What's wrong with Scunthorpe?

A requirement that I’ve seen appear a few times recently is for:

Automatic filtering of content that is deemed to be harmful, threatening, unlawful, defamatory, infringing, abusive, inflammatory, harassing, vulgar, obscene, fraudulent, invasive of privacy or publicity rights, hateful, or racially, ethnically or otherwise objectionable is filtered.

Yes, we all love a good jolt of lawyerspeak that is completely unconnected with what is possible for a computer to do ‘automatically’. I mean, let’s ignore that the court cases for what counts as, well, any of those things can be less than clear cut – if they were, cases would be short, and lawyers would earn less.

(As it happens, I am legally an ethnic minority in England – which I object to!)

Anyway, obviously complex decisions that require the judgement of human intelligence (let alone Judges) are beyond the wit of a computer program – hell, if it can’t find my printer, how’ll it find a bit of text ‘defamatory’?

But perhaps they requirements are for something a bit simpler – such as the screening of rude or politically incorrect words? Well, my memories of when this had gone wrong made me laugh! The first time I heard of this was AOL banning “Scunthorpe” ‘cos it’s name contained an Anglo-Saxon four-letter-word. Their suggestion was that the UK might like to call it “Sconthorpe” instead. Idiots.

In researching this, I found there’s actually an entry on Wikipedia about the problem – it’s called the Scunthorpe Problem (what a lovely name!) It’s got a lovely list of examples, but missed some:

  • A long time ago the Times had to write a correction to an article about economics – where the line said “and this will put America back in the Afro-American” should have read “and this will put America back in the black”
  • Recently the sprinter Tyson Gay was ‘outed’ by the American Family Association with the headline “Homosexual eases into 100 final at Olympic trials”. Nice. I mean, what’s wrong with ‘Gay’? Or are they objecting to its use as meaning ‘weak or ineffectual’? It’d be a brave man called Tyson Gay weak or ineffectual!
  • Discussions of Peter Jacksons plans to remake the Dambusters were hampered – because the squadron’s dog in the film was called ‘Nigger’. Definitely not politically correct now, but it was at the time. I guess I agree with Peter Jackson though – the dog’s name isn’t that important.

I found that the more I thought about it, the more examples I could think of, especially in media and music. What about mention of ‘Never mind the bollocks, here’s the Sex Pistols’?

I suppose the thing is that context matters, and computers don’t understand context. It would be far better to just flag items that have possibly objectionable words in them, and then have a human brain review them. And that’s what I’ll keep saying to people who want “Automatical content filtering”.

Edit: Jeff Atwood is now asking if they’re an ‘Incredibly Intercoursing Bad Idea‘. And he likes the Scunthorpe Problem too.

What's wrong with Scunthorpe?

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.