# Posted 3:41 PM by Patrick Belton  

FREE STUFF! FREE STUFF! I've been an ardent supporter, for some time, of all attempts to create online libraries across subjects on the internet which are vast, useful, and free to access. In this vein, I've been happy to expend praise on Project Gutenberg and Wikipedia's online encyclopaedia, on Bartleby's compilation of reference works and the Harvard six-foot-shelf, and on attempts by H-net to create academically worthwhile communities of research and scholarship conducted entirely online. (Sneak peak: we're incidentally in the process right now of beginning a listserv there for research on democratization and the promotion of rule of law and human rights, to be called, surprisingly, H-Democracy...)

I've only recently discovered Wikisource, brought to you by the same people who created Wikipedia. Both are free to access, and both, significantly, are completely collaborative efforts, with no editing other than that provided by other users, and depending entirely upon the contributions of volunteers. So, you'd obviously think it'd suck, wouldn't you? Who'd ever realistically think you could leave a blank sheet somewhere accessible to everyone in the world, invite them to write an encyclopaedia on it, and it would end up filled with anything other than graffiti at best, and hate-filled diatribes in a more likely case? Who indeed.

On Wiksource, I've recently been perusing night time reading (yes, Josh) from the Qur'an, from the Iliad and Odyssey (Butler translations both, with Greek available as well), and from the complete Shakespearean corpus. All of Leaves of Grass is on there, along with every state of the union and inaugural address, and several Upanishads. As a new project there are lots of blank spaces, but if you hit one, you just might be able to do something to fill it.....

UPDATE - AND STILL MORE FREE STUFF!: Bo Cowgill by email points out something I hadn't yet known existed - Google's new portal, in beta testing, for academic research at http://scholar.google.com. So if you've gotten so accustomed to reading Professors Drezner, Volokh, and Reynolds online that it seems wrong to read their academic work in print, now's your opportunity!

Also, our friend MB points to a TCS piece by former Britannica editor Robert McHenry, who argues that regression to the mean should apply for Wikipedia. Interesting argument, and I think it bears research. I'm not sure he's right, though, because: (a) regression to the mean involves random distributions, whereas wikipedia writing is an intentional activity; (b) even construed as such, prior versions aren't deleted, but are available for resurrection by subsequent editors if intervening instances constituted retrogression, therefore the relevant probabilistic question becomes whether a retrograding editor will be followed by one perspicacious enough to revert to an earlier version; finally, (c) regression to the mean is generally a better explanation of the behaviour or characteristics of individuals than of groups of individuals. But I'd be interested to hear thoughts on this point. (2) opinions -- Add your opinion