Content Moderation Case Study: GitHub Attempts To Moderate Banned Words Contained In Hosted Repositories (2015)
from the word-filters dept
Summary: GitHub solidified its position as the world's foremost host of open source software not long after its formation in 2008. Twelve years after its founding, GitHub is host to 190 million repositories and 40 million users.
Even though its third-party content is software code, GitHub still polices this content for violations of its terms of service. Some violations are more overt, like possible copyright infringement. But much of it is a bit tougher to track down.
A GitHub user found themself targeted by a GitHub demand to remove certain comments from their code. The user's code contained the word "retard" -- a term that, while offensive in certain contexts, isn't offensive when used as a verb to describe an intentional delay in progress or development. But rather than inform the user of this violation, GitHub chose to remove the entire repository, resulting in users who had forked this code to lose access to their repositories as well.
It wasn't until the user demanded an explanation that GitHub finally provided one. In an email sent to the user, GitHub said the code contained content the site viewed as "unlawful, offensive, threatening, libelous, defamatory, pornographic, obscene, or otherwise objectionable." More specifically, GitHub told the user to remove the words "retard" and "retarded," restoring the repository for 24 hours to allow this change to be made.
Decisions for GitHub:
- Is the blanket banning of certain words a wise decision, considering the idiosyncratic language of coding (and coders)?
- Should GitHub account for downstream repositories that may be negatively affected by removal of the original code when making content moderation decisions, and how?
- Could banned words inside code comments be moderated by only removing the comments, which would avoid impacting the functionality of the code?
- Is context considered when moderating possible terms of service violations?
- Is it possible to police speech effectively when the content hosted isn't what's normally considered speech?
- Does proactive moderation of certain terms deter users from deploying code designed to offend?
Unfortunately for GitHub, this drew attention to its less-than-consistent approach to terms of service violations. Searches for words considered "offensive" by GitHub turned up dozens of other potential violations -- none of which appeared to have been targeted for removal despite the inclusion of far more offensive terms/code/notes.
And the original offending code was modified with a tweak that substituted the word "retard" with the word "git" -- terms that are pretty much interchangeable in other parts of the world. The not-so-subtle dig at GitHub and its inability to detect nuance may have pushed the platform towards reinstating content it had perhaps pulled too hastily.
Originally posted on the Trust & Safety Foundation website.
Thank you for reading this Techdirt post. With so many things competing for everyone’s attention these days, we really appreciate you giving us your time. We work hard every day to put quality content out there for our community.
Techdirt is one of the few remaining truly independent media outlets. We do not have a giant corporation behind us, and we rely heavily on our community to support us, in an age when advertisers are increasingly uninterested in sponsoring small, independent sites — especially a site like ours that is unwilling to pull punches in its reporting and analysis.
While other websites have resorted to paywalls, registration requirements, and increasingly annoying/intrusive advertising, we have always kept Techdirt open and available to anyone. But in order to continue doing so, we need your support. We offer a variety of ways for our readers to support us, from direct donations to special subscriptions and cool merchandise — and every little bit helps. Thank you.
–The Techdirt Team
Filed Under: code, content moderation, repositories
Companies: github
Reader Comments
The First Word
“PC: It's what basic human decency is called by people who have none.
Subscribe: RSS
View by: Time | Thread
Being familiar with the history of that specific program, I can say this: They were decidedly not using that word in the “inoffensive” way.
[ link to this | view in thread ]
Re: PC Retard...
So I'm not allowed to complain about the retarded idiot that put a 20V nominal capacitor into a spot in the circuit where it will see 30V????
I suppose I could use stupid, mindless, or addled instead, but i think all of those terms have been applied to those with mentally difficulties in a formal medical diagnostic context, so what pejoratives are allowed???
I'm still adjusting to everything I write becoming public and permanent!
[ link to this | view in thread ]
Re: Re: PC Retard...
It doesn't matter when you clearly don't give a fuck. So don't worry about it.
[ link to this | view in thread ]
As long as they don't filter my pron. :)
[ link to this | view in thread ]
Re:
Don't worry, they are. After all it's the most offensive thing they can try not to think of.
[ link to this | view in thread ]
Re: Re: PC Retard...
Saw this happen with the wine repos recently. Made things confusing because they couldn't use the terms: master, slave, whitelist, or blacklist.
PC of course, would have the entire project just obliterated from existence on two marks:
It's namesake is an addictive substance that harms and kills people.
So obviously, wine shouldn't exist.
/s
[ link to this | view in thread ]
Re:
Shouldn't matter either way.
But then again, this is the road that was chosen. Letting a bunch of PC idiots dictate to the world what can be said and done because their right to contract makes all other rights irrelevant.
[ link to this | view in thread ]
Re: Re:
Who would have thunk that github were spelling Nazis?
[ link to this | view in thread ]
I host master files for many ebooks I produced for Project Gutenberg on GitHub (copyright expired, so that is no issue). Some of these use words that are pretty offensive to some people today. Filtering them out would be a falsification of history, so would force me to take those ebooks away from GitHub. Not seen a notice yet. Lets see where this ends.
I wonder, would they be able to host the code that checks for those words on GitHub, or would that be banned as well.
To prevent comment spam, I have a list of bad words on my personal website as well, but that is mostly filled with the names of luxury brands and certain medicines.
[ link to this | view in thread ]
Re: Re:
Sounds like you're referring to some Christian fundamentalists' talking points:
"My right to be a bigot against gay people trumps your right not to be discriminated against and fire you for acts that harm nobody and take place outside office hours in the privacy of your own home."
"My right to be an idiot gives me the right to force doctors to provide false information on reproductive health to women in a medical emergency situation."
[ link to this | view in thread ]
I realize you’re being facetious and all. It’s a bad look for you.
Buuuuuuuuuuuut…
I have a couple of lists for alternatives to “stupid”, since people differ on whether the word is ableist. Besides, some of these are sincerely better alternatives because they can denote agency of some kind. “Stupid” and its “cousins” (including “idiot”, “crazy”/“insane”, “hysterical”, and the r-word) generally do not.
absurd
asinine
awful
banal
baseless
clueless
dull
erroneous
fallacious
fals e
fatuous
foolhardy
foolish
frivolous
gullible
ignorant
ill-advised
ill-considered
illogical
imperceptive
imprudent
inaccurate
inane
incompetent
inept
indefensible
in judicious
insipid
irrational
laughable
ludicrous
misguided
misinformed
mistaken
naï ve
nonsensical
obtuse
oblivious
outrageous
pathetic
pig-headed
pointless
preposterou s
puerile
rash
ridiculous
senseless
shallow
silly
simplistic
stubborn
substandard
thoughtless
trifling
trite
trivial
unfounded
uninformed
unreasonable
unsound
unten able
unwarranted
unwise
useless
vacuous
vapid
waste of time
wasteful
willfully ignorant
worthless
Other similar replacements:
don’t be stupid → try to make good decisions
You may also want to think about the variety of uses/meanings when people default to “stupid” and examine the implications of what a particular culture attributes to intelligence. (Basically, it comes down to low intelligence being inherently bad and undesirable, and with the sense that a person can choose their intelligence.) You have far more descriptive and detailed ways to use language than relying on ableist slurs.
And if anyone reading this comment is thinking “BuT iT’s ToO hArD tO cHaNgE tHe WaY i TaLk!” or something along those lines? Yeah, it can be difficult — at first. I’m not perfect when it comes to avoiding “stupid” and its “cousins”, either. But when you warm up to the idea of using the alternatives, replacing shitty language with personally considerate language is a habit that’s easier to form than you might believe. It also doesn’t cost you anything but a few words on which you were likely overrelying.
[ link to this | view in thread ]
Alternative terms to the “master/slave” and “whitelist/blacklist” dichotomies already exist. Not using those alternatives is a decision only the Wine devs could make. Nobody could force the devs to use the alternative terms — and the devs couldn’t force a third party to host their code if they chose to keep using the original terms.
As for your other two assertions: Your inanity is hilarious in a pitiful way.
[ link to this | view in thread ]
Sounds like you’re upset that the language and ideas you prefer aren’t “politically correct” and you wish they were. That’s life, fam: You don’t ever get everything you want.
[ link to this | view in thread ]
I would assume GitHub mods can recognize the difference between novels from decades-to-centuries ago using such language and a person in the here and now using such language as part of a README, project name, or comment.
[ link to this | view in thread ]
Re:
You assume a lot.
[ link to this | view in thread ]
Re: Re:
"bunch of PC idiots dictate to the world what can be said and done "
Are you complaining that our language is dynamic? What sort of ignorant rant would proclaim such silliness?
Yes, language changes. Our environment is dynamic and our language is used to describe said environment, therefore it too must change. This is rather obvious.
[ link to this | view in thread ]
Re:
wait .. there's pron on github?
[ link to this | view in thread ]
Re: Re:
If you find pornography to be offensive, perhaps you are watching the offensive porn.
[ link to this | view in thread ]
Re:
Do they allow encrypted files?
[ link to this | view in thread ]
Re: Re: Re: PC Retard...
"couldn't use the terms: master, slave"
I prefer the terms primary and secondary as they are much more descriptive of their purpose. But do not let that interfere with your bigotry.
Why do you not complain about black friday? Seems like the obvious target to me.
[ link to this | view in thread ]
Re: Re: Re: Re: PC Retard...
Primary / Secondary are not equivalent to Master/ Slave.
In M/S the slave does nothing until the master initiates a contact. It is a question and answer sort of relationship
In P/S the secondary could be a backup service or a number of other scenarios where both could operate independently
The problem with this is the intent. Bigotry implies that the terms are being used with a hurtful intent. If this is not the case i.e. it is purely descriptive, the problem is not necessarily with the user. The article was about using the word retard - was there an intent to upset? No, so it is ok. Why does everyone have to bow down to the most offended denominator or risk being called a bigot/racist/<insert others>?
[ link to this | view in thread ]
Re: Re: Re: Re: Re: PC Retard...
parent/child, server/client, controller/worker. There are plenty of alternatives that are equivalent in one context or another.
Even if the offensive terms aren't used with intent, remember that the purpose of documentation is communication and that an important part of communication is engaging with your audience. Using words that make your audience think you're an aging boomer who is unable to adapt to the times isn't a good look in the tech industry that values youth and adaptability.
[ link to this | view in thread ]
Re: Re: Re: Re: Re: PC Retard...
Specifically, in wine, what is this master/slave usage that you claim primary/secondary are not usable?
"In M/S the slave does nothing until the master initiates a contact. It is a question and answer sort of relationship"
That is one usage. Similar to NTP, yes?
In NTP there is a server and there are clients.
There is no need for the continued use of the master/slave terminology as it is easily replaced, many times with more descriptive terms.
[ link to this | view in thread ]
Re:
Plus it can be fun to use high-falutin' words like vacuous, asinine or preposterous to emphasize your point, rather than the vapid vocabulary usually seen in such comments.
[ link to this | view in thread ]
You can say bigoted things, and language can imply bigotry, without bigoted intent. Think about the “blacklist/whitelist” dichotomy: Why is “black” the bad one and “white” the good one? (This is why I prefer “allow list/deny list”: They’re less fraught with that issue by virtue of not being a bit of metaphorical language.)
You’ll find similar phrases/dichotomies that use “black” as a stand-in for “bad” (and “white” as a stand-in for “good”) if you think hard enough (e.g., “blackball”). It’s a subtle effect of language that I’ve heard called the “bad is black effect”, and it’s especially pernicious in the U.S., where race relations have always been…troublesome, to say the absolute least.
As for the “master/slave” dichotomy, “primary/secondary” is not the only suggested replacement. You could always try these on for size:
It really isn’t hard to alter your language patterns in ways that are personally considerate to others.
Oh, and a few of other things:
The original name of the program was “WebM for R⸻ds”. The program still has that name on a different code repo site. The name implies that the program is so easy to use, even an intellectually disabled person could use it. (The subtle backhand there: If you couldn’t get it to work, you were…well, I think you get the picture.) That doesn’t strike me as an “inoffensive” use of the word, regardless of intent — and I’m pretty sure the intent wasn’t “inoffensive”, either.
In case you didn’t know already, I use “personally considerate” instead of “politically correct” for two reasons. One, it’s more descriptive of that kind of language and the thought processes behind it. Two, it sometimes pisses off people who don’t like being reminded of their inconsideration for others.
Do I think the idea of “P.C.” language can go a little too far? Yes. Overzealousness is a thing; it can affect everyone — including the personally inconsiderate/politically incorrect. But what do you have to lose from being a little more personally considerate?
[ link to this | view in thread ]
PC: It's what basic human decency is called by people who have none.
[ link to this | view in thread ]
Re: Re: Re: Re: PC Retard...
Let me blackball this comment! ;->
[ link to this | view in thread ]
They’re a private corporation. They can do as they please, right?
[ link to this | view in thread ]
Once again: Having the right to do a thing doesn’t make it the right thing to do.
Although in this instance, yes, it was the right thing to do.
[ link to this | view in thread ]
Re: Re: Re:
No the complaint is that their usage is being interfered with by goddamned morons. Programmers get quite enough of idiotic tweaks from management at work
To say it is complaining about language being dynamic is disingenuous. Language usages are also allowed to diverge in practice. Last I heard there wasn't a crusade to standardize color/colour or if pants should refer to a garment which should be worn on the outside or not. The changes pushed don't actually help those affected by it but a bunch of refugees from the Saphir-Worf hypothesis's discredittint are acting like they are saving society by being the equivalent of church ladies or Home Owner Associstion members from hell, or stereotypical liberal arts grad students vying for tenure looking for any excuse to discredit competitors. Regardless of unflattering comparison they all come across as shoving their nose into other people's business and transparently trying to grab any ounce of social power and control they can get. It is more than reasonable to have an opinion that these sorts should just fuck off.
[ link to this | view in thread ]
Re: Re: Re:
Offense does not exist in a vacuum it is always taken by the beholder. Cutting someone's guts open on the top of an altar and gladiators fighting to near death and having their life decided upon a whim were once "not offensive".
In this case them not having control over the sexuality of others is whar is judged as offensive.
[ link to this | view in thread ]
Re: Re: Re: Re:
Wow, ok.
[ link to this | view in thread ]
Re: Re: Re: Re:
why you so mad, bro
[ link to this | view in thread ]
Re: Re:
Another alt-righter complaining that other people have private property rights. It's strange that so many of them are communists.
[ link to this | view in thread ]
Anyone who genuinely believes the use of personally considerate language is “saving society” can be rightfully dismissed as ridiculous. The use of that language is but a tiny step forward in changing society for the better.
You’re not legally beholden to using it. But it’s not a bad idea only because you’re not the one who gets to decide what language is “personally considerate”.
Given your crack about “liberal arts grad students”, I can safely assume you must be unfamiliar with Republicans’ attempts to do what you accuse “liberals” of doing. The GOP, after all, is the party that wants to criminalize a medical procedure (and possibly jail the women who undergo that procedure).
[ link to this | view in thread ]
PC : It's meaning
"Politically Correct" ... simply meaning that it's worded in such a way to be as inoffensive as possible while white-washing it's actual public revulsion.
Basically, if you want to be PC, you go ahead an be one of the "many bloodsuckers" re-writing what is "correct"
[ link to this | view in thread ]
In my experience, those who whine the loudest and longest about “political correctness” tend to be the people who want to say lots of heinous shit without facing any consequences.
[ link to this | view in thread ]
Re:
In my experience, those who speak in such a "politically correct" way are the ones who have the most to hide, usually about the subject they are downplaying by being politically correct.
[ link to this | view in thread ]
I don’t misgender or deadname trans people. That doesn’t mean I’m hiding some sort of secret trans identity.
…although if cyborg bodies a-lá Ghost in the Shell were a thing… 🤔
[ link to this | view in thread ]
Re: Re:
What is it you think they're hiding? I'm not sure if you're saying people who object to gay slurs are secretly gay, and people who object to racial slurs are... secretly non white? Or what.
[ link to this | view in thread ]
The Cloud
"The Cloud" is just someone else's computer and all the BS that goes along with it.
[ link to this | view in thread ]
Re: Re: Re: Re:
No, it was offensive.
Just that fearless leader at that time disallowed dissent.
[ link to this | view in thread ]
Re: Re:
Defeats the purpose of wanting to share things, right?
[ link to this | view in thread ]
Re:
My objections are entirely upon the people forcing the change and their methods not the changes. By all means use or suggest a clearer term than master/slave for the context to clarify it without the awkwardness and misunderstandings. Leader/follower, boot drive/expansion drive. Instead they aggressively take their context and shove it in everywhere else, assume everything is in the worst possible context and flip out and summon their flying monkeys.
I actually know what the liberal arts are unlike you. The toxic academic culture which finds novel statements and standards to try to bully their way to power and say the dumbest shit in its service of a "cause" which looks awfully like a tool. Noticably the worst among liberal arts majors where the fight for tenure is the most intense.
Your whataboutism is a demonstration of the prepostorus with us or against us mentality. It is the same insanity as a QAnon asserting that you support pedophiles for pointing out that the pizza place doesn't even have a goddamned basement. Lets not have standards set by complete fucking morons - I thought recent events made it obvious why doing so is dangerous.
[ link to this | view in thread ]
Re: Re:
Do you ever get tired of comstant lies and projection like that?
[ link to this | view in thread ]
That sounds like a Trump rally.
[ link to this | view in thread ]
Re: Re:
I've never seen anyone who whined about OpenZFS removing the derogatory language, who was even remotely affected by it.
Only little shits looking for something to cry their triggered racist eyes out over.
[ link to this | view in thread ]