Giant who started the silicon age

He gave us the transistor, but he was also a vigorous advocate of eugenics. Philip Anderson recalls a flawed miracle worker

Genius? In the community of theoretical physicists all are geniuses - just ask a mathematician. But William Shockley, who invented the transistor, was something else again.

Joel Shurkin's biography follows the career of this remarkable, disturbingly flawed man. It charts his lonely childhood, the 20 years during which he built up Bell Telephone Laboratories' research to the fabulous level it maintained for 50 years, the period in which he provided the impetus for the phenomenon of Silicon Valley but lost his own semiconductor company and his moorings, and finally the self-inflicted tragedy of his final campaign as the majordomo of the eugenics movement, which will for ever colour his public image.

I was hired by Shockley in 1949, at a time when, according to histories of the transistor breakthrough, he should reasonably have been consumed by that role. In fact, he was simultaneously involved in exciting research in other fields, for which he was recruiting spectacularly (four of us who went on to substantial careers of our own arrived at Bell on the same afternoon). Whatever scientific challenge Shockley picked up, he seemed well on the way to a solution before others had even grasped the problem. In addition, he was carrying a full load advising the military.

Shurkin's access to detailed records made available by the Shockley family, including some previously unopened safes, as well as his use of recorded interviews with many key figures, makes Broken Genius revealing about Shockley's personal life as well as his scientific achievements. To his credit, Shurkin does not presume to come to any pat psychoanalytical conclusions, but allows us our own view of the man based on voluminous data.

Shockley was born in 1910, the only child of a mature and highly intelligent woman, married to a considerably older mining engineer whose money-making talents did not match his charm and erudition. The result was a somewhat itinerant childhood with few friends of his own age.

Fortunately, the family finances extended to sending him to the young but already vibrant California Institute of Technology (Caltech) as an undergraduate, where he learnt first hand about the revolutions taking place in modern physics.

In 1932, he and classmate Fred Seitz drove across the country together in his somewhat battered DeSoto to graduate school at the Massachusetts Institute of Technology and Princeton University, respectively. In these two centres, the new field of the quantum theory of solids was being born.

Even as students, both Shockley and Seitz became full, even key, participants in the field.

Here, too, began Shockley's lifelong near-obsession with physical fitness - he even earned the odd sum modelling, although he was no Arnold Schwarzenegger. His chief sport was rock climbing, with routes named after him in the Shawangunk Mountains and elsewhere. At the same time, he also became an accomplished magician, a skill he used to charm children.

By 1936 Bell Labs, recovering from the Depression, had begun to recruit. Its young scientist-executive, Mervin Kelly, on the advice of Phil Morse, Shockley's mentor at MIT, brought in Shockley and his friend Jim Fisk (a future president of Bell Labs). Others hired then included stars-to-be such as Charles Townes (lasers) and John Pierce (comsats). The idea was to see what quantum physics could do to replace the awkward technology used in Bell's telephone system, in particular the ubiquitous vacuum tube. A number of veteran Bell scientists, such as Walter Brattain, joined these "young Turks" in a research department that was the forerunner of the postwar Bell Labs. Shockley immediately assumed a leadership role.

In an odd interlude, for a month in 1940, Shockley and Fisk were withdrawn from their lab benches and given the ultrasecret assignment of assessing the consequences of uranium fission. Their analysis so resembled the design of the first atomic pile built by Enrico Fermi in 1942 that it appears to have trumped all later patent claims, although it is unclear whether it ever broke through the fog of military secrecy enough to have been useful to Fermi.

The Second World War broke up this unique research group. Shockley was recruited, again by Morse, to join the Weapons Systems Evaluation Group (WSEG), then desperately trying to counter the German U-boat threat in the Atlantic. Shockley was second in command. The achievements of this group in improving the performance of the Navy boggle the mind - and for this alone both the US and the UK owe Shockley a real debt of gratitude. The commitment of time and energy, however, led to increasing strains in Shockley's marriage; during the war his young family received little attention.

After 1945, Kelly, then president of Bell Labs, with Shockley set out to create a group that could revolutionise the technology of communications. Its members included John Bardeen and Brattain - the two who would share the credit with Shockley for the transistor. Other noteworthy recruits were acquaintances from the WSEG, such as Conyers Herring and Charles Kittel. It is worth emphasising that semiconductors were not the group's only interest; this misleading impression arises from the natural ex post facto bias of historians and is repeated in this book to some extent.

By Christmas 1947, Bardeen and Brattain had created the first semiconductor amplifier. And from that point on, the triumphal story of Bill Shockley the wonder-worker began to unravel.

Shurkin describes the Bardeen-Brattain phone call to Shockley announcing their breakthrough. It elicited not joyful triumph but blinding anger that the amplifier, which Shockley had expected to be his crowning achievement, had been stolen from him by scientists he saw as subordinate, even inferior. Shockley moved quickly to fill, and had witnessed, notebooks with extensions and generalisations of the original Bardeen-Brattain device in order to ensure a separate patent claim for his (undoubtedly superior) version. He then isolated Brattain and Bardeen from the group he set up to exploit the breakthrough and saw to it that the publicity photograph of the first transistor showed him at the bench experimenting, with Brattain and Bardeen in the background looking on.

It is ridiculous to argue, as some have, that because Bardeen and Brattain did not predict some details of the device they had created Shockley's behaviour was partially justified. Their device by its very existence revealed the basic principle of "transistor action", the use of a forward-bias junction for injection, and started the semiconductor industry on its merry way. Bardeen's depth of understanding of semiconductor physics is manifested by a review he wrote in the next year with Gerald Pearson, which is as much a basic text for physicists as Shockley's book was for engineers.

For a while things went swimmingly for Shockley. His group, after the excision of Bardeen and Brattain (and a few misfits such as Kittel and myself), achieved steadily increasing control of their materials and developed Shockley's improved design into the junction transistor. In 1951, Bell Labs hosted a noteworthy symposium at which transistor technology was thrown open to the world, and the media began to notice. In all this, the visible leader was Shockley. However, he had violated the cardinal rule of Bell Labs management that one should not compete against a subordinate, and it was beginning to look as though his future at Bell was limited. He still had influence - it was Shockley who pushed through the expansion of Bell's salary range in 1954-56, a significant move - but he would never rise to a higher level of management. He began to spend more and more time advising the Pentagon and, ominously, developing his theories of "human quality".

Some of his colleagues had made profitable moves to outside competitors, such as Texas Instruments, that were springing up in the now lucrative transistor industry, but only Shockley saw the enormous potential of what they had begun. In 1955, he began to form Shockley Semiconductors as an independent company (funded by Arnold Beckman) in a small building in Mountain View, California, just south of Stanford University. His company was Silicon Valley's first start-up.

Shockley's marriage now finally broke down, as recounted by Shurkin in rather excruciating detail. Enormous solicitude for his wife Jean's uterine cancer was followed by abandonment. There was a brush with suicide by Shockley, and some episodic flirtations; but in the end Jean was replaced during the same busy year by Emmy, a plain, no-nonsense, bright nurse-educator who went with him to Stockholm in 1956 for his Nobel prize and stuck with him through his final unravelling.

True to form, Shockley's first California team was brilliant, with such legendary figures as Bob Noyce and Gordon Moore. But within little more than a year, the boss's management style had alienated and challenged the top eight of his new staff, who left to form Fairchild Semiconductors, which was then bought out from under them, providing funds to found Intel, from which defectors founded the next start-up. Meanwhile Shockley Semiconductors languished and died. In 1965, its founder returned to a much diminished role at Bell Labs.

The demise of his company left Shockley time to think about his ideas on "human quality". He believed that it could be quantified and that, in the absence of some better measure, one should rely on IQ, with which he was somewhat familiar because of his mother's early acquaintance with the Terman "gifted children" programme at Stanford, in which high-school students from the area were given special educational opportunities. (Oddly enough, her son was twice rejected for this.) In a sense, Shockley's philosophical approach came from his wartime work with the WSEG, much of which concerned finding "a quantitative measure of success" and then acting on "the obvious consequences".

To Shockley, the obvious consequence of quantifying human quality was eugenics in all its easy fallaciousness: opposition to remedial programmes such as affirmative action and Head Start, support for sterilisation of the "unfit", the lot. He spent his last 25 years at the barricades, pursuing such a programme with characteristic singlemindedness, exploiting every opportunity for publicity afforded by his Nobel prize. The story is not pretty. I well remember the disruptive sessions at the National Academy of Sciences, described by Shurkin, at which Shockley broke for life with his old friend Seitz (then president of the NAS). As Shurkin describes, it was indeed difficult to spend time with Shockley, but one could still sense somewhere within him a reasonable, likable, even admirable man - if one could avoid his obsession.

How can one summarise such a man? Perhaps his physics is the clearest clue.

Shockley called his method of thinking "try simplest cases". But that is really what many good theoretical physicists do when confronted with a complex problem; they try to find a simple model. Only Shockley elevated it to a mantra. He could see his way through the first few stages of any problem very quickly, but he rarely employed his talents to look further below the surface or to check his models against reality. Those who had this ability - for whom Shockley undoubtedly had a good eye - would almost inevitably go off on their own because Shockley could never allow them to follow their own instincts. He saw them as competition and a threat to his authority. I know, because it happened to me in a small way in 1950, and, like others, I survived and may have been the better for it.

Shurkin is a good storyteller, and better still as a researcher of the personal facts. But Broken Genius needed more copy-editing. Not all the facts are reliable: Bardeen never ran a theoretical group at Bell - one did not exist as such until 1956 - and Shockley's secretary in 1947 was not Betty Sparks. Perhaps more important, it seems inevitable that a historian is not the best explicator of science; I would have liked a lot less talk about the mystery of quantum theory and at least a mention of energy bands.

Such quibbles are perhaps unfair - few biographies are reviewed by someone who was on the scene for much of the time. What comes out of the book well is Shockley's importance. He arguably saved Britain from the U-boats during the battle of the Atlantic. He is certainly the true father of the age of silicon. He is even the inventor of the graphite-moderated nuclear reactor. Yet the public will remember his name as that of the nutty Nobellist who donated his sperm to a genius bank. Who would not want to hear his whole story?

Philip W. Anderson is emeritus professor of physics, Princeton University, New Jersey, US. He was awarded the National Medal of Science by the US Government and is a Nobel laureate in physics.