Tokyo Tales: Wikipedia in "not always 100% accurate

I'm a big fan of Wikipedia; I know better than to rely on it as a sole source for anything, sure, but in general it's an excellent resource, especially when you want an introduction to, or overview of, a particular topic. I find myself pointing people towards it all the time.

So I was interested to read this Guardian article, entitled Can you trust Wikipedia? They have experts from a number of fields give their analysis of the respective Wikipedia entries for their specialist subjects. It's an interesting piece, and as you'd expect, some entries do better than others (8/10 for Bob Dylan, 5/10 for the article on encyclopedias) but with mainly sixes and sevens out of ten. The guy reviewing the encyclopedia entry is particularly sniffy, in fact, calling it "a school essay, sketchy and poorly balanced". Mee-ow! Saucer of milk, table five. Possibly this is something to do with the fact that he's a former editor-in-chief of the Encyclopedia Britannica.

Now of course, yes, a suitably qualified expert will often be able to pick holes in any given Wikipedia article, or find something to disagree with. But all of the criticisms implied in the Guardian's critique are missing one very obvious point, completely aside from the "well it's only written by amateurs, so what do you expect" argument. And that point is, Mr. Robert McHenry, that if you find something inaccurate, poorly written, or otherwise unhelpful, you can fucking well log in and fix it yourself. Prick.

The point of Wikipedia is not so much that it's written by amateurs, but rather that it's written by anyone who wants to contribute. And that includes slumming professionals as much as zealous but imperfectly resourced amateurs. Furthermore, the fact that anyone can create an account and make changes to the page means that, broadly speaking, over time it's just going to get more and more accurate - not less.

To test this, I performed a simple check. I picked an error highlighted by one of the experts (in the Samuel Pepys article, the name of his wife is misspelt), and checked the corresponding Wikipedia entry. Oh - look at that: fixed already. Another one (a typo in the publication date of the Encyclopédie): fixed. The muddled intro to the Steve Reich entry: rewritten. And another. And another. I got bored after my first half dozen random picks all turned out to have been addressed already, but I'm willing to bet that each article on that list has already been given a good going over, if not by the Guardian experts themselves, then by other interested parties who took it upon themselves to contribute.

Don't get me wrong, I'm not possessed of some messianic fervour when it comes to all things Web 2.0. But it struck me as highly ironic that the Guardian experts, via their list of nitpicks, have inadvertently highlighted Wikipedia's greatest strength instead: it will take those nitpicks, assimilate them, and be all the better for it. It may not be perfect, but it's surely just going to keep getting better and better.

Comments

I'm going to play devil's advocate a bit here... but purely in the interests of debate as I think wikipedia's great too, I just happen to agree with some of the naysayers sometimes.

Chris, surely the only reason those entries that you checked were corrected was because of the Guardian article. How long had they and many others, been sitting there, full of mistakes until someone knowledgeable pointed them out?

As you say, that's kind of the whole point about wikipedia and part of it's strength, but not everyone who uses it for reference is going to be discerning enough to always go back to proper source material (think bored teenagers doing essays), which just propagates the mistakes.

Yes, you can point out mistakes and fix them, but you have to know they are there in order to do that! Most people won't know, and that's the problem. The old guard, Britannica and the like at least have a very stringent editorial policy that should pick up most mistakes like that so that it's not left to the user to do it.

I guess what I'm trying to say is that I don't know if wikipedia is always going to get better just by virtue of people fixing mistakes as they find them. It might improve, slowly, but at the virtue of mis-informing a whole bunch of people in the process with the mistakes that haven't been corrected yet.

Standards, as well as collaboration. That's what we need.

Just my 2 pence...

I concur; Wickerpedia is clearly the (woven rattan) bomb.

As for number-of-edits counters, I followed up a hunch and checked the "Discussion" and "History" links that sit at the top of each article. Sure enough, the latter gives a complete revision history, so you can see the complete progression of the article from day one.

Jo, I take your point about the dangerous potential for spreading misinformation into teenage homework assignments around the world, but I think pretty much all textbooks or reference works will have errors in them somewhere (except the ones Dave edits, of course). I'm sure the Britannica is hardly immune to mistakes itself. That's why it has teams of editors, after all. :)

In any case, these bored teenagers simply shouldn't be using Wikipedia (or any source, for that matter) as their sole source. If people want to blithely copy from any one source (Wikipedia or elsewhere) without cross-referencing things themselves, thereby replicating errors, is that the source's fault, or their own?

I think if Wikipedia claimed to be authoritative, that would be unforgivable, but then it never does. It's very upfront about what it is. The haute couture article, for example, actually had a big "This article is in need of attention from an expert on the subject" disclaimer at the top of it, last time I checked.

Plus, I would suggest that the Wikipedia collaborative method is more flexible and far faster than whatever review systems the deadtree encyclopedias use.

But yes, I certainly see your point that mistakes have to be spotted before they can be fixed. If I remember rightly, though, there's a comprehensive system of e-mail notifications and watchlists alerting interested contributors as and when articles are created or edited. So there is a kind of review system in place to try to pick these things up. Plus the odd snooty Grauniad piece, of course.

Anyway, my broader point stands - it's just going to get better and better as more people contribute. That's why I have no patience for people who would sit at their computers and moan about the inaccuracy of a given article, when the power to correct that inaccuracy is sitting right in front of them in the form of an "edit this page" link.

Right, I have to dash off and put together a venture capital presentation for Meebkogami, my new Web 2.0 company selling a geotag-based search engine via api mashups, apparently. (Nice find!)

Further reading on Wikipedia: Phantom authority (A bit dry but some good info on the inner workings.)

October 25, 2005

Wikipedia in "not always 100% accurate - yet" shocker

Comments