Sure, given twitter's design it makes perfect sense. That doesn't make it a good or unavoidable problem though; you can imagine easy changes to twitter that wou

Sure, given twitter's design it makes perfect sense. That doesn't make it a good or unavoidable problem though; you can imagine easy changes to twitter that would make handling this stuff much easier. Detecting that you're sending a very similar response to a message to other previous responses and warning you, for example, in the same way stackoverflow does to catch duplicate questions, would solve this exact problem almost totally. More widely, better handling of abusive content generally would help. Automatic trolling and abuse detection has actually had quite a lot of research done (e.g. http://sentic.net/do-not-feel-the-trolls.pdf http://sentic.net/do-not-feel-the-trolls.pdf), and while it's not perfect you can definitely imagine an approach where you start to reduce notification spam from tweets likely to be needlessly offensive, or generally unlikely to be adding much to the conversation. Obvious repeatedly detected offenders, or people who almost always get their messages flagged, could result in notifications for the target only occasionally (instant notifications for interesting responses, daily/hourly batched notifications for less-interesting responses), or no notifications at all (but potentially still visible if you actively go look at the conversation). Twitter et al could drastically improve this issue. While being a public figure is difficult, this kind of thing isn't an inevitable aspect of it: there's clear concrete things you could do that would substantially lessen it.