I'm writing another article for Cutter about the changing definition of what a document is. As a consequence, I've been thinking a lot about metaphors and how we use them in computer science and content management. I think it can safely be said that the desktop metaphor first popularized by Apple Computer, and later adopted by Microsoft, made computers more accessible to more people than any other single innovation. Instead of command prompts, directories, files and "rm" commands, the user was confronted with a familiar desktop, fitted out with folders, documents and trashcans.

When I was a child, I thought that the flapping of the leaves of the trees in my backyard was what caused the wind to blow. Each leaf, held aloft by sturdy branches, waved in the air (the same way that church bulletins were fanned by my mother on steamy Sunday mornings) and generated the delightful golden breeze. Of course, I had made a fundamental (and all too common) error: I mistook the consequence for the cause. I know now that it is the breeze itself that causes the leaves to flap in the wind, and not the other way around. Had I been a more astute kindergartner, I would have limited my conclusion to the much more precise observation that there was clearly a high degree of correlation between the presence of a breeze and the flapping of the leaves of trees, but that any assertions of causation were premature.

To this day, I still like to sit on my front porch and pretend to myself that it is the flapping leaves that drive the wind. I know it's not true, but I like the idea because of the feeling it evokes in me. It is meaningful to me because it recalls a time when the world was alive to me, when it was life itself emerging out of the verdent ground that was the animus that moved the wind and stirred my imagination. It serves to describe not the way world is for me; rather it describes the way the world feels to me.

This is the purpose of metaphor, to use one idea in place of another because it makes it easier to understand, or evokes a sense that would otherwise be difficult to understand using more accurate, yet abstract language. Hence the desktop metaphor. There aren't really documents sitting in folders on your computer, but it certainly helps to pretend that there are.

Having an effective way to think about a computer is more important to the end user than actually understanding what is actually going on inside with all those electrons. The desktop metaphor, and the idea of a document certainly helped us to think about computers, but this is a metaphor that is beginning to show signs of strain. It's a little less effective at describing the kind of communication (and encoding of data) that takes place now in this Web-2.0-ish world. It used to be that applications created documents; now documents are beginning to look much more like applications (likewise, our creative outlets now tend to assist in the discovery of knowledge as much, if not more so than serving as merely a representation of knowledge already gained. We have a document object model (DOM) that is scriptable). Certainly a scriptable document is something new?

The document metaphor pervades computing. Web sites have pages; browsers are shaped like pages, PDF documents viewed on the screen are formatted as pages (I hate having to print things out to read them, but I often have to because of this excessive focus on printed output). Microsoft Word documents are documents, even though they have made efforts to add functionality that make them more computer-screen friendly for reading. The most problematic issue, though, is that word processors focus too much on formatting and layout and are ideologically tied to the idea that what you see on the screen should look like what will be sent to the printer. In a world of smartphones, PDAs, laptop, desktops, wall-mounted plasma screens and print-on-demand, What You See Cannot Possibly Be What You Get Because You Don't Know What Device It Will Be Viewed On (WYSCPBWYGBYDKWDIWB).