Direction of Irish immersion
I wonder if team Irish will be allowed to add both directions of immersion to the Irish duolingo.
Obviously translators usually translate into their native language and the translations done on immersion are better for being translations to the known language from the language being learned. However, it's unlikely there will ever be a English for Irish speakers course and it would be sad if that meant there was never an English to Irish immersion. English to Irish immersion would be so useful, not least for adding articles to the Irish language wikipedia, it is missing so many articles even on subjects particularly relevant to Ireland. Perhaps English to Irish immersion wouldn't produce perfect translations but is fearr Gaeilge bhriste ná nothing at all!
I have to say I disagree with this. I think there's way too much room for error, and that is a mistake. There's a lot of translations native English speakers would miss that I feel could be harmful unless they were corrected. We'd end up with a lot of Béarlachas in the stories, instead of actual Irish. Now, this wouldn't be an issue if there were enough people with highly fluent Irish to correct that. However, there isn't, so these translations are likely to slip through the cracks and be upvoted - after all, they are literally correct. I just see this leading to too many errors that hinder the learners, and think it should stay with translating from the language you're learning.
I agree with you, at least as far as official, sold translations go. Translating into a language you only have an intermediate level in is very unprofessional. Sure, I could translate something into Japanese (my most proficient non-English language) as an informal favor in order for a native Japanese to understand it, but I would never ever consider that a quality translation that would hold up to a native-level writer. A very important translation skill is not just your understanding of the languages, but being a decent writer. And Irish, much like Japanese, would be even worse, I imagine, since it is structured and phrases things so differently from English. A literal translation between English and German can sound quite natural sometimes, I've found, whereas that's rarely the case for Japanese.
What would your thoughts be, though, if Duo had an English->Gaeilge translation section that was specifically not professional (so people would just upload, say, Wikipedia articles for fun)? Would that be helpful to people learning still, or would that likely just reinforce bad phrasing, habits, etc.?
I still think it'd reinforce bad habits, but that's just me. I understand the desire to translate English -> Irish, as well as how it could be more helpful. But I don't think it will be unless there is someone to correct and, and, more importantly, explain why it is wrong. I just don't think we have enough people to do that, and I think immersion gives too much room for it to be abused by just reverting to the original, Béarlachas translation, as opposed to a more idiomatic Irish one - solely because it's more word-for-word and the person might not get the idiom or whatever.
So, really, if we could get a good core of people checking it and explaining it, yes, it could work. If not, I think it's best just to stick Irish -> English.
I don't understand the concerns about "professionalism" or that the immersion feature might get "abused". I look at it as being primarily for learners, not for the general public. I don't care so much about any document I upload or work on ever making it into Wikipedia or elsewhere. I would just welcome the exercise of shifting my thinking to Irish, as well as the exchange of ideas with others. There would be a growing core of people who would take it seriously enough to really stretch for being able to express ideas in natural Irish. I know it would motivate me to dig in even more. Again, I don't get the downsides.
Well, the issue is how Duolingo uses its immersion. Basically, that's how they get paid. That's why there's a focus on accuracy and professionalism.
As to the abuse: I've seen it happen several times where a better translation was changed to something less idiomatic because it helped the current person. Sadly, it's all too common.
> . There would be a growing core of people who would take it seriously enough to really stretch for being able to express ideas in natural Irish.
But the issue is they'll never learn from their mistakes unless they have someone to correct them. Don't get me wrong, I love the idea of translating into Irish. But I wouldn't do it unless I could get someone to also tell me where I messed up, and, more importantly, why I was wrong. That won't be available if it were to come (which, sadly, it is not).
For example, though: How would you translate: "He had a nightmare?" I would hazard that most people (and Duolingo at first, though it has been corrected) would say Bhí tromluí aige. However, this is not correct, and instead you would use Tháinig tromluí air. Why? Because it's idiomatic. But it's not the first thing people would assume, and I can easily see someone changing it back to Bhí tromluí aige, thinking it's correct.
It's those types of issues that I was worried about, since there aren't enough people here to accurately do it. Again, I love the concept of English -> Irish, but there needs to be a check to explain where it went wrong, and why, or else I feel we'd end up with a lot of Béarlachas and not as many Irish idioms.
I better understand your concerns now, and I still think it's possible to set it up in a way that encourages accuracy and discourages abuse. Wikipedia itself is a great example of that, with tiers of editors, the ability to lock an article to only being edited by people at a certain tier if less skilled people are making a mess, etc. It's a crowd-sourced effort that actually works.
Ar aon nós, it looks like this is all somewhat rhetorical since it sounds like this is not a feature coming anytime soon to the English -> Irish course. I'm glad to have gotten a better understanding, though. Thanks for your contributions to that.
As my old Gàidhlig teacher said, if people just want to speak English with Gàidhlig words, what's the point of learning Gàidhlig? Same is true with Irish. I don't see allowing a free-for-all English-to-Gaeilge translation "immersion" encouraging people to actually learning Irish, at least not in a good way.
"Perfect" and "useful" are not synonymous. The Duolingo Irish course itself is not "perfect", but I don't think many would argue it is not useful. It is improving all the time, and I have no reason to think that an immersion feature wouldn't do the same. Most of us who have persisted through the entire tree are committed to improving our understanding and use of the language and we need encouragement and resources to practice, practice, practice. If nothing but perfect were good enough, who would ever learn anything? We need the permission to make lots of mistakes, exam them and learn from them.
I am surprised that there is such pessimism about immersion; from the point of view of learning the language I would guess that nothing would be so effective at learning the inner logic, the hidden music, the beauty of the language than seeing your English-with-Irish word attempt corrected to true Irish by a helpful member of the community who enjoys perfecting other people's stumbling translations and supplying useful comments and explanations. However, it is possible that that wouldn't happen, there'd be too many articles and people would translate and move on, never commenting or correcting, with too many beginners, too few expert and corrections likely to be minor or wrong. I don't know, my experience of French immersion lies somewhere between the two.
So maybe the issue here is that immersion doesn't quite work, they have certainly fiddled with it quite a bit without managing to monetize it and @alexinIreland says below they've stopped adding it to new languages and maybe they are thinking about how to make it work. Perhaps immersion sentences need to be added to ordinary lessons further down the tree, as translations or as "which is correct", so sometimes you get a mark for answering but learn later that another version was preferred. In other words, some sentences would be marked, ones that already had a consensus translation, and others would not since your answer would be part of the process of discovering the translation.
Also, perhaps there needs to be fewer articles, perhaps it needs to mean more: obviously lots of people want different things from the site, learners want to learn, the owners wants it to make money or get bought, and others hope it will help support vulnerable languages. At the moment a wikipedia language that gets translated in immersion usually just disappears, it might work better if the site supported a shadow wikipedia of translated articles that could be easily copied over to the real wikipedia: wikipedia seems much better at correcting grammar than creating articles!
You can certainly hope for native speakers correcting wrong translations for a language like French, even though you say that your experiences even there have been mixed. But for Irish, keep in mind that its Duolingo learners outnumber all of the native Irish speakers on this planet, and even more so those on Duolingo.
Your idea is interesting, and we proposed this to Duolingo a while ago (though we are aware of the issues and limitations that English to Irish translation would include). Unfortunately, Duolingo doesn't have plans to expand any type of immersion to newer courses in the near future.
That's really a shame. Translating, especially to the target language, is such a good activity for increasing one's understanding. Without it, there's not a whole lot left to do on Duolingo once you've finished a tree, other than to do strengthening exercises for review and to participate in the forums. The immersion feature seemed like such a natural next step within the Duolingo framework. On to other things, I guess.
Well, the main point of immersion is to earn income for Duolingo and since that isn't working out brilliantly right now, Duolingo are instead going to focus their resources on other projects (though immersion will still remain and the Duolingo users will still do paid translations for CNN and Buzzfeed)
Well, that's not exactly what I meant. There are no plans for immersion for the newer courses right now (so that includes Irish, Dutch, Danish and Swedish). This doesn't necessarily mean it won't be launched eventually :) (Though I can't be certain, I'm just going by what Myra told me)