NRC Report Sparks Debate Among Computer Scientists

A report from the National Research Council outlining an agenda for the future of computer science has attracted stinging comments from nearly 1,000 computer scientists, including some of the major names in the field. The critics say the report completely ignores large areas of computer science, such as artificial intelligence (AI). They are calling for the document to be withdrawn and revised. On September 4, the detractors sent a petition to the E-mail message boards of the United States' largest university computer science departments. The message brands the NRC report as "misleading and even harmful as an agenda for future research." It asks those with similar feelings to "join in requesting the report's withdrawal." However, NRC chairman Frank Press refuses to consider any major changes in the report, entitled Computing the Future: A Broader Agenda for Computer Science and Engineering. The document, released July 13, was produced by NRC's Computer Science and Telecommunications Board (CSTB) to define the priorities and policies the field needs in order to successfully enter the 21st century. The conflict began as sharp disagreements were voiced concerning the report's three focal points. Computing the Future asserts that sustaining the field's core research agenda over the long term depends upon strengthening the theoretical and experimental science base--generally rooted in academic institutions--on which computer application programs are often fostered. The two other priorities set forth by the document are "broadening the field" and "improving undergraduate education in computer science and engineering." Perhaps the main concern of the petitioners, according to John McCarthy, a computer science professor at Stanford University and the main author of the dissenting appeal, is that the report places too much emphasis on applied research while ignoring basic research. He says that the document "completely ignores" computer science for science's sake. According to the petition's authors, at press time 960 computer scientists had sent E-mail notes requesting that their signatures be added to the document. The 16-member committee that wrote the NRC report was chaired by Cornell University computer science professor Juris Hartmanis. It included 12 computer scientists from academe and four from industry. The committee's work was overseen by CSTB, which is an NRC standing panel of 20 scientists and engineers. CSTB also includes representatives from both academe and industry. The E-mail petition was initially sponsored by computer science professors Nils Nilsson, John Mitchell, and McCarthy of Stanford; Robert Boyer at the University of Texas in Austin; and Jack Minker of the University of Maryland. Four highly regarded American computer scientists who have joined in sponsoring the petition are Stanford's Don Knuth, a world- renowned algorithm scientist; IBM Corp.'s John Backus, who invented the computer language FORTRAN; Harvard University's Barbara Grosz, incoming president of the American Association of Artificial Intelligence; and Massachusetts Institute of Technology professor Marvin Minsky, a prime mover in AI research. The stature of the petition's signers is making many computer scientists take notice, says Peter Lee, an assistant professor in computer science at Carnegie Mellon University. "These are people I have great respect and admiration for," he says. "They are not radical or reckless people. The fact that they wrote against [the NRC report] is making me take the document much more seriously." "A report from the National Research Council constitutes essentially the advice that the National Academy of Sciences gives the government," says McCarthy. "It seemed to me that it was bad advice." The petition does not attack the NRC's entire 260-page document: only the preface, the executive summary, and two subsequent chapters. The strongly worded petition decries what its endorsers see as four main inadequacies in the content of Computing the Future: a narrow definition of computer science that ignores areas such as AI and theoretical analysis; a suggestion that computer science teachers are "troops on the ground," and therefore government agencies should set research agendas; support for applied science to benefit the nation's economy, while basic research is ignored; a lack of adequate research goals. Additionally, the petition criticizes the report for merging computer science and computer engineering into one structure, saying that the result "abolishes" the former and "seriously narrows" the latter. By ignoring basic science, the petition says, Computing the Future represents "a pre-emptive surrender to anti-science tendencies" in the U.S. government. McCarthy was given an hour to speak at the CSTB meeting on September 21. He said he tried to persuade Press to withdraw the report. However, Press would not back down from a September 16 statement concerning McCarthy's petition in which Press stated that NRC "will not withdraw or revise the report." "McCarthy is upset by the tone of the report because he is misreading a good deal of it. He is overreacting," says committee chairman Hartmanis. "Someone wanting to support science for science's sake simply does not live in the real world," Hartmanis says. "Much of computer science is supported by the military. DARPA [the Defense Advanced Research Projects Agency] and the Office of Naval Research do not support computer science for the sheer fun of it, but because it is relative to their missions--and they surely expect some benefit," he adds. Hartmanis says one intention of the NRC committee was to acknowledge the field's history, determining how basic research can be sustained while expanding the field's applied aspects. Computer science, when compared with biology or physics, is a young field. The first electronic stored-program digital computer was invented only 50 years ago, and not until the late 1960s could one receive a computer science degree. Computing the Future states that its authors' "first and foremost judgment was that computer science and engineering is coming of age." The core areas of basic research have now developed into a solid technical discipline. Many universities feature computer science as its own department, Hartmanis points out. Both sides recognize that the debate, largely a conflict between basic vs. applied research ideologies, is not over. McCarthy says that even though the petition was not successful in forcing NRC to withdraw the report, it did bring the concerns of a large number of American computer scientists into the limelight. Hartmanis agrees: "As long as we keep the debate on the issues of the science and away from the personalities involved, hopefully this will be good for the field."

NRC Report Sparks Debate Among Computer Scientists

Interested in reading more?

Become a Member of