Dan Cohen's #DigDil09 Puzzle, or The Crowd versus Search
There is nothing like a game to get folks interested; so when Dan Cohen, as part of his talk "Scholars and the Everywhere Library" at the Digital Dilemmas Symposium today, proposed a game, and I was free (read: should have been working on the dissertation), I couldn't resist. Details are available on his blog. But those who visit the link will be presented with an image of what looks like a piece of handmade, Native American jewellery, a brief bit of context (crucial for those of us who will use search to solve this mystery), and the question, "So: What is this?" I was, of course, not attending #DigDil09, and so did not hear the talk, for which this game was an illustration. To judge from the live-tweets, part of the goal here was to illustrate the benefits of crowd-sourcing over "static machine search." Or, as Cohen puts it:blog comments powered by Disqus
Confused about what this was, the anthropologist brought the object back and presented it [to his] colleagues. I would like to reproduce that activity digitally by presenting the object online, to see what readers of this blog and my followers on Twitter can make of it, individually and by talking to each other.According to this model, Twitter would reproduce the experience of a scholarly meeting on a massive, immediate, and globally distributed scale. Such a mode of scholarly interaction would be quantitatively different from the meeting of a scholarly society, but not qualitatively so. (Of course, when making such distinctions, go in fear of Hegel.) In some sense, then, it was crowd-sourcing versus search (or, Twitter versus Google). That this separation between the two modes was intended seems manifest from @dancohen's tweet after the game:
Note on solving the mystery: I was hoping for tweets more like @chandlerhora's http://is.gd/sP0Y than googling for it. #digdil09The tweet from @chandlerhora, reads:
@dancohen agree with @jamescalder looks like a spider image on a gorget, likely cherokee? I remember seeing spider before #digdil09And some did play the game this way: consider this tweet or this one. They represent the "pure" crowd-sourcing model--smart people putting their heads together to answer a question. For my own part, I was participating from home, and the first thing I did was fire up Google, and use the search terms culled from the original post to find a reference to this object on-line. I think I started with:
1882 "Clair County" Illinois shellI wasn't the only one. As the 28ish minutes it took for the puzzle to be solved passed, people traded links via twitter to books and museum archives--search was the clearly the dominant strategy. To the extent that the game was meant to demonstrate the power of crowdsourcing, I wonder if it didn't demonstrate something else. Imagine another version of the game, with exactly the same starting point. But this time, let's do it with teams; on one team, a single person with full-text search (Google, Google Books, archive.org, and so on); on the other a team of folks who can communicate with each other but who can't use Google (a unlikely scenario indeed!). Who would give you the better result after an hour? It all depends, of course, on who these folks are, and so and so on. But if I had to put money on it, I'd probably bet on full text search. Full text search, it seems to me, is really the game changer here. Of course, full text is the game changer if the item we're investigating is indexed or archived somewhere. I wonder--how did Dan Cohen find this image, and this example, for the game in the first place? Did he google around, go to Google Books, archive.org, or some other digital archive? I ask, because, of course, something one finds on-line, is necessarily findable on-line. There remain, of course, plenty of undigitized objects of scholarly inquiry, which might thwart the search strategy handily. Which of course only reinforces the point that people have been making for so long: as the human record is being digitized, it is imperative that scholars be involved, so that the record is as rich and complete as possible. In any case, great fun all around. Thanks again to Dan Cohen for a stimulating (and educational!) distraction from diss work.