Hateful words and actions directed toward various groups are nothing new. But in our digital age, social media sites must increasingly face the fact that with their services, a friendly discussion and exchange of ideas can turn into an ugly undercurrent of nastiness.

The question for sites such as Twitter, which on Tuesday responded to a petition to make reporting abusive behavior easier, is how to police hundreds of millions of people, providing a safe environment for some users while respecting the free speech of others.

At the heart of the problem are the mechanics of policing. The sheer number of users means that flagging misbehavior is like playing a vast, never-ending game of Whac-a-Mole.