Return of Joe Isuzu

Fine. Great. I get about a comment a day through the summer and then the first day of classes, when I have no chance of keeping up, I do something that brings a boatload or two of people to my blog. So, with apologies for the shotgun response…

1. You killed Wikipedia. You bastard!

Well, not really. I did waste the time of a few editors, but in the case of the only one I’ve heard from, they didn’t really mind. Why are they so sanguine about having to clean up after me? First, because I was a tiny blip on their cleanup screen.

For whatever reason, I had in mind a roving pack of browsers who might check the Recent Updates once in a while, or keep up on their own pet areas. In fact, it seems there is an army of people watching an RSS feed of that page, just waiting for the error they can pounce on. It shouldn’t surprise me: I keep a reasonably tight lid on my own wiki (that half of China seems to consider their own personal advertising venue), but I guess I thought the content-to-watcher ratio would be much higher than it is.

So, anyway, now some people have something to point to that shows that errors really do get corrected, and it cost just a few corrections. It is important for anyone to be ethically reflective, and that the question at hand is important enough to counterbalance any potential harm. In this case, I think that the balance was well struck. I am happy, however, that there is an interest in the ethical use of community projects, and I welcome the criticism, even if I don’t agree with it.

2. If you really wanted to mess up Wikipedia…

Yes, yes, it seems everyone is wanting to tell me about how dumb it was that I not only made the changes from the same IP, I even attached them to a user name. And furthermore, I actually admitted to making the change on my blog, which no doubt has a readership to rival the New York Times and would quickly give me away to the careful Wikipedia police.

I realized this the moment I was “caught.” If I had it to do over again, I might have settled for less changes and restricted myself to the less traveled articles (e.g., my Hesston, KS change). But, you have to remember that I thought I was going to get away with it. I didn’t realize I needed to be that sneaky. I suspect that changes made to the obscure pages might last a bit longer, but I am convinced that there is a reasonable amount of rapid fact-checking going on at Wikipedia.

(And I think the marginal return on redoing the experiment is not worth more time wasted by reviewers.)

Update (9/4): Looks like a more subtle approach does indeed yield less satisfactory results. All the more reason to have an “endorsed” page (or an informal endorsement on talk pages).

3. Because of your 14m3 attack, you have proven nothing!

I don’t think there is any way to “prove” the effectiveness of Wikipedia to such a degree that it will satisfy everyone (see #4, below). I have, to quote a commentator, provided “one more data point explaining how the system works to people that aren’t familiar with it, and one more data point to use with naysayers who think that having a resource be freely editable means that by definition it can’t also be authoritative.”

4. Do you know how many 4th graders you misled in those two hours? Any error on the site, left for any length of time, calls into question its validity.

This is just silly. I have both written for a “real” ink-on-dead-trees encyclopedia, and refereed for two of them, and I would trust these publications about as much as I trust Wikipedia. It rates a “pretty darn good,” in my book, and that’s good enough for me. Anyone who takes Wikipedia for gospel truth (hell, anyone who takes the gospel for gospel truth) deserves whatever they get, and that’s equally true for those who believe that Britannica is 100% error free. If it was, there would be little need for new editions.

In other words, this doesn’t prove anything, but to my mind it refutes the assertion by a librarian that Wikipedia is a site that appears authoritative but is not. One would think that there are ==< ==partisan dig ==>== plenty of other websites ==< ==/partisan dig ==>== that could better exemplify this.

5. When is a preprint available?

This was some Sunday curiousity, not part of an NSF-funded project. You’ve seen the write-up. I did take some preliminary notes, but they basically say why I picked particular pages on which to make changes. The Syracuse page was an obvious reflection of the question that spurred the effort. I also intentionally left an error on an article linked to the main page, to see if it was discovered more quickly (it was). The rest were a result of “random page” results, and included adding chunks of information or adding new pages from stubs. Most the changes can be found here (note that two new pages: one for the chemical element Alexium, the other for the Middle Earth Dwarf Gamil Zarik were deleted from the site and do not appear on that list now), and the early responses from alert readers are on the talk page.