, , , ,

Navigation is a central aspect of interaction design on the web. If the navigation of a site is bad, the user won’t find his content and experiences a bad product. Luckily there are several proven tools that help the user to find his way on the web: the top menu, the sub menu, and the breadcrumb.

But there is one aspect that interaction design has not tackled yet: how navigation works outside of the web in the real world. Or better yet, how the brain works with regard to navigation. The brain switches to “navigation mode” when it has to travel through an environment that is larger than the direct perception. That makes sense, because if you have can see the destination of your travel, you can just navigate on sight. For example, if you want to travel from one side of a open field to another side of the field, you can see your destination and are able to continually adjust your movement towards it.


But if you can not see your destination, you can not travel on sight anymore. For example, in a maze you can’t rely on sight anymore, but you have to rely on your conception of the maze in order to travel through it. This is when the brain’s “navigation mode” kicks in: it no longer relies on visual data but rather on a mental model of the environment.

Our brain uses two strategies when constructing and using this mental model: the route and the map. The route is a mental model in which the travel from point A to point B is stored as a sequence of directions. In the maze example, this would result in a route like this: right, left, left, right, straight through, right, left, goal. In real life, this sequence is often extended with references to landmarks: take the first left turn after the white building with the red sign, then go straight through until you come at a sigar store, and then walk to the right about 50 fifty meters.

This strategy has its disadvantages: when one or more landmarks disappear suddenly, the route doesn’t work anymore. And the routes is also weak in travelling back from point B to point A: sequences such as “walk until you come at a sigar store” will only work in one direction.

The route is fairly efficient when you only have to remember one or two routes in an enviroment. But if you have to travel accross an environment from all sorts to places then it is more efficient to construct a mental map. A mental map is more than the sum of all the routes. If you rely on routes, you may know how to come from point A to point B, and from point A to point C. But if you have to go from point B to point C, you will first have to go back to point A. If you have a mental map of the environment, you are able to go from B to C, even if you haven’t travelled this route before. That is because a mental map is an objective model of locations and their relations, and a mental route is a subjective model of the traveller and his environment.

How to apply this knowledge to the web? The web is also a kind of space, but not the same as real space. We speak of websites, home pages, disk space, information high ways, and sitemaps, but how does virtual space relate to physical space? The most important difference is that in web space, any page must be immediately accessible from any other page. Hyperlinks link up all the web pages together in a hyperspace. So in this situation it is impossible to construct a reliable mental map of the website, because the pages have no objective relation to each other. So basically all you can do is construct a large collections of routes of the web space.

The current navigation tools are all designed for this strategy. The breadcrumb is a prototypical routefinding tool, because you can see the steps you have made from the starting point of the site. This starting point is always point A, and you always have to come back to point A. There is no route directly from point to H or K. The top menu and sub menu are designed to skip going back all the way to A, but only to the step directly after A in case of the top menu, and the step directly after the top menu in case of the sub menu.

This works fine for most internet websites. Most websites are relatively small with regard to content pages, so they can apply certain restrictions. The best restriction is having so little pages that they could all fit in the menu structure. That way the user could navigate on site, and doesn’t have to construct a mental model at all. Another restriction is not being more than 2 steps deep from the root page. This ensures that the routes to content items are short, and easy to remember.

But for the larger website, this causes problems. If (semi-)governmental agencies are trying to use the web (extranet) for offering a number of services like tax payment or social security services, they have to build large and complex content systems, that are too large for the above-mentioned restrictions. And moreover, these are websites that the user needs to visit and return to, and not websites that the user surfs by for fun and leaves without the obligation to revisit. For these kind of websites, the standard navigation tools are not sufficient because they are oriented towards mental routes.

Also most intranet sites have these problems. They just contain to much information the oversee in one glance, and the routes to the desired content is too long. In these case it could be worthwhile to explore the possibilities of mental map oriented web navigation. This is a kind of navigation in which the user has gained experience of the spatial structure of the website, and is able to go from one place of content to another on the basis of his expectation instead of having to go the startpoint and begin a search quest all over again.

There are not much concrete implementations yet of this kind of web navigation, but taking a look at how Prezi works could be inspiring. Also, videogames have been addressing the problem of navigation for years. Like websites, videogames consist of a different spatial structure than real life. Videogames could also have hyperspace-like feature, with warp holes and “skipping over” large parts of the environment. But they are more oriented toward the player as a returning subject to the space, and all kinds of smaller and larger gamemaps are used to provide the player with extra navigation tools. Not that websites have to be like videogames, and the internet has to become one large Second-Life-like world, but navigation designers on the web have to be aware that there is more than one cognitive navigation strategy, and that the cognitive route has its limitations.