Hypertext: writing that branches

The word is his, from 1963. So is the idea that a text need not be a line.

Nelson's definition in Literary Machines is short: by hypertext he means non-sequential writing, text that branches and allows choices to the reader, best read at an interactive screen. The prefix, he insisted, refers to structure and not size. A hypertext is not a big text; it is a text with more than one way through it.

He coined hypertext and hypermedia in 1963 and put them in print in August 1965, in a paper for the ACM 20th National Conference called Complex information processing: a file structure for the complex, the changing and the indeterminate. The title is the thesis. Ordinary files impose an order that human thought does not have; the paper proposed a file structure that could hold changing, interconnected, indeterminate material, and introduced zippered lists as the mechanism.

The history around him is worth knowing. Vannevar Bush had imagined the memex in As We May Think in 1945, a desk of microfilm joined by associative trails. Douglas Engelbart's NLS, demonstrated in December 1968, put linked text, the mouse and shared screens in front of an audience. Nelson himself consulted on the Hypertext Editing System at Brown in 1967 with Andries van Dam, one of the first working hypertext systems. The World Wide Web arrived in 1989-91 and made one narrow version of the idea universal.

What the Web kept was the jump. What it dropped, in Nelson's view, was almost everything else: links that point both ways, connections you can see, quotations that stay attached to their source, versions that remember each other. He has spent the decades since saying so, sometimes loudly. The site that hosts these words uses one-way links and cannot, in his sense, quote him properly. That is his point.

Sources: Wikipedia, Literary Machines; Wikipedia, Hypertext; ACM Digital Library. Where this page paraphrases him, the paraphrase is ours and the source is his.