Moral responsibility, reputation, and trust

That Orlowski has gone off on one isn't actually very surprising. Andrew Orlowski is an Internet journalist who gets paid to write on (and about) the Internet. He writes for The Register, a British website which has developed a reputation for edgy and controversial journalism. It lives not only by reporting on the ephemeral phenomena the Internet throws up, but also by attacking the emerging centres of power of the information age. Andrew Orlowski is, on occasion, just as attacking of Oracle, of Microsoft and of Google as he now is of Wikipedia.

A quick Google (if you'll excuse me, Mr Orlowski) is sufficient to demonstrate that Mr Orlowski is a good thing. He puts people's backs up. The first page of hits shows us

• Andrew Orlowski, Sloppy Journalist or Bold Faced Liar?
• Andrew Orlowski is a hack
• Noted Register troll Andrew Orlowski

This gives us a feel for Orlowski's standing in the Internet community: of his reputation. Orlowski is a gadfly. Someone who winds up so many people can't be all bad; particularly when you consider who he winds up. We need journalists prepared to confront the new powerful.

And the person behind Wikipedia, James Wales (he'd rather I called him 'Jimmy' or 'Jimbo' - so much more cuddly, don't you know?) is an unlikely hero. He made his first fortune as a futures trader, but then went into the Internet porn business, and also ran websites which 'scraped' content from other websites and added advertising, using their content under his branding to drive revenue into his pocket.

Futures trading is considered respectable under capitalism; and I am in no position to criticise people who create porn. But content stealing is sleazy in anyone's book (ironically, Wikipedia itself must now be the most screen-scraped site on the Internet).

Furthermore, it's easy to paint Wales as an egotist. Again, I'm in no position to criticise someone who has his own website, or runs his own 'blog' (hideous neologism). But Wikipedia has an entry on Wales, as does WikiMedia, as does WikiQuote... Wales repeatedly describes himself as the 'founder' of Wikipedia, and emphasises his own role to the exclusion of all others:

'About two years before I founded Wikipedia I had founded another project called Nupedia. It was based upon the same concept as Wikipedia, which was that it was a freely licensed encyclopedia that was written by volunteers. Unfortunately, we didn't use the Wiki software and it was a very top down model, which ultimately wasn't very successful. It was difficult to manage and when all was said and done, it wasn't very much fun for the volunteers. We found the Wiki editing software and began using that, which turned out to be quite a success.'

There seems no doubt that the money behind Wikipedia is his. But he seems to find it easy to forget the contribution of (e.g.) Larry Sanger, who apparently did most of the work and seems to have done a lot of the creative thinking.

And, of course, Wikipedia does, at present, have a problem. It is too easy for ill-intentioned or merely mindlessly destructive people to edit articles on Wikipedia, as the Seigenthaler incident amply demonstrates. And in the particular (rather unusual) circumstances of the Seigenthaler incident, where the saboteur edited the article from a machine on a fixed IP address, Orlowski is correct in observing that requiring editors to log in, rather than showing their IP address, actually reduces transparency.

Nevertheless, Orlowski is wrong to attack Wikipedia. Wikipedia is far more than the ego vehicle of James Wales. It is an experiment. It is new. It has got teething troubles. This is normal. It is nevertheless at the very least a most interesting experiment, and it is also, already, a very useful resource - albeit one which should be used with some caution.

In mounting a defence of Wikipedia I will start with the main thrust of Orlowski's recent article, and go on to mount a more general defence by pointing out what I think is interesting and valuable in the Wikipedia experiment.

So: Orlowski claims that "There's no Wikipedia entry for 'moral responsibility'". In fact, at the time that he wrote that, he was literally correct - the entry was created at quarter to midnight last night. But the thrust of the first point that he's making is that, while Wales had claimed that he'd sought to unmask the Seigenthaler saboteur, neither he nor anyone else at Wikipedia had done so; and that it had in fact proved very easy to do so.

Well, yes, egg on Wikipedia's face. In an age when most Internet users are either subscribers on dynamic IPs or posting from behind the firewalls of large corporates, the Seigenthaler saboteur was - quite exceptionally - posting from the fixed IP address of a small organisation. An IP address which happened to be the same IP address as the organisation's website, allowing it to be trivially identified. Full disclosure: you could track me down the same way. But this way of accessing the Internet is increasingly rare, and if I wanted to go and sabotage someone else's project I'd go round to an acquaintance's house and borrow their line.

And that's the point. No-one who was seriously attacking Seigenthaler (or Wikipedia) would have covered his tracks so badly. The saboteur was easy to find not simply through the sheer chance of his happening to use a fixed IP, but also because his puerile, mindless vandalism did not seem to him sufficiently important to merit taking even fairly basic precautions.

It's not really surprising that Wikipedia's admins didn't think to check something as simple as this. Yes, they should have. But given the combination of unlikely circumstances that made this particular saboteur easy to discover, I'm not surprised they didn't.

So next for Orlowski's second claim, that requiring Wikipedia editors to log in makes the situation worse, not better.

No, sorry, Andrew; that one won't wash. Static IPs are very much in the minority these days. Tracing the user behind a particular dynamic IP is hard, and as it isn't possible to tie together the edits from a particular dynamic IP and say that these represent the same person, or to say of edits from two different dynamic IPs to say these represent two different people, using IPs to identify users is normally pretty useless.

By contrast, while Unattended Terminal Syndrome is sadly a fact of life, it is reasonable to say that all the edits done by a particular login are all the responsibility of one person (since if they left their terminal unattended, that's their responsibility too). We can't, of course, say of two different logins that they represent two different people, since malicious users may register 'disposable' user names in exactly the same way spammers used to register disposable hotmail accounts.

In discussions on Wikipedia, you will often see users signing themselves off 'Joe Bloggs [Contribs]' where 'Contribs' is a link to the user's contributions page. Why do users do this? It's a way of saying 'look, you can check here and see I'm a reputable person, a member in good standing'. This is a convention, not a rule, and it's not even yet a widely adopted convention. But behind it is the notion that users build up standing in the system by making useful contributions. It's needed because there are many too many users for everyone to know everyone else. And it's needed because Wikipedia does not provide any automated way of tracking users standing.

But automated standing trackers have their problems too. I know that if I post four or five reasonably good posts to Slashdot, I can get away with one truly outrageous troll about the stupidity and lack of education of Americans without losing significant karma. My Slashdot karma reflects the fact that I'm usually a good citizen, and conceals the fact that I'm a part-time troll. If the cost to my karma of trolling were higher, however, I would be forced to set up a separate identity to hyde behind when trolling. This identity would never gain much karma (if any), so its posts wouldn't get read, so trolling wouldn't be fun, so I probably wouldn't bother.

This is what Wikipedia needs to do. It needs to make reputation valuable, so people are motivated to keep and develop single logins. It needs to prefer edits by reputable editors to edits by less reputable ones. And it needs to make the cost of sabotage high. But it needs to do all this while still allowing 'newbies' to contribute and begin to climb up the reputation ladder.

The problem at present behind most of the criticisms which aren't sheer spite is that Wikipedia currently values all edits equally, irrespective of what it knows (or doesn't know) about the contributer; and that it doesn't provide any simple means of tracking the reputability of a contributer.

It's even fair enough to point to Wales as the person behind the present laissez-faire attitude to editing. Personally, I think that says more good about Wales than ill. People tend to expect other people to be like themselves. Untrustworthy people tend to distrust others. Greedy people tend to fear the acquisitiveness of others. Wales expects that contributers to Wikipedia will be public spirited, generous with their time, honest, and conscientious. The fact that Wikipedia is generally reliable on most of the material it covers illustrates that, in general, he's right. But it also tends to suggest that these qualities characterise Wales himself.

And it's unfair to suggest that the Wikipedia community doesn't care about quality and accuracy. It does care, deeply, and the ongoing discussions on policy and on cleanup clearly illustrate this. Wikipedia is, largely successfully, developing its own civil society, its own governance, and its own etiquette. This seems to me to show precisely the right attitude.

Again, something that makes me feel good about trusting Wikipedia is its advice on choosing a user name:

"The best username is typically either your real name, or a longstanding Internet pen name. Please pick a username that helps us to write an encyclopedia. That means picking a name that you're comfortable writing under, but it also means picking a name that others are comfortable seeing and collaborating with."

Speaking as someone who has always used his own name (almost) everywhere on the 'net, this seems good advice to me. It means you carry your identity with you from project to project. It's pretty easy to find me on any of the sites I regularly visit, and assess what my standing is there; and pretty easy to track me back to my real-life existence. Encouraging others to do likewise is encouraging them to care about their reputation, to act reputably. Using a real name is a step towards making contributers accountable, which directly addresses Orlowski's criticisms. It's not a sufficient step, admittedly; but, again, it shows the right attitude.

In summary, Wikipedia is still young; it is suffering the pains of extremely rapid growth; it has admitted teething problems. But it is growing because it is valuable, it is developing its own systems of self-regulation with commendable thought and civility, and it promises to be an extremely, perhaps eventually pre-eminently, valuable source of human knowledge. Orlowski dislikes Wikipedia's grand vision; he sees it as pretentious, as overweening. He's wrong to do so. Without setting our goals high we cannot achieve great things. Wikipedia aims to achieve great things, and shows every possibility of achieving them. As it grows towards them, it will, of course, become very powerful.

We need gadflies like Orlowski who are not afraid to challenge the powerful. But on this occasion, he's wrong.