200 people answered this survey, and they are deeply split on the meaning of “proactive” -- a word that’s at the center of many intermediary liability policy discussions right now. That’s a big problem. 1/ https://twitter.com/daphnehk/status/1253060304289685505

The problem’s not that one side or the other is “wrong.” It’s that the we mean different things when we speak, and misunderstand each other. I’ve seen that happen with "proactive" a few times, which suggests it’s happening a lot out there. 2/

I think the biggest split is between exactly the people who most need to speak clearly to each other: workers who do content moderation on the one hand, and policymakers on the other. Having them unable to communicate (without lobbyists to translate) is a bad situation. 3/

Here’s my theory of the split:

If you do COMO ops, “reactive” means “in reaction to a notice” and “proactive” means “not in reaction to a notice.”

If you pass laws, “proactive” might come to mean “not in reaction to a legal requirement.” 4/

That oversimplifies, but you get the idea. 5/

One solution is to stop saying “proactive” or to at least always pair it with another word that clarifies which meaning is intended.
We could say voluntary/compulsory for the policymaker meaning. Or notice-based/platform-initiated for the COMO ops meaning. 6/

The problem is that “proactive” is a useful and concise word – for both its meanings. We may be stuck with the ambiguity. But at least we can be aware of it and correct for it, if we want to be understood clearly. 7/

There used to be a similar problem with the word "filter." I used it to mean "automatically detect and block specific content." But some people used it for any kind of blocking, or for partial blocks like geoblocking.
I think usage has converged thanks to Article 17 debates. 8/

That's the end of the intermediary liability observations, the rest of this thread will just be language nerdery. 9/