Thesis: The Computer Science Department and Entrepreneurial Culture

This is one of several parts of my undergraduate thesis at Stanford entitled “Academic Revolution and Regional Innovation: The Case of Computer Science at Stanford 1957-1970”. It was submitted on May 17, 2011, and the text here remains unchanged and unedited since then.

Introduction

Computer science researchers across the nation faced the difficult task of building up programs without the academic legitimacy afforded to traditional disciplines in the university. Computation and its related science at universities in the United States typically developed as applications of other disciplines, particularly physics.1 There was thus an immediate question whether computer science was itself a discipline, or merely a component of other fields.

Establishing academic legitimacy was typically even more difficult because computation and computer science were really composed of two separate groups of activities. On the service side, there was the provision of computing resources to other departments and research centers at the university. At campus computation facilities, programmers would assist professors in writing programs for use in the computer — acting as office staff as opposed to faculty. On the academic side, however, there was the investigation and pursuit of fundamental knowledge of how computers worked as well as the theoretical mathematics that underpinned their operation. There was generally incredible tension between these two sides, and as Atsushi Akera has shown, they “disintegrated” at MIT and the University of Michigan.2

The story at Stanford was radically different. Computer science grew out of the interests of two professors of numerical analysis, George Forsythe and John Herriot, who believed that computation would reshape the field of mathematics. From the beginning, they developed a program with heavy emphasis on theory. For example, the courses when the department was founded in 1965 included only a handful of applications-based courses such as “Computer Simulation of Cognitive Processes” and arguably “Data Reduction and Control Programming.” Other courses focused on introductory programming skills, numerical analysis, formal languages, artificial intelligence, and digital systems.3

Furthermore, the disintegration of the service and academic wings of computer science never occurred at Stanford. Quite the opposite, Stanford’s Computation Center would actively facilitate the development of the Computer Science division, and after 1965, the Computer Science department. There was tension between the two organizations over the center’s financial subsidization of the department, particularly in the later years of the 1960s when revenues became more difficult to secure. However, the Computation Center continued to provide 17% of the the Computer Science department’s budget in 1970, and many personnel held joint appointments between the two organizations.4

Why is the story of computer science so different at Stanford? This chapter answers this question by exploring the institutional factors that led to the rise of the Computer Science department, including its success in expanding the Computer Science faculty as well as building academic legitimacy within the university. It provides a primarily historical institutional lens of the expansion of the department, analyzing the people and organizations that shaped its rise. However, it also evaluates the culture of both the department and the university administration as a critical element that assisted in the department’s success.

This chapter looks at four intersecting institutional factors and patterns that were crucial for the success of the department. First, a warm and productive relationship characterized the connections between the Computation Center and Computer Science throughout most of the 1960s. Forsythe headed both groups in the early years of 1962-1965, building a culture of openness and collaboration between the center and the division that continued even after he left the Computation Center to chair the newly formed Computer Science department.

Within the department, Forsythe was aggressive in offering new courses and expanding student enrollment, a second important factor in the development of the department. When Forsythe arrived in 1957 at Stanford, there was barely a notion of computer science, and certainly no academic courses in the field. Within a little more than a decade, the Computer Science department had enrollments at the undergraduate and graduate level of more than 2,500 students combined per academic year.5

The need to expand student enrollments and the leadership of Forsythe helped to inculcate a culture of entrepreneurship among the Computer Science department’s faculty that is reminiscent of today’s start-up companies in Silicon Valley. The department had strong and confident leadership, an ability to boot-strap new programs and classes on limited budgets, a disregard of bureaucratic constraints (much to the exasperation of senior university leaders), and fundamentally, an expansive vision for the field of computer science. Taken on a holistic level, Forsythe’s papers show a leader who was supremely confident in the importance of his discipline, and took advantage of every opportunity to push the university administration for additional funds. He was simultaneously forceful, energetic, and cleverly spirited in his pursuit of building the field.

Despite his administrative talents, Forsythe was not an intellectual heavyweight when he arrived at Stanford, unlike some of the faculty that opposed his expansion plans as seen in the last chapter. However, despite the stature of the opposition arrayed against him, the university administration continued to support the development of the Computer Science department, matching the growth of students with new funds remarkably well over the course of the 1960s. This chapter will argue that a third factor important to the Stanford case was a culture among university administrators of organizational flexibility and passive facilitation toward computer science, partly due to their own educational and academic backgrounds.

University provost Frederick Terman would play a crucial role in providing some of the initial funding of the department as well as in providing advice from his own experience in developing the Electrical Engineering department. The various deans of the School of Humanities and Sciences, who had direct budget authority for Computer Science, were generally supportive if not always leading champions of the department. This passive facilitation is apparent in numerous examples in which Forsythe would add classes or make commitments to spending without prior approval from the school. The school would find the funds to meet the obligation, admonishing Forsythe to avoid the problems in the future without truly seeking to change his ways. Thus, the school accommodated Forsythe’s ambitious expansion strategy, while never actively endorsing the goal itself.

The fourth and final factor that assisted the department’s rise is rather counterintuitive. The focus of the faculty on theory and their desire for academic legitimacy encouraged a strong focus on building up the department’s doctoral program at the expense of undergraduates and industry students. In fact, it would be decades before the department would offer an undergraduate major. However, the department ended the 1960s with thousands of dollars of annual revenue from industry developed through joint university-industry programs, and it would expand its course offerings at the undergraduate level, teaching a record number of undergraduates in one year. Why did the department change?

A combination of negative financial trends, including a growing deficit at the Computation Center, nationwide inflation, cuts to defense research agencies due to the Vietnam War, and declines in endowment funds, forced the department to seek new sources of revenue to sustain its operations and expansion plans. Dissenting from the typical interpretation of university financing, this chapter argues that it was actually the lack of funds, both for its budget and for a new academic building, that forced the Computer Science department to engage with industry and create the regional innovation networks that would create Stanford’s “steeple of excellence” in computer science.

The Computation Center

The Computation Center was the primary means of receiving computing at Stanford University. As such, it played a crucial role for researchers and students who were conducting computer research. However, its role also expanded to subsidize the Computer Science department, both directly and indirectly. The close relationship between the two organizations — the service wing with the center, and the academic wing with the department — facilitated this subsidization despite increasing concerns from the university administration about the benefits of the arrangement. Given the department’s tight budget within H&S, the center provided a crucial financial lifeline, while also serving as a venue for engaging other academic departments.

The development of the Computation Center and the Computer Science division in the early 1960s was facilitated by Forsythe, who served as the head of each. He effectively, if unofficially, merged the accounts of the two organization, financing people from whatever budget currently had funds. This particularly annoyed the meticulous Terman, who complained in late 1963 that “funds for running the Computation Center, and budget accounts of the Division of Computer Science should not be regarded as interchangeable.”6 The center’s subsidization included not just funds for faculty (many of whom were jointly appointed between the division and the center), but for graduate students as well. In 1964, 10 graduate students received their stipends from the center’s general funds.7

Financing the Computation Center in these years became increasingly difficult because of a lack of revenues. Due to the novelty of the technology, funding agencies and the university were not entirely accounting for the costs of computing time. In the 1962-1963 academic year, the center received revenues for slightly more than half of the utilization of the center. Unsponsored research, essentially time donated to faculty and students, accounted for 26% of the total computation time in the center, with the remaining 22% of the time being used for class assignments.8 The latter two were effectively subsidies to the Computer Science division’s research and education program. By 1967, the direct subsidy for the department represented approximately 21% of the department’s budget for faculty — almost equal in size to external research grants.9 The budget pressure created by these subsidies started to affect the number of graduate students the center funded, with some of the slots eliminated in order for the center to purchase new equipment.10 Furthermore, there was increasing faculty dissent about the queueing system at the center, which gave no priority to funded versus unfunded computing tasks. That dissent encouraged the creation of a Committee on Computing that would develop policies for creating a prioritization system for running jobs at the center.11

Much as with the Computer Science department, the center’s unstable revenues encouraged a “start-up mentality” to secure new sources of funding. Forsythe would later remove himself from directing the Computation Center with the formation of the department in 1965. The center continued its rapid expansion, eventually reaching $1,500,000 in operating expenses in 1967, with plans to increase to $2,500,000 by 1970. The same report cautioned that “expenses are extremely high, and income is never certain,” and thus the center is continually “on the edge of bankruptcy.” If the center suddenly experienced a drop in revenues, the fear among the faculty was that it “could force drastic retrenchments in the entire university.”12 In response, the center began reaching out to other parts of the university and to external organizations that might find computing useful. Some examples include local law enforcement (for crime statistics), Stanford’s law school (for joint degrees on privacy and statistical studies of the legal practice) and special training programs for minorities.13 Even though the funding was precarious, the Computation Center continued to subsidize the Computer Science department to the tune of $107,059 in 1967, only slightly behind H&S in funding, which stood in the 1967-68 academic year at $133,000.14 Thus, the fate of the two continued to be intertwined throughout the decade.

While subsidization of the department continued, the worries about revenues were justified just a few years later when nearly all academic units at Stanford were hit with budget cuts. Various funds began running out, including a grant from the National Science Foundation and the Provost’s Computing Funds (which until then had attempted to cover some of the unsubsidized research expenses at the center). By 1970, the center was facing a deficit of a half million dollars, and in 1972, the deficit was likely to double to one million dollars.15 The university responded by raising user fees and cutting some unsponsored research work, but staff members felt that the center’s administration were “contemplating their navels” in solving the budget crisis.16 These financial pressures forced the center to aggressively reduce its subsidy for the Computer Science department. In 1970, this subsidy was reduced to $28,000, out of a budget of $165,000 (or roughly 17% of the budget).17 Forsythe described the loss of personnel support from the center in 1970 as “a natural consequence of the Computation Center’s tight finances and consequent need to cease furnishing unreimbursed services to our Department.”18

The Computation Center actively worked to subsidize the programs of the Computer Science department, both directly and indirectly. Forsythe and the later directors of the center used it to build up computing at Stanford, with little regard for the finances of how the whole enterprise might work out. For its part, the university administration took a relatively hands-off approach to the subsidies, despite the increasing risk of a budget deficit at the center. These subsidies made the department’s budget easier to handle for H&S, and thus there was an incentive to ignore the subsidy despite its cost to the users of the center.

Building a Budget and an Entrepreneurial Culture

An introduction to university budgeting is critical to understanding the department’s desire to expand course enrollments. The main revenue streams for Stanford in the 1960s were tuition, research grants, endowment income, and auxiliary revenue (including industry grants and other sources of income). Tuition was the most flexible source of funding, and as such it was often the first source for new initiatives at the university. Research grants were normally given by the federal government and foundations for specific proposals, and are rarely convertible to other university needs.19 However, most of them do include some notion of “overhead” or discretionary revenue that can provide the university some budgetary flexibility. Endowments were provided by donors and were usually constrained in scope when given, making them flexible within a given domain but not flexible across the university. Finally, auxiliary income often stayed with the source of revenue (for example, revenue from clinical treatments generally stays with the medical school).

Thus, the primary issue facing the Computer Science department was finding fungible funds. Competition for these funds is typically fierce between all academic departments, and computer science’s lack of stature should have proven difficult to overcome. However, academic legitimacy is not the primary criterion for receiving funds from the university. Instead, course enrollments often drive the year-to-year changes in budgeting. Increasing course enrollments in a department will generally encourage the university administration to invest new resources in that area. It is this relationship that Computer Science hoped to exploit as part of a multi-pronged strategy to secure more funding.

One method used to receive more funds was simply to spend more. Forsythe disdained the budget restrictions placed on the department, and regularly misspent the money and spent more than allowed as he believed best fit the goals of the discipline. In the early years when Forsythe was head of the Computation Center and head of the division of Computer Science, he appeared to spend from both accounts freely, ignoring the university’s bureaucracy and budget processes.20 Another example came two years later when the department forgot to request funds for the salary of its department secretary in its budget request, asking the university administration for additional funding to cover her. Royden would move money from one of the department’s industry funds, writing to Terman that the move “may cause the Department to be a little more careful about budgeting in the future, but this is probably wishful thinking on my part.”21

As the Computer Science division developed, so did the demand from students for classes. Much to the exasperation of senior Stanford administrators, Forsythe and other Computer Science faculty continually allowed greater numbers of students to take their classes and often offered more sections than were budgeted. Halsey Royden, an associate dean of H&S, wrote to Terman that the school had not given approval for extra sections for one course, but had “slipped up” in watching the number of sections that the division had created.22

Forsythe’s enthusiasm to have Computer Science teach more students assisted the department’s claim for more money. The 1960s were a time of expansion at Stanford, and the School of Humanities and Sciences was no exception. However, the breakdown of this increased funding was constantly debated. These teaching funds were essentially zero-sum: gains made in the Computer Science division would have to come at the expense of the growth of other departments. Developing high enrollment numbers was thus a useful strategy for the division, giving Forsythe a cudgel to use in budget negotiations against other departments. The large enrollments and small size of the faculty also meant that class sizes were necessarily quite large, and this helped in the yearly budget negotiations.

Using this marginal analysis, Forsythe continually believed that computer science was underfunded by the university, and demanded more funding in the years following the 1965 creation of the department. This notion of finances fit into the philosophy developed by provost Frederick Terman, who told Forsythe that “Dollars will be scarce in coming years. It will be essential to accomplish a lot with a little money.”23 Taking the lead, Forsythe argued in 1970 that the department was underfunded by $65,000 to $200,000 a year in teaching funds. He compared Stanford’s budget numbers with those for Computer Science at Cornell, which received roughly double the amount of funds while teaching fewer students.24 The issue would continue into the 1970s, with a 1975 report noting with deep concern that the Computer Science faculty had a sense of “inadequate university support,” and it also observed that none of the new 75 chairs in the school were assigned to the department25

Throughout the 1960s, there was a growing fear that other departments at Stanford might try to teach classes in computation, particularly given the large enrollments students experienced in computer science courses. Despite the inherent interdisciplinary nature of the field, the department was mostly successful in maintaining centralized control of the teaching of computer science within the university. One of the major concerns with this development was that students would lack a coherent understanding of the new field. In response to a question of allowing other departments to teach introductory courses, Forsythe wrote that “If we left the introductory course to the user departments, I think the students would fail to see computing as a changing subject, much in need of development by the students themselves.”26 Forsythe saw the students as potentially the next generation of computer scientists, and he viewed introductory classes as the means to proselytize the new discipline.

While educational concerns were certainly one element of the fear, financial concerns were at the heart of the Computer Science faculty’s concerns. In a report generated in 1967 on the future development of the department, the department’s faculty noted that they felt threatened by the rise of computer-based classes in other academic fields in mostly financial terms. “Given relatively unlimited funding, we would have no objection to parallel and even competing projects in our field,” but this was obviously not the case, and they complained to the university administration that “it seems quite unjust to see funds and space made available for computer science in other parts of the university,” when its academic home was so poorly funded. They also made it clear that only the Computer Science department should have the sole authority to appoint computer scientists at Stanford.27 These views were later summarized by John McCarthy, the artificial intelligence researcher, who wanted to ensure that the department would “monopolize” computation at Stanford — an ironic choice of words given the later nature of the computer industry.28 The lack of funds for the department encouraged more centralization of the computing curriculum than might be expected for such an interdisciplinary field.

Underlying these trends of centralization was a belief that computation at Stanford would be intrinsic to the university in just a few years, and that the university was uniquely positioned to take leadership of the field. Forsythe believed that computer science would become one of the most popular fields in the university, and he envisioned a department that might even reach 100 faculty members to handle the growth in demand.29 For the faculty at the time, there was already recognition of the differences between Stanford and its peer departments. Forsythe believed that Stanford had succeeded well with limited funding, but acknowledged the immediate advantage of better-funded peer departments. “Contrast Michigan, with a big infusion of Ford money in the engineering computing line, MIT, with project MAC costing ARPA millions of dollars, or Carnegie Tech, with Mellon money, or Cornell, with recent Sloan money. Hence Stanford is a very fruitful place to plant some big bucks.”30 Despite such funding, Terman believed that other universities handled their finances in computer science poorly, and thus “a reasonable amount of work will pay off very well.”31 This optimism played a crucial role in the approach that the Computer Science faculty took in expanding the department.

CS-taught Units for the 1960s

However, there were disadvantages to the the growth of student enrollments and the department’s teaching budget. Figure 4.1 shows the growth in the number of units taught over the course of the decade.32 It is important to note that the 200-level courses, which were graduate level but accessible to undergraduates, had greater than linear growth throughout this period. 100-level classes, which are most typically designed for undergraduates, grew quickly before the creation of the department in 1965, but remained steady afterward. The late growth of the 000-level classes was due to the creation of introductory classes in the department, which greatly expanded the number of students who could study Computer Science. All of these courses were not created equal: some cost as little as $5 per unit and others as much $170 per unit, depending on the extent of programming assigned in the course.33 This full coverage of computer science may have forced the teaching budget higher, but it placed greater demands on the faculty than for professors at peer departments, and Forsythe observed that this “really hurts” the competitiveness of the department.34

University budgeting is generally characterized by its slow year-to-year change. Throughout the 1960s, Computer Science at Stanford faced the problem of securing money in its teaching budget commensurate with the growing numbers of students in the field it served. Figure 4.2 shows the growth of the CS budget over the 1960s, and Figure 4.3 shows the growth in the number of students enrolled in at least one class over the same time period.35 The number of students, both undergraduate and graduate, increased at a rapid pace throughout the decade — almost doubling every two years.36 However, the university managed to expand Computer Science’s teaching budget roughly commensurately with the growth in student numbers: both student enrollments and the budget grew by about five times from 1961 to 1970.

CS Budget for the 1960s
CS Enrollments for the 1960s

In summarizing the Computer Science department’s growth strategy, the old adage of “if you build it, they will come” would seem to apply. Its faculty aggressively expanded the number of courses offered, and students swarmed the department’s offerings. The increasing visibility of the department forced the university administration to budget more teaching funds to computer science. In a way, the department was setting the financial agenda of the school, and it is this sort of passive facilitation of the growth of the department that proved critical for its success.

Support of the Administration

The Computer Science department benefited from university administrators who varied in their support and acquiescence in handling the financial growth of the department. Despite the vigorous protests of some faculty in the natural sciences, as explored in the last chapter, there was remarkably positive support among the administration for the growth of the field throughout the 1960s, a strong benefit for a nascent discipline with little academic stature. University provost Frederick Terman proved instrumental in ensuring the health of the Computer Science division’s budget before 1965, and later helped the Computer Science faculty develop the department’s approach to industry. Toward the end of the decade, during the tough financial years of the Vietnam era, Computer Science would become the sole department in the university to continue expanding amid budget cuts.

The backgrounds of Stanford’s administrators proved quite beneficial for the department. Terman earned a ScD in Electrical Engineering from MIT under the direction of Vannevar Bush, who became the “father” of the National Science Foundation and led the policy development of the government’s approach to financing science.37 The two kept in contact throughout the years, and Terman used his access to Bush to receive advice on securing grants for Stanford. Bush was an early and strong supporter of computing, and it is quite probable that Bush’s support encouraged that of Terman as well.38

Another administrator was Halsey Royden, who served as an associate dean, acting dean and later dean of H&S. He was a professor and former chair of mathematics, but did not join his colleagues in their dissent to the Computer Science department’s growth. Royden’s interests were in complex analysis, particularly Riemann surfaces, but he also held a strong interest in undergraduate education, leading the Mathematics department’s overhaul of its major in the 1950s. Perhaps most importantly for this study, Royden started at Stanford as a member of the Applied Mathematics and Statistical Laboratory, a research center with important connections to computation.39

That background was similar to the dean of the Graduate Division, Albert H. Bowker, a professor of Mathematics and Statistics and founding chair of the Statistics department throughout the 1950s. He worked in the same laboratory as Royden, and Bowker was interested in applying mathematical statistics to issues related to engineering. He would go on to a substantial administrative career after Stanford, becoming chancellor of the City University of New York and University of California – Berkeley.40

The dean of H&S in the early years, Robert Sears, was a notable child psychologist who expanded the research on IQ tests —an area pioneered by Lewis Terman, father of Frederick Terman. Sears continued Lewis Terman’s long-range study of children with high IQs, publishing several significant studies. While his research did not connect to computation directly, his personal connection to Frederick Terman likely influenced his deference to handling computer science.41

It was Terman, undoubtedly, who proved the most influential and supportive of Computer Science among these senior administrators, especially in the years before 1965. He quickly recognized the value of the new field, and Forsythe felt that the provost understood the situation faced by Computer Science as early as 1963.42 Later that year, Terman began supplying his own discretionary funds to the division, because he believed that the division was underfunded.43 The budget transfer eventually reached $5,000 a year, enough to appoint an additional joint faculty member to the division.44 Terman wanted the division to quickly grow and reach department status, and he thought that early funding would greatly assist the effort. Terman’s timeline was ambitious. Concurrently with the formation of the division, Terman remarked that department status was likely only a year away provided everything continued to go well.45

Terman’s desire to move quickly on computer science was likely colored by his own experience expanding the Electrical Engineering department at Stanford. There, he greatly increased the academic calibre of the department by securing greater government funding and industrial revenues.46 Terman called the idea of building strong academic departments “steeples of excellence.” In the case of electrical engineering in the postwar years, Terman had to build up an academic program against strong competitors who were already well-established, such as MIT. In the case of computer science though, there were no established departments, and Terman likely saw an opportunity for Stanford to easily build a steeple of excellence in the discipline if it was willing to move resources. His approach in the university was later adopted by the Computer Science faculty in positioning their own department — they argued that the university should unevenly supply additional funds to “selected areas” that the university felt worthy of expansion.47 Later, Terman advised the senior leadership of the department on how to create a strong program, urging them to build “steeples of excellence that are in the mainstream of the future development of the subject. You don’t need many steeples, but they should be high.”48

In addition to Terman, the Computer Science division benefitted from the strong encouragement and passive facilitation of its most immediate administrators. The associate dean Halsey Royden supported the division in its relationship with other parts of the university, and he was generally enthusiastic about expanding the division’s budget. During the conflict with the Mathematics department over the appointment of Marvin Minsky (see last chapter), Royden lobbied for the appointment behind the scenes with Terman, who appreciated that Royden was “[...] sticking your neck out on this one with your mathematics colleagues [...].”49 In late 1963, Royden also supported providing the new division with more discretionary funds, and he lobbied Terman to provide them from provost’s funds, arguing that the division needed more “hard money” to be able to hire additional faculty members.50 Terman did give the funds just a few months later.

In the years after 1965, Computer Science benefited from significant donations from corporate sponsors that removed much of the pressure on Stanford administrators to find funding from university sources. However, even the new funding did not smooth the difficulty of the funding picture in the late 1960s. A combination of factors, including high inflation, decreasing endowment returns, and a murky government funding picture due to protests over the war in Vietnam, brought tremendous pressure on universities across the United States, and Stanford was no exception. The president of the university wanted to make base budget cuts in the 1970-1971 academic year, and instituted a hiring freeze throughout the university.51 However, the constant prodding of Forsythe and the relatively enthusiastic support of university administrators allowed the department to thrive even in this inauspicious fiscal climate. By 1974, the H&S School recognized that the Computer Science department was underfunded, and Computer Science became the only department that year that did not have to plan for an actual budget reduction.52

Together, we see that the cultures of the department and the administration complemented each other surprisingly well. Despite the displeasure of administrators, Stanford did have the funding necessary to finance computer science’s growth, and Forsythe never stopped demanding, and continuing to secure, a greater share of resources. Stricter university governance over budgeting and finance would have almost certainly retarded the growth of the department. Likewise, a less bold and more timid leader of the Computer Science department would not have sought out the opportunities and gambled so many times to push the discipline forward.

Developing New Venues

A perennial problem facing the Computer Science division and later the Computer Science department was a lack of physical space. Budgeting in university environments may be slow, but physical space growth can be nearly glacial. It would take more than a decade before the Computer Science department would receive a dedicated building, and before this construction finished, the faculty and staff were spread across more than a dozen buildings on the peninsula south of San Francisco. It was clear that the department needed a permanent home, ideally in a single building.

The need to fund the construction of a computer science building prompted a search for potential sources of funding. One group targeted was alumni, and the department engaged them in several ways. More importantly, the department began to develop industry partners, creating connections and networks to industry that would prove crucial to its success over the years. The department would form the Computer Science Advisory Committee to connect senior executives into the department, and this committee also helped to shape the direction of the department in the years to come.

The department’s growth in faculty and students put incredible pressure on the space available to the department. The department’s main office space was in Polya Hall, and the division began with 2,610 sq. ft of space in 1961. Over the course of the decade, the office space expanded briefly to as much as 4,200 sq. ft., but would end the decade with only 3,380 sq. ft. in the building. The expansion of the Computation Center eventually cut into the space available to the department. Thus, on-campus space grew about 30% over the course of the decade (compared to a fivefold increase in the size of the budget and students in the department).53

The problem is that the Computer Science department had no space of its own on campus, and merely leased space (for free) from the Computation Center’s home in Polya Hall. This required faculty to share offices and prevented the department from providing space to graduate students. For instance, in 1967 the Computer Science faculty noted that there were a total of two to four researchers in the department who had worked out of the chair’s office over the past few months, and that they were “desperately short of space.”54 In response to the space problem, the university added 12,500 sq. ft. in additional space in the off-campus Powers building, but that space was 15 minutes away by driving, making it inconvenient for academic use.55

Forsythe began a lobbying campaign in earnest for the building, working with senior administrators to plan its financing and construction. In 1966, he discussed the issue of raising needed funds for the building, which were estimated at the time at $600,000. Government sources were not likely to support financing of the building, and so Forsythe was encouraged to create a visiting committee of people with interests in computation that would assist the department in setting its direction.56 Forsythe continued to argue for the building through both official channels and other, more creative means. For instance, when one professor asks for a visiting professorship, Forsythe responded that “We have the degree of tolerance you need, but you can’t imagine how bad the space situation is here,” and told him that ”I can almost promise you a desk.” Forsythe blind carbon copied the letter to the president, provost and H&S dean.57 Later, Terman would provide assistance in laying out an approach to raising funds for the building, whose costs were quickly increasing. He told Forsythe to break up the building construction into smaller phases, and he suggested that corporations would be willing to donate as little as $25,000 to as much as $800,000 toward a building. He also believed that raising the funds would be tiring and would require a lot of energy for success58

One approach to getting financial support was reaching out to alumni. Forsythe encouraged the Computer Science faculty to increase their participation in alumni conferences, writing that showing up to the morning receptions “might even bring us money for a building.”59 Later, Forsythe would join the Committee on Education for Alumni, where the committee developed an idea for a “Portable Stanford” volume that would include articles written by faculty, such as on cybernetics. The committee also created educational programs for alumni such as speakers bureaus in major cities.60 Interest from alumni played a role as the building construction got underway in the mid-1970s. The Alumni Association asked the department to teach several short courses on computers and society to alumni. For the department, the courses provided the opportunity to showcase the work of Computer Science to potential donors.61

In addition to alumni, the department created the Computer Science Advisory Committee in 1967 to assist it in setting a direction while engaging potential donors in the work of computer science. In developing the committee, Forsythe was told by the administration to place “strong computer scientists” on the committee. However, most of the people desired were alumni, trustees, “people with big money,” and “managers of large corporations like banks.” Even so, Forsythe was told by university administrators that the committee should be focused on the department’s mission, and that it should not be used directly for fundraising.62 Developing the membership definitely focused on the money aspect: one example included Ross Perot, who was both interested in computing and had “plenty of money.”63

Forsythe and the university were relatively successful in their approach, and in 1970, the 15-member committee was composed of six academics, and several wealthy individuals, including the president of Varian Associates, the executive vice president of Bank of America, and the chairman of the Fireman’s Fund Insurance Company.64 Invitations to businessmen were often rejected due to potential conflicts of interest, but others like Fred Merrill of The Fund American Companies believed that the assignment was interesting and “can be of considerable value to our companies.”65

While the committee may have been created to bring in donors, it also shaped the department’s policies and educational program. The committee’s primary mission was to advise the university president on all matters related to computer science and computation at Stanford.66 The committee’s activities varied widely, from looking at the undergraduate and graduate programs to analyzing the finances of the Computer Science department. For example, the 1970 meeting included discussions of the time to PhD, the selection of introductory courses, the development of a professional master’s program and admissions policies.67 Later that year, Forsythe wrote in his report on “where you have helped us,” that “your advice has encouraged us to move in certain directions.” Those directions included creating a Computer Engineering master’s degree, creating the Computer Forum, restricting the admissions of foreign students, and delaying the creation of an undergraduate major. Beyond this influence though, Forsythe also thanks the committee’s assistance in terms of “actual gifts of money.”68 Thus, the committee had fulfilled both its official and unofficial missions: to direct the department and to serve as a source of revenue for major projects.

Conclusion

The Computer Science department would spend nearly the entire decade pushing for more resources from the university. While the funds were never enough, the department managed to put together one of the strongest programs of computer science in the nation. Its success largely flowed from four important institutional factors: a strong relationship between the Computation Center and the Computer Science department, an entrepreneurial Computer Science faculty led by a bold leader, a flexible and generally supportive university administration, and a need to seek out additional funding to pay for a new academic building and an expanded budget.

The desire for additional funding led to a pragmatic Computer Science faculty that created new programs to engage industry and admitted additional high-paying industry students into the department’s educational programs. Along the way, such programs built the first university-industry networks between professionals in computation in the area known as Silicon Valley, networks that would build the world’s most notable regional innovation hub. It is this part we turn to next.

Continue to Chapter 4, "The University-Industry Nexus" or return to Chapter 2, "Academic Politics and Legitimacy"


  1. For a survey at the beginning of the 1960s, see Louis Fein, “The Role of the University in Computers, Data Processing, and Related Fields,” Communications of the ACM, Vol. 2, No. 9 (Sept. 1959), pg. 7-14.↩︎
  2. Atsushi Akera, Calculating a Natural World, MIT Press, 2008.↩︎
  3. Courses may have had applications of theories, but the main intention of the vast majority of courses in the department was clearly the teaching of theory. Forsythe to File, “Meeting 5 January 1965,’ 19 Jan. 1965, Forsythe Papers, SC98/15/1.↩︎
  4. Forsythe to Sears file, “Conference of 12 Nov. 1969,” 13 Nov. 1969, Forsythe Papers, SC98/14/7.↩︎
  5. Forsythe to Computer Science Advisory Committee, “Report on Computer Science Department,” 19 Oct. 1970, Forsythe Papers, SC98/1/19↩︎
  6. Terman to Royden, “Computer Sciences,” 8 Nov. 1963, Terman Papers, SC160/3/12/“CS and Center: 62-63.”↩︎
  7. Forsythe to Virgil Whitaker, “Graduate Aid Available and Desirable,” 16 July, 1964, Sterling Papers, SC216/1/1/26.↩︎
  8. “Computation Center,” Sterling Papers, SC216/1/1.5/27.↩︎
  9. ”The Future of Computer Science at Stanford,” 8 Feb. 1967, Lederberg Papers, SC186/10/4↩︎
  10. Forsythe to Virgil Whitaker, “Graduate Aid Available and Desirable,” 16 July, 1964, Sterling Papers, SC216/1/1/26↩︎
  11. Forsythe to Bowker, “Request for more parting advice!,” 6 Sept. 1963, Forsythe Papers, SC98/2/19.↩︎
  12. ”The Future of Computer Science at Stanford,” 8 Feb. 1967, Lederberg Papers, SC186/10/4.↩︎
  13. John Ehrman to Susan Kolasa, “Ideas for Funded Programs in SCC,” 23 Sept. 1971, Forsythe Papers, SC98/2/38a.↩︎
  14. Feigenbaum to Ron Jamtgaard, “Approximate Annual Subsidy of CSD by SCC,” 17 Nov. 1967, Miller Papers, SC208/3/1.↩︎
  15. E. Howard Brooks to Budget Files, “Budget Conference with Associate Provost W.F. Miller,” 6 Jan. 1969, Lyman Papers, SC215/1/“Comp. Center 1964-1969.”↩︎
  16. Paul Armer, “Brief Overview of the Stanford Computation Center,” 12 Mar. 1970, Lederberg Papers, SC186/4/1.↩︎
  17. Forsythe to Sears file, “Conference of 12 Nov. 1969,” 13 Nov. 1969, Forsythe Papers, SC98/14/7.↩︎
  18. Forsythe to Computer Science Advisory Committee, “Report on Computer Science Department,” 19 Oct. 1970, Forsythe Papers, SC98/1/19.↩︎
  19. At the time, defense research grants were relatively more undirected than similar grants today.↩︎
  20. The issue was part of the concern seen in the last chapter over separating the Mathematics and Computer Science budgets. Terman to Royden, “Computer Sciences,” 8 Nov. 1963, Terman Papers, SC160/3/12/“CS and Center: 62-63.”↩︎
  21. Royden to Terman, “Missing salary for Diana Saunders,” 21 June 1965, Terman Papers, SC160/3/12/2.↩︎
  22. Royden to Terman, “Facts about Computer Science,” 17 Oct. 1963, Terman Papers, SC160/3/12/1.↩︎
  23. Forsythe to Terman file, “Fred Terman on computing plans,” 29 Apr. 1969, Forsythe Papers, SC98/14/9.↩︎
  24. Forsythe to Raymond F. Bacchetti, “Request for study of instructional costs and class sizes,” 27 May 1970, Lyman papers, SC215/1/“CS68-71.”↩︎
  25. “Report of 1975 President’s Advisory Committee on Computer Science,” Feigenbaum Papers, SC340/13/23.↩︎
  26. Letter from Forsythe to Daniel Bailey, 19 Nov. 1965, Forsythe Papers, SC98/3/7.↩︎
  27. ”The Future of Computer Science at Stanford,” 8 Feb. 1967, Lederberg Papers, SC186/10/4.↩︎
  28. Forsythe to Senior Faculty File, “Meeting of 18 March 1968,” 18 Mar. 1968, Forsythe Papers, SC98/14/15.↩︎
  29. Forsythe’s number would place Computer Science as almost a quarter of the faculty of the school. Forysthe to Reddy, “Computer science in the future,” 1 May 1968, Forsythe Papers, SC98/14/13.↩︎
  30. Forsythe to File, 4 Mar. 1966, Forsythe Papers, SC98/14/20.↩︎
  31. Forsythe to Terman file, “Fred Terman on computing plans,” 29 Apr. 1969, Forsythe Papers, SC98/14/9.↩︎
  32. Date comes from Academic Planning Office data. “APO Data for Years 1959-1970,” 19 Oct. 1970, Forsythe Papers, SC98/1/19.↩︎
  33. Forsythe to Sears and Moses File, “Conference 14 Apr. 1966,” 15 Apr. 1966, H&S Papers, SC36/8/“CS 65-66.”↩︎
  34. Forsythe to File, 4 Mar. 1966, Forsythe Papers, SC98/14/20.↩︎
  35. A contract was an external funding grant provided to the department. The data for both charts comes from Forsythe to Computer Science Advisory Committee, “Report on Computer Science Department,” 19 Oct. 1970, Forsythe Papers, SC98/1/19.↩︎
  36. Forsythe to Sears and Moses File, “Conference 14 Apr. 1966,” 15 Apr. 1966, H&S Papers, SC36/8/“CS 65-66.”↩︎
  37. C. Stewart. Gillmor, Fred Terman at Stanford, Stanford University Press, 2004.↩︎
  38. G. Pascal. Zachary, Endless Frontier: Vannevar Bush, engineer of the American Century, The Free Press, 1997.↩︎
  39. Brad Osgood, Ralph Cohen and Albert Hastorf, “Memorial Resolution: Halsey Royden.” 1993, http://histsoc.stanford.edu/pdfmem/RoydenH.pdf↩︎
  40. “A.H. Bowker’s Home Page,” http://stat-www.berkeley.edu/users/bowker/↩︎
  41. Alfonso A. Narvaez, “Dr. Robert R. Sears, 80, Is Dead; Child Pyschologist and Educator.” The New York Times, 26 May 1989, http://www.nytimes.com/1989/05/26/obituaries/dr-robert-r-sears-80-is-dead-child-pyschologist-and-educator.html↩︎
  42. Forsythe to Files, “Final Conversation with Bowker,” 25 Sept. 1963, Forsythe Papers, SC98/2/17.↩︎
  43. Forsythe to Files, “Conference with Bowker,” 3 Dec. 1963, Forsythe Papers, SC98/2/17.↩︎
  44. Terman to Royden, “Computer Sciences,” 8 Nov. 1963, Terman Papers, SC160/3/12/“CS and Center: 62-63,”↩︎
  45. Terman to Sears, “Mathematics and Computer Science Division,” 6 Dec. 1963, Terman Papers, SC160/3/12/1.↩︎
  46. C. Stewart. Gillmor, Fred Terman at Stanford, Stanford University Press, 2004.↩︎
  47. ”The Future of Computer Science at Stanford,” 8 Feb. 1967, Lederberg Papers, SC186/10/4.↩︎
  48. Forsythe to Terman file, “Fred Terman on computing plans,” 29 Apr. 1969, Forsythe Papers, SC98/14/9.↩︎
  49. Terman to Royden, “Marvin Minsky,” 3 Dec. 1963, H&S Files, SC36/8/“CS1964-65.”↩︎
  50. Royden to Terman, “Computer Science Division,” 7 Oct. 1963, Terman Papers, SC160/3/12/1.↩︎
  51. Forsythe to Sears file, “Conference of 12 Nov. 1969,” 13 Nov. 1969, Forsythe Papers, SC98/14/7.↩︎
  52. Floyd to CSD Faculty, 18 Dec. 1974, Forsythe Papers, SC186/5/2.↩︎
  53. Forsythe to Computer Science Advisory Committee, “Report on Computer Science Department,” 19 Oct. 1970, Forsythe Papers, SC98/1/19.↩︎
  54. ”The Future of Computer Science at Stanford,” 8 Feb. 1967, Lederberg Papers, SC186/10/4.↩︎
  55. Forsythe to Computer Science Advisory Committee, “Report on Computer Science Department,” 19 Oct. 1970, Forsythe Papers, SC98/1/19.↩︎
  56. Forsythe to Money-raising file, “Meeting of 18 May 1966.” 18 May 1966, H&S Files, SC36/8/“CS: 65-66.”↩︎
  57. Forsythe to Formen S. Acton, 13 Sept. 1967, Sterling Papers, SC216/C1/14.↩︎
  58. Forsythe to Building Financing File, “Discussion with Terman, 7 Nov. 1969,” 10 Nov. 1969, Miller Papers, SC208/3/1.↩︎
  59. Forsythe to File, “Alumni Conference, May 1968,” 28 Dec. 1967, Forsythe Papers, SC98/14/15.↩︎
  60. Thomas Newell to Committee Members, “First Meeting of the Committee on Education for Alumni,” 22 Apr. 1971, Forsythe Papers, SC98/2/2.↩︎
  61. “CS Department Faculty Meeting, Apr 13, 1976,” 13 Apr. 1967, Lederberg Papers, SC186/5/2.↩︎
  62. Forsythe to Money-raising file, “Meeting of 18 May 1966,” 18 May 1966, H&S Files, SC36/8/“CS: 65-66.”↩︎
  63. Forsythe to Building-finance file, “Meeting: Miller, Forsythe, Ruetz - 7 Nov. 1969,” 11 Nov. 1969, Miller Papers, SC208/3/1.↩︎
  64. Forsythe and Miller to the University Computer Facilities Committee members, “Invitation to Computer Science Advisory Committee meeting,” 12 Jan. 1970, Lederberg Papers, SC186/4/1.↩︎
  65. Letters of rejection from chair of Texas Instruments and President of IBM, along with Merrill are in Sterling Papers, SC216/C1/15.↩︎
  66. Robert Lamar, Press Report. Stanford University News Service, 4 Oct. 1967, Forsythe Papers, SC98/1/14.↩︎
  67. Forsythe to CSAC File, “Impressions of advice to the Computer Science Department gained from meeting of 18-20 January 1970,” 30 Jan. 1970, Forsythe Papers, SC98/1/18.↩︎
  68. Forsythe to Computer Science Advisory Committee, “Report on Computer Science Department,” 19 Oct. 1970, Forsythe Papers, SC98/1/19.↩︎

Continue to Chapter 4, "The University-Industry Nexus" or return to Chapter 2, "Academic Politics and Legitimacy"