Brown CS News

Where Meter Meets Mainframe: An Early Experiment Teaching Poetry With Computers

    by Kevin Stacey (Science News Officer, Physical Sciences)

    In the early 1970s, two Brown University professors devised a radical idea for the time. English professor Robert Scholes and computer scientist Andy van Dam wanted to use Brown’s room-sized mainframe computer to teach a course not in physics, engineering or computer science, but in poetry.

    The idea doesn’t sound so revolutionary now, in an era when computers are everywhere and can help people to do nearly anything. But some 45 years ago, Scholes and van Dam met with resistance. Many in the humanities thought it folly to try to explore the deeply personal experience of poetry in silico. Experts on the technology side — including the gatekeepers who allocated precious time on Brown’s mainframe — thought poetry was an improper use for the machine.

    “There was a cultural attitude that said that computers were for number crunching,” said van Dam, the Thomas J. Watson, Jr. University Professor of Technology and Education at Brown. “The idea of using a multimillion-dollar mainframe for manipulating text was hard for scientists and engineers to swallow.”

    But van Dam and Scholes pressed on, eventually getting their computer time, as well as a grant from the National Endowment for the Humanities (NEH) to help administer the class and evaluate its effectiveness. Late last month as part of the NEH’s 50th anniversary celebration, a group of scholars and NEH leaders including the director, gathered at the University of Maryland to discuss the project and its legacy. The event included a screening of a short documentary film about the project made in 1976 that hadn’t seen the light of day in decades.

    Parts of the film, which was recently digitized and is now available online, seem as quaint as one might expect from a film about 40-year-old computing technology. There are plenty of shots of room-sized computers, large magnetic disk drives and green-and-black monitors with telephone receivers affixed to the top.

    But the project itself seems to have had an uncanny prescience.

    “I really believe that we built the world’s first online scholarly community,” van Dam said in an interview. “It foreshadowed wikis, blogs and communal documents of all kinds.”

    The course was organized to create new ways for students to interact with texts, their instructors and each other. Units centered around a single poem, which students could scroll through on a mainframe terminal. Sections of the central poem were marked by descriptive “tags” used for searching that we would call hyperlinks today. Students could select these links to find critical reviews, related poems or other supplementary and contextualizing material. Students could also add their own thoughts and commentary into the system — all of which would then be available to their professors and classmates through new links from the central poem.

    In the documentary, van Dam describes the approach as a means for students to contribute to a “creative graffiti.” The idea was to create “a new kind of community with professors and students working together to develop a true process of learning.”

    To see if that process actually helped students learn, van Dam and Scholes approached the project as a formal experiment, placing students into two experimental groups and a control group. One group took the computer-based course, while another group used paper versions of the same materials. The control group took the same course in a traditional classroom setting. The project showed that the computer group came away with the deepest understanding of the material and the highest satisfaction with the course. It also created a lively dialog between students, instructors, professional critics and the texts themselves.

    “One of the things that we saw is that people gave roughly equal weight to their classmates’ and their instructors’ opinions as they gave to the professional critics,” van Dam said. He attributes that to the fact that all of the commentary from professional critics and classmates alike looked the same on the screen. There was no fancy typeface to give the appearance of professional heft. As a result, students weren’t intimidated by or overly reverent of the professional critics.

    “It was a leveling,” van Dam said, “a democratization.”

    But perhaps the most important outcome — certainly the outcome that was critical to Scholes as an English professor concerned with the decline of student writing — was the fact that using the computer encouraged students to write more. On average, the students in the computer-based group wrote three times as much, however informally, as the students in the other groups. Van Dam attributes that outcome, at least in part, to the ease with which the students could compose their thoughts in the software system used in the course.

    The system was dubbed FRESS, which stood for File Retrieval and Editing System. However, the name was not originally an acronym — van Dam added that later. Initially, the name came from the Yiddish word “fresser,” meaning a gluttonous eater. FRESS gobbled up a quarter of the Brown mainframe’s 512-kilobyte memory capacity at the time.

    Van Dam developed FRESS in 1968 with the help of mostly Brown undergrads and a few master’s students. It combined word processing with the creation of hypertext — text with links to other texts. It was the successor to the Hypertext Editing System that was developed by van Dam and Ted Nelson, coiner of the term “hypertext.”

    “From the beginning we wanted FRESS users to be both readers and writers, consumers and authors,” van Dam said. “So the fit to the poetry course was perfect.”

    FRESS had many attributes of a modern word processor. The system made it possible for users to move a cursor to different parts of a document and use commands to add, delete or move text. There was an additional feature — now ubiquitous in modern word processing — that van Dam is particularly proud to have developed for FRESS.

    “We had, I believe, the world’s first ‘undo’ [command], which for me is the most important element of any system,” van Dam said. “It’s critical because it takes the fear of experimentation away. You can always just undo it.”

    The development of FRESS has other very direct connections to modern word processing. One of van Dam’s students who worked on the development of FRESS would go on to be a program manager for Microsoft Word 1.0.

    But van Dam and Scholes had little of that in mind when they created their poetry course. They simply set out to show that digital technology had a place in the humanities and that an understanding and appreciation for poetry could be aided by streamlined human-computer interaction.

    By that measure alone, the project was an unqualified success, van Dam said.

    “We learned that students gain a tremendous amount of insight from having vast amounts of information at their fingertips.”