Computer science is what computer scientists do
Louis Fein's conception of computer science as an the academic discipline.
In a footnote to last Monday’s post, I quoted Ron’s Law, stating computer science is “whatever computer scientists are doing today.” It turns out Ron’s Law has always been true. And while perhaps there are good parts to maintaining pride in this adage, many of the venal side effects of the discipline have also been with us from the beginning.
According to the OED, the earliest use of the term “computer science” is in a 1956 ad by the Remington Rand corporation in the New York Times, advertising R&D positions on the UNIVAC offering “unparalleled opportunities to associate with the prominent pioneers of computer science at outstanding salaries.”
At the same time, on the other side of the country at the Stanford Research Institute, Louis Fein was compiling evidence to support the establishment of computer science as an academic discipline. After a few years of work, Fein presented a case for “The role of the University in computers, data processing, and related fields” in the September 1959 issue of Communications of the ACM.1 Using scare quotes throughout, Fein wielded the industry buzzword to make the case for university graduate schools of the same name.
Fein argued for a school of computer science using the lingua franca of the modern university: demand.
“The milieu of our society is characterized by a demand that more phases of our traditional culture be subjected to a critique guided by rationalistic and scientific rules of evidence.”
Demand, of course, is an economic term, and universities in the post-war period were excited to find new innovative ways to market themselves as suppliers. The US government demanded computing expertise for national security endeavors. Outside the university, they had created the RAND Corporation to carry out contracts of interest to national defense. They were subsequently also experimenting with establishing research relationships directly with universities, such as the Army Mathematics Research Center at the University of Wisconsin. More broadly, the federal government had begun handing out grants to university researchers for computing projects, in the still early days of universities getting a taste for the sweet overhead funds that came with such awards.
Contemporaneously, industry had a demand for people with computing skills. Industry itself was establishing schools to train professionals in computing technology. As Fein puts it,
“IBM has a manpower problem now; they know it will be severe in ten years. Their problem is two-fold. They need professionally trained people to help sell their product. They want customers to have professionally trained people to use their product properly. IBM has ‘presented’ 650s to over 50 universities by now under the condition (among others) that a couple of courses in data processing and numerical analysis be given.”
Fein emphasized that IBM’s “gifts” always came with a catch. “It is fair to say that, in many cases, to the extent that a university computer activity has a purpose at all, it has been made for them by IBM.” Fein also noted that this led to widespread faculty debasement: “The scramble to get in on a ‘free’ 650 computer from IBM is a disgrace in some cases.”
Fein’s view was that if you could have a physics department without a cyclotron, you should be able to have a computer science department without a massive mainframe. While many academics saw computing as nothing more than infrastructure (or, “plumbing” if you want to be pejorative) and resisted the idea of a research department, Fein saw a bigger picture beyond the computing center. The center and those who used it would be more effective if collected into a graduate school. There, workers could be trained for industry, government grants could be deposited for intellectual advancements, and the associated tuition and overhead could be co-deposited. Though other disciplines were able to hide behind earlier noble traditions of the life of the mind, the cultivation of inquiry, or commitment to the citizen, computer science was built in the post-war age where the university had turned its focus to industrial shilling and government billing. Fein’s computer science was designed to embrace Vannevar Bush's idealistic bureaucratic visions.2
Obviously, there had to be some intellectual core to build a graduate school. So Fein stitched one together based on Ron’s Law. Faculty could get better resources and pay in the private sector, so how could they be enticed to stay at the university? A department that let them work on whatever they wanted seemed in order. The unifying intellectual foundation of Fein’s computer sciences was the research people who used computers were doing. Fein pined that once the departments were mature enough, an underlying universal theory might emerge. But he didn’t want intellectual cohesion to get in the way of the pressing need for a training program.
Fein’s article frames computer science as akin to “management science” and “library science,” primarily training programs connected to university infrastructure that would be agile to changes in the economic and technological substrates on which they were built. He compared such research infrastructure to mathematics: multidisciplinary tooling that needs its own self-study.
“Computer science is not isolated; it is inter-disciplinary—like a library, or mathematics, since library science or mathematics are both disciplines in themselves as well as service tools to other disciplines.”
Fein’s proposed computer sciences thus included the things we associate with core computer science now: architecture, programming languages, artificial intelligence, information retrieval. Fein also viewed what some academics now call “data science3” as computer sciences: Statistics, forecasting, probability models, simulation, data gathering, retrieval, analysis. All applications of computers, be they in business, physical science, or biological science, were envisioned as part of computer science. Computer scientists were whichever faculty wanted to join the department. Computer science was whatever computer scientists were doing. Maybe they’d come up with a unifying theory. Maybe they wouldn’t. But that was secondary to filling the administrative coffers by meeting external demands.
Computer scientists have spent the last two decades putting forward Fein’s marketing and PR campaign for new and bigger Schools of Computing or Data Science. The only difference is Fein argued near the beginning of the information age, and our modern deans are lobbying at the end. The university has a rather different relationship with government and industry than it did in the 1950s. Industry claims they don’t need the university, arguing that it can now automate software engineering. The federal government is waging an outright assault on the university system and is slashing direct and indirect cost payouts. What’s the case for a computer science department in 2025? What is it that computer scientists will do? It will be interesting to see.
h/t Dawson Verley, who found this during a class discussion.
It’s worth it for all CS professors to read this and then laugh at the notion that computer science courses can be apolitical. No matter what op-eds published in the New York Times say, computer science is as political and ideological as economics.
Like Fein, I’m using scare quotes to distinguish from the now unfashionable industrial buzzword of the 2010s.
How do Numerical Analysis and numerical methods more generally fit into this view of Computer Science? Numerical methods are widely used in many practical applications, but AFAIK generally aren't viewed as part of Computer Science. I don't know where Numerical Analysis fits in the academic universe now Applied Math departments are disappearing from universities.
Really enjoyed this historical perspective. I'm sure you've read it, but just in case, "How data happened" (https://wwnorton.com/books/how-data-happened) also gives a very interesting history mid-to-late 20th century computer science. In particular, there's a very interesting discussion about why machine learning and data science developed in CS departments and not in statistics departments (because the former was worried about practical uses, aka engineering stuff, while the latter was more interested in theoretical stuff).