index
Kinds of textual adjacency
click on images for full-size:
A map of intentional hypertext connections (the "Tourism" region of this text)
Books, their physicality and connections
Sprawl physical adjacencies (Austin, Texas)
Books provide an intermediate case for two kinds of connection. Parts of the book will be "linked" by intentional connection to distant parts: one page may contain a reference to a distant page, or a sentence contain a footnote number linking to a note in the back of the book. Then there is the linear sequence of materials from page to page, which is both a physical and an intentional connection. There is, also, the physical availability of the edge of the book; I can stick my finger in two-thirds of the way through, or turn thirty pages at once to see what comes up; this allows abrupt non-intentional juxtapositions of different parts of the book, somewhat like finding something uncontrollably next-door to where you build. Hypertexts seldom have any analogue to this physical availability, though something like it could be built in, for instance, by providing a map of the hypertext that allows one to jump to another part of the text without following any intentional link. Such maps are, however, labeled, so that the jump is not quite as unintentional as jumping through book pages can be. Something more like the abrupt non-intentional adjacencies of physical space could be built into a hypertext by introducing randomized contacts, or a skein of built-in connectors independent of meaning or order of creation, etc. Such connectors could be random, or partly intentional, as in a library or bookstore where there is some order but unexpected encounters can still happen. Or, there could be automatic link creation. There is another paradigm of hypertext, not as linked chunks of text, but as fields of text from which search engines and algorithms create links on the fly depending on user interest. For instance, imagine software that watches what you are writing and adds links to other texts based on statistics about your use of words, or questions you ask in the text, or metaphors you create. We don't have intelligent enough software to do this well, though there are beginnings in some proposals for extending the capabilities of the Web, and there have been demonstrations of such link-on-the-fly programs. This type of hypertext does not have a fixed armature of links made intentionally by an author, though its links do represent priorities that were jointly set up by the authors of the software and the behavior of the users. (It is possible, of course, that such a system might suggest a link based on regularities in my writing or word use that I was totally unaware of and might find very helpful.)