From the Boidem - 
an occasional column on computers and information technologies in everyday life

January 28, 2006*: Dr. Hierarchy and Mr. Associative

Over the years (again?) I've bemoaned the retreat of the web from what would seem to me to be it's logical associative "structure". I've gone as far as admitting that my vision of the web has become little more than an historical artifact. Sometimes, after a long day (or week) of encountering only straight and narrow links, I'm ready to admit defeat. That being said, I also have to admit that there are times when I see the hierarchical use of hypertext as totally legitimate. And not just legitimate - I often use hierarchical hypertext myself.

Cyberspace can be confusing. I've seen novices close to tears when they first encounter Yahoo!'s main page (or the directory, which seems, each time I click over to it to take a glimpse, to take up less and less space on their main page). Why would I ever want so many links, they seem to ask. And I suppose that we should admit the truth - we really don't need them all. Nobody, after all, is really expected to cover all of cyberspace. Each of us has his or her own small corner, and if we get to know it well, we're satisfied. And, strange as it may seem upon first glance, most of Yahoo!, and particularly the more traditional Yahoo! directory, is purposefully hierarchical. The entire idea behind a catalog structure is that we'll be able to find something within a given, predefined, and yes, there's that word again, logical, framework. Precisely because it really is so easy to get lost, navigational aids (somehow, the phrase itself seems to suggest, perhaps even assume, the existence of a hierarchical structure) are desirable.

A number of attempts have been made to make it easier for the web surfer to navigate the reading experience. It's nice to know that some web designers see this as an important service, though Steven Johnson claims that the fact that advanced navigational tools aren't built into the browser is a major shortcoming of web browser design. John December, whom I've read pretty much since I first started reading on the web, designed a number of simple, and basic, buttons that he pastes into his texts after links. These buttons identify for us what sort of link we've encountered, so that we can decide, before clicking, whether we want to follow it. But it's not only a question of the destinations to which various links lead us, but also of the sort of reading experience that various navigational tools imply.

Mark Bernstein in Patterns of Hypertext suggests numerous possible narrative/navigational patterns. A number of these (cycle, contour, counterpoint, tangle, sieve) he illustrates with pleasant graphics. Bernstein's goal isn't to outfit us with aids that help us identify where a link will lead us. Instead, he seeks to classify or categorize various hypertextual narrative playing fields and in that way help us to understand the different functions that a link can perform. He doesn't juxtapose associative and hierarchical structures (he only mentions hierarchies when discussing sieves) in order to choose sides, but to help us become more aware as readers. In a manner somewhat similar to Bernstein, Nicholas C. Burbules in Rhetorics of the Web: Hyperreading and Critical Literacy suggests various functions that links in a text can perform, functions that may at first be taken for granted but that critical readers should take note of. Among Burbules' categories we find: metaphor, metonymy, synecdoche, hyperbole, antistasis and more. Burbules' paper is from 1997, and it's my guess that I first read it back when it was new. For some reason I remembered that, like Bernstein, his paper had graphic representations (I learned long ago not to expect links) of the different functions that links fulfill, but I discovered that I was mistaken.

Somewhere along the line, basically with the rise of the WWW, associative linking won out over hierarchic linking. Perhaps it wasn't the associative linking that Bernstein and Burbules and others theorized about, but it was certainly different from the menu-based systems that had been the standards of an earlier internet. As a few commentators have noted, the now almost totally defunct Gopher systems did an excellent job of hierarchical organizing. Though the vast majority of internet users have probably never heard the name Gopher, there are still some people who long for those good old days. But the sheer simplicity of associative hypertext, our ability to create a link within a document so that it jumps out at the reader (or into his or her mind) precisely where we want it do, defeated Gopher hands down. As has been proven true in numerous other contexts, too much success can often lead to failure. Since that stunning victory, a victory that grew out of the obviousness of associative linking within the web, that same associative hypertext has been in retreat. As though it was embarrassed by such a complete victory, associative linking has since then been busy apologizing, has been limiting itself to "see also", and "reference" links. To a large extent it has abandoned the central chunks of text that can be found on a web page and has instead retreated into the margins, and has allowed hierarchical menus to restake a claim on the page.

As should have become obvious even well before this point in this column, I let my associations overtake me this time. More than overtaking me, they seem to have taken over the entire column. I seem to recall that I my original intent was to examine how my preparing of web-based information is a constant struggle between two different and distinct approaches toward linking - the associative and the hierarchical. But it appears that my associative side wasn't interested in coexistence, and it pushed the hierarchical side almost totally out of the picture. That being the case, it seems rather obvious why I chose the associative aspect of my linking personality as the doctor in my title.

That's it for this edition. Reactions and suggestions can be sent to:

Jay Hurvitz

back to the Boidem Contents Page

Return to Luftmentsh