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.
Stanford’s Computer Science department continued to grow in size and influence throughout the following decades. With the rise of the software and internet industries in the 1980s and 1990s in Silicon Valley, the department remains one of the most important stories in the history of computer science and research into university-industry partnerships. What were the major factors behind Stanford’s success? Speaking just a few years after the founding of the department, William F. Miller wrote about the university’s success in the field: “Being one of the early departments in getting our program announced, we got the cream of the crop for a few years and success continues to make more success.”1 Miller is certainly correct that speed was an important element in Stanford’s success, but many other factors played an influential role in Stanford’s ability to be competitive.
This study analyzed the development of the Computer Science department as an academic discipline within the milieu of Stanford’s School of Humanities and Sciences. It first examined the politics of creating a new discipline within the academy, as well as how the consequences of these debates shaped the direction of the department. Next, this study explored the four factors that were primarily important to the rise and prominence of the Computer Science department, including a strong relationship with the Stanford Computation Center, an entrepreneurial culture among the faculty of the department, an organizational flexibility among the university administration, and finally, a need to engage with industry. The latter factor led to the creation of new networks with industry that were explored in the final part of this study, including the development of corporate relationships of mutual benefit and the development of new venues of engagement with industry, such as the development of the Honors Co-op program and the Computer Forum.
How an Academic Revolution Shaped a Region
These different components, while separated thematically in the study, are intimately related. One major pattern that flows throughout the development of the Computer Science department relates to the politics of knowledge, a concept that examines the social and political factors that shape the everyday construction of knowledge. This concept is particularly pertinent to the analysis on the development of computer science as a discipline. In the early years, the field was merely a part of the Mathematics department, one subfield of the larger area of numerical analysis from which Forsythe was hired. Computers may have been gaining importance in society, but a discipline had not yet developed with the mission of researching problems associated with their design and operation. However, this would soon change in universities across the country. At Stanford, Forsythe worked almost immediately to build up a program when he joined in 1957.
There were several factors that allowed the division to grow quickly in the early years. First, the leadership of Forsythe as both the head of the Computer Science division and the Computation Center provided a means of coordinating the two activities to positive effect. Unlike the debates at some other universities detailed in Akera,2 computer scientists at Stanford only lightly engaged in the service/academic debate, believing that both elements were crucial for computer science to succeed. Thus, the faculty that would eventually coalesce into computer science was already relatively unified in their approach, which strengthened the discipline’s case.
When the division began searching for faculty members outside of the traditional areas of mathematics, a disagreement erupted between the faculty members of mathematics who feared the encroaching of a new discipline and the computer scientists who desired to expand to new domains of human knowledge. In a way, the notion of Kuhn’s scientific paradigms appears, as well as his analysis of the stages of a scientific revolution.3 Mathematics as a discipline was opposed to the notion of computer science for a host of reasons, but ultimately, its opposition was part fear and part ignorance. The theories developed by computer scientists could radically alter the field of mathematics and its many centuries of history, and it simply takes time to adjust specialists to a new mode of thinking.
The same issue is visible in the letter written by the chemist Paul Flory. He worried that a new discipline like computer science could be “lethal” to the university and that the “bulwark of the disciplines” that is H&S, would crumble with its addition. Flory was responding in a particular moment of postwar scientific culture. Vannevar Bush’s essay on science, with its separation of basic and applied science, represented the blueprint of that era. Computer science, though, does not cleanly fit into Bush’s system (which is surprising given his personally strong connection to the development of computing). Like other engineering disciplines, Computer Science benefits from connections to industry that are developed through programs like the Computer Forum. Yet simultaneously, the discipline is highly theoretical with important core results. It is this hybrid nature that makes the field so controversial, and yet, enriching to analyze.
How then did computer science navigate the political landscape of knowledge within Stanford? Two factors are important. First, the organizational flexibility regarding the appointments in artificial intelligence shows the forward-looking nature of the university administration. Part of the reason for this culture was certainly Frederick Terman, who is among the most powerful and influential administrators of Stanford in its entire history. However, Stanford benefitted from a system of governance that deemphasized the power of individual faculty members to the importance of building the university in directions of future promise. Faculty members like Flory who were concerned about the development of applied science were essentially powerless to stop the trend.
Forsythe particularly benefitted from a university administration that was by education or by vocation already relatively connected to the issues of computing. Frederick Terman’s mentor, Vannevar Bush, made major contributions to the development of computing that became very well-known among the general public. In addition, Terman’s own experience as chair of the Electrical Engineering department likely assisted his decision in pushing for the development of a strong department in computer science.
Other administrators with direct budget authority over the Computer Science department had interests in fields that made them supportive of the department’s expansion. Halsey Royden and Albert Bowker both conducted work in applied mathematics and statistics, a field with important connections to computing. Robert Sears, the dean of H&S, continued the work of Frederick Terman’s father, and this close connection likely gave Terman influence over Sears.
Considering Forsythe’s actions toward the university administration, including ignoring budgets and effectively merging the budgets of the Computation Center and the Computer Science division in the years before 1965, it is remarkable that the administration never once forced the department to halt its expansion. There is little question that the administration’s enthusiasm for the department had a key role in its development.
Second, the Computer Science division quickly connected to other departments through interdisciplinary initiatives and research, building legitimacy with colleagues by spreading the value of its peculiar kind of research. One particularly useful component of this outreach was adding classes in other departments into the Computer Science curriculum. Few departments will say that their own classes are not a legitimate form of study, and few will criticize a particular collection of classes, especially when they have a defensible intellectual coherence.
In the end, Computer Science’s spirited trajectory both assisted and harmed its development. The attention the growth of the field provided likely increased budgets earlier, thus reducing the time required to establish its place in the university. However, the field’s underdevelopment certainly made the issue of academic legitimacy more palpable. The short span from its original conception to full department status is remarkable — only six years. It seems quite possible that a more mature field would have faced less controversy from other faculty members.
The political gauntlet faced by computer science did have repercussions. The Computer Science department at Stanford would not have an undergraduate degree for decades, largely because of the department’s attention to graduate work. That focus is at least in part a response to the debates about the academic legitimacy of the field. Along these lines, initial faculty hesitation at programs to engage with industry may also have been affected by this debate. The Computer Science department would never truly fit in with the departments of H&S, but it would take almost another twenty years before the department transferred to the School of Engineering in 1985. For undergraduates interested in Computer Science, the department recommended a Mathematics major, and later an interdisciplinary applied mathematics major. Only by 1986 would Stanford authorize a bachelor’s degree in Computer Science.
Thus, the combination of the politics of the academy and the need for more revenues created a strategic direction for the department that emphasized theory but also encouraged the practical application of results. The networks that were formed between Stanford, IBM, HP and the dozens of companies that joined the Computer Forum demonstrate the strength of this model. The Computer Science department is intimately related to the development of dozens of companies in Silicon Valley, and the creation of these linkages can be largely explained as a contingent development of the department’s maturation in the 1960s.
The politics of knowledge then has had a large transformative role in how the department organized itself, what activities it engaged in, and how it approached its future development. Since Stanford was a leading department, and its model was a blueprint for other universities, the politics faced in H&S had large effects on the course of computer science throughout the country.
Areas for Further Research
As more sources become available, historians of science are now increasing their attention to the development of computer science as a discipline. However, there remains large holes in our current understanding. First, a better study of the political economy linking computers, computer science and defense funding would allow for a closer examination of the effects of science policy on the development of an academic field. This is particularly important given the significant effect of defense research policies on computation.
Second, further research must be made at the university level to understand the politics of knowledge that existed at different institutions. This study provides an archival-based history of the developments at Stanford, but this information must be put into a comparative framework to analyze the varying experiences of universities including MIT, Harvard, Stanford, Carnegie Mellon and the University of Michigan. Along this line, additional analysis is required of how these departments evolved, and why universities specialized in particular areas.
Third, the connections between mathematical research and computer science needs to be further explained. The biographies of many of the pioneering computer scientists include a mathematical preparation, and this background provides a particular perspective on the development of the field. However, the development of a notion of computer science changed this biography in later generations. What are some of the problems associated at the interface between these two groups, and were there different approaches to evolving the field?
Finally, many of these components can be seen in the publication of the model curriculum developed by the ACM. Forsythe himself was highly involved in such efforts and was a passionate advocate for education in computer science. However, there has not been significant research to see how universities adapted the model curriculum to their own institutions. Such research would connect with the politics of knowledge framework, and could provide a strong comparative model from which to analyze.
The history of computer science provides a remarkably untapped area of research that can and should be explored. Our growing understanding of the ways in which social and political factors influence the course of research has the potential to illuminate the important origins of computer science, a field that will shape the world for years to come.
In terms of regional studies, there is significantly more work to be done on understanding the development of regional innovation hubs. First, there is an important new strain of research in spatial history that needs to be further explored. The co-location of industry and universities has become a commonplace in the development plans of most regions, but the exact nature of the relationship between them remains unclear. Particularly with the advent of online communication tools, the research on geographic proximity has the potential to be of immense value.
Second, along similar lines, there needs to be a more encompassing study of the ways in which public policy can influence the development of innovative economies. The research on military funding in the 1950s has become the default answer for scholars to answer how Silicon Valley began. However, the region was comparatively small compared to its Boston counterpart, and received fewer government grants as a result. Therefore, the connection between government funding and innovation is significantly more complex.
Third, and one of the main areas where this study has attempted to shed light, is how the development of new disciplines on the edge of existing academic fields can be used to coalesce a new industry in a particular region. The development of entire new fields happens continuously as our understanding of new domains of knowledge increases. Regional planning authorities should take advantage of these new disciplines as spearheads to build new industries with limited competition. However, to do so requires an understanding of their formation, and this is the role that research can fill.
In addition to these research strains, there is also a need for a more expansive look outside of the United States across all of the literatures connected to this study. Significant comparative research remains to be conducted on the approach of foreign universities to new disciplines, and how the politics of knowledge are influenced by local academic culture. In addition, the pursuit of regional innovative hubs has been conducted mostly outside of the United States, and has been largely a disappointment for the governments that have sponsored them. The connections between industry and academia, and the underlying ecology between them, needs to be closely studied.
Regional studies as a whole has a remarkable future. As cities work together to solve joint problems like traffic, land use, housing and economic development, researchers will have significantly more effect on the direction of public policy. Some of the most complex policy problems are regional in scope, and further research into any element of them will create a stronger, more vibrant economic picture of the United States – and the world.
- Miller to Manuel Rolenberg, 23 Oct. 1967, Sterling Papers, SC216/C1/14.↩︎
- Atsushi Akera, Calculating a Natural World, MIT Press, 2008.↩︎
- Kuhn, Thomas S. “The Structure of Scientific Revolutions.” Third Edition. University of Chicago Press, 1996↩︎