Newer View 2: Constructivity
Near the very beginning of hypertext research, Michael Joyce distinguished exploratory hypertext, in which the reader follows links to explore what an author has written, from constructive hypertext — hypertext that invites the reader to extend the hypertext. This was an extremely inviting theoretical idea in 1988, since postmodern ideas about reading the writing were then new in mainstream American scholarship. As Joyce memorably suggested, constructive hypertexts are "versions of what they are becoming, a structure for what
does not yet exist.” The idea has had less impact on the engineering side of hypertext research, partly because computer science has never quite known what to do with postmodern thought, and partly because early efforts were not entirely convincing.
In Tinderbox, we typically make one note for an idea, and then make a new note for the next idea. That’s not the way cities are built: we make cities street by street, tower by tower, subdivision by development. That means a fast-growing city is bound to have some vacant dwellings awaiting new residents. That’s also common in note-taking.
1. When starting a new scientific paper, we often know at the very outset what the major sections of the paper will be. In many fields, this is canonical: you have an introduction, a review of the literature, a description of the experiment, a discussion of the results, and conclude with ideas for further work. Why make this piece by piece, and not all at once?
2. In a big piece of nonfiction, sources don’t really come one by one. Today at the library, you sit down to find out about why historians think History can inform contemporary opinion. That’s not going to be a single source, but you’ve only got space for a paragraph or two, and you don’t want to read the whole library. You’re going to need a handful of sources to satisfy Referee #2, and you probably don’t have time for more than a dozen. Why not make a dozen notes, and fill them up as you find books and articles with reading?
In map view, you don’t want to have many “vacant” notes because, even on today’s screens, you can only fit 100 notes or so on the screen before things start to feel crowded. Because the Information City doesn’t label every note all the time, each note can use fewer pixels. By making different building look different, we may be able to remember what things are and find them when wanted. Vacant notes have black roofs, while notes that are in use have red roofs. An area in which we plan to work is simply a neighborhood of vacant buildings, ready to be occupied by new sources.