A Pioneer in Participatory Design Turns 80

In the mid-1970s, Christiane Floyd was a computer programmer with software development company Softlab in Munich, Germany. One project on which she worked was the development of a system to automate work in a datacenter of the steel company Halbergerhütte, 250 miles away in Saarbrücken. Floyd, together with her Softlab colleague Rita Nagel, spent many nights and weekends testing the new system during off-hours, when the computers were not being used for the day-to-day business of the steel plant.

"There was so much coal dust in the air that we had to shake our packages of punch cards, or else the card reader would get clogged." she recalled. "There was an awful little canteen nearby where we could get watery coffee and bad-tasting soup until 10:30 p.m., and then we were on our own until 6 a.m. next morning. Not a nice place for two young women to spend their nights and weekends." The demands of the job were especially difficult for Floyd, who at the time was a single parent of a small child.

For Floyd, who turned 80 years old last month, the work at the steel plant became a seed for her pioneering research in participatory design, a methodology that incorporates the needs of all stakeholders as crucial ingredients in the creation of software systems. As the first professor of software engineering in the German-speaking world, as well as the first woman professor of computer science in Germany, she had a major influence on the development of that subject as an academic discipline.

Floyd had a vision of how computers could impact society that was "really forward-looking ," said Albrecht Schmidt, professor of computer science and head of the Human-Centered Ubiquitous Media laboratory at the University of Munich. "It was quite amazing at the time because computing was not ubiquitous. It was basically special-purpose machines. They were not on everybody's desk." Floyd advocated the notion that computer science "has a responsibility to interact with society."

Born Christiane Riedl in Vienna on April 26, 1943, she was something of a child prodigy and showed a talent for mathematics, learning arithmetic at the age of four so she could help her older sister with homework. She also learned to read at that age and developed wide interests, including history and philosophy.

After earning a doctorate in mathematics at the University of Vienna in 1966, she moved to Munich and became a programmer for Siemens, where she was part of a small team building a compiler for the ALGOL60 programming language. "ALGOL60 never became a language that was important in use, but it was sort of the grandmother of most languages that we know," Floyd explained. The ALGOL60 compiler project was quite prestigious and enabled her next opportunity, at Stanford University.

There, she was for a short time part of Edward Feigenbaum's DENDRAL project, which created the world's first expert system. Her main work, which dealt with compilers, was at the Stanford Linear Accelerator in the group of David Gries, who had also worked with ALGOL60 and had studied in Munich. After her marriage to computer scientist (and, in 1978, ACM A.M. Turing Award recipient) Robert Floyd and the birth of their son, she taught computer science part-time at Stanford.

Upon their divorce in 1973, Christiane Floyd returned to Munich and joined Softlab. In addition to her work at the steel plant, she worked on Maestro, the world's first integrated programming environment for software, which eventually saw more than 20,000 installations worldwide. Because she was in charge of developing and teaching programming methodology to new Softlab employees and to customers, Floyd developed an intimate understanding of what people needed from computers and how computers affect work environments.

By 1977, she had married Danish computer scientist (and, in 2005, ACM A.M. Turing Award recipient) Peter Naur, and had given birth to a daughter. The following year she moved, without Naur, to Berlin, to become Germany's first professor of software engineering at the Technical University of Berlin (TU Berlin).  It was not easy; her new position was extremely demanding, and she was raising two children on her own, in an unfamiliar city. "In spite of many difficulties, I immediately knew that this was the right place for me to be," she said. "I loved to teach, and my research actually evolved out of my teaching practice."

For many years, Floyd was Germany's only woman computer science professor, so "she had huge relevance as a role model in computing," said Eva Hornecker, professor of human-computer interaction at Germany's Bauhaus-Universität Weimar.  Hornecker remembered not only Floyd's supportive and encouraging presence at many conferences and in personal interactions over the years, but also "her stance on software design, which was completely innovative at the time."

At TU Berlin, Floyd and some of her colleagues tackled the challenge of creating Germany's first academic program in software engineering. Based on her work in industry, Floyd believed students would learn best by working on actual software projects. However, she noted, this approach was at odds with that of some of the other professors, who preferred "conventional, old-fashioned classroom teaching, with the students listening and going home and doing homework."

This was also a time when students were questioning authority figures and academic traditions. In California, Floyd had witnessed campus protests over the U.S. war in Vietnam, an experience that gave her insights useful in her work in Berlin. "One reason why I was hired as a professor in Berlin was that they expected me to mediate between the conservative professors and the riotous students, which is, in fact, what I did for years," she said.

Teaching and talking with many students, Floyd took their concerns seriously. She reflected on her work at Softlab where, as she put, "I was a prophetess of structured programming until I saw the limitations of this approach." She absorbed ideas from like-minded critics of structured programming, such as Naur and David Parnas. Inspiration came also from researchers farther afield, such as the anthropologist Gregory Bateson and the physicist and philosopher Heinz von Foerster. With the latter, Floyd developed a warm and lively friendship and collaboration.

Structured programming mandates a top-down approach. "You should start with an abstract statement of the problem, and then refine it," said Floyd.  However, she added, "The process in time is much more complicated. You go back and forth, you think here and there, you experiment." Her ideas became the basis of extensive collaborations with colleagues at TU Berlin, and later, at the University of Hamburg, where she moved in 1991, as well as with researchers in the U.S. and Scandinavia. This collaboration produced Software Technology for Evolutionary Participatory System Design (STEPS), one of the first process models for software development. STEPS provided a human-centered approach to software engineering with an emphasis on architecture, object orientation, and technical soundness. Floyd wrote a brief description of STEPS for the June 1993 issue of Communications.

"In retrospect, [STEPS] can be called a precursor of agile methods," said Floyd.  "Agile methods, of course, are now very much in and rely on supporting teamwork and a lot of feedback from users and the like…  So on the software engineering and project organization side, I was a precursor of that, but I went far deeper into the user context" than is typical today in agile methods.

The arc of Floyd's intellectual journey led naturally into philosophical and ethical concerns in computer science. She has been active in the Bremen-based group Computer Scientists for Peace and Social Responsibility (Forum InformatikerInnen für Frieden und gesellschaftliche Verantwortung, or FIfF). She has longstanding ties to Ethiopia, where she helped found a doctoral program in computer science. In recent years, she has been advising on the development of a smart phone-based app for monitoring pregnancies, to reduce infant and maternal mortality in Ethiopia.

A few decades back, it was common for computer science departments in Germany to ignore the social and ethical dimensions of their subject. "They would say, 'It's not our business, we are just teaching programming'," observed Hornecker. Floyd, she said, "always had that ethical impulse and was always thinking about the consequences of research. …Seeing a senior person in the field having this naturally integrated into her thinking—that has made her a huge role model."

 

Allyn Jackson is a journalist specializing in science and mathematics, who is based in Germany.