When Edsger Dijkstra got married in Amsterdam in 1957, he answered the question about his profession: programmer. The municipality did not accept that, because that profession did not exist at all then. So it became theoretical physics, the subject Dijkstra had studied in Leiden before embarking on the completely new path of computer science. “So much for the slowness with which I saw the programmer’s profession arise in my own country,” he wrote later in the ironic tone that so characterizes him.
Two years earlier, Dijkstra’s promoter Adriaan van Wijngaarden had convinced him that computers were here to stay. And why shouldn’t he be the one to make programming a respectable discipline? He carried out that mission with heart and soul until his death in 2002, twenty years ago. This year also marks exactly fifty years since he won the AM Turing Prize, the Nobel Prize in Computer Science, as the only Dutchman ever.
Dijkstra was a kind of guru. He attracted a small, loyal group of followers
Krzysztof Apt professor emeritus
Dijkstra played a key role in the establishment of computer science as a scientific discipline in the 1960s and 1970s, with a strong emphasis on provable correctness of computer programs. Anyone who plans a route with a navigation system uses the algorithm that Dijkstra devised in 1959: the shortest path algorithm, an efficient method for determining the fastest route from A to B, one of his seminal concrete contributions to computer science.
“Dijkstra was a kind of guru,” says Krzysztof Apt, professor emeritus at the Centrum Wiskunde & Informatica in Amsterdam, showing a series of handwritten letters from Dijkstra to himself at home. Apt is co-author of the newly published book Edsger Wybe Dijkstra: His Life, Work and Legacy, where dozens of computer scientists look back on Dijkstra’s life and work. Apt: “Dijkstra attracted a small, loyal group of followers who even imitated his writing style, notation, mannerisms and type of sandals.”
Humble and incredibly honest
Apt worked as a professor with Dijkstra at the University of Austin, Texas (USA) between 1987 and 1990. “On the one hand he was domineering, intimidating, rigid and with a negative sense of humor, but on the other hand he was also modest, honest and incredibly honest. The latter was decisive for me.” While working on the book about Dijkstra, Apt discovered how many colleagues Dijkstra corresponded with in letters and how often he let them stay at his home. Apt: “When he lived in Nuenen, he took many foreign colleagues on a bike ride with his tandem.”
Apt calls Dijkstra “the man who carried computer science on his shoulders”. “He felt responsible for computer science, as if it depended on him,” explains Apt. “At all the lectures he attended, he did not hide his opinions and was often merciless towards the speakers. As if he felt: this is my territory and it must not be polluted by a wrong approach.”
“Dijkstra’s most important legacy is rigorous mathematical thinking about programming,” says Turing Award winner Leslie Lamport. “He looked at computer code as if it were a mathematical object.” That is why he was a supporter of learning to program without a computer. He only bought a computer very late in his life, with fresh reluctance.
Dijkstra also had a weakness, and that was that he wanted others to think like him
Leslie Lamport Winner of the Turing Award
“However, Dijkstra also had a weakness,” says Lamport hesitantly, “and that was that he wanted others to think like him.”
In 1984, Dijkstra exchanged TH Eindhoven for Texas. He had become disillusioned with Dutch higher education in computer science. In the TV documentary Thinking as a discipline (2001) he said: “It was a time when the main concern of all departments was: is my curriculum watery enough?”
Dijkstra had a special way of disseminating his ideas. Yes, of course he occasionally wrote scholarly articles, mainly as sole author, but only if he felt his ideas were mature enough – and his bar was set high. Much more often he wrote his famous EWDs: 1,318 notes with his initials and a serial number. They are pieces of research, mathematical proofs, conference reports, commentaries and statements.
“He then mailed it to about twenty colleagues,” says Apt. “And they could then make copies of it themselves and send it on. With his EWDs, he was, so to speak, the first blogger in computer science.”
Apt: “They are often gems. He also showed how clearly you can write about real problems without using formulas. And he could perfectly summarize ideas in a single sentence and formulate tantalizing one-liners.”
At first Dijkstra wrote his EWDs on his Olivetti typewriter, but from 1979 he wrote them only with his Montblanc fountain pen. His handwriting became so iconic that computer scientist Luca Cardelli designed a Dijkstra typeface in the 1980s.
Lamport and Apt knew Dijkstra well. But how does the younger generation view Dijkstra? Felienne Hermans has recently become a professor in computer science education at VU and has developed the programming language Hedy for children. She states by email that Dijkstra had just passed away when she started studying.
Read an interview with Felienne Hermans: ‘The storm of the Capitol is also software’
Hermans: “When the EWDs began to circulate, we read them avidly as students. It was a kind of blogging, or twitter, full of sharp, simplistic statements against everything and everyone that “Cobol cripples the mind‘. It sounds funny, but of course it’s also a very ugly comment because Cobol, a programming language from the 1960s, contributed so much as the first machine-independent language. Then you might think: well, it was a different time, it’s true, but for me Dijkstra belongs to a generation of computer scientists who, with their nasty communication style, have ensured that people in computer science have long believed that we can work together like that . and that being rude is part of being a nerd, as long as you’re clear.”
Hermans believes that Dijkstra was so convinced of his own right that he really believed he had to convert everyone. “But when I think about what makes creating software really hard, I think about people things: collaboration, documentation, changing requirements and a changing world. People who found it interesting were driven out of the field by Dijkstra with his rigid way of thinking and often dull way of communicating.”
More extreme views
Dijkstra increasingly became a philosopher of computer science. Apt: “As computer science grew, its influence diminished. For many people, his principles became too idealistic. And as he grew older, his opinion hardened and his views became too extreme and therefore unpopular.”
Dijkstra’s thinking was a good fit in a time when computer programs were still fairly manageable, when not everyone had a smartphone yet.
But computer science has had to adapt. Computer science in the 21st century requires collaboration not only with mathematics, but also with engineering, social sciences, and the humanities. Apt: “At the height of his career, Dijkstra was probably the most influential computer scientist in the world. But now, fifty years later, the legacy of his most memorable is his concrete contribution to computer science and not his emphasis on a mathematical methodology.”