Scott McCloud, in his brilliant Understanding Comics (which I recommend for any artist, regardless of accustomed media) talks about detail and audience identification. The image that he uses as himself is deliberately stylized and information-poor specifically because it allows the reader to mentally redraw it as her- or himself.
The lack of detail invites the reader to become the character. That is a principle that I have adapted to my own work in several ways, but most consciously in my description of the settings for my stories.
Where does Catskinner’s Book take place? Well, if you are familiar with the St. Louis, Missouri metropolitan area you can recognize enough landmarks to figure out most of the locales (which I usually based on existing places) but I don’t ever come out and say that’s where James lives.
That’s deliberate. What I do say about the city is that there is a river along side it and that it has a botanical garden, some long-term resident hotels near the airport, a street called Manchester, a street called Lindbergh, and a baseball team that has a stadium downtown. I mention a city sales tax and the fact that there is an interstate outer-belt. That’s about it.
I suspect that those facts apply to quite a few cities in the US, although I haven’t set out to research the issue.
I did not want my readers having to stop and try to imagine the places where the action was taking place. I wanted them to simply watch the show. If a reader was familiar with Memphis or Cincinnati or Oklahoma City or any of a dozen other Midwestern towns then she or he would have all of the “stock backgrounds” to plug my characters into.
This is why I avoid mentioning specific store names. I say that James is going through a drive-thru window at a fast food place, but I don’t specify “McDonald’s” or “Jack In The Box”. That’s not important to the story and I don’t want you to have to spend mental energy on imagining the details of unimportant things. Whatever image comes first to your mind when you read the words “fast food drive-thru” is fine for narrative purposes. It is almost certainly not the same image that I had when I wrote those words, but it doesn’t have to be.
In Cannibal Hearts I do specify that the city is St. Louis because I have the whole riverboat thing and it became too tiresome to keep avoiding specificity. I kept the general principle, though. I mention a few architectural details on some of the buildings–Olde Towne Tower and Stone Edge Medical Plaza–but I tried to use design motifs that were generic enough that most readers could readily recall a building that fit the general criteria.
Most places are just thumbnail sketches. A machine shop. A clothing store. A converted fast food restaurant. Again, your mental image isn’t going to be the same as mine, but there is no reason that it should be. I don’t want you focused on the store, I want you to be thinking about James and what is going to happen to him next.
Where details are important–for example, the space-warping interior of The Good Earth–I put the critical points in early. In The Worms Of Heaven the Orchid’s labyrinth is described in some detail because the nature of the space is important to both the mood and the mechanics of the story. The same with Village Green and Zenith in Gingerbread Wolves–the altered towns are themselves characters and in those cases I want to specify the mental image in the reader’s mind–but even then I am trying to work within the reader’s existing set of mental images, just mix them up some.
It sounds paradoxical, but I believe that a studied vagueness (what Douglas Adams called “rigorously defined areas of doubt and uncertainty”) makes things more real for the reader. Because I don’t want you thinking, “This is happening just down the street from Misha’s house”, I want you thinking, “This is happening right down the street from my house.”
I want to tell you just enough so that you can take my story and bring it home.