When I was in college a professor told us a story about an architect (Developer) who would build all of his buildings, but put down no sidewalks. He would just plant grass. 6 months later he would come back and put sidewalks down where all the paths were worn. In this way, he assured that the walks would be where the people were mostly likely to walk. The point of the story was that we should observe how people do whatever it is we are trying to model in software and then build it to work that way, thus creating “user friendly” software.
Is this an urban legend? Anyone else know the story? Anyone know who the developer or architect was?
I really doubt that story is true, at least pertaining to UCI. There are two sets of cocentric circles which don’t seem like something cows would do. UCI was a planned campus, so it is better laid out than most schools are. You can see a PDF map of the campus here.
The story has been around for decades and is so logical that it’s in books on design. I can’t provide any specifics but it would stun me if it weren’t true.
I know places that have retrofitted their sidewalks to cover over the bare lanes made by people cutting through the grass, so by now it’s undoubtedly occurred to enough people to try this from the beginning that it happens regularly.
A co-worker a few years ago told the same “Pave the paths!” story. She left shortly afterward to become president of a new college in Winnipeg (IIRC), and had been studying design techniques for quite a while, so she might have a better cite. Or she might have been listening to the same professor.
IIRC: It appears in Douglas Normans “Design of Everyday Things”, commonly referred to as the bible of Computer Interface design. You might trying checking that to see if you can find the original source.
It seems like an inefficient way to design sidewalks to me. Really, it’s not that hard to figure out where to put a sidewalk. If people make paths off of an existing sidewalk it is because the sidewalk is in the wrong spot. No thought was given to where people will actually walk. Chances are they picked the cheapest or most ascetically pleasing spot for the sidewalk. If you care about matching the placement of the sidewalk with traffic patterns, then it’s not such a big challenge to predict where people are going to actually walk.
To get from A to B across a lawn, people will take the shortest route. You don’t need to wait for the path to appear to tell you where that is. If it really isn’t obvious where that will be than I doubt a clear path will form in the future anyway to help you out.
You’re making big assumptions here. If A and B are on opposite corners of a square, then fine. But if they’re at the mid-points of adjacent sides of the square, the ‘average’ route chosen will quite possibly be an arc. Individual humans may be rational, but the non-concious choice of route isn’t a rational decision. Plus, you may need to design paths on a far more complex layout than a simple geometric shape, and you may also have points C, D, and E through Z to deal with.
Yes, if the layout is complex, then the resulting paths probably will be too. In other words; all over the place. So you won’t be able to match the sidewalk to the existing paths and still keep the amount of sidewalk you build to a reasonable level.
If an obvious pattern forms, than I feel a little thought would allow you to predict it beforehand. You don’t have to be perfect. Only if you get it drastically wrong will people ignore the sidewalk and still walk on the grass. This would only happen if you haven’t given any thought at all to possible traffic patterns.
There is a cost to designing sidewalks this way. The construction crew has to return a year or two later to complete the job. This is more expensive than just doing it immediately. Also you have to put up with a year or two of no sidewalks, with the resulting dirt and mud being tracked through the buildings. I wouldn’t want to be the architect (sidewalk architect?) who has to explain this inconvenience by saying “I don’t know where to put the sidewalks until I see where people actually walk.” It makes him look like he doesn’t know what he is doing.
I’m sure many design decisions are made during the construction of the buildings that make where to put the sidewalks, look pretty simple by comparison.
Since people pretty quickly figure out where and who they want to go, you could probably delay it by only a week or two on a college campus. Actually, one problem I see all the time is corners. People don’t walk on corners. They don’t make right-angle turns, given the choice. So making sidewalks with right angles in the midle of nowhere is weird and usually winds up with a chewed-up bit of grass,
I lived in the neighborhood when UC Irvine was being built. It was a completely planned construction. The interviewee in the article you linked to may have been thinking about UC Santa Cruz, which utilized this natural planning concept for some areas.
Your missing one subtle point: paths which are cleared are slightly easy to walk over than uncleared land. This means the “optimal” path not only depend on where you are and where you want to go, but also where every single person before you has wanted to go. I’m not sure whether the observation lead to the analysis or the analysis lead to the observation but this is actually a hugh, non-trivial area of research in mathematics with many other practical applications.