Sunday, November 5, 2017

Tracking “Xanadu” around the web

IMGP6920rd

I'm gearing up to write a review of Franco Moretti's latest; well, actually, not just Moretti, but Moretti plus thirteen others: Mark Algee-Hewitt, Sarah Allison, Marissa Gemma, Ryan Heuser, Matthew Jockers, Holst Katsma, Long Le-Khac, Dominique Pestre, Erik Steiner, Amir Tevel, Hannah Walser, Michael Witmore, Irena Yamboliev (notice the alphabetic ordering). N+1 is putting out a collection of papers from the Stanford Literary Lab, Canon/Archive: Studies in Quantitative Formalism, and I'll be reviewing it for 3 Quarks Daily.

I first learned about Moretti just over a decade ago when I was writing for The Valve. We had a "book event" organized around his Graphs, Maps, Trees (2007). A number of people contributed posts and Moretti responded. I contributed two posts, a short one, and a somewhat longer one, One Candle, a Thousand Points of Light: Moretti and the Individual Text. The format's gone bizarro and there's a lot of link rot, but the discussion was long and interested. What I did was google the term "xanadu" and investigated the fascinating results. Who would have thought that such an exotic word would get two to three million hits? That was 2006; now it gets almost 10 million hits.

Anyhow, sometime later I took that old post, refined it a bit, and published it as a working paper: One Candle, a Thousand Points of Light: The Xanadu Meme. You can download it here:
It’s rather different in spirit and technique from the text mining/corpus linguistics techniques that are currently raising the bar in computational humanities. But it is computationally intensive. It’s just that the intensity is not on my desktop, it’s in Google’s servers. I used Google to troll the web for “Xanadu” and identified two contexts, other than the Coleridge’s poem, where I got a lot of hits.

Here’s the abstract:
I treat a single word 'xanadu', as a 'meme' and follow it from a 17th century book, to a 19th century poem (Coleridge's "Kubla Khan"), into the 20th century where it was picked up by a classic movie ("Citizen Kane"), an ongoing software development project (Ted Nelson's Project Xanadu), and another movie and hit song, Olivia Newton-John's Xanadu. The aggregate result can be seen when you google the word, you get 6 million hits. What is interesting about those hits is that, while some of them are directly related to Coleridge's poem, more seem to be related to Nelson's software project, Olivia Newton-John's film and song, and (indirectly) to Welles' movie. Thus one cluster of Xanadu sites is high tech while another is about luxury and excess (and then there's the Manchester Swingers Club Xanadu).
We can call the first cluster of sites the cybernetic context while the second is the sybaritic context. This simple diagram shows the evolution:

Xanadu Cladogram

The Citizen Kane line is, of course, the sybaritic one, while the Ted Nelson line is the cybernetic one.

My point is simple and, I hope, obvious. That this one term, “Xanadu,” has taken on two different valences, which are active in two different contexts. The cybernetic doesn’t seem to exist in Coleridge’s text at all. Ted Nelson created that when he provided a new context for the poem by making it the inspiration for his hypertext system. Though it wasn’t the poem that was the context as much as it was Coleridge’s alleged inability to remember the whole thing. Nelson imagined his hypertext system as one where nothing would ever get lost.

The sybaritic context, however, is there in the poem—“pleasure dome,” “demon lover,” “damsel with a dulcimer” and so on. Welles’ film, in effect, stripped the rest of the poem away and then magnified the cultural presence sybaritic element by putting it on the motion picture screen. If we examine the historical record more closely, though, we have hints that that job had started before Welles made his film. Prior to 1940 The New York Times, for example, makes mentions a yacht, or perhaps two, named “Xanadu.”

Now, imagine contexts for thousands of words in thousands of documents. That’s what is examined in topic analysis. Documents provide contexts for topics, and topics provide contexts for works. Algorithms for topic analysis examine words in large collections of documents and infer the topics in which they belong.

* * * * *

Here's the table of contents:
Introduction: “Xanadu”— 2
Googling for Memes — 3
A Thousand Points of Light, a Metaphor — 4
Xanadu: A View from the Wikipedia — 5
Xanadu: A Google View. — 7
Examining the Xanadu System — 12
Beyond the Meme — 18
Beyond Interpretation — 19
Appendix 1: Googling Oedipus — 19
Appendix 2: Xanadu in Google Books — 21
References — 23

No comments:

Post a Comment