NRC Break From Past Produces Payoff

By John Makulowich, Senior Writer



The National Research Council's decision to ask a historian to chair a sweeping study on computer science has yielded an illuminating account that should go far toward deepening understanding of the contributions of federal funding to computer technology development, several experts said.

And by questioning the heavily mission-oriented focus of much of the computer research authorized by Congress, the report offers policy-makers guidance on ways to broaden their vision in selecting and supporting future research projects, they said.

The report, "Funding a Revolution: Government Support for Computer Research," notes that the gap between federal and industrial funding for computer research ? which narrowed earlier this decade ? is widening again.

Carried out from July 1996 to June 1998 and released by the research council, the study was chaired by Thomas Hughes, the internationally respected and award-winning historian of technology and a professor emeritus at the University of Pennsylvania. It also included Paul David, professor of economics at Stanford University, among its 13 committee members and one special adviser.

"This study shows the importance of taking a historical approach to the role of federal funding," Hughes said. Rather than simply project the present into the future, he said, the historical approach imparts the ability to see long-term developments.

"We could follow what was intended, and then see what came from it. It allows policy-makers to see the broad overview," he said.

The study took as its starting point the belief among technology and industry experts that the double-digit improvement rates for computer-based technologies experienced over the last two decades can continue for at least another decade if the appropriate investments are made.

The fundamental questions still unclear are what those investments should be and on what they depend. Further, according to the study, there seems to be little understanding of how to relate a "seemingly strong and steady flow of new technology to the slower and more diffuse processes of assimilating new technology into the economy."

The committee took a two-pronged approach to the study. First, it researched the history of computing and communications to identify the key trends in each historical era, as well the significant government activities that aided the industries' development.

Second, it created case studies in five areas representing a broad range of federal roles in the innovation process.

Along the way, the committee debunked some urban myths, such as the Internet's origins ? it was not intended as a network to withstand nuclear attack; that was the research interest of an early pioneer of packet switching, Paul Baran of the Rand Corp. ? and the role of government research in critical technologies such as artificial intelligence, virtual reality and relational databases.

The study's director, Jerry Sheehan, senior program officer for the National Academy of Sciences, noted this was the first time a historian was featured as chair. That decision, he said, was based "on the effort to build on the 1995 Brooks-Sutherland report [on the dynamic nature of innovation].

"We wanted to bring in a group of people who could teach us about process and content, about how to look back at the history of computing, to address questions about the limitations of what you can do," said Sheehan.

For Sheehan, the historians on the committee brought not only the standard analytic and quantitative approach, but a historical one. In the former, the future generally is projected from the present by extrapolating from quantitative data. In the historical approach, empirical evidence of the success and failure of past policies allows patterns to be uncovered that can inform future decisions.

Hughes, in chairing the committee, said he was greatly impressed by the importance of the federal government in stimulating the development of computer and communications in the 1950s, '60s and early '70s.

"We would not be where we are today if it had not been for federal funding in those early decades. The government played the primary role," Hughes said. "I was also impressed with the importance I found in the interactions among government, industry and the university. I heard so much about the military industrial complex and much has been critical. What I saw was the remarkable and fruitful results of this interaction with the government playing an oversight role."

Hughes said one of the values of the report is it shows what worked in those early years.

For example, the role of project managers is noteworthy. Some were masterful leaders in bringing principal investigators together. He recalls one who drew out what should be done, then came back with his own ideas.

Hughes' appraisal was echoed by Paul David, another committee member and professor of economics at Stanford. He commented: "After studying the historical record of the program directors and project managers involved in the federally sponsored research initiatives in this area, one comes away with the conclusion that the government could be remarkably successful in 'picking winners' ? not by targeting certain technologies for funding, but by recruiting and providing sustained support for talented research leaders."

According to David, the committee's work shows that by following the approach of historians who are interested in process ? in the way in which research is conducted, the way in which information and knowledge percolates through the economy and affects research in other areas ? one gains a wide view of relevant events and a deeper understanding of the connections between publicly funded exploratory research and the development of commercially successful technologies.

One example David cited was the field of artificial intelligence, a potentially controversial area, considered by some members of the research community to be the Holy Grail.

"You can look at [artificial intelligence] skeptically from the standpoint of the decision maker who sees that federal funds were spent and wants to know whether or not the programs achieved their announced goals," David said. "Of course, that is interesting, but to stop there misses the larger process to which government funding of AI projects contributed.

"What the historian naturally wants to do is to follow the ideas and the researchers that emerged from the program, to see where they went and to understand the many impacts they had on applications-oriented research and development results in both the public and the private sectors," he said.

"By asking whether the organization and management style of those pioneering research programs had something to do with their exceptional productivity, in terms of creating technological bases for so many new products and branches of industry, the committee was able to point to a richer and perhaps more widely applicable set of lessons about the government's role in the making of the information revolution," said David.