Why palimpsest?

In July 1999, a kind person who will remain anonymous as I've not asked her permission to drag her into this, wrote:

Not only am I interested in the history and statistics of the site, I am curious as to how/why the name "palimpsest" was chosen. I went to the glossary and looked this term up to make sure that I understood it, and now I'd like to know about its ties to the site.

The cheap answer is that I've always been drawn--before I ever got seriously involved with conservation--to the notion of palimpsest both as metonym and as real object, carrying traces of its own history as signs, scars, accretions, adornment. The palimpsest is the palpable, deducible mystery, exposing itself to the reasoning of the detective, whether Sherlock or scientist, and as such is an analog of all texts, which carry with them the traces of their past, their dialogue with past texts, with the history of language, (what shortly after my time in school came to be called intertextuality). For reasons I no longer recall (my own text having been overwritten so many times since), the notion of palimpsest became linked in my mind, many years ago, with that of periplum, early navigational texts in which the mapping of place is given as instructions for a voyage of discovery, and a mapping between palimpsest and periplum still seems to me apt, as the vertical (deep, down into layers) and horizontal (wide, exploring outward, sideways through the physical or virtual world) axes by which knowledge and understanding are navigated. For anyone way too curious about this, a film I did in the 1970's, named palimpsest periplum is still available from Canyon Cinema. It probably will shed considerably less light on the topic than the above paragraph, (assuming, most overoptimistically, that it can shed light at all, that it hasn't succumbed to the thousand natural shocks that film is heir to).

So, when the time came to name a machine, I think I think I chose palimpsest half without thinking, as one might name something after an old friend or fondly remembered uncle. It seemed nevertheless, and continues to do, entirely the right thing to call a digital environment, whose contents are overwritten continually, sometimes even continuously, one layer overlaying the next, over, and over. That the traces of these obliterated texts can be as difficult to resurrect as their medieval antecedents is brought to mind almost daily. CoOL itself has always been a palimpsest of sorts, beginning as a collection of WAIS databases, then being overlaid with Gopher then Web servers and inevitably, with time, will gain another crust, with whatever wonderful new means of organizing and navigating information uncovers itself. Traces of WAIS and gopher are still deducible in CoOL, if one looks hard enough.

I do have regrets, and often wish I could erase the name if I could, and cover it with something easier to spell (my phone calls would be a lot shorter without all that "m as in michael, p as in peter" business). But I dissemble, perhaps to gain your sympathy and patience (if you're on the other end of that phone); in truth, I'd not change the name, and if there were ever a second machine, would probably name it pentimento, which is a sure sign that I've not repented at all.