One day in the late 1970s, Bardeen came by the office of Karl Hess, a young postdoc who had recently come from Austria to study with Bardeen. Hess had studied under Karlheinz Seeger, who had worked with Bardeen some years earlier. Hess asked Bardeen what he looked for in choosing a good physics problem.

Speaking slowly and using his fingers to count off his points, Bardeen told Hess that there were three important requirements. “First of all,” look at “whether there is a technological basis” for the work. “If you think of something in some theory but it can never be realized because there is no technology there, you are working in empty space.” Second, the problem needed to be challenging, “because if it’s so simple that you can do it on the back of an envelope, well, the project is over.” Third, the research should have applications potential. “That’s what most people in basic science overlook. If you do something and you want it to have importance, it has to mean something to the people.” Asked once what he did with a problem that showed no promise of having practical applications, Bardeen smiled and replied, “I choose another problem. You don’t want to pick one that’s too difficult, but you must pick one that will lead to significant results.”

With such a high value on the practical side of physics, it was natural for Bardeen to want to maintain strong ties to industry,

even after his return to academia. His friend Nelson Leonard, a chemist, said that Bardeen’s “face always lit up” when he spoke about the interplay of science, technology, and industry. Charles Gallo, a scientist at 3M, recalled that when Bardeen would speak about high-temperature superconductors at Minnesota Mining and Manufacturing, he “never once mentioned a theoretical, abstract or philosophical motivation. He only stressed the myriad practical applications and the economic viability.”

Bardeen’s longest industrial association began in 1952, when he was contacted by a small company based in Rochester, New York. The Haloid Company had invested in an electrophotographic copying process to become known as xerography. The concept was still new and relatively undeveloped. Its scientific and engineering problems were daunting. No one could have predicted then how far the commitment to xerography would take Haloid.

In 1906, at the time Haloid was founded, the company specialized in high-quality photographic paper. Its business existed in the margins of the markets occupied by its giant competitor down the road, Eastman Kodak. In the mid-1930s, Haloid acquired the Rectigraph Company, and with it the copying process known as the photostat, which used Haloid’s photographic paper. From Rectigraph came John Dessauer, who became Haloid’s director of research.

In 1945 Dessauer noticed a brief reference in Kodak’s Monthly Abstract Bulletin to a process called electrophotography. Immediately struck by the commercial potential of this process, Dessauer brought the work to the attention of Haloid’s president, Joseph Wilson.

The concept of electrophotography was then less than a decade old. It had been invented by Chester F. Carlson, a patent lawyer, who wanted to develop a process useful for copying documents. After studying the available photographic methods, including Rectigraph’s photostat process and Kodak’s verifax process, Carlson came upon an electrographic method that used photoconductivity, a process pioneered by Paul Selenyi. The Hungarian physicist had been dissuaded from pursuing the industrial application of his idea. On the basis of Selenyi’s description, Carlson began to investigate how to form electrostatic images on photoconducting insulating

layers. The work led him to invent electrostatic electrophotography in 1937. His application for a patent was accepted in 1943.

With the hope of developing his idea into a marketable product, Carlson contacted various corporations, including IBM, A. B. Dick, and RCA. None of the corporations were interested, for the working model that Carlson had constructed failed as often as it worked. But the idea interested Russell Dayton at Battelle Memorial Institute, a small nonprofit research organization based in Columbus, Ohio. In the article that Dessauer read in Kodak’s Monthly Abstract Bulletin, he learned that Battelle wanted to supplement its research into the electrophotography process by contracting with a larger industrial or university laboratory.

Haloid’s president, Joe Wilson, agreed with Dessauer that the electrophotography would dovetail with Haloid’s product line. Haloid then entered into an agreement with Battelle. A year later, in 1946, Haloid bought the patents for electrophotography from Battelle. In 1947 this process was renamed “xerography,” after the Greek words for “dry” (xeros) and “writing” or “drawing” (graphein). Part of the way in which Wilson supported research on Carlson’s electrophotographic process was by securing military contracts. The process could be used in copying microfilm held in military archives.

Bardeen’s connection with Haloid arose from a decision that Dessauer made several years later to draw on the expertise of outside scientists. Harold Clark, a member of Haloid’s staff, suggested asking Bardeen whether he would be interested in consulting for the firm. Clark had taken a course from Bardeen at the University of Minnesota and thought of him when he heard that Bardeen was leaving Bell Labs. Dessauer considered it unlikely that a scientist of such high stature would want to associate with such a small company, but felt it would not hurt to ask. Intrigued by the concept of an electrostatic copying process, Bardeen readily agreed to meet with Clark and Dessauer in Cambridge, Massachusetts, where he was giving a seminar.

As the three men strolled along the banks of the Charles River, they talked about the relationship between science and industry. Bardeen may also have recalled the exciting years he had spent in Cambridge as a Harvard junior fellow. As he listened to the Haloid scientists explain the copier project, he may well have recalled the monumental task he had faced making the two copies that

Princeton had required of his doctoral dissertation. He was undoubtedly also intrigued by the vast potential social and economic value of such a technology. In later years he would add that he was immediately attracted to the philosophy of the company’s president, Joseph Wilson, who considered research “a key to our development.”

Bardeen admired the young company’s self-image as an institution willing to take a risk that would “challenge its short-run position in order to buttress the long years ahead.” He accepted Dessauer’s offer. His agreement, formalized in October 1952, called for six to eight visits per year to Haloid’s research laboratory, which at that time was based “in a few laboratory rooms in a converted frame residence.” In 1961, as Haloid poised itself to become an international giant, the company would change its name to Xerox Corporation. Its Xerox 914 copier would be hailed by Fortune magazine as “the most successful product ever marketed in America.”

In the years when Haloid was growing into a Fortune 500 company, Bardeen consulted with its researchers on many scientific issues. For example, after selenium was selected as the photoconductor material, Bardeen made useful suggestions regarding particular design processes, such as reducing the time the photoconductor plate rests between copies (the “fatigue delay”) by doping the selenium to better control its electrical properties. This innovation, for which he filed a patent in February 1958, was an important contribution to the research process but was never directly incorporated into commercial Xerox machines.

On a typical consulting visit Bardeen would deliver a lecture in the morning. After lunch, he would visit several research teams in their labs or offices. Between visits he wrote letters and made phone calls about areas in which he felt the company ought to be up-to-date. He commented on possible scientific hires and gave his opinions about who to appoint to positions of leadership. In its early days Haloid had found it difficult to attract the best Ph.D.s. Bardeen’s presence as a consultant, and later as a member of the board of directors, helped to allay job candidates’ fears that they might not have the freedom or the resources to engage in cutting-edge research.

Xerox scientists found that Bardeen “would listen intensely to what we were saying,” giving the researchers “a chance to talk

things out.” The frugality of his verbal responses led some to “wonder whether he was seriously interested” or possibly so alarmed at the poor quality of work “that he was being kind by not probing too deeply with his questions.” But even the scientists who had been nonplussed by Bardeen’s unresponsiveness found that “at the end of the day, he provided critical guidance.”

Contact with one of the most prominent solid-state physicists in the world also boosted morale at the laboratory. Xerox scientist Charles Duke said that Bardeen projected a philosophy that “the business of research was the discovery of new insights rather than the refinement of the old” and that the best approach was to “hire the best, equip them to the cutting edge of the state-of-the-art, and stay out of their way.” In addition, because Bardeen was “unfailingly kind, considerate, and courteous,” he contributed toward setting “a standard of humane behavior in fiercely competitive and arrogant surroundings.”

Bardeen’s reports to Xerox offered guidelines for successful industrial research. For example, in 1978 he specified that only two “levels of effort” were viable: “a major effort at the forefront that is recognized worldwide” or “a minor effort sufficient to keep in close contact with developments made elsewhere.” For Xerox’s facility in Webster, New York, he advocated the latter, encouraging its scientists to choose “side problems,” such as photoelectron spectroscopy, that would bring Xerox into cooperation with other major research centers, such as the surface program of the University of Pennsylvania. He explained that such a strategy would make Xerox “part of a leading research group at a fraction of the cost of the total program.”

Bardeen’s talks at Xerox often conveyed his philosophy that “invention does not occur in a vacuum.” Most advances “are made in response to a need, so that it is necessary to have some sort of practical goal in mind while the basic research is being done; otherwise, it may be of little value.” He recalled the period at Bell Labs before the invention of the transistor as a productive time, citing teamwork in the semiconductor subgroup as “the sort of basic research which should be done in industry.” Referring to the Bell Labs mission of improving communication and to Mervin Kelly’s goal in the late 1930s of replacing the relay and vacuum tube, Bardeen said, “Those doing basic research should be well aware of the long-range goals of the company and of the specific research

programs in which they are involved.” In 1973 he worried that budgetary requirements imposed by Xerox’s “bean-counters” were dictating “too much opportunism, jumping on the latest band-wagon.” He felt that the lack of long-range planning created “much of the uncertainty over missions and goals as emphases are changed from year to year.”

The most fruitful problems, Bardeen would stress repeatedly, lie at the intersection of science and engineering. The transistor was an example. When Bell Labs engineers were frustrated in the 1930s by the inadequacies of vacuum tubes, their scientist colleagues suggested that it was “possible, theoretically at least, to control the flow of electrons in a semiconductor,” even though at that time “no one knew how to do it.” For a few years, “semiconductors became one of the most popular fields of physics.” The invention of the transistor then “opened up an intensive period of device development and of basic research on semiconductors.” The example illustrated an important feedback loop in which product development feeds basic research and vice versa.

Bardeen’s many talks and writings about the relationship between science and industrial research express his commitment to the idea that the highest use of science is for the public good. He believed that it made sense to look first at the technological base and then work on developing the corresponding science, rather than “finding something in science and then looking around for applications.” He emphasized the difficulty of transferring a scientific discovery that had not been made in the context of any technological mission “into useful products, particularly in competition with products which already exist.” In the case of the transistor, where the technological need led the science, it was possible to reach social usefulness directly.

Bardeen disagreed with the usual distinction made between basic and applied research. “Basic research is defined by the National Science Foundation as that directed toward fuller knowledge and understanding rather than toward practical application. I prefer not to stress the last phrase of this definition, since I believe that much good basic research is done with applications in mind.” He would remind his audiences that the research program resulting in the transistor “was basic in that it was directed toward understanding the electrical properties of semiconductors, but everyone working on the program was aware of the long-range goal and of its

importance.” He believed fundamentally that “there is really no sharp dividing line between basic and applied research.”

Bardeen also stressed the important role of basic research in the advancement of technology. During one Xerox symposium that he helped to organize, he listed four interrelated reasons why companies should support basic research. (1) Research makes industrial research less a function of cut-and-try by suggesting “areas where significant advances are possible.” (2) It can provide the background for innovations. (3) It can help develop personnel in such a way that specialists will be available to consult on thorny problems. And (4) it ensures that industry is in communication with science. In this way, by supporting fundamental research, corporations could avoid the “scientific dust bowl” that Brian Pippard alluded to in his famous 1961 talk “The Cat and the Cream.”

Above all, Bardeen emphasized the human aspects of science in industry, saying that “the most difficult to find are the people— the required leadership and qualified scientists.” He encouraged industries to offer their best scientists freedoms comparable to those in academia. Bardeen pointed to Bell Labs’ “enlightened research philosophy” as a desirable model. Within the confines of the organization’s mission, he believed, scientists ought to be free to select their own research problems and “to publish their results, to attend scientific meetings, to visit and give lectures at universities, and, perhaps, occasionally have a sabbatical to get refreshed. In other words, they should participate fully in the scientific community.” Bardeen also thought that industrial scientists ought to have the opportunity to see their work come to fruition in marketable products. “Creative work is difficult,” he pointed out, “and motivation is required for sustained activity.”

In 1961 Bardeen was elected to Haloid’s board of directors. This was also the year Haloid changed its name to Xerox and was first listed on the New York Stock Exchange. As a board member he continued to offer advice not only on existing research programs but also on long-term goals, including management policies. That same year he pointed out to Dessauer that “the outstanding problem in computers is a good rapid-access memory,” an area to which “Xerox might contribute in the future.” Moreover, the company’s increasing size meant that its electronics research program “will depend more and more on overall competence rather than patents for protection.”

In the 1970s Bardeen also served on Xerox’s Technical Advisory Panel (TAP), a committee of scientists that met twice a year to review research activities of the different Xerox laboratories and to advise senior management on the overall quality of research efforts. Charles Duke believed that Bardeen’s advice had helped Xerox “build one of the finest industrial laboratories in the world in the fields of organic and disordered solids during the late 1970s.”

When Dessauer retired in 1968 Bardeen pushed Xerox to bring in Jacob (Jack) Goldman as the company’s chief of research. Bardeen and Brattain used to play poker with him at American Physical Society meetings. Bardeen had admired Goldman’s work during his thirteen-year stint as the director of Ford Motor Company’s scientific research laboratory, especially his ability to attract top-notch researchers and keep them productive.

Bardeen supported Goldman when he became the driving force behind the creation of Xerox’s Palo Alto Research Center (PARC). From his position on the board of directors and as chair of TAP, Bardeen vigorously argued both for PARC and for the continued development of the Webster, New York, facility.

PARC had barely opened when the company faced new challenges. Xerox’s key patent on selenium had expired, and the company’s near monopoly on copy machines was threatened by the entry into the copier field of several other companies, including IBM and Eastman Kodak, who boasted new technologies that Xerox had yet to integrate. With the impending expiration of other important patents, it was clear that Xerox, now a huge corporation, would shortly be plagued by antitrust suits. The fact that the company had not reached its profit targets in November and December of 1970 added to the atmosphere of crisis at Xerox in the early 1970s. Another problem came from the acquisition, against Bardeen’s advice, of the computer company Scientific Data Systems (SDS). Xerox lost well over a billion dollars when SDS had to be closed in 1975.

Xerox’s research program came under threat from the “bean-counters,” accountants and MBAs who seemed more interested in improving the bottom line in the short term than in supporting scientists whose research was a long-term investment. Goldman sat fuming at a 1971 board meeting while the rest of the members debated whether to abandon PARC. It seemed a reasonable step, given the company’s small fixed investment in the new center and its questionable profitability.

Bardeen and another scientific consultant, Robert L. Sproull of Cornell, dissented; they argued that PARC’s $1.7 million budget was but a drop in the bucket of Xerox’s overall spending. It would be shortsighted and irresponsible to cut off the company’s best chance of conducting innovative research that might lead to new products and markets. Opinion turned, and PARC was saved. In the subsequent two decades the computer scientists at PARC developed much of the foundation for the coming computer revolution, including the Alto computer, the first graphics-oriented monitor, the first simple handheld “mouse” inputting device, the first home word-processing program, one of the first local area communications networks, the first object-oriented programming language, and the first laser printer. Why Xerox did not then succeed in leading the production and marketing of computers in the 1980s is an important story that is yet to be fully analyzed by historians of science and technology.

The 1973 decision to relocate its office products and information system development facility in Dallas limited the company’s ability to bring digital office technology to market. Texas Instruments, the leader in transistor development and production, was in Dallas, but that did not help Xerox. “Dallas turned out to grow a culture that was completely orthogonal to and independent of the digital world in general and PARC in particular,” said Goldman, who considers the move to have been the greatest error in Xerox’s history. He and most of the scientific and technical staff at Xerox fought the decision, but executives overruled them, citing cost considerations such as labor, taxes, and transportation.

Bardeen was among the last to learn about the decision on the new Dallas facility. He dashed off a handwritten postscript to a letter he was about to send to Xerox:

Bardeen said he hoped that such an important decision might be postponed until it could be discussed at the upcoming board meeting, scheduled in a little over a week. He was so concerned that he drafted a personal and confidential letter later that day to

Peter McColough, chairman of the board of Xerox, detailing his reasons for opposing the scheme. Bardeen said it would “be difficult to attract to that area the key innovative people required for a successful operation.” Moreover, Dallas was “far removed from the major centers of activity which are in the East Coast Boston-Washington complex and on the West Coast.”

It was too late. By the time McColough received Bardeen’s first letter, the decision had been finalized. Bardeen was deeply troubled that Xerox had made a major decision on the basis of short-term cost-effectiveness without consulting its board of directors or its senior scientific and technical consultants.

Bardeen retired from Xerox’s board of directors in 1974, but he continued to consult for the company for another six years. He retired from consulting as well in 1980, after almost thirty years with the company. He wrote, “The problem is not so much the time spent on visits but to try to keep up with the various areas of science so that I can talk intelligently about the problems.” Bardeen maintained lifelong friendships with Dessauer and Goldman. In 1976 he received word that Dessauer and his wife were anonymously donating money to fund scholarships at Clarkson College. They hoped Bardeen would consent to have the scholarships named after him. Bardeen wrote Dessauer, “It is your name that should be commemorated. Your great contributions in opening up and developing from scratch a whole new area of technology are not well enough appreciated.” Dessauer, however, insisted that the awards be named after Bardeen. John responded that he would “be proud to have the Scholarships named after him in recognition of our long years of close association and friendship.”

Bardeen consulted for relatively brief periods for a few other firms. In 1954 he began consulting for General Electric (GE), visiting its labs as often as once a month in the late 1950s and early 1960s. He met there primarily with scientists in areas related to his own interests. When Nick Holonyak left the army in 1959 and went on to work at GE’s research center in Syracuse, New York, he occasionally ran into Bardeen there when both were visiting GE’s Schenectady laboratory. Holonyak noticed his mentor’s name appeared on the same third-floor office door as the director of the semiconductor group. Bardeen resigned from his GE consulting position in 1961 when he joined the Xerox board of directors.

Bardeen also consulted for and sat on the board of Supertex, a company founded in 1976 by one of Bardeen’s former students, Henry C. Pao. Supertex specialized in the design and manufacture of integrated circuits and electronic components. Viewing this consulting as part of mentoring a former student, Bardeen accepted only token compensation for it.

On at least two occasions Bardeen’s advice saved Supertex from jumping into risky ventures. He advised the firm against accepting incentives to set up a plant in Indiana to provide semiconductors for the automotive industry. On a recent visit to a General Motors facility in Indiana, he had found the employees both extravagantly paid and resentful of working conditions created for their own safety, such as wearing protective gear. In contrast, Silicon Valley workers accepted the protective gear without complaint. Bardeen warned Supertex that shortcomings in worker relationships would last for the life of the plant. Pao never regretted turning down the Indiana offer. Similarly, Supertex heeded Bardeen’s advice against accepting what on the surface appeared to be an attractive opportunity to establish a plant in Vancouver. Bardeen thought that Canada’s social democratic government might make operations more expensive.

Also on the board of Supertex was Pao’s father, a successful manufacturer. The older Pao and Bardeen began to follow every board meeting with a game of golf. Pao himself took up golf so that he could play with them, but never managed to beat either of the two. In playing, Pao noticed that while John’s shaky hands made it difficult for him to hold a fork at lunch, on the links the palsy seemed to disappear.

Bardeen never officially consulted for Sony Corporation, but he enjoyed a warm relationship with the company over the years. On October 6, 1989, Sony paid tribute to Bardeen and his scientific contributions to the electronics industry by endowing a $3 million chair at the University of Illinois in his name. Bardeen’s old friend Makoto Kikuchi, the research director of the corporation, told reporters that the transistor was responsible for igniting modern technology. “John Bardeen is a great figure in science, and this is our way of honoring him. Because this (university) has been such a good place for him, we hope (the fellowship) will attract others here.”

After a long deliberation, the search committee named Nick Holonyak the first recipient of the Sony professorship: the John

Bardeen Chair of Electrical and Computer Engineering and Physics. “Dr. Bardeen once told me (in a low voice) that he wanted to have Nick on the Bardeen Chair,” Kikuchi told Jane. “He was such a humble person that he added, ‘but I am not related to the Search Committee so this is just my hope.’”

Bardeen’s last visit to Japan, in May 1990, was made in connection with Sony’s contribution to his chair at the University of Illinois. Sadao Nakajima recalled that on Bardeen’s “last day in Tokyo, he invited a number of Japanese friends, and he asked me and my wife to join him. The atmosphere at this dinner was just like a family gathering. Just very quiet and peaceful talking. I think he loved to meet other people in that way, very informal.”

To foster relationships between local industry and the University of Illinois, Bardeen helped to establish the Midwest Electronics Research Center. The goal was to assist industries in developing their research capabilities and bring them in touch with research at universities. As Bardeen explained to the Illinois Governor’s Office, the center sponsored symposia on various topics and an industrial research visitors program in which scientists encouraged university relations with midwestern firms such as 3M, Honeywell, and Delco.

“When I took John along,” joked Paul Coleman, a colleague of Bardeen’s in the electrical engineering department, on business trips “we’d go in and talk to the president. If I went to the company [without him], we’d talk to the janitor.” Coleman was well aware that other scientific luminaries would be “too busy with their Nobel Prize work than to get in an airplane and fly through a thunderstorm over to Company X, to help Paul and his colleagues to interact with industry.”

A typical visit began with a tour of the facility. Then Bardeen would give a short talk, after which the group discussed the firm’s problems and goals, often with company scientists and engineers present. The visitors and company executives would then have lunch and usually head off to the golf course. On these trips, as elsewhere, Bardeen was “low keyed, a little on the shy side, and spoke very slowly. But on the golf course, my gosh, that guy was competitive… . He would try his hardest to win.” Sometimes the company would engrave a plaque with the date and names of the

players to commemorate a Bardeen visit and its inevitable game of golf. George Russell, who had been a colleague of Bardeen’s at the University of Illinois, kept one of the plaques on his wall even after he became chancellor of the University of Missouri.

Bardeen and his engineering colleagues would advise the companies in planning their research and development strategies. After a visit to 3M in 1985, Chuck Gallo wrote Bardeen a letter of thanks. “The fact that you think the ‘search for high temperature superconductivity is worth pursuing’ carries great weight with all of us at 3M.”

The university visitors could also help the companies hire promising students. And if the company was having problems with its computers or other equipment, Coleman would bring along a technical expert from the university. The university, in turn, gained jobs for its students and some incidental benefits, such as the occasional donation of a perfectly good piece of research equipment, abandoned because the company was updating its technology.

Bardeen never stopped believing that free communication between academic and industrial scientists would result in the most rapid advancement of science and the greatest benefit to society. He carried this conviction not only into his industrial consulting but into the many advisory positions he assumed in government and professional societies.