Unfortunately I don't have time to respond to all of this, but you make some really good/interesting points. On the original point, I kind of agree with you, b

Unfortunately I don't have time to respond to all of this, but you make some really good/interesting points. On the original point, I kind of agree with you, but there are two edge cases that immediately come to mind. The first is where we see an individual (heated) comment thread where someone takes deep offense and/or fears for their life. There's probably a point in that situation that should constitute actual illegal harassment. The second is kind of the reverse, where we see someone 'follow' someone else around, but not deliberately. They may have continued arguments/disagreements over the same topic across the internet by virtue of the fact that they are both heavily invested in that topic. So yeah, vague lines somewhat moderated by intent and how vitriolic the discussions are. Second, your "human mind model"/"internal representations" idea is well studied. I don't know how much you know about this, but generally this is grouped under the idea of "Mental Models", and the idea has been around for decades. I've studied them in the context of how students learn to code. They're crucial in programming because we have to form a whole lot of correct mental representations of an inherently abstract concept (in which, yes, we compartmentalise). It's an incredibly interesting area of work. In terms of how they form the basis of human interaction, that's a little more complex (and well outside my 'expertise', if you can call it that), but intuitively I doubt you're far off.