Four Hosts Were Not Yet Everyone
ARPANET in 1969 and the later work of internetworking
ARPANET began connecting research computers in 1969, with four hosts linked by year's end. It was a publicly funded research network, not an immediately universal public service.
Imagine researchers at two universities sharing access to a costly instrument through their computers. The connection broadens what their teams can attempt, but a nearby school without approved access still remains outside. Sharing resources within a network does not define its membership fairly.
A major motivation was sharing expensive computing resources across institutions. Packet switching divided communications into units, supporting a different approach from dedicating a circuit to each exchange.
Ideas and implementations came from multiple communities, including work in Britain and the United States. BBN engineers built packet switches, while participating institutions developed host software and applications.
The network supported collaboration among people as well as machines. Documentation, directories, and agreed procedures made connections useful; these tasks required sustained organizing and technical labor.
Its initial community was limited and institutionally selected. Access to a research network did not mean that every household, discipline, or country could participate on equal terms.
Two neighborhood associations keep separate contact lists and message rules. Connecting them requires agreement about addresses, delivery, and responsibilities, not simply adding another wire. Linking functioning networks is a further organizational achievement beyond connecting a few devices inside one system.
ARPANET should not simply be described as a fully decentralized Internet built to survive nuclear war. Its resource-sharing history differs from separate survivability research often folded into that story.
The later Internet required connecting networks of different designs through common protocols. TCP/IP development and ARPANET's 1983 transition were further milestones, not features completed in 1969.
Open technical discussion helped development, but openness of documents was not the same as universal access. Funding, equipment, administration, and network rules still shaped participation.
Distributed communication did not abolish concentrated authority. Operators, standards bodies, governments, and later businesses continued to influence how connections functioned and whose interests received priority.
ARPANET helped change the possibilities of shared computing without predetermining today's online world. That world emerged through additional inventions, institutions, policies, and choices about inclusion.
Further reading
What technical and institutional choices are needed for a network to expand participation without treating connection itself as proof of fairness or shared authority?