Web sites as architectural spaces

Matt Jones discusses "A web site is a public place", an article in Ignition Design's Journal of Design Science which compares the environment of Web sites with architectural spaces. In the article, they say, A Web site is like any other public place where people come to look, to learn, to search, or to experience. And as in any public place, visitors will succeed in what they came to do only if the site gives them a clear indication of * where they are * where they can go * what they will find there. A good navigation system uses the site’s information structure as the basis for a visual hierarchy that guides the user experience. ... There is no one “right” path through an interactive product. Ideally, viewers should be able to call up information in whatever sequence seems logical to them. A product that’s easy to use will accommodate any reasonable request at any time. On BlackBeltJones, Matt ponders: If there's one problem I've struggled with over and over and over, then it's the old chestnut of whether a piece of content or information* should have:

  1. a single 'location' in the structure of a website,
  2. or have a single 'location', but multiple access routes,
  3. or be able to be accessed from *anywhere* and have no 'location'.
Does the child have many parents, or just one (but perhaps lots of friendly aunts and uncles...!) In digital environments I think it probably helps to reinforce a sense of position because as we experience life in the physical world, we often rely on cognitively mapping objects to their respective places in order to recall their position later. The concept of creating an information architecture that's built around a category/topic/subject hierarchy with a system of cross-references and thesaural term management has great advantages for aiding the information seeker. I think, most importantly it aids in navigating an information-use environment by:
  1. allowing cognitive mapping of the site's domain of content which aids in understanding the scope of content on the site and in the elimination of unwanted search paths
  2. allowing browsing, which makes possible serendipitious discovery of related documents within a category/topic
I think Yahoo! does it right with the way they treat categories. In their category listings, they list categories with an @ symbol that do not reside in that branch of the topic hierarchy. It's called a cross-reference, and it's a concept borrowed from information science. Take a look, for example at their Computers and Internet > News and Media section and you'll see that at that level they list several subcategories you can drill down into, and several categories that reside in other parts of the site's hierarchy. I believe that, especially in large information collections, hierarchical systems, where each entity is classified under one node, continue to be sucessful as effective systems for navigating content. However, there are other ways to make use of indexing methods so that hierarchical browsing isn't the only way to find an entity. For instance, I am still compelled by the concept of using Ranganathan's Facet Analysis as an alternate method for information discovery. I think most IA's would want to see alternative methods of navigating content used successfully and repeatedly before considering them, but there is a great opportunity now to add to the user experience by adding new models of navigation to the Web. In meat-space, this would be like inventing the elevator in a world where only stairs exist.