> I do have one theoretical problem with hypertext. It seems to be
> founded in the assumption that knowledge can be organized in the form of
> a hierarchy, and that the best way to study a hierarchy is to go down
> one long path through it, without ever seeing the whole hierarchy.
> It also seems to be founded in the assumption that hierarchy is the
> most general form of knowledge organization.
I don't think that this is not a property of hypertext itself, but in the
way some hypertext systems have been implemented. Used creatively,
hypertext can subvert the whole notion of a hierarchy by providing
multiple ways of organizing, or viewing, the same information. What is
seen as the "right" levels of importance to one view or organizing
principle need not be the same set of things as another view, while both
can be easily accomodated in the same system.
For instance, we have created fairly large collections of information
about particular projects in hypertext databases. These databases are
structured so that there are multiple organizing views to the same
information. These include (among others) "project management" views,
which are concerned with timelines, deliverables, and status; "modeling"
views, which organize information about the problem domain in structured
maps; and "issue-based" or topic views, which contain or point to
free-form discussions about issues of importance to the project. All the
views contain appropriate links into each other, so that there are many
ways to get at the information that the project team has produced.
The degree to which a hypertext database appears hierarchical is due to
its design, rather than a property of the medium itself.
-- Al Selvin Decision Support Systems Group NYNEX Science & Technology selvin@nynexst.com