When thinking about the infrastructure of the Internet, it is important to consider the role of government in this infrastructure.

This is a question that involves two aspects: the role of government, and the role of the computer scientists who are part of the needed government structure or institution. Reviewing the history of the development of the Internet helps to highlight the importance of some role for both government and for computer scientists.

The history of the Internet, as Charles Herzfeld pointed out in the March 1999 issue of iMP, involves the role played by Department of Defense Advanced Research Projects Agency (ARPA)‘s Office of Information Processing Techniques (IPTO). In 1962, J.C.R. Licklider was invited to develop a computer science program at ARPA and began the IPTO. Licklider had a vision of an “intergalactic” network which would make collaboration possible among computer scientists working on different computers and using different operating systems.[1] When he came to ARPA, he recognized that it was too early to start research for such developments, but only seven years later, by 1969, research was funded by IPTO to create the ARPANET, a prototype packet switching network.

Robert Kahn, now president of the Corporation for National Research Initiatives (CNRI), joined IPTO as a Program Manager in 1972, after being part of the team at the consulting firm Bolt, Beranek, and Newman (BBN) that designed and built the Interface Message Processor (IMP) subnetwork for the early ARPANET. After Kahn came to IPTO, he began research on connecting a ground packet radio network and a satellite network with the ARPANET. This required research on a new protocol, a protocol that would allow the internetworking of diverse packet switching networks.

Kahn invited Vinton Cerf, now at WorldCom, who had also worked on the ARPANET, to join him in designing this new protocol, subsequently known as TCP/IP. By September 1973, they had a paper describing their new protocol which they presented to networking researchers from several countries at a meeting in the U.K., sponsored by NATO at the University of Sussex in Brighton. And in May 1974 their paper describing this new protocol was published in IEEE Transactions on Communications, titled “A Protocol for Packet Network Intercommunication”.[2]

It took six more years of implementation and development work on the new protocol before a decision was made to replace the ARPANET protocol with TCP/IP. The cutover from the early ARPANET protocol to TCP/IP was made on January 1, 1983. And the ARPANET was split into an ARPANET and MILNET in October 1983, or into an early Internet.[3]

The reason this background is important is that networking developments have required an attention to future needs. These needs require long term research commitments and the ability to take a long term view of what will be needed and to prepare for those needs. Also, this research requires the leadership of computer science researchers who are able to provide the needed technical expertise. IPTO as an institution within the U.S. government, within the U.S. Department of Defense, was able to provide the support for research to identify the need for a new protocol to support internetworking, to develop and implement this protocol, and then to oversee the cutover to the new protocol on the ARPANET. Without such a research institution within government, it would be much more difficult for figures like J.C.R. Licklider or Robert Kahn to have been able to give the leadership and support to the research community to be able to work on future developments. It would have been equally difficult for them to have the kind of grassroots input that they required to be able to provide far-sighted direction for the research.

Thus, government has a role, a very special role, in the support of computer science research. This role is demonstrated by the creation and development of the Internet. The federal role is to provide an institutional form for supporting leadership in the research that can look forward 10 or 20 years and do the work needed to prepare for that future. Also, government is needed to involve computer scientists with a vision for the future in the support of those whose research will serve future needs.

Government and infrastructure are, in this way, interrelated. Infrastructure is something that forms a foundation to build on. It is that which the rest of the structure is built upon. There is a need for appropriate government institutions or structures because of the long term view that is required to build and adequately maintain an infrastructure. The development of the Internet shows a similar interrelationship between government and infrastructure. The creation and development of the Internet demonstrates the splendid achievements possible when there is a government research institution led by leading computer scientists to provide both a vision and the needed support for future-oriented research.

There are other aspects of the IPTO that are important to understand. One of these aspects is the interrelation between the computer science research community and the IPTO staff. During much of its existence, IPTO made it possible for the grassroots computer science research community to have the needed funding and administrative oversight, expedited contractual arrangements, and scientific leadership to explore the research questions that advanced the theory and practice of computer science.[4] IPTO provided a means to develop the new concepts and new principles on which new practices could be built. Without such basic research foundations set by many researchers, those exploring new problems have fewer of the needed resources to build on. Thus a research infrastructure is needed for the Internet infrastructure. And such a research infrastructure needs the lessons learned from the creation and experience of IPTO toward fashioning a government/computer science institution that will serve the need for continued computer science research to scale and continue the development of the Internet.

——-

[1] See “The Internet: A New Communication Paradigm”

[2] Com-22, in No. 5, May, 1974. Pgs. 637-47

[3] “From the ARPANET to the Internet: A Study of the ARPANET TCP/IP Digest and of the Role of Online Communication in the Transition from the ARPANET to the Internet”

[4] “Computer Science and Government: ARPA/IPTO (1962-1986) Creating the Needed Interface”