This problem is actually part of a bigger problem, the one of insufficient moderation. If we could solve this, not only would the swearing no longer be a problem, but hacking, scamming, etcetera, would all be history. There is really no other way to stop this, without getting rid of chat altogether. Swear filters? Use 1337, or misspell the word. There are thousands of spelling permutations of any given word, and no way to filter all of them. The only reasons these people don't swear their heads off in the real world is because they care more about their identity with real people, people who they actually know, and their actions have much greater repercussions. In real life, you can't make a new account.
The chief problem is finding trustworthy people to do the moderation. My solution? Make it a hierarchy. Have the most trustworthy people in the company running the game each watch over a small group of people, spot check them every once in a while (to keep them from abusing power), and assign each moderator a specific location to watch. If the world is big enough, then those people might also watch subordinates instead of doing the patrolling themselves, and this could even get down to player moderation. As long as each person is fairly trustworthy, and as long as the untrustworthy people are weeded out quickly, this system should work. Theoretically.
Slashdot Mirror
The chief problem is finding trustworthy people to do the moderation. My solution? Make it a hierarchy. Have the most trustworthy people in the company running the game each watch over a small group of people, spot check them every once in a while (to keep them from abusing power), and assign each moderator a specific location to watch. If the world is big enough, then those people might also watch subordinates instead of doing the patrolling themselves, and this could even get down to player moderation. As long as each person is fairly trustworthy, and as long as the untrustworthy people are weeded out quickly, this system should work. Theoretically.