Duolingo: Censoring Great Literature?
Would it be possible to reconsider the policy that is evidently prohibiting uploads of some of the greatest literature ever written? This spring I have repeatedly observed that the work of the greatest authors in the Spanish language (Iberians and Latin Americans) is blocked from uploading for group translation because "DuoBot thinks [it] might not be appropriate for language learners of all ages." I can't even upload classics that I read in school in other languages when I was 8, 10, 12, 14 years old! At least to me, it's ironic, if not ludicrous, to block such classics when the full range of mediocre and downright trashy content is readily available on the internet. Nothing wrong with Wikipedia, but it's not a bastion of creativity, and when we rule out great literature, we rule out some of the most challenging and educational content. Wouldn't it be simple to enter the names of specific distinguished authors (Cervantes, Perez Galdos, et al.) as overrides to the stupid bot? And what about regional variations: is it worth forfeiting anything that refers to "catching the bus" in Castilian Spanish simply because the same Spanish words used in Mexico would be understood as "f---ing the bus?"
This comes up frequently, and the Duo staff knows that it's a problem. It's on their ever-growing to-do list, but I wouldn't expect it to be addressed any time soon (that's just a guess on my part, though). They implemented a filter that rejects content based on a blacklist of words / phrases and a lot of innocuous articles run afoul of it.
As for translating classic literature, I'm not sure that I'd place much value on translations of bilinguals that aren't themselves experts on the subject matter being translated, much less the crowdsourced translations of novice language students. This type of writing is generally far too nuanced to be translated well by those who lack an intimate familiarity with the subject matter.
It would be excruciating to read, I hope it would never get put on the internet somewhere else! But it might be an interesting translation exercise.
The issue with the filters goes way beyond a wonky filter with innocent articles being rejected, though, otherwise we wouldn't have all the tacky porn actor articles showing up.
I've not encountered any of those myself, since I don't do much translating currently (tedious edit wars turned me off). I've seen them mentioned a time or two on the forum, though. Aren't those commercial pieces? I imagine that the for-pay material from Buzzfeed and the like aren't subject to the same (or any) submission filters that user submitted articles are.
No, they're not - though the Buzzfeed stuff is pretty grim.
Oh, wow. I'm no prude, but I'm surprised that this article, or others like it, could survive a report to the Duo staff. I can only assume that articles added by Duobot are not screened in the same way that user submitted ones are checked for inappropriate content.
But they do survive, repeatedly. And that's by no means the worst we've had, though those haven't all been Duobot uploads. By and large, though, if it was a porn actor, it was Duobot. And it wasn't random, it wasn't as if it was working through Italian or French Wikipedia in alpha order or generally uploading all the actors. It does seem to have settled down recently. But the point is that the rules still say nothing with sexual content and it still primly rejects completely sex/drugs/violence -free articles as 'inappropriate for all ages', while Duobot and other users upload this sort of stuff: https://www.duolingo.com/translation/64b6ec93b0a2b66bb87f78fa5e36148c There's no logic to any of it.
I won't go on, but a quick site search on censorship may intrigue you.
@Luscinda No, I can see that they clearly do manage to get through and, even more surprisingly, stick around. I was just speculating on how that could happen. Clearly they have some serious gaps in their screening system. I wonder if checks are skipped for Wikipedia articles, assuming that its content is OK.
It would be an interesting experiment to try to upload the same article from another source (say cut and pasted into a tumblr or blogger.com entry). I'm just not sure I want my user name to be associated with those experiments haha.
It's not a gap in the system that doesn't delete them when reported, though. Nor is it a Wikipedia exemption, everything I get rejected is on Wikipedia. I do recommend a look at the past threads for comedy value.
Meanwhile, I'm pondering whether it is this conversation that has just encouraged some loser with an American comedy user name and a sadly inadequate flasher photo for an avatar to 'follow' me. To which I can only say sod off back to school, you silly child. :-/
@Luscinda Yes, the fact that they're not taken down when reported is really surprising, but that's a breakdown in a different system I suppose.
I've just started translating and have been dismayed by the deadly dull range of articles I've encountered so far. If I wouldn't waste time reading it in English, I sure will not want to spend days translating it from Spanish. I was wondering where the literature was. So I'm alarmed to hear this. I'm doing Neruda and Paz at home but that gives me no credit on Duolingo.
If you don't like what others have uploaded, upload something yourself. It's as much your responsibility as anyone else's.
I am wondering about needing to get author permissions for blogs and things like that. And how to figure out if something online is public domain or not. Are there guidelines for this here?
I agree this issue needs to be addressed. One of the problems arising from the modern expectation that children must be allowed everywhere is that there is nowhere left where adults can actually be adults.
I can only share your pain.
On my personal experience I can only say this: On my German practice, I put the "tree" aside after a while and only came to it one in 10 times I was using DL. The other 9 I was reading uploaded blogs that talked about local culture, usage and some Germans seeing the world (and describing it) from their perspective. As you see, not precisely great literature classics, just a cultural insight.
That was quite some time ago and then DL brought the censorship and the "good practices" and at the same time a myriad of different efforts filled the page... without any specific aim, just make it more "playable" (apparently) The day I could not find more than Wikipedia articles I stopped using DL for learning German.
I do not want to go on with the low points from DL or what I infer from their policies, but I cannot but be amazed of several facts or just perhaps only one: Censorship. Why? When one wants to get the culture of a country or a population through their language, they must immerse in it. And there are "bad" words, because the culture and the day to day life is full of "bad" things.
It is my understanding that certain cultures do ban words, making them taboo words. Watching the world through a coloured polarized glass, might save your eyesight, but you won't see much.
And that seems to be the opposite of what DL claims to be doing.
:Snorts: Have you looked at the various threads on what is and isn't barred from being uploaded by the auto censor? You might find it entertaining.
You need to report rejected legitimate uploads. Through the support tab, with a note pointing out why it's more acceptable than the rubbish Duobot uploads.