A Brooklyn beginning
Robert Elliot Kahn was born in December 1938 in Brooklyn, New York, to Beatrice Pauline Kahn, née Tashker, and Lawrence Kahn, in an Ashkenazi Jewish family. Through his father he is related to the futurist and strategist Herman Kahn. He grew up in New York City and, like many children of the city’s Jewish families in that era, went to the tuition-free City College of New York, where he earned a bachelor’s degree in electrical engineering in 1960.
He went on to Princeton University, earning a master’s degree in 1962 and a doctorate in 1964, both in electrical engineering. His dissertation, supervised by Bede Liu, dealt with problems in the sampling and modulation of signals, the mathematics of turning continuous signals into data that can be sent and received.
From theory to building networks
After Princeton, Kahn joined the technical staff of Bell Labs in 1964, and in 1966 he became an assistant professor of electrical engineering at the Massachusetts Institute of Technology. Wanting to spend more time on applied research, he took leave from MIT to join Bolt Beranek and Newman, the Cambridge engineering firm known as BBN, as a senior scientist.
In 1968 the Pentagon’s Advanced Research Projects Agency, ARPA, later renamed DARPA, asked for proposals to build an ambitious new computer network. Kahn’s background was in signal processing, but his knowledge of communications networks made him central to the proposal that won BBN the contract. The resulting ARPANET used packet switching, sending data in small blocks across many routes, so that very different computers at universities and laboratories could share information. Kahn is widely described as one of its principal designers.
The project brought together the ideas of many people, including Paul Baran at RAND and Donald Davies in Britain, who had independently proposed packet switching, and Leonard Kleinrock at UCLA, whose laboratory hosted the network’s first node in 1969. Kahn’s role was to help turn those ideas into a working system that could be built, tested and run.
Showing the world it worked
By 1972 the ARPANET existed, but few people outside a small research community believed in it. In October of that year Kahn organised a public demonstration at the first International Conference on Computer Communication in Washington, D.C., connecting some twenty different computers to the network on the conference floor.
The demonstration has been called the watershed event that made people suddenly realise that packet switching was a real technology. Later that year Kahn moved to DARPA itself, to its Information Processing Techniques Office, where he would oversee the ARPANET and start new programmes in networking by radio and satellite.
The problem of many networks
At DARPA, Kahn faced a problem that had no precedent. Besides the wired ARPANET, the agency was building a packet radio network for mobile units and a satellite network called SATNET to link to Europe. Each had different speeds, packet sizes and error rates. The ARPANET’s existing protocol, called NCP, assumed a single, reliable network and could not tie these together. Kahn began sketching a way for separate networks, each built however its owners liked, to connect and pass traffic between them. He called the idea open-architecture networking.
In the spring of 1973 he brought in Vint Cerf, a young computer scientist then at Stanford, to work on it with him. Together they designed what became the Transmission Control Protocol. In May 1974 they published it in the IEEE Transactions on Communications under the title A Protocol for Packet Network Intercommunication, a paper that laid out the architecture of the internet.
How TCP/IP works, and why it matters
The design rested on a few simple principles. Networks would be joined by special computers, first called gateways and now called routers, whose only job was to forward packets. No single part of the system would be a point of failure or be able to control the whole. Every piece of data would carry a sequence number, so the receiving computer could put the pieces back in the right order and notice any that were missing. The receiver would send back an acknowledgement for what arrived safely, and anything not acknowledged in time would be sent again. A checksum on each piece would reveal any damage along the way.
The new protocols were not adopted overnight. They were tested across all three of DARPA’s networks, the wired ARPANET, packet radio and satellite, before being made the agency’s standard. Cerf later explained that TCP was split into two parts partly to handle real-time data such as voice better. In 1981 the network engineer Jon Postel published a plan to move the ARPANET’s roughly 400 host computers from the old NCP protocol to TCP/IP, with a hard deadline: any machine not converted by January 1, 1983, would be cut off. Cerf recalled that when the day came there were no grand celebrations, only relief among system administrators who had raced against the clock, and pins proudly worn by those who had come through it, reading I survived the TCP/IP switchover.
The protocol was later divided into two layers: the Internet Protocol, or IP, which handles addressing and moving packets from network to network, and TCP, which makes sure the conversation between two computers is complete and in order. On January 1, 1983, the ARPANET required every connected machine to switch to TCP/IP, a date often called the birth of the modern internet. Because the protocols do not care what hardware or software a network uses, anyone could connect. That openness is why the same internet now links phones, laptops, hospitals, banks, schools and satellites, and why billions of people can reach one another regardless of the device they use.
Leading research and looking ahead
Kahn became director of DARPA’s Information Processing Techniques Office and started the Strategic Computing Initiative, a billion-dollar programme that was the largest computer research and development effort the United States government had ever undertaken. After thirteen years at DARPA he left in 1986 to found the Corporation for National Research Initiatives, a non-profit research organisation, which he has led as chairman, chief executive and president.
In the mid-1980s he coined the term National Information Infrastructure, an idea later popularised as the information superhighway. At CNRI he has worked on the Digital Object Architecture, a system for identifying and managing digital information over long periods. In 1992 he and Cerf co-founded the Internet Society to provide leadership on internet standards, education and policy.
Jewish background
Kahn grew up in an Ashkenazi Jewish family in Brooklyn, part of the large Jewish community of mid-century New York, and like many of its children he was educated at City College, the free public institution that served as a ladder into the professions for generations of the sons and daughters of Jewish immigrants. He is listed among Jewish computer scientists by jinfo.org, alongside his fellow internet pioneers Leonard Kleinrock and Paul Baran.
Kahn has not made his religious life a public theme, and this profile does not claim more than the record shows: a Jewish family background in Brooklyn and an education in the city’s public institutions that set him on his path.
Honours
With Cerf, Kahn received the National Medal of Technology from President Bill Clinton in 1997 for creating and sustaining the development of the internet protocols. The two shared the 2004 A. M. Turing Award, often called the Nobel Prize of computing, for pioneering work on internetworking, and Kahn received the Presidential Medal of Freedom in 2005. He was inducted into the National Inventors Hall of Fame in 2006 and the Internet Hall of Fame in 2012, shared the inaugural Queen Elizabeth Prize for Engineering in 2013, and received the IEEE Medal of Honor in 2024.
Kahn has also been generous in sharing credit. Remembering Paul Baran, who first developed the packet idea at RAND, he called him one of the finest gentlemen he had ever met.
Legacy
Most people have never heard of Robert Kahn, yet they use his work every time they send a message, stream a film or look something up. The internet is not a single network but thousands of independent networks that agree to speak one language, and that language is the one he and Cerf designed in 1973 and 1974.
Looking back almost forty years after the paper he wrote with Kahn, Cerf said that while the two of them had high hopes, they did not dare to assume that the internet would turn into the worldwide platform it became. The key was a decision Kahn made early: not to build one perfect network, but to let many imperfect, independently run networks join together on equal terms. Packet radio, satellites, local office networks, and later mobile phone networks and home broadband all fitted into the same design without anyone needing permission to connect.
His own account of his role is modest. A lot of people were thinking about these ideas, he has said; he was simply in a position to propose building it for real. That combination of engineering vision and the will to turn it into working systems, from the ARPANET demonstration in 1972 to the TCP/IP switchover in 1983, helped give the world an open, shared communications system that no single company or country owns.
Sources & further reading
Robert E. Kahn, National Inventors Hall of Fame · Robert Kahn, Internet Hall of Fame · Marking the birth of the modern-day Internet, Google (The Keyword) · Jewish computer and information scientists, jinfo.org · Bob Kahn, Wikipedia · Researched from the sources above and checked against them. Spotted an error? Tell us.
A life of tikkun olam. This is part of tikkun olam, the Jewish idea of repairing the world, and every one of us can add to it.