China has made these claims.

Turkey has called the Kurds terrorists. America may view them differently.

I mean, China's influence was enough for companies like Activision Blizzard to clamp down on people expressing support for Hong Kong, so it's entirely credible.

Turkey's president literally just ran an op-ed in the Wall Street Journal defending his invasion of Syria and telling tall tales of Turkey's humanitarian mission and unique effort to fix migration, violence and instability.

Twitter is swamped with "bots" (be they actual automated accounts, throwaways, state actors or regime-friendly humans) posting and re-posting fraudulent news stories, lies and plain old propaganda in favor of all kinds of national interests and political leanings.

I wouldn't be surprised if Twitter in an attempt at "corporate impartiality" would be more inclined to go after the Kurds or Hong Kong protesters than Turkish and Chinese slurs, dehumanisation and calls for violence. After all, there are probably more "people" reporting these accounts than reporting the various propaganda outlets.

We can cherry pick a few examples of some of the protestors acting violently, then use that to claim the protestors in general are violent. That's what they've successfully done to claim Pepe the frog is racist after all.

There's a video from Hong Kong that went somewhat viral showing a protester jump kick a police officer in the face who was trying to drag and arrest another protester lying on the ground.

The video was widely shared by protester-friendly accounts celebrating the act because several arrested protesters had recently turned up dead or disappeared.

It's very easy to construct a narrative portraying the protesters as violent criminals if you start from the assumption that all police violence is justified and resisting any form of police violence is not.

EDIT: Oh and the problem with Pepe wasn't that "Pepe is racist because someone used it that way", it's that Pepe became popular on 4chan and was then co-opted for racist imagery and embraced by actual racists who then used it unironically, so it eventually became a calling card.

Likewise the "ok hand" gesture and emote was initially planted as a "hoax" white supremacist symbol by 4chan members but then actual white supremacists started using it, first ironically as a nod to the "hoax", then unironically to demonstrate their affiliations to others.

The problem with these "dog whistles" is that they're intentionally ambiguous and can be entirely innocent in the right context. Of course not everyone posting "funny green frog memes" or using the "ok hand" gesture is a nazi, but if someone who's already suspiciously copying far-right rhetoric randomly flashes the ok hand in a photo op, or uses Pepe (or the less subtle NPC meme) they're most likely trying to say something while retaining plausible deniability.

I see Pepe used a lot on 4Chan in not racist contexts even now.

What would you suggest is the minimum evidence necessary to support the view that an ideology promotes violence?

- Isolated incidents of violence on behalf of the ideology?

- A "sufficiently large" number of incidents on behalf of the ideology? If so, could you elaborate on "sufficiently large" as you see it?

- A general tacit approval of violence?

- An explicit approval of violence?

- Something else?

The problem I think that many see is there tends to be extensive cognitive dissonance. People apply one set of standards to groups they like, and another to ones that they don't. I think you'll find that in trying to answer my above question it becomes quite difficult to include the groups you don't like, while excluding the groups you do like. That's even when my framing above which is unrealistically softball. For instance I didn't even mention things like harassment, destruction of property, etc which I think we should also generally be intolerant of - even when it's a group we like engaging in such things.