Recently, the Coca-Cola company came under fire for being anti-gay. The company had launched a marketing campaign in which customers could personalize a virtual can of Coke with whatever words they wanted. The site disallowed profanity, but included the words "gay" and "homo" in it's list of banned words.
I'm a web developer. My company is tiny compared to Coke, of course, but I couldn't help but envision how this might play out if I were asked to design such a site for my company, because profanity is something that developers run into from time to time. At my previous job, we shipped our product on CDs and required the customer to input a randomly generate product key. A customer called in to inform us that their key included the letters 'FUCK', and one of our engineers had to scramble to change the key generator to filter profanities.
I can't help but feel that I would have done a better job.
Most developers would probably choose to use a software library called a wordfilter to perform the profanity scan. Think of a software library as being like an app for your phone. You just download it and use it - you don't need to know how it works. In this case, think of an app that lets you input some text. You then press 'Done' and an alert pops up if there's profanity in the text.
Most of these libraries would come with a built in library of profanity, including multiple languages and alternate spellings, possibly even more devious entries (like 'f.u.c.k' instead of fuck) that wily users might enter to circumvent the filter.
Of course no software is perfect, and a wordfilter is no exception. It can only split words into two categories - non-offensive and offensive. That's good enough for a smaller company such as mine. But a company as massive as Coke should have at least attempted to define a tertiary set - words that could be offensive. Entries with words in that set would have to be approved by an actual human before being allowed.
Let's take a non-controversial example: suck.
"I love sucking down an ice cold Coke!"
"IF U DON'T LOVE COKE U SUCK!!!1!!1111"
Clearly only one of those examples represents the Coke company in a positive light. 'Homo' and 'Gay' could easily be plugged into the second example, and would be caught by a human, though missed by a software library.
If someone entered a word that could be offensive, just display a message like this:
Your message contains word(s) that could, in some cases, be used in a derogatory manner. It will be manually reviewed and, if the message is not offensive, you will be allowed to use it.
It's not a perfect solution. Some might still take offense. And a *cough* unenlightened employee might refuse the word 'gay' if it's NOT used in a derogatory manner.
Y'know...hmm...this entire idea sounds like a PR disaster waiting to happen, and probably should have been nipped in the bud. Perhaps they should have talked to someone from Nike.
Those who do not learn from history are doomed to repeat it.