Week 10: Readings and Comments

by Zach Tomaszewski

for LIS 699-2, Fall 2001, supervised by Dr. Luz Quiroga

This week was Chapter Four of Wayfinding in Architecture. This book just keeps getting better!

The first part of this chapter concerned signs. It defines three types: directional (pointing to a destination), identifying (identifying a location), and reassuring (providing information about the current path). There should (and often is) a different form or graphic identity or even location for wayfinding signs than there is for commercial signs. And this form should then be consistently enforced. Signs are skimmed by hurrying users and so need to be brief; yet they have to be complete and ambiguous (this is no easy task). A sign's intended audience (e.g. drivers verses pedestrians) needs to be clear. At the very least, a sign should make clear whether it is directional, identifying, or reassuring in nature.

This discussion of signs reminds me of labeling in a navigation system. Navigation is all directional signs--how to get to another location. Some sites have identifying signs, such as a "bread crumb trail" or navigational path stated somewhere at the top or bottom of a webpage. And page titles and headings may be considered identifying signs. But there are even fewer reassuring signs on the web than there are offline. Yet since there is very little transit between pages (you're either on one page or another and rarely on a path between), this is not surprising. Still, I think we should try to include more reassuring signs were possible and appropriate. (The title attribute of links might be considered a reassuring sign--it appears a second or so after a user hovers her mouse over the link and provides further information. After reading Nielsen, it might be that web users need reassurance before they click a link!)

The second part of the chapter was on architecture, space, and how people form a mental image of their environment. There are two factors to this: legibility and imageability. Legibility is the ease with which one can get information from the environment. Imageability is how easily that information can be formed into a useful image. A dark, empty room has low legibility, but once the environmental information is accumulated (by feeling around in the dark), imageability is pretty high (there are only 4 walls and a floor).

Passini examines the work of Kevin Lynch on the imageability of cityscapes; he finds similar parallels in his own work regarding buildings. There are 5 features that seem to be used in forming an image of a space: paths (the channels of transportation, such as roads and corridors), landmarks (physical reference points, such as buildings or certain shops in a mall), nodes (entry points and important foci, such as squares and intersections), edges (linear elements not used as paths; very often serve as boundaries, such as coastlines or walls), and districts (large areas that share a recognizable flavor or distinction). Interestingly, Passini does not feel that imageability can be increased by emphasizing these features, which one notable exception: landmarks. While people can operate well when most of these 5 features are sparse or nonexistent, they complain when there are no landmark. Landmarks seem to anchor their cognitive maps to the world. (For example, the 3rd floor of POST has well-defined corridors, walls, and districts such as classrooms, offices, labs; yet it lacks distinguishing landmarks. I have adapted and begun to use room numbers and bulletin board content as landmarks.)

So, aside from improving landmarks, how do we increase imageability if not through these 5 environmental features. Passini identifies three other aspects. The first is spatial organization. Making clear the order behind the architecture is very important. Large, multi-floor spaces are very helpful in this. He also examines city street layouts, such as the American grid, the medieval circle, and the Islamic hierarchy. Attempting to navigate these streets without understanding the basic, underlying organizational scheme is incredibly confusing.

The second aspect is spatial enclosure. This forms boundaries that define the space and domain. Also, the externals give a clue to the internals of a space. For example, a rectangular building with a triangular internal pathway was extremely disorienting, except for the minority that realized the inside was based on a triangle.

Finally, spatial correspondence is important. This means anchoring the space to the surrounding environment or connecting spaces. For example, in most buildings, people have difficulty relating to directions of the city outside. (Definitely a problem for me with POST; I frequently come out on the wrong side of the building and have to walk around.) Also, they have difficulty combining models of separate wings or areas in the same building if they do not clearly correspond. (I still have this problem with the new Hamilton Library layout.)

Applying all this to a Web environment is particularity interesting. Paths likely correspond to links. (This is not an entirely satisfactory translation though, since links have no transit length; if each webpage is a room and each link is a door to another room, there are no hallways between them, unless you count the page load time.) After some thought, I think each webpage may be considered a node. Again, it's not an entirely satisfactory translation. But it seems clear with a main page, with many links to and from, this is a central area from which to base wayfinding operations. It's rather like a food court in a mall. Hopefully a solid, consistent site identity provide edges and some spatial enclosure. Some sites have subsite districts. Page titles, images, and content are probably the closest things to landmarks--something that stands out as a recognizable object or location. (I can't decide whether pages are nodes or landmarks, but I think this may be due to the fact that the two factors can overlap in the real world, one being an important space and one being an important object.)

Spatial organization, probably the most important aspect of wayfinding for Passini, seems to correspond beautifully with Rosenfeld's notion of organizational schemes and structures. It may help to think of websites as spaces/rooms/landmarks/pages with paths/doors/links between them. How do you identify landmarks to someone who hasn't seen them yet, and how do you make the relationships and paths between them understandable. Sites are certainly more like buildings than cities--following a link is not like driving down a road, where you get a chance to look around and see your relation to other landmarks. Rather, you're in one room at a time and can see only the (well-marked) doors leading out of it.

Maps, the last major part of this chapter, examine how to convey information for both decision-making and comprehending a space. When using a map, a person has to understand how it related to reality, where he is on the map, where his destination is on the map, and how to get there. Passini's description of map users reminds me of all the usability reading I've done: people glancing at information, refusing to slow down or give their full attention, with astonishingly short attention spans. Maps have to abstract and provide all the important information for wayfinding (organizational structure, available paths, notable landmarks, etc.) without adding anything extraneous and without overwhelming or confusing the impatient user. A tall task indeed.

This chapter concluded with a short section on verbal directions and information booths. I have heard of a couple e-commerce sites including such a feature.

I believe the next and final chapter will provide some clues as to how to go about this.

Post-discussion Issues:

To Week 12 →