Hacker News new | past | comments | ask | show | jobs | submit login
Adventures in Twitter Censorship [pdf] (efn.org)
101 points by oskarth on Oct 18, 2015 | hide | past | favorite | 80 comments

All of you are misinterpreting Twitter's "features" here.

Most likely the root cause of this is an internal feature at Twitter called "per-country takedown."

If someone from Germany asked (and had a valid court order) that the Tweet be taken down for German users, then Twitter has the ability to disable the Tweet for people coming from German IPs.

I don't agree with this, I lobbied heavily against it while I worked there to no avail, but it's a solid explanation for why you wouldn't be able to see a Tweet in different regions.

when that was announced they said it would show that the tweets had been withheld: https://blog.twitter.com/2012/tweets-still-must-flow ”Q: What will people see if content is withheld? A: If people are located in a country where a Tweet or account has been withheld and they try to view it, they will see a alert box that says “Tweet withheld” or “@Username withheld” in place of the affected Tweet or account.”

Did that policy change?

epeus is right; I'm not at Twitter anymore either, but the last time I looked at this, per-country takedowns were clearly labeled as such (ie. you hit the tweet from the blocked country, you get clear "this isn't visible to you" messaging).

My honest guess? Cache coherency is hard, man.

Amusing side-conversation we're having right now: apparently my reply above didn't show up immediately for some folks.

Obviously, Hacker News is censoring me.

Thanks! You spared me from downloading the PDF...

One of my main concerns with mainstream social networks is this kind of subtle censorship/filtering. Some people think their Facebook stream is an accurate representation of their friends and surrounding reality, they are ignorant of the algoritms that decide what they will see and what not.

This so true and a major concern. The next wave of highly effective propaganda and thought control could very easily manifest itself through this. Today, people know better than to blindly believe the news media. If information is coming from a single entity, it could be pushing an agenda, so take it with a grain of salt. We all know that. But when it appears to come from a community, that's entirely different. It seems completely genuine and trustworthy.

But what happens when Zuckerberg meets with German political leaders to discuss the silencing of anti-refugee posts on Facebook? We know that the technology exists. If you've ever advertised on Facebook, you've seen how easily one piece of content can be given massive priority while others can go quietly ignored. But that's not how it seems to the user. It just seems like all of their friends are thinking one way....so they should too.

There are massive implications to this.

Germany does not intend to silence anti-refugee posts. Germany wants to stop illegal incitement of hatred. Such as promoting or even threatening to burn down asylum centers, to hurt or even kill refugees and those that help them.

Those are not just words, asylum centers are being burned to the ground, refugees are attacked, just recently some were killed on the border the EU, just yesterday someone tried and almost suceeded in killing a politician who helped refugees.

You have to be insane to support this as free speech unless you support these nazis.

The trouble with fighting for human freedom is that one spends most of one’s time defending scoundrels. For it is against scoundrels that oppressive laws are first aimed, and oppression must be stopped at the beginning if it is to be stopped at all.

— Widely attributed to H. L. Mencken

Would HN be better if personal insults and death threats were allowed, would it improve the exchange of ideas and the expression of opinions?

Yes. Your idea of a personal insult is my idea of spirited debate. I've been warned twice in the past few weeks by HN moderators because I've used similes that they've found offensive. Honestly? I find that offensive. They'd better censor their censorship.

I'm no stranger to insults, I can take what I can dish, but I also think this is true: everybody can argue their point without being personally insulting (some, like me, take getting hellbanned a bunch of times, others just do it straight away), but not everybody can look past the insults and focus on the points made. So a free-for-all this excludes people who simply don't like being insulted while gaining nothing when it comes to the content of the discussion, the arguments.

FWIW, I'm not a huge fan of downvotes, and I don't even like things being ranked by upvotes. But the rough spirit of focusing on substance and backing up claims, that I like.

I think you will find many people disagreeing with you on that. In Germany it's even illegal to insult people because in Germany dignity is considered to be a human right just as much as freedom of speech is.

Clearly, Germany has failed - If people had any dignity, they wouldn't need a government to protect it.

The minister in talks with Facebook called the content "ideological trash" and "right wing nonsense" so it's pretty clear it's not just limited to incitement of violence.

The incitements of violence in question definitely qualify as ideological trash and right wing nonsense.

It's already happening.

This is what the US internet looks like, post-Arab Spring.

Recently a Chinese copy of twitter, Weibo, decided to implement similar functionality, but their ability in machine learning is very limited, so their lame work has only promoted paid members and threads. In the first week of its introduction, most users has noticed that there's roughly 50% less threads were read by other users.

Actually I think theres's big future for this kind of censorship, if performed well, can be hard to detect and hard to test(think about A/B testing). So I disagree with 'subtle', it's just our enemy get smarter.

Is this censorship or editorial/curation?

I've been infuriated by Facebook because I have a local events group on high priority in my feed, but FB has decided I only want to see about one post in five. So I miss a lot of news I want to see.

Meanwhile I still get porn posts on the various technology groups I read, even though I keep asking not to see similar content.

I'd have no objection to a good AI editor curating my feed. But FB has a very bad AI editor making bad choices for reasons that ignore my interests, and I really don't want that at all.

The difference is completely in the eye of the beholder.

But a huge centralized curator that decides what everybody reads is something people feared for centuries. It does not really matter if its intention is benign.

Absolutely. And even more dangerous is the perception that there isn't a curator, when there really is one. That is the case with Facebook today.

Does switching it from "top posts" back to "most recent" still fix it, or is even that filtered? They clearly want you to use their "curation," but a bit of fast scrolling and skimming still does a much better job.

Don't use networks that don't share their code. Don't use networks that hold your data and don't give you raw access to it. Share directly with your friends instead of a large corporation.

Edit: Instead of downmods, please post reasons why you should use networks run with secret codes and which suck your data into their proprietary databases.

I didn't downvote you, but I'm not aware of any "network" (by which I assume you mean website or application) that matches the requirements you set. Open source websites simply don't exist - as it's impossible to actually validate that the code running on a server matches whatever source code a company might distribute, and every site and application attempts to limit or control access to your data in some way.

Hacker News itself explicitly doesn't share their fork of the Arc forums, and don't allow raw access to your data.

GNU Social is Open Source and federated.

Your criticism is still valid, obviously there is no way to validate that a GNU Social instance is not running modified code.

But the federation itself mitigates this to some extent. If you don't trust one instance, you can always fire up your own.

I'm not sure I really need much code running on a server that I don't control. I could just share my data directly with those people I want to, instead of handing it off to FB to catalog and exploit.

There are plenty of networks that fit this requirement!


Fair enough, but the entire web doesn't, which makes jsprogrammer's imperative practically impossible if taken at face value, unless one wants to avoid most of modern technological society altogether.

No, the entire web doesn't. One can still reject secret code and proprietary databases while still accepting that they exist. One can even still use them.

HN is known for its inability to downvote properly. Gave you an upvote. :)

Thanks. I know :) Unfortunately, others have since come and bumped the comment back down.

I've documented evidence of Twitter censorship; what likely amounts to millions of tweet-replies. This seems to be used to shape perceptions and opinions on twitter in at mass-scale:


I've now registered my username on Kloak as @greg, and am recommending to all my friends to move off of Twitter to similar networks.

Here are some networks to ditch twitter for:

- SpiderOak's Kloak (in beta): https://spideroak.com/solutions/kloak

- Indie's Hearbeat (in alpha): https://ind.ie/

- Twister: http://twister.net.co/

- Vole: http://vole.cc/

Bye Orwell, won't be seeing you in the land of free speech!

>Register for an invite to the private beta (iOS only)


GNU Social is a great alternative.

Im more concerned with social censorship and the mob mentality that seems to be determining justice. If you say something that is not politically correct or against the current narrative, you will get fired/bullied/harrassed/silenced/all three.

Without honesty, there will be no progress. I shouldnt be afraid of posting my opinions and expressing my freedom of speech because i might lose my livelihood. But its the sad reality.

You are in essence saying there shouldn't be social consequences for speech.

The internet has created a weird environment where repercussions for socially unacceptable speech (even if taken completely out of context) can be dramatically more severe than any harm that was caused by the original speaker. In general, a distasteful tweet or Facebook post shouldn't cost someone their job, make them a target for anonymous threats or make them fear interacting with society at large any more than it would have if the message had just been spoken. There was a good piece on NPR recently that explored a number of examples:


There has always been repercussions for socially unacceptable speech. What the internet has affected is who can hear or read your speech.

If I go into a supermarket and shout, "I hate black people and they shouldn't even be in America!" I am going to suffer the consequences, but probably only from the people who heard me. Everyone else will have heard it second-hand and may due to social circle pressure decide to sever ties with me. The only way I'll actually suffer long-term (other than being beat up by a mob) is if I was famous since people like it when famous people do things.

The problem is, people continue to think that "The Internet" is just some kind of throwaway void where they can spout whatever they want. Justine Sacco probably isn't a racist, and was probably making a joke about "white privilege" or "white bubble". But she showed poor judgement in assuming that writing that in a public space would be benign. Even with her lack of followers. Text is not a good way to facilitate a joke that requires context. Her joke about AIDS and Africa and white people not being affected is based on some context no one but herself had.

If _I_ am at the supermarket and I hear some old cranky dude yell that out. It's a story I'll probably tell my friends about. They in turn may decide to use that as an inside joke/catchphrase of our circle of friends. But there is no way I'll repeat that to someone else outside of the group of friends that know what the context is. I definitely won't post it on Twitter, not without at least having posted the story about the crazy old man in the supermarket.

So I kinda laugh when people say they can't say anything that's not "PC" anymore. That to me just means they want to really spout some really offensive opinions to a public audience without paying the consequences of doing so. Then of course when there is kickback, they blame it on people being too "PC"

You're acting like the networks we use are just neutral conduits through which we engage conversationally exactly how we do in person, and that's just not true. They change how we interact. For one thing, these networks make it so that you can reach out and help mob somebody from the safety of your home, which is completely asymmetric compared to how such interactions would work in person.

Part of the problem that I see, is that in modern America, everyone is a victim. Anything can be considered offensive. And when that happens, we find it acceptable to utterly destroy people's lives over 140 characters...in the name of "social justice". It's ok to call Justine Sacco a fucking bitch and wish brutal rape upon her, so long as you do so in defense of social justice, right? It's ok to shame her, cause her to lose her job and go into hiding, so long as you're part of the "progressive" and "forward thinking" mob, right? In fact, one may even consider it comedic, right?

That is what you call justice?

We have effectively returned to the days of witch hunting and public hangings, but this time our medium isn't the town square. It's the Internet.

Justine Sacco's tweet was a joke. And she even subscribed to the very "liberal" thought that was shaming her. Her joke was tongue-in-cheek. But her shamers weren't liberals. They didn't care about actual reason and discourse and open-mindedness. They weren't the liberals that we think of today; the ones that were always on the right side of history. The ones that marched alongside MLK. The ones that stood up for women's rights. No, they were "progressives". Pseudo-liberals that fight for the oppressed by oppressing others. The open-minded, forward thinking, altruistic group that wants everyone to be heard...until they disagree with their views. Then they must return to the mob-mentality of the stone ages, in the name of "progress".

These people don't care about progress. They don't care about social justice. They care about the entertainment of destroying someone's life, the feeling of dehumanizing another human being, and the gratifying thought that in doing so, they were being a model human...an enlightened individual...better than the rest. But they aren't. They are the most simple, selfish, uncompassionate, and closed-minded beings of our generation.

And we support them. We encourage them.

We have no idea how stressful and oppressive of a world we are creating for ourselves. And all in the name of pseudo "anti-oppression". We must remember the importance of true freedom of speech. The importance of unpopular thoughts and discussions. The importance of open discourse without the threat of condemnation. Where everyone can be heard and we can build a better world together. Where the enlightened truly care about helping the unenlightened to see things differently. Where true progress is bred.

I agree there is a lot of mob mentality going on. But I feel like you are only viewing this when you perceive it to be started by the "left". Mob mentality is not a left or right problem. "Progressives" are not the one's doing this. Mob mentality is a human thing.

Is your reasoning that only people attacking Justine Sacco were "progressives" and "liberals?" Does that mean everybody, to put for a lack of a better term, "on the right" understood the context of Justine Sacco's tweet and did not join in and pile on?

I mean in the context of just those tweets, what she said was pretty racist and offensive. So the implication that only liberals attacked her means that either conservatives didn't attack her because they already understood what she meant by those tweets (hard to imagine anyone could in that moment since one of the biggest things about it was that she was on a plane and couldn't defend herself) or they didn't believe those tweets were racist (which they were when taken without context).

Dividing people by framing this as a left/right issue is not the way to go if you really want to stop the internet mob.

What it should and should not do is a matter of opinion. The fact is, putting something in writing is permanent. Writing casually/carelessly can haunt you, like it always has. The confusion comes when regarding things like Twitter as 'speech' and calling the authors 'speakers'.

But it is speech in the legal sense. It's a usage that's been with the language for quite some time. If we mean literally "spoken," we should be clear about it.

Generally it shouldn't, sometimes though it certainly should.

I didn't get that at all from the post. I read that GP is sad that the 'social consequences' for speech have created an environment where people are afraid of voicing benign, unpopular opinions.

The kind of people who are oversensitive to speech don't care about what's being said; they care about how it's being said. I don't think there should be social consequences for saying something that isn't PC. And I think OP falls into the same category. This is far from something so general as "consequences for speech" which would include things as abstract as judging the level of education of the speaker. I don't make this point lightly either. It is a phenomenon I've observed too many times in my life. There are people that are actually willing to listen to others, that actually try to get the context of who this person is that they're speaking with. But most people are not willing to listen to someone else within the context of who that person is, and are more than willing to run with their early assumptions of that person. Why is it that the people who are oversensitive always fall into the latter category? Being sensitive to political correctness is a way of being ignorant and pretentious to the world, to the billions of people who are not like yourself.

Big Think - People Will Misjudge You Unless You Manipulate Them: https://www.youtube.com/watch?v=EX83nf6t6Vg

Its all about context. With your bro's after the game, sure. But in Twitter, where the whole world (anybody who cares to look anyway) can see it? Its expected to have a different voice there. Folks who get all chummy on Twitter are forgetting where they are.

"Its expected to have a different voice there."

Expected by who? The "expectation" you think everyone has, everyone does not have. In this discussion we're referring to this expectation as "PC" or "political correctness".

Normal social conventions. You don't hum in church. You don't yell "Theatre!" in a crowded firehouse. Etc.

The consequences for speech should resemble an asymptote, always approaching zero.

Consequences of speech are people's thoughts or reactions to speech.

Where does someone's right to say what they want without consequence end and my right to have a thought or reaction to what I heard begin?

Your right to speak your opinion coexists with people's right to have a reaction. This might be socially inconvenient, but that has always been the reality. Nothing has changed because the avenue expression is online, just the audience and availability of your speech.

You are in essence saying that ad hominem is a valid debating technique.

Go on

Expressing a dissenting opinion against the mainstream shouldn't have to the extent of threatening ones livelihood because it suppresses honest discussion. That is the point 0x49 is trying to make. You are inferring with your comment that dissenting from mainstream opinion should be allowed to have the equivalent social consequences of e.g. advocating genocide. But with that line of reasoning you are allowing people with unpopular opinions to argued by the character of the people making them.

Everything isn't to be treated as a debate in good faith. It is very, very rare for a person to be making inflammatory comments in an earnest attempt to start a debate in good faith. I agree, if the platform is a debate-like context, it would be bad form to do that.

But if you walk into a bar or post on Twitter talking about how much you dislike ___, don't be surprised if someone has a reaction. I certainly won't be inclined to start a Socratic dialogue exploring why you dislike ___ in an attempt to persuade you.

You wouldn't spout a racist manifesto on a talk show, don't expect to be able to do it on a medium that everyone can see and think nothing will come of it. You need to be acutely aware of your audience and reach when you speak.

In this context, do not be surprised if people cannot separate their judgment of the comment from the person making it. Especially when the context is "I just read this awful comment on my screen by a person I don't know and already dislike."

>You need to be acutely aware of your audience and reach when you speak.

The problem 0x49 referred to is the polarization of discussions, where any attempt of moderation and nuance of the arguments is punished immediately with personal attacks and accusations of siding with the enemy. So it isn't just audience and reach that you should be acutely aware, it is also if you should speak on some matters at all. If you step in and say something in a polarized issue that doesn't fit someone's sensibilities in the extremes it makes you a target.

Is there a difference between social consequences for speech and harassment? I'm reminded of the old freedom fighter vs. terrorist debate.

Yes. The first can simply be a reaction someone has to speech and the other is a criminal offense.

Online harassment is almost always a reaction to speech. If ten thousand people send hateful messages to someone because they tweeted something they found offensive, is that harassment or social consequence?

How about some concrete examples: did Justine Sacco suffer social consequences for speech, or was she harassed by an online mob? How about Anita Sarkeesian?

We had a guy fired basically because he was an islamophobe. He was annoying people with his "opinion" all the time. That did not work out pretty well with all the other colleges so he did not get jobs. Nobody wanted to work with him. His chargeability was messed up and then he was gone.

I'm pretty OK with that.

Generalizations never work.

Someone getting fired for annoying people with unprofessional opinions is vastly different than a guy making a dick joke to his friend loudly enough that someone else heard it, was offended by it, and told his employer she demanded his resignation.

I worked for a company thatnwas owned by 2 christians. An employee, who talked about his athiest ideas all the time, was fired for the same reason (many coworkers were christian) Are you also okay with this??

My brother works with a guy that talks about islam contantly, to the point that it makes many people in the company feel uncomfortable. Should this person be fired?

i would have no problem if we gave equality to all, but its just not happening.

Yes. I'm OK with it for the same reasons.

Guy is interrupting business. If people don't want to be around him. He needs to go. The company is not there for you to spread your theist or atheist ideas. Keep it to your private life where it belongs to.

That sounds like a pretty standard HR situation. Can't do that shit at work. But your personal life should be different (no matter how inconvenient that is in some cases), and the social justice movement is looking to change that. They're causing people to lose their jobs over something that they did in their personal life; on their own time, separate from work. That's fundamentally different and it isn't justice. It isn't freedom. It's oppression.

I'm not sure it's that they're looking to change it, vs. take advantage of something for their own political ends that already exists (and has for a long time). The traditional "rule" in the US is that you can do whatever you want on your own time, unless it's something that your employer finds out about, and finds sufficiently offensive, embarrassing, or inconvenient. Then, since employment is at-will, they can fire you for it, as long as it isn't from a specific list of things that would constitute unlawful termination. The main change, as I see it, is that using it as an organized tactic (compiling lists, contacting employers, etc.) has mainly been done by conservatives in the past, directed at communists, the LGBT movement, flag-burners, etc. Those more on the left using it (or at least, using it with any success) is a somewhat newer development in the US.

To me the fundamental place to tackle this would be at the mechanism: the problem is that employers can and do fire people who have opinions they either dislike or find embarrassing. That will continue to be true, unless labor laws are changed to make it more difficult.

It sounds like that he would be fired even if he was annoying people in a different, politically-correct way.

That doesn't sound as that much of a problem.

There's a more straightforward censorship here as well: abusing the report functionality. Most of these are automatic so anyone with enough followers can easily exploit this mechanism and get other people locked out.

You and anyone interested in this should read Jon Ronson's book "So You've Been Publicly Shamed". An excellent, rational, balanced take on this issue. Also check out the works of Dr. Gad Saad. If you'd rather not take the time to read their work, they've both given speeches on this and participated in several big podcasts as well (just search their names on YouTube or iTunes).

Awareness of this issue is dramatically increasing by the day and these public figures (among others) are leading the charge.

It is incomprehensible to me that someone would be more concerned about being able to, for example, make fun of fat people without being treated as an asshole, than censorship of our mass media on issues of national security. Sorry but that's like first world dumb teenager problems to me.

You just arent thinking long-term. It starts with things like fat-shaming.

Its also not just being treated like an 'asshole'. Its a complete destruction of ones personal life, abusive behavior such as harassment, and many times the mob moves to the persons place of employment and the business doesnt want the negative publicity, and fires the employee.

Does a person deserve this sort of treatment? Arent we trying to stop online bullying?

Mobs are fickle. While today, you are on the side of the mob...tomorrow the mob might be against you. At that point, it will be too late to protest because it is already an acceptable practice in our society.

The government doesnt even really need to pass laws to control the people. The mobs will take care of it for them.

Its the wet dream of the nsa and should scate you as much as it does me.

That is some mental gymnastics you are displaying. It is not the argument 0x49 is making. It is allowing people to dissent from mainstream opinion in e.g. gay marriage without a threat to their livelihood.

This is exactly correct. You arent allowed to disagree with pretty much anything mainstream these days publically without suffering huge consequences.

The people that do successfully have the power, which means enough money that a firing doesnt matter.

You're talking about two different issues / results of censorship here, both of which are legitimate and need to be discussed. Don't silence one with the other.

That's not the argument he's making.

“You can say what you like but no one will hear you. And also, you'll think no one cared, so you'll give up trying.”

Isn't this what JTRIG's "Online Covert Operations" is all about? https://theintercept.com/2014/02/24/jtrig-manipulation/

This is not at all surprising in that censorship is one of the few viable business models for a 'free-to-play' social network.

Defusing and diffusing social movements before they get traction is something many organizations with money are going to find appealing whether or not it works...

Facebook understands this quite well and is the canonical example of the model; but if you don't think that "Steering the conversation" is a core value proposition for privately operated social networks ( including Hacker News ) that are free to the general public; you are naive.

Yet another situation to highlight http://indiewebcamp.com/webmention

Circumventing that would be near impossible.

I'm seeing this Retweet from the US. What's going on?

The writer seemed to be suggesting the censorship could be temporary, and lifted after no longer fresh and less likely to be retweeted. Also, though probably not true in your case, it sounds like actively following the account means these rules don't apply.

It seems like Twitter is experimenting with hell-banning. This is hardly a new evil.

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact