The term 'end-to-end' has become a familiar characterization of the architecture of the Internet, not only in engineering discourse, but in contexts as varied as political manifestos, commercial promotions, and legal arguments. Its ubiquity and opacity cloaks the complexity of the technology it describes, and stands in for a richer controversy about the details of network design.

This essay considers the appearance, in the 1970s, of the term 'end-to-end' in computer science discourse, and how the term became a point of contention within disputes about how to build a packet-switched network. I argue that the resolution of some of those disputes depended on the transformation of the term from descriptor to 'principle'. This transformation attempted to close specific design debates, and, in the process, made the term dramatically more useful in those discourses beyond engineering that eventually took a keen interest in the design of digital communication networks.

The term, drawn from common parlance and given not only meaning but conviction, was shaped and polished so as to be mobile. As such, it actively managed and aligned disparate structural agendas, and has had subtle consequences for how the Internet has been understood, sold, legislated, and even re-designed.