were five or more years younger than their immediately older sibling. Typically, first borns represent 50% or more of the community of active scientists. The firstborns also surpass their later-born colleagues on various criteria of success. They are referred to more frequently in the scientific literature of their field. They receive higher creativity ratings at the hands of experts in their discipline. The firstborns even have a better chance of earning the Nobel Prize. Nor is this primogeniture effect restricted to science. In A Study of British Genius by Havelock Ellis, the same advantage was found for those who won an entry in The Dictionary of National Biography. Other studies demonstrated that the advantage holds for domains ranging from classical composers to politicans. One recent inquiry found that nearly half of all notable creators, leaders, and celebrities of the twentieth century were firstborn children.

Although the firstborn appears inevitably to come out on top, this is not always the case. Creative writers, for example, are more inclined to be later-born children, as are political revolutionaries. These exceptions would seem to cast into doubt any generalization about how birth order relates to creative genius. Yet this pessimistic conclusion would be very much mistaken. A child’s ordinal placement in the family may have a powerful influence on personality development, including the nurturance of traits that determine creative behavior. As mentioned earlier, investigators in the area of behavior genetics have shown that the nonshared environment has much more impact on personality development than does the shared environment. And birth order constitutes nonshared environment par excellence.

Many researchers have tried to describe the developmental processes by which birth effects occur. But from the Darwinian perspective, the most provocative by far is the recent theory advanced by Frank Sulloway in his 1996 book Born to Rebel: Birth Order, Family Dynamics, and Creative Lives. What makes this theory so noteworthy is its explicit use of a Darwinian framework to explain why not everyone in Darwin’s day was sympathetic to the doctrine of evolution by natural selection. In a sense, Sulloway used Darwinian concepts to explain the history of Darwinism. To pull off this feat, Sulloway was obliged to devise a complex theory that details how the development of creative talent is shaped by specific family circumstances, and especially by ordinal position.

Given that Sulloway required a hefty volume to expound and document his evolutionary theory, I do no more than provide a bare outline here. Again, the interested reader can always consult Born to Rebel to get a deeper understanding of this rich theory.

Openness and Birth Order

Frank Sulloway is a historian of science who has had a long fascination with Darwin and Darwinism. One fact about the early history of Darwin’s theory greatly impressed him, namely that the Origin of Species provoked considerable controversy upon its publication, even among scientists. Ideally, the truth or falsity of a scientific idea is decided by fact and logic, but such rational considerations often played a minimal role in the case of Darwin’s theory. Indeed, some of the most notable scientists of Darwin’s time immediately blasted these newfangled ideas. These staunch opponents included Louis Agassiz, widely viewed as the greatest living naturalist; Adam Sedgwick, the distinguished geologist and Darwin’s former teacher at Cambridge; and Pierre Flourens, the famous experimental physiologist and perpetual secretary of the French Academy of Sciences. To offer specifics, Agassiz branded Darwin’s theory as “a scientific mistake, untrue in its facts, unscientific in its method, and mischievous in its tendency.” Likewise, Sedgwick accused Darwin of deserting “the true method of induction,” producing “a dish of rank materialism cleverly cooked and served up.” And Flourens needed to write a whole book to inveigh against Darwin’s Origin: “What metaphysical jargon clumsily hurled into natural history! What pretentious and empty language! What childish and out-of-date personifications!”

To be sure, Darwin’s revolutionary theory had its early adherents as well. When T. H. Huxley first grasped the Origin’s core thesis, he at once propounded: “How extremely stupid not to have thought of that!” Even so, what struck Sulloway is how little the arguments focused on scientific issues per se. The controversy seemed more a conflict of personalities than of data or deduction. After scrutinizing more closely the participants in this vitriolic debate, Sulloway suggested that the differential response to Darwin’s ideas may reflect individual differences in openness to experience. Those who have this openness are more favorably disposed to revolutionary ideas, whereas those who lack this openness are inclined to attack such ideas, doing what they can to conserve tradition or the status quo.

Given the impact of this personality trait, the next question is where individual differences in openness come from. How much of it is nature? How much nurture? Reviewing the behavioral genetic research regarding this character, Sulloway noted that around 30% or 40% of the variation may be ascribed to genetic endowment. Shared family environment accounts for an additional 5% of the individual differences. And, finally, the nonshared environment may explain seven times as much variance as the shared environment: fully 35%, or about as much as the genes. Sulloway then asked what aspect of the nonshared environment figures most prominently in the development of an open personality. After an extensive consideration of many possible influences, he concluded that birth order may represent the single most critical factor. The eldest children are more prone to identify strongly with parents and other authority figures. This identification with parental and societal authority is normally associated with being conforming and conventional, a tendency that makes the eldest children more disposed to be defensive about challenges to traditional ways of viewing the world. In short, the firstborn is less likely to exhibit openness to revolutionary ideas. Sulloway was thus led to the proposition that a scientist’s ordinal position in the family may have been a significant developmental factor in determining who accepted and who rejected Darwin’s theory.

In making this developmental connection, however, Sulloway was careful to distinguish between biological and functional birth order. The first category concerns the actual order that a child was brought into the world. Because of various extraneous circumstances, biological birth order has very little developmental interest. Siblings may die, and new siblings be adopted or acquired through remarriages. So what really matters is functional birth order, which pertains to what a person actually experienced growing up. If the firstborn sibling of a second-born child dies in infancy, the latter child becomes a functional firstborn. This distinction is crucial for using birth order to predict openness to scientific revolutions. Agassiz, for example, was biologically the fifth-born child of his parents (with three younger siblings). But none of his older siblings survived infancy, converting him into a functional firstborn.

If Sulloway had decided that biological birth order were more critical than functional birth order, then he would have to look for purely physiological causes for the effect of ordinal position. Perhaps he would have had to investigate the changes that might take place in the intrauterine environment in successive pregnancies. In contrast, the decision to focus on functional birth order located the causal process in the distinctive social environment occupied by each successive child. Given this locus, the issue is to discern the family dynamics underlying ordinal position. One of most brilliant aspects of Sulloway’s theory is that he looks for those dynamics where many would least expect it—in the fundamental principles of Darwinian evolution.

Sibling Rivalries and Family Niches

Survival of the fittest—such is one of the cardinal concepts in Darwinian theory, primary or secondary. As Darwin deduced from the cruel realities of Malthusian population growth, not all offspring will survive to reproduce. Offspring without superior adaptive fitness will fail to make a contribution to the next generation’s gene pool. We tend to conceive of the struggle for existence as taking place in the population of sexually mature adults. We conjure up images of adults competing for scarce food, shelter, and mates, some winning and others losing. Sometimes these images can include aggressive, even vicious and cruel behaviors—as in Tennyson’s “Nature, red in tooth and claw”—but the struggle takes place among adults who have the wherewithal to contend on roughly equal ground. Yet the reality of evolution is that it sometimes inspires competition between the big and the small, the strong and the weak. And one place where this unequal contest can occur is the family.

To illustrate this point, Sulloway describes the phenomenon of siblicide. In the ovoviviparous sand shark, for example, the young engage in a life-or-death struggle in the usually protective confines of their mother’s oviducts. The young eat each other until only one well-fed shark has emerged supreme. Albeit not nearly so dramatically, siblicide occurs frequently among birds as well. Typically, the eldest nestling, which is almost always the biggest and the strongest, will either peck a younger sibling until it bleeds to death or push it out of the nest to succumb to starvation or exposure. Almost always, the victim of this ugly competition is the youngest, the last of the brood to have hatched from its egg. In the more benign cases, this atrociously vicious form of sibling rivalry is only triggered when the eldest’s body weight drops to a level that threatens its future fitness. In the more grim instances, the siblicide is inevitable. The parents produce a brood expecting only one to survive to maturity. The parental strategy apparently is to let “sibling selection” decide which of the offspring is most fit to compete in the world beyond the nest.

Although sibling rivalries certainly do not operate at this intensity among Homo sapiens, our species does exhibit more subtle forms. Each child must compete for parental attention and resources, and certain children have advantages over the others. The firstborn certainly has the biggest edge. The firstborn will have had the first shot at parental affection and investment, and will usually retain a privileged status so long as the child lives up to parental expectations. Furthermore, the firstborn will normally be more mature mentally and bigger physically relative to the later-born siblings. What firstborns cannot win by outsmarting the later-borns they

~ Development ~                         137

can gain by physical intimidation. In a sense, the eldest have both brain and brawn on their side.

The firstborns are thus ensconced in a special family niche that their younger siblings can only envy. This means that later-borns must find some other niche that will allow them to win the parental involvement that they need for their own development. Trying to imitate what the eldest is doing will seldom gain the youngest much advantage. The eldest has already preempted the most privileged spot and will not take kindly to requests to share that position. Also, later-borns are not as able to act as a surrogate parent, which is typically part of the firstborn niche. Accordingly, the later-borns must employ some alternative strategy. According to Sulloway, the younger siblings must optimize their adaptive fitness by exhibiting Darwin’s “principle of divergence.” Just as competition forced Darwin’s finches to find new ecological niches to exploit on the Galapagos Islands, so must the later-born siblings be willing to diversify, to depart from the mainstream norms. The later-borns must be open to alternative possibilities, new options. Moreover, the later in the birth sequence the children fall, the more likely that their older siblings will have filled up the more obvious niches. Hence, the later the child comes in ordinal position, the more intense the pressures to diversify, to be sensitive to novel opportunities, to do what no one else has done.

Before further detailing Sulloway’s Darwinian argument, I must pause to highlight the interesting parallel between birth order’s developmental progression and an analogous transition that occurs during creative problem solving. Back in chapter 2 I noted how when individuals first encounter a particular problem, they determine whether the problem has a straightforward solution. Perhaps there is some easy algorithm that will dispatch the problem in an eye blink. If not, people will rummage around through a collection of heuristics that can help narrow the range of the search to settle on a solution. If even that approach fails, the problem solver may have to regress into primary-process thinking, into trial and error, into remote association and divergent thinking. As one attempted solution after another falls by the wayside, future attempts become ever more haphazard, ever more desperate, ever less justified, increasingly divergent. In a word, repeated failures force the adoption of increasingly blind ideational variations.

Sulloway essentially affirms that an analogous process occurs in development. The most obvious adaptive strategy for a child is to live up to parental expectations and to act as a surrogate parent. The child should identify with the parents, conform to their norms and values—do everything possible to become a “chip off the old block.” This rule is so obvious as to be almost algorithmic. But if the firstborn has already claimed this family niche, the

second-born must identify some other, less obvious and less conventional niche. Yet as each successive child enters the increasingly crowded family ecology, the fewer customary and conforming roles remain. The later-borns have no other option but to diverge ever more, to become increasingly unpredictable in their developmental trajectories. In short, the later-born siblings must engage in ever more blind behavioral variations to maximize adaptive fitness.

This trend toward ever more Darwinian development can help us explain the most important features about the relationship between birth order and achievement, both everyday and exceptional. Firstborns are more likely to do well in school, just as parents expect. In fact, the gifted children in Terman’s longitudinal study were predominantly the eldest. And just like these children, firstborns are more likely to get higher degrees and to become professionals, such as doctors, lawyers, and college professors. The firstborns thus make their parents, relatives, and teachers proud. The best of these firstborns, moreover, advance to become presidents of the United States or Nobel Prize-winning scientists. Meanwhile, their younger siblings struggle to find a niche by aspiring to unconventional, high-risk, and often radical careers. The later-borns are overrepresented among those who participate in physically dangerous sports, for example. Of course, the least successful among them may lose out in the competition, eventually descending to the ranks of alcoholics or the homeless. However, the most successful will become revolutionaries who overthrow the tyranny of their oldest siblings or bohemian artists who defy all social and aesthetic norms. If they cannot find a secure niche in the home and in polite society, they will discover a place in anarchist politics or avant-garde culture. The later-borns, thus, exhibit more Darwinian variability in their career paths.

Revolutionaries and Reactionaries

Sulloway did far more than offer unconstrained speculation about the relationship between birth order and receptiveness to revolutionary ideas. To test his developmental theory, he applied quantitative techniques to “121 historical events, which encompass data on 6,566 participants. These 121 events include 28 revolutions in science, 61 reform movements in American history, 31 political revolutions, and the Protestant Reformation” as well as “a database on U.S. Supreme Court voting behavior, which includes biographical information on the 108 justices to date.” Hence, despite the fact that Sulloway is a historian of science and began his studies with the proponents and opponents of the Darwinian revolution, his theory and method drove him into social, political, and religious history as well. He thus endeavored to demonstrate that the same theory applies just as well to American civil rights leaders, French revolutionaries, advocates and enemies of the Protestant Reformation, and Supreme Court justices as it does to those who chose to accept or reject the latest scientific advances. In general, firstborns seem most likely to oppose revolutionary ideas, whereas later-borns appear more likely to accept them.

In the case of the Darwinian revolution, for example, later-borns were nearly five times more likely to accept the theory than were firstborns. Indeed, the opposition included firstborns like Louis Agassiz, Elie de Beaumont, John Herschel, Roderick Murchison, William Dwight Dana, and John Stevens Henslow, whereas the proponents included later-borns like Ernst Haeckel, T. H. Huxley, Joseph Dalton Hooker, and George Bentham. Moreover, not only was Darwin himself a later-born, but so was the cofounder of evolutionary theory, Alfred Russell Wallace, who was the eighth of his parents’ nine children. In fact, pre-Darwinian advocates of evolutionary ideas also tended to be later-borns, including Benoit de Maillet, Erasmus Darwin, Jean-Baptiste Lamarck, and Etienne Geoffroy Saint-Hilaire.

This is not to say there exist no exceptions to the rule. To Sulloway’s credit, he actually named and discussed numerous aberrant cases. For instance, later-born opponents of evolutionary theory included Adam Sedgwick and Richard Owen, while firstborn proponents included Charles Lyell, Asa Gray, and August Weismann. But rather than shrug his shoulders, Sulloway responded by elaborating his theory to the point that the number of exceptions is dramatically reduced. Furthermore, these elaborations are not post hoc, but rather they ensue directly from his Darwinian theory of family niches. In particular, Sulloway scrutinizes the developmental repercussions of innate temperament, the spacing of births, early parental loss and parent-offspring conflict, as well as gender and membership in a minority group. For example, eldest children who do not get along with their parents, for whatever reasons, tend to become functional later-borns. Similarly, but with a different underlying cause, children born with an innate inclination toward shyness and introversion are going to find it difficult to retain dominant status when confronted by a younger sibling who is predisposed toward extroversion.

Sulloway also took great pains to consider the specific historical nature of a particular scientific revolution. To begin with, he showed that the later-born predominance in the early phases of a revolution may be replaced by a firstborn hegemony in the later phases. The firstborns learn how to reassert their authority by deflecting the revolution toward different ends. Sulloway provided detailed quantitative evidence how this usurpation occurred during the French Revolution, the guillotines of the Reign of Terror being the tragic consequence. In addition, Sulloway noted that not all scientific movements have the radical ideological implications of the Darwinian and Copernican revolutions. Some actually reflect more conservative notions that work in favor of the status quo. A good example is the highly idealistic systems of biological classification, such as that of the Quinarians, which seemed to endorse the belief in an all-powerful, all-knowing Creator who conceived all living forms according to a wonderfully logical and eternal scheme. Advocates of these intellectual movements were more likely to be led by firstborns than by later-borns.

Once more, Sulloway was not content to speculate. Instead, these diverse complicating factors were subjected to scientific scrutiny. The jewel of these empirical analyses is a single mathematical equation that predicts the odds that a scientist in Darwin’s time would come out as an evolutionist. Not only does this equation make the correct prediction in 84% of the cases, but additionally birth order comes out as the single most important predictor. The effects of ordinal position are 1,000 times more influential than socioeconomic class and 37 times more powerful than national loyalties, for example.

Moreover, those who played leading positions in the emergence of evolutionary theory tended to have the most impressive predicted probabilities. Darwin’s predicted likelihood of endorsing evolution was 94%, that for Wallace, 96%. In other words, if the odds are high enough, one might predict that a person will originate a revolutionary theory rather than just support a revolutionary theory someone else proposed. Consistent with this is the predicted probability that Patrick Matthew would support evolutionary theory. Back in 1831, this obscure dilettante with an interest in botany had proposed a theory that was for all scientific purposes equivalent to that of Darwin and Wallace. Darwin himself wrote to Wallace that Matthew “gives most clearly but very briefly ... our view of Natural Selection. It is a most complete case of anticipation.” Wallace agreed, saying that “he appears to have completely anticipated the main ideas of the Origin of Species” Unfortunately, Matthew published this theory in an appendix of the book On Naval Timber and Arboriculture, which did not attract many readers, and even fewer enthusiasts. This was an evolutionist who was clearly ahead of his time. Nevertheless, it must be considered a striking confirmation of Sul-loway’s developmental model that Matthew possessed a 97% probability of becoming an evolutionist.

These last findings show that Sulloway’s is not just a theory of who appreciates revolutionary genius but a theory of the origins of revolutionary genius besides. It is a remarkable testimony to the power of Darwin’s theory that it should provide the foundation for explaining not only why some scientists were more receptive to his arguments, but also why it was only an elite triumvirate of intellects who could so much as propose such revolutionary ideas.

Integration

No doubt Sulloway’s theory will have to undergo radical transformations as it encounters new facts and new critiques. This is the way of science. As pointed out in chapter 2, evolutionary epistemologies presume that scientific progress is itself Darwinian. Just as life never stops evolving, so do scientific theories undergo an incessant process of formulation, refutation, and modification. But also like organic evolution, the future fate of Sulloway’s theory is not easily forecast. As is apparent from the book’s dust jacket, Darwinian sympathizers have proclaimed Born to Rebel a masterpiece of the highest order. Edward O. Wilson called it “one of the most authoritative and important treatises in the history of the social sciences.” Ernst Mayr ventured that “every once in a long while a book is published which changes a whole field of scholarship, perhaps even everybody’s thinking. Such a book is Born to Rebel” And Sarah Hrdy proclaimed that this “book will have the same kind of long-term impact as Freud’s and Darwin’s.” Other scholars were not nearly so generous, and a few have condemned it to the class of intellectual fashions that will not survive the struggle for continued existence in the history of scientific ideas. A systematic study of the birth orders of Sulloway’s adherents and opponents remains to be carried out.

Nonetheless, whatever the ultimate fate of Sulloway’s monograph, I believe it provides a useful model of how to resolve some of the controversies that have plagued developmental theories since their inception. Certainly Sulloway did not take an extreme stance on either side of the naturenurture debate. A later-born himself, he may have been too open ever to have adopted a one-sided stance. Instead, he tried to integrate nature and nurture in a coherent developmental model. Moreover, nature and nurture do not operate in a passively independent fashion, heredity providing this, genetics providing that. On the contrary, in his integrative thinking, the diverse factors, whether genetic or environmental, interact in complex ways, moderating their respective influences. Furthermore, these interactions take place dynamically over time, so that the impact of a factor may change with a change in circumstances. For example, the sudden death of a parent may dramatically influence the degree to which the firstborn must identify with parental authority. Yet this consequence is itself contingent on the family’s socioeconomic class as well as the eldest’s gender and temperament. The trajectory of talent development is always capacious, as fortuitous events may without warning deflect a youth toward success or failure.

Eventually Sulloway’s framework—with all necessary modifications— may be expanded to encompass all of the central facets of talent development. The childhood and adolescent emergence of creative genius will then be described as an evolutionary process, a process of variation and selection. After all, each year millions of human beings are born into the world with a certain set of potential talents and with a specific set of conditions in which those talents will either wither or grow. Each of these humans enters the world with the same ultimate purpose—to find that niche where he or she can be fruitful and multiply. For some individuals, this task is easy, the solution to life’s greatest riddle practically dictated to them on their first birthday. These are the bright, extroverted firstborns, especially males from middle-class majority backgrounds who will grow up to become doctors, lawyers, engineers, ministers, and professors. But for many others, discovering the right niche entails a true quest, with many trials and almost as many errors. And each error has some chance of proving fatal. The developmental lineage goes extinct when the youthful talent pursues an irreversible course that leads to a cul-de-sac of lifelong unfulfillment. Perhaps only a small percentage will survive the cruelties of this variation-selection process. Only a lucky elite will find that perfect match between an ever-evolving configuration of interests, abilities, and values and the constantly transforming familial, educational, and occupational niches.

Darwin illustrates well enough the vagaries of evolutionary development. His whole biography is a litany of “what ifs”—any one of which might have deflected his latent talent into oblivion. What if his elder siblings had all died in infancy leaving Charles to assume the place of the first child? What if his mother had not died when he was eight years old? What if he had been more proficient in languages and mathematics or too sickly to enjoy strenuous outdoor activities? What if he had encountered some Cambridge professor of theology who had taken him on as a protege instead of having his botanical mentor Professor Henslow? Most important, what if Darwin had not become naturalist on the Beagle?

Darwin himself recognized the central place that this “chance of a lifetime” had in his personal and professional development. He left England an aimless youth, headed in the direction his father most feared—the life of an idle sportsman. He returned five years later a mature scientist, having “discovered, though unconsciously and insensibly, that the pleasure of observing and reasoning was a much higher one than that of skill and sport.” As Darwin continued,

the voyage of the Beagle has been by far the most important event in my life, and has determined my whole career.... I have always felt that I owe to the voyage the first real training or education of my mind; I was led to attend closely to several branches of natural history, and thus my powers of observation were improved, though they were always fairly developed.... That my mind became developed through my pursuits during the voyage is rendered probable by a remark made by my father, who was the most acute observer whom I ever saw, of a sceptical disposition, and far from being a believer in phrenology; for on first seeing me after the voyage, he turned round to my sisters, and exclaimed, “Why, the shape of his head is quite altered.”

In fact, from this point onward Darwin aspired to make his mark on the world of science, an ambition from which he never wavered. “I am sure that I have never turned one inch out of my course to gain fame,” he admitted.

Yet Darwin also marveled that this life-changing experience was by no means inevitable, for “it depended on so small a circumstance as my uncle offering to drive me thirty miles to Shrewsbury, which few uncles would have done, and on such a trifle as the shape of my nose.” In this he refers to the fact that it was Darwin’s uncle who persuaded his father to let him go, and that Captain FitzRoy of the Beagle, a physiognomist, doubted that any one with Darwin’s nose “could possess sufficient energy and determination for the voyage.” Fortunately for Darwin’s hopes, both his father and the captain changed their minds, unleashing a chain of events in the development of Darwin’s talent that would eventually alter the course of history.

Naturally, some might argue that Darwin’s talent would have realized its potential by some other means. Cousin Galton might have argued such, at least when he wrote Hereditary Genius. Yet in terms of a bona fide Darwinian theory of talent development, this does not seem likely. Most biological lineages end with extinction, and most promising talents do the same. In each case, the niches needed to support further evolution vanish at some fatal juncture. Of course, even the biggest alteration in Darwin’s developmental trajectory would have left only a modest trace on that of Wallace, the codiscoverer of natural selection. So evolutionary theory might still have appeared, only under a different name. This book would then be discussing (awkwardly) Wallacian perspectives on creativity. Nevertheless, from the standpoint of Charles Darwin—born on 12 February 1809, with a unique configuration of genetic and environmental traits—the outcome would have been tragically different. One of the world’s greatest potential talents would have become as extinct as the dodo.

This page intentionally left blank

'-'•J

If the genius of modern European civilization is largely defined by such creators as Newton, Shakespeare, Michelangelo, and Beethoven, so are these geniuses themselves mostly defined by the creative products on which their respective reputations are founded. Take away the Principia, the Opticks, and the mathematical essays, and Newton would become just a paranoid and emotionally volatile aficionado of alchemy and esoteric biblical exegesis. If it is ever proven that Shakespeare was not the true author of the plays and poems currently attached to his name, all that would remain would be a meager residual of mundane facts about an obscure Elizabethan actor. If all of Michelangelo’s artistic and architectural achievements were destroyed in some tragic holocaust, there would survive only a minor writer of Italian sonnets. And Beethoven, in the absence of his concert, chamber, church, and theater compositions, would become a once promising piano virtuoso sadly reduced by progressive deafness to an impoverished alcoholic.

The genius of Darwin, too, ceases to exist without the Voyage of the Beagle, Origin of Species, Descent of Man, Insectivorous Plants, Formation of Vegetable Mould, and his other scientific works. He would be obliged to become that very person his father most feared he would become, a member of the idle well-to-do. Needless to say, from the standpoint of the Galtonian definition of genius, these assertions are all truisms. If genius is defined as an exceptionally high IQ score, then it matters not one iota whether the stratospheric intellect devotes every spare moment to reading cheap detective novels. But if genius is defined in terms of eminence in a domain of creative activity, then there must be some concrete evidence of achievement. Creative products provide just that evidence.

Given the definitive nature of these products, no account of genius can be considered complete without a detailed theoretical analysis of creative productivity. I will begin my analysis with a general treatment of the creative careers from which these products emerge. This first section will apply to all forms of creativity, whether artistic or scientific. However, the following two sections will concentrate on issues pertaining to specific types of creative products. I first look at artistic creativity, with the specific focus on the determinants of artistic style. I then turn to scientific creativity, examining a distinctive phenomenon that almost devastated the life of Charles Darwin. I close with a general discussion of why creative products—even a masterpiece like Darwin’s Origin of Species—must originate in a Darwinian process.

Productive Careers

The first step toward understanding a creator’s productivity is to recognize that it has two separate aspects. First, as pointed out in previous chapters, creative individuals vary in regard to the number of products that can be credited to their names. There are one-idea men and women, and there are the extremely prolific whose life work is truly phenomenal. So we are looking at individual differences, or cross-sectional variation, in the total lifetime output of creative ideas. Second, the career of any given creative genius is by no means a homogeneous entity, but rather it consists of life phases or periods. In other words, creative productivity may wax and wane over the course of the career. Thus arises the issue of longitudinal variation, or developmental changes, within a single career. In Darwin’s case, for example, we can ask how Darwin’s productivity differed from that of his fellow scientists, and we can ask how his output fluctuated from the time he ended the Beagle voyage to the day he last laid down his pen.

Individual Variation

According to Darwin’s theory of natural selection, individual differences in adaptive fitness cause individual differences in reproductive success. It turns out that this seemingly simple idea is riddled with all sorts of complexities. One problem is conceptual: It is often difficult to conceive a gauge of fitness that is independent of reproductive success. Thus, evolutionary explanations can sometimes become circular. If fitness is synonymous with the capacity to reproduce, then it cannot serve as a cause of the differential reproductive success of various members of a species. Another problem is empirical: It is not always easy to measure reproductive success. To be sure, we can always count the number of copulations, or the number of offspring born or eggs laid, or the number of progeny that survive to maturity, and so forth. But each potential measure is laden with potential difficulties. For example, in species in which the females have multiple partners during their sexual cycle, clandestine sperm competition and cryptic female choice may decide which male is actually successful. Even in species that form pair bonds, such as birds often do, the females will occasionally cuckold their partners to increase the genetic diversity of their broods. Fortunately, modern techniques of DNA fingerprinting can alleviate many of these ambiguities by helping to identify the paternity of offspring. Nonetheless, these methods are far from error-free, especially when the breeding populations are quite homogeneous genetically.

As already noted, productive success may be considered the sociocultural analogue of reproductive success. Those individuals who make the most contributions to their disciplines are also those who have the greatest claim to the appellation of creative genius. In a sense, eminent creators are those who have proven themselves to be best adapted to their sociocultural milieu. However, the assessment of individual differences in productive success is fraught with all sorts of conceptual and empirical difficulties. Perhaps the biggest problem concerns measurement. What counts as a creative product? Should Newton’s substantial but unpublished work on alchemy be counted along with his obviously more accomplished research on mathematics, physics, and celestial mechanics? Or should his alchemical speculations be put in the same class as copulations that yield not a single fertilized ovum?

Even if we concentrate on products that have been exposed to the scrutiny of contemporaries, should we consider all published works or only those that actually had an impact on the field? Often in biological reproduction only a small proportion of the offspring in any given breeding season will be sufficiently fit to produce offspring themselves, either because they fail to reach maturity or because they cannot successfully compete for mates as adults. The sociocultural system is no less cruel in the application of its selectionist winnowings. In the sciences, for example, between one-third and one-half of the articles published in technical journals are apparently ignored by colleagues; the typical article receives absolutely no citations in other publications appearing in the scientific literature. This sociocultural neglect holds even for publications by distinguished individuals. One study scrutinized the total output of 10 highly eminent psychologists, including figures as diverse as Donald Campbell, J. P. Guilford, B. F. Skinner, and Wolf-

gang Kohler. Fully 44% of their publications received no citations during a five-year period. In other words, nearly half of their output was essentially ignored by the scientific community. An even more drastic selectivity was found in a study of 10 famous composers, which included such notables as Bach, Mozart, and Beethoven. Only 35% of their total output continues to have any active place in the classical repertoire today.

The response I adopt here is to use two different definitions of what counts as a product. The inclusive definition considers any published work as a product, regardless of its reception by contemporaries or future generations. This conception is closest to a purely behavioral definition of the phenomenon. However, we can also consider a more exclusive definition, which takes into consideration the actual impact of the product. By this latter, more sociocultural conception, a scientific article that no one cites or a musical composition that no one performs cannot qualify as a creative product. Because the exclusive definition depends on the reactions of others, such as colleagues, readers, and audiences, this conception is one step removed from objectivity. Indeed, to the extent that these social judgments are unstable, so will the identification of products be unreliable. Nonetheless, the exclusive definition comes closest to fitting our definition of creativity. Products that have had an influence on contemporaries and posterity must be considered adaptive in an certain sense. These are the entities most closely analogous to a biological variant that has managed to survive and reproduce its kind.

Given these definitions, I wish to examine two main features of individual differences in the output of products. I begin by looking at the specific shape of the cross-sectional distribution. Next, I turn to the functional relationship between those products defined inclusively and those defined exclusively.

Normal versus skewed distributions. Darwin was more interested in the existence of variation than in the specific form of that variation. Perhaps this lack of interest reflects his dearth of mathematical acumen. Nonetheless, many researchers before and after 1859 have preoccupied themselves with the specific shape of individual differences. For instance, back in 1835 Lambert Adolphe Jacques Quetelet established the importance of the normal distribution—the distribution defined by a bell-shaped “Gaussian” curve. Hence, for a trait like height, most people would exhibit average height. The probability of individuals being taller or shorter than this population mean would at first decline rapidly as the departure from the mean increases. But eventually the likelihood of encountering someone either extremely short or extremely tall will approach the zero point asymptotically. At this outer edge the distribution has tails that approximate the rim of a bell. A distinctive feature of the bell curve is that it is symmetric around the mean or average level of a trait. The probability of finding a person who is one foot taller than average is the same as that for a person who is one foot shorter than average.

Francis Galton, in his 1869 Hereditary Genius, extended the bell curve to the distribution of natural ability. The further out an individual stood on the right-hand tail of the normal distribution, the greater that person’s claim to genius. Later still, the pioneers in the psychometric measurement of intelligence made the normal curve almost an article of faith. This credo is obvious in the very title of the controversial book The Bell Curve, which discusses the origins and consequences of intelligence. Yet the same symmetric distribution has now been used to describe individual differences on virtually any psychological attribute, including creativity. In fact, this probability distribution is currently so ensconced in the behavioral sciences that the overwhelming majority of statistical methods posit normal (and multinormal) distributions as essential preconditions for all inferences!

If the traits on which adaptive fitness is based are normally dist ributed, it seems reasonable to suppose that reproductive success would itself be normally distributed. Yet in the biological world this is seldom the case. On the contrary, it is commonplace for a very small percentage of the individuals to generate the most substantial percentage of the offspring in a given breeding season. This disparity is especially conspicuous when reproductive opportunities are contingent on the individual’s placement in a dominance hierarchy or when they require the possession and successful defense of a suitable territory. Indeed, in some species many members of a single cohort—most commonly males—may not breed at all in their entire lives, while the top, “alpha male” may sire most of the offspring. To illustrate, one inquiry examined the reproductive prowess of a single tomcat with an easily traceable genetic trait. Scrutiny of all the kittens in all the litters produced by all the female cats in the study area revealed that this dominant male fathered more than 95% of the offspring! Hence, the cross-sectional distribution of reproductive success frequently departs substantially from the normal curve. Rather than a symmetrical, bell-shaped curve, the distribution may be highly skewed to the right; the lion’s share of offspring is produced by a small percentage of organisms, while the vast majority reproduce themselves but once if they succeed at all.

The same highly skewed distribution is characteristic of productive success. In any given domain of creative activity it is typical to find that around 10% of the creators are responsible for 50% of all the contributions. Stated in terms of the length of the curriculum vitae, only one creator in a hundred will make at least 10 contributions, and only one creator in a thousand will make at least 100 contributions. On the other hand, those creators who make up the bottom half of the population in terms of output are usually responsible for merely around 15% of the total supply of products. Indeed, the most frequent (or modal) level of lifetime output is usually a single product. These one-idea contributors usually constitute more than half of the total number of creators in the field. In contrast, the most productive members of the discipline are at least too times more prolific than members of this one-idea group. Actually, these statistics understate the magnitude of the productive elitism, for only those individuals who have generated at least one creative product are included. Needless to say, the population of individuals who have failed to make even a single contribution—whether patent or painting, article or book, poem or composition—is far, far greater. As in many biological populations, many individuals fail to pass this minimal test for creativity, leaving absolutely nothing for posterity to ponder or appreciate.

The skewed distribution of total output has been replicated so consistently that it has been described in terms of two mathematical laws. The simplest principle is the Price law. This affirms that if k represents the number of creators active in a given domain, then V¹ k gives the number of those who can be credited with roughly half of the products in that domain. For example, if there are roo individuals active in a field, then just 10 of them, or 10%, will be responsible for half of everything published. A more complex mathematical principle is the Lotka law, which states that the number of individuals who contribute exactly n products will be inversely proportional to n². In formal terms, if /(«) is the count of the creators who have made n contributions, then f(ri) = c/n², where c is some constant that is contingent on the particular domain of creative activity.

Most of the research on skewed distribution has used an inclusive definition of the creative product. For instance, the investigator might tabulate all of the papers published in a given set of scientific journals or all of the books recorded in the Library of Congress. Nonetheless, the same cross-sectional distribution appears when an exclusive definition is used. For example, all of the works that make up the standard repertoire are the product of approximately 250 classical composers. The square root of this number is 16, when rounded off to the nearest integer. In actual fact, just 16 composers account for half of all the pieces performed. Moreover, the 150 at the bottom of the heap can be credited with only one work each. All of their compositions together represent 6.0% of the repertoire, which is less than Mozart’s figure of 6.1%, and only slightly greater than the 5.9% each contributed by Bach and Beethoven.

Given the solidity of this finding, the obvious next question is where this skewed distribution comes from. Why do the top creative geniuses form such a conspicuous elite from the standpoint of productive success? One early suggestion can be dismissed at once, namely that the distribution merely represents the truncated upper tail of the normal curve. At first glance this seems a reasonable explanation. If creative output reflects underlying individual differences in natural ability, and if there exists a minimal level of ability necessary to generate a creative idea (e.g., an IQ of 120), then all of the members of the population with capacities below the threshold level will merely disappear from the tabulations. Unhappily for this explanation, the right-hand tail of the productivity distribution extends out much too far for this to provide a plausible account. To provide a dramatic illustration, if IQ scores were distributed the same way as creative productivity, the current human population could boast a half million intellects with IQs of 340 or higher! This figure easily exceeds by more than 100 points the highest IQ score ever recorded.

This overly simple explanation thus ruled out of court, I can turn to the following three alternative interpretations of the phenomenon:

• 1. Cumulative advantage—Many sociologists of science have argued that the observed individual differences in productivity can be explained according to the reward structures of the scientific community. The creative output of a scientist is rewarded each time a paper is accepted for publication in a major journal and each time a grant proposal is approved by a major funding agency. Because journal editors and program directors must be selective, the odds of success are usually very small. Hence, those scientists who just happen to be successful early in their careers will have their creativity reinforced sooner, giving them an edge over those of their cohorts who at first failed. This process will often be accentuated by the fact that those who are successful earlier will find themselves affiliated with more prestigious institutions that provide more support for scientific research. All of this differential reinforcement can then accumulate over time to generate large inequalities among scientific workers, even when all began their careers with equal creative potential. The rich just get richer while the poor get poorer. The outcome of this process will be the highly skewed distribution observed in the empirical literature. Although this model was first proposed with respect to scientific creativity, it is not hard to extend the same explanation to individual differences in the output of artistic products.

• 2. Multiplicative influences—William Shockley, the Nobel laureate who was not bright enough to make it into the Terman sample, has provided an interpretation that recognizes that many different factors determine creative productivity. Furthermore, these several determinants may participate in a multiplicative rather than additive fashion. That is, the capacity for producing creative ideas may be the product rather than the sum of the various components. Thus, people who have an insufficiently high intelligence will be uncreative no matter how high they may be on the other participating factors. This multiplicative model has a critical implication with respect to the predicted probability distribution of productivity. If the diverse components are normally distributed in the population, the distribution of their product will not assume the same form. On the contrary, the expected distribution would be “lognormal.” This is a distribution that would be normal were the productivity counts subjected to a logarithmic transformation. This asymmetrical distribution is highly skewed right. Most of the cases are collected at the lower end of the distribution, while the upper end of the distribution is characterized by an extremely long tail. This upper tail, in fact, is indistinguishable from that observed for creative output. Thus, according to the multiplicative model, the most prolific creators will be extremely rare.

• 3. Combinatorial explosions—According to the analysis of the creative process discussed in chapter 2, the generation of ideational variants depends on the richness of a person’s associative network. The larger the number of concepts in that network, and the more diffuse their interconnections, the greater is the potential supply of recombinations. However, it is necessary to recognize that the number of available ideational variants would not increase as a linear function of the number of ideas that can be freely permuted. On the contrary, the number of possible combinations would increase explosively with increases in the number of elements being combined. As a rough but reasonable approximation, we might suppose that the potential supply of ideational variants increases exponentially with the amount of raw material. Let us further assume that the size of the reservoir of ideas available for recombination is normally distributed in the population. It immediately follows that the cross-sectional distribution of total potential variants would again be a lognormal distribution. Once more we can obtain a highly skewed distribution that permits the emergence of a supremely productive elite.

There is no particular reason why we must adopt one of these explanations to the exclusion of the other two. It is conceivable that all three processes collaborate in the determination of creative output. Even so, the three interpretations have variable implications for a Darwinian conception of creative genius. The combinatorial-explosion account is clearly the one most consistent with Campbell’s variation-selection theory of creativity. Moreover, the multiplicative-influences interpretation has obvious affinities with the behavioral-genetic model of emergenesis. It has the additional advantage of being applicable to both productive and reproductive success. Biological fitness may also be a multiplicative function of a large number of normally distributed traits. This would explain why the cross-sectional distribution of reproductive success tends to exhibit the same skewed form observed for productive success. In fact, both Price and Lotka’s laws may apply to the output of progeny. To verify this conjecture, I reanalyzed the data from a 15-year study of 32 male rhesus macaques and found that the six (=^32) most prominent progenitors sired 48% of the offspring. This is in close agreement with the prediction of the Price law. So both cultural and biological fitness may be governed by a deeper multiplicative process.

In contrast, the cumulative-advantage explanation would appear to sit uneasily with a Darwinian view. In particular, advocates of this theory frequently argue that individual differences are imposed from the outside rather than being intrinsic to the creative person. Variation in associative powers, personal disposition, and early background—such as treated in the prior two chapters—may be simply irrelevant. Instead, the contrasts between genius and nonentity can be ascribed to the capricious influence of luck and timing. Indeed, this notion has been christened the “Ecclesiastes hypothesis” after the biblical passage “The race is not to the swift, nor the battle to the strong, neither bread to the wise, nor yet riches to men of understanding, nor yet favor to men of skill; but time and chance hap-peneth to them all.” Consequently, productive success becomes decoupled from cultural fitness. Such a hypothesis is as anti-Darwinian as would be the analogous claim that variation in reproductive success had absolutely no correspondence with adaptive fitness.

Hence, if cumulative advantage were the only plausible explanation of individual differences in creative output, a Darwinian theory of creative genius would be seriously jeopardized. Even worse, a psychology of creativity would appear to be a meaningless enterprise. The more productive creators would not differ in any important manner from their unsuccessful colleagues. It is fortunate, therefore, that this sociological model fails to accommodate all aspects of the phenomenon. A recent investigation has shown, for example, that this interpretation cannot account for how productivity changes across the career. According to the model, the output in adjacent periods of the career should be more highly correlated than is the output in noncontiguous periods. Instead, individual differences in the output in the several periods that define a career appear to reflect crosssectional variation in an underlying factor. This latent factor has been styled “creative potential.” The particular skewed distribution of this potential may be then explicated in terms of both the multiplicative-influences and combinatorial explosion theories. Thus, there is no need to assume that the distinctive cross-sectional distribution of creative productivity is inconsistent with a Darwinian conception of genius. Nor do we have to dismiss the discussion in chapters 3 and 4 as being irrelevant to a theory of creativity.

Quality versus quantity of output. As noted earlier, reproductive success has more than one operational definition. Although this might seem worrisome, the existence of alternative definitions is only a problem if the various conceptions lead to dramatically different conclusions. Yet the contrary appears to be the case. Those sexually reproducing organisms that copulate more also tend to parent more offspring, and to have more offspring survive to reproduce themselves. So alternative measures all converge on the same basic assessment. But does the same convergence hold in the case of productive success? Do individual differences in the output of works inclusively defined correspond to individual differences in output exclusively defined? Is it true that those who generate the most products are also those who generate the most products that actually exert an influence? And are those who had virtually no impact on the field also those who ventured very little in the first place?

Certainly we can conceive two types of creative careers that strongly contradict the correspondence between inclusive and exclusive definitions. First, there may exist perfectionists who produce very little, but every single product is a masterpiece of the highest order. Second, there may appear mass producers who originate numerous products to no effect, not one of their offerings making any noticeable impression on their colleagues or the world at large. This latter career type is especially problematic, for mass producers would be creative by the inclusive definition but uncreative by the exclusive definition. They are cultural counterparts to a sterile beast who incessantly copulates but still conceives not a single offspring.

Fortunately, study after study has found that those creators who are the most prolific by the inclusive definition are also the most prolific by the exclusive definition. In other words, productive quality of output, or socially certified creativity, is positively correlated with productive quantity, or mere behavioral output regardless of consequence. For example, U.S. Nobel laureates publish two times as many scientific papers as do scientists still worthy enough to make it into American Men and Women of Science. The number of citations that a scientist received in the work of fellow scientists is strongly associated with the total output of publications. In fact, the total number of publications predicts the amount of citations received by a scientist’s three most acclaimed works. Moreover, this correspondence between quantity and quality holds over the long haul. For instance, the total length of the bibliography of a nineteenth-century scientist predicts how famous he or she is today. Thus, a scientist who was then in the top 10% of the most productive elite has a 50-50 chance of earning an entry in a recent edition of the Encyclopaedia Britannica. In contrast, their less prolific colleagues have only three chances out of a hundred of earning that distinction. Mendel could make a lasting impact on science with only a half dozen publications, but cases like his are not frequent enough to undermine the correspondence.

The positive correlation between quantity and quality has a provocative repercussion. If the number of influential works is directly proportional to the total number of works produced, then the creators with the most masterpieces will be those with the most ignored and neglected products! Even the most supreme creative geniuses must have their careers punctuated with wasted efforts. W. H. Auden put it well: “The chances are that, in the course of his lifetime, the major poet will write more bad poems than the minor.” This holds for major creators in other domains. In Einstein’s crusade to overthrow quantum physics, he often made embarrassing mistakes. Once after an extended debate with Niels Bohr, Einstein conceived a sophisticated argument that he thought would demolish the Copenhagen school. Bohr found a fatal flaw nonetheless: Einstein had neglected to consider the implications of his own theory of relativity. Yet Einstein would defend his mistakes with the maxim “Science can progress on the basis of error as long as it is not trivial.”

Naturally, from the Darwinian perspective these results are not surprising. Owing to the harsh struggle for existence, only a small proportion of even a successful organism’s progeny will survive to populate the next generation. Many offspring die before even leaving the nest or den, and still more fall victim to predation, disease, starvation, or exposure once they start to fend for themselves in the wild. Even should they become mature adults, they may succumb to the rigors of territorial defense, dominance hierarchies, and mate competition. Yet none of this matters. The more potential descendants generated, the more actual descendants may survive—and that is all that Darwin’s theory demands. Adaptive fitness does not signify that every nestling or cub or child will become a sexually active adult. Fitness only permits the organism to make many trials, and many errors, with the implicit hope that at least one variant will carry its genes into the next generation.

One final Darwinian feature of the quantity-quality relationship deserves emphasis. If the creative genius is generating failures as well as successes, this seems to support the assumption that the creative process is to a certain extent blind. Even the greatest creators possess no direct and secure path to truth or beauty. They cannot guarantee that every published idea will survive further evaluation and testing at the hands of audiences or colleagues. The best the creative genius can do is to be as prolific as possible in generating products in the hope that at least some subset will survive the test of time. To be sure, a critic might take the existence of the mass producers and the perfectionists as exceptions to this rule. Perhaps these careers differ in a qualitative manner from those whose successful works are a predictable proportion of their total output. On the one hand, perhaps the mass producers are not really generating variations, but rather they are one-idea individuals applying the same technique or theory over and over. On the other hand, maybe the perfectionists have grasped some special trick that allows them to circumvent the process of trial and error, allowing them to produce masterpiece after masterpiece.

Nonetheless, the existence of mass producers and perfectionists does not really introduce problems for the Darwinian interpretation. After all, the relationship between quantity and quality is far from perfect. As a consequence, there will always exist considerable scatter (or residual errors) around the regression line that predicts quality as a function of quantity. Moreover, in all empirical data sets so far published these departures are distributed around the prediction line in the usual statistical fashion. Most of the errors will be close to the line of best fit, with very few errors far away. There apparently exist no dramatic outliers that would represent the truly problematic cases. Thus the mass producer may be considered a person who was simply unlucky, whereas the perfectionist may be considered a person who was merely lucky. Chance alone will randomly distribute luck both good and bad, and a small percentage* of individuals will appear unusually lucky or unlucky. If I repeatedly flip a coin to times in a row, I will eventually obtain 10 straight heads or 10 straight tails. But when this happens, those with savvy in probability theory do not suddenly jump to the conclusion that the coin is weighted.

The same pattern applies to reproductive success. Two organisms may have identical adaptive fitness, but one may produce more descendants than the other owing to totally extraneous factors. Just by chance one will become a victim of a predator, an epidemic, a dispute over territory or mate, a flood, or fire, or volcanic eruption. For this reason fitness cannot be equated with reproductive success. If they were really equal, then identical twins would invariably have exactly the same number of offspring, which is no more true in the animal world than it is in the human world. Accordingly, fitness must be defined as a propensity to survive and reproduce. Individuals with the identical genetic constitution living in equivalent ecological niches will have the same propensity, but the actual success exhibited by these individuals will be scattered around some mean value that represents the predicted reproductive success. Because an individual’s fitness represents solely a probability or likelihood, two individuals with identical outcomes in the struggle for existence may actually have had unequal levels of adaptive fitness. Thus, in organic evolution the linkage between fitness and success is stochastic rather than deterministic.

I realize that the parallel just drawn between natural and cultural selection runs counter to a recurrent myth about creative genius. There exists a strong tendency to idolize historic creators, to see them as infallible in their capacity to generate one magnum opus after another, to deem them all perfectionists. The list of their contributions by the inclusive definition is thought equivalent to the list according to the exclusive definition. In brief, they are all hits and no misses. Yet we must never forget that we inevitably view the achievements of eminent creators through the selective processes of sociocultural evolution. The successful products are what survive, while the unsuccessful products are usually condemned to oblivion. The bad books go out of print while the surviving copies rot in storage rooms. The inferior paintings sit in crates in museum basements. The poor compositions collect dust in obscure musical archives. These are cultural variants that suffered extinction, buried until almost inaccessible benea th cultural strata. But the assiduous biographer, just like the meticulous paleontologist, can often dig up these forgotten creations from the depths of the creator’s and history’s past.

Darwin himself illustrates what can be dug up with only a little effort. For many Darwinists, he appears to represent the model scientist, the bona fide perfectionist. Hit seemed to follow hit in a most admirable progression. But if we delve carefully into his lifetime output, this idealized portrait begins to reveal many blemishes. He was capable of publishing erroneous interpretations and even silly conjectures. An early paper provided such a completely mistaken explanation for a particular geological formation that it came to cause Darwin considerable embarrassment. Later, despite his extremely detailed work on the cirripedes, he was forced to admit that he had “blundered dreadfully about the cement glands.” Later still, he introduced the erroneous theory of pangenesis that contaminated his evolutionary theory with the Lamarckian notion of acquired characteristics. Yet all of these mistakes and numerous others are forgotten and forgiven. His geological paper on Glen Roy is politely ignored by geologists, and his work on the barnacles has been superseded by more accurate monographs. Darwin’s theory of pangenesis has been reduced to a tiny footnote in the history of evolutionary theory. What remains in posterity’s eyes is a sanitized Darwin whose career seems quite un-Darwinian—no variation and selection, no trial and error, no hits and misses. Yet I hope that this misperception will eventually enter the historical record as just another false idea that did not survive cultural selection. This unjustified glorification of genius must be buried and fossilized along with the dinosaurs.

Developmental Change

Darwin’s scientific career was impressively long. While still a teenager attending Edinburgh University, he had read two papers before the local Plinian Society. At age 22 he began the voyage on the Beagle, and soon after his novel observations began to be read before various scientific societies. Shortly after his return to England, he began writing up his findings, and he started collecting notes on the transmutation of species. Before he turned 30 he already had several publications to his credit, beginning the pattern of consistent output that was to continue throughout his life. In the last year of his life, at age 74, he read before the Linnean Society papers on the effects of ammonia carbonate on plant roots and chlorophyll bodies, besides sending a note to Nature on the dispersal of freshwater bivalves and writing a preliminary notice to a colleague’s paper on Syrian dogs. The day after the latter piece was read before the Zoological Society, Darwin died. All told, Darwin’s scientific life lasted well over a half century.

During that long interval he underwent many profound developmental changes. Moreover, as an evolutionist with a deep fascination with development, he could not resist making his own observations on how human beings change across the life span. For instance, he published a pioneer study of the early development of one of his own children. Darwin’s writings also contain many insightful statements about adulthood and old age. In the Origin of Species, for example, Darwin predicted an antagonistic response from his scientific elders. Specifically, he did not “expect to convince experienced naturalists whose minds are stocked with a multitude of facts all viewed, during a long course of years, from a point of view directly opposite to mine.” Instead, Darwin looked “with confidence to the future— to the young and rising naturalists, who will be able to view both sides of the question with impartiality.” In private conversations, Darwin expressed this opinion in stronger terms, as is evident from what he once told Lyell, who was a dozen years his senior. “What a good thing it would be,” Darwin told his colleague, “if every scientific man was to die when sixty years old, as afterwards he would be sure to oppose all new doctrines.” When Lyell finally converted to evolution after having already reached his 70s, he told Darwin that “he hoped that he might be allowed to live.”

Almost a century later, Max Planck made a similar claim with respect to the reactions to the new quantum theory: “A new scientific truth does not triumph by convincing its opponents and making them see the light, but rather because its opponents eventually die, and a new generation grows up that is familiar with it.” Despite Darwin’s patent priority, this proposition is now styled “Planck’s principle.” Nonetheless, the first empirical test of this developmental hypothesis was based on the reception history of the theory of evolution by natural selection. Not only was Darwin’s conjecture proven correct, but other studies have shown that the phenomenon is truly a general one. Older scientists and scholars are somewhat disinclined to accept new ideas. In fact, this age effect seems generally operative in the affairs of everyday life, too, such as in business and farming.

Given his powers of observation, it is unfortunate that Darwin did not offer any definite hypotheses about how creative productivity changes across the career. The only such speculation I can identify is his conclusion that elder scientists are often prone to propose foolish theories. So strong was his belief, in fact, that he promised himself he would not theorize after he had turned 60 years old (a promise he failed to keep). Despite this apparent silence, longitudinal changes in the output of products must be considered by any comprehensive theory of creativity. Below I examine two aspects of the phenomenon: career trajectories and the equal-odds rule.

Career trajectories. Curiously, the first scientist to study the crosssectional distribution of human characteristics was also the first to investigate the longitudinal fluctuations in the output of creative products. Back in 1835, Quetelet looked at the plays produced by the leading dramatists of France and England and tabulated the number of works produced in consecutive age periods. Many other investigators have followed this pioneer effort, making the necessary methodological improvements. This research has established the following four empirical generalizations:

• 1. The output of creative products tends to increase as a curvilinear, single-peak function of age. Figure 5.1 gives the typical age curve. Needless to say, any particular career will depart noticeably from this idealized trajectory. Physical illness, war, and other extrinsic events can disrupt the creative process. Even so, if tabulations of output are averaged across many careers, so that these random shocks can cancel each other out, the correspondence between this function and the data becomes quite striking. Usually about 95% of the career fluctuations in productivity can be ascribed to this inverted backward-J curve.

• 2. This trajectory in output is a function of career age rather than chronological age. Thus, late bloomers who delayed launching their careers will have the predicted peak displaced by a proportional amount. By the same token, early bloomers will have their career high points advanced.

• 3. The specific shape of the curve varies according to the discipline. In domains like pure mathematics and lyric poetry, the curve approaches the peak rather rapidly, and the decline after the optimum level may be dramatic. In fields like geology and philosophy, the rise to the maximum output point is more gradual, and the subsequent descent less precipitous.

Career Age

Figure 5.1 Typical curve describing the annual output of creative products as a function of age according to a Darwinian model of the creative process.

• 4. Once allowance is made for these interdisciplinary contrasts, individual differences in lifetime output do not affect the overall shape of the curve. Instead, this cross-sectional variation determines the overall height of the curve. The more prolific creators are simply generating more products per un