Google Translate and Mini Stories: How we can do half the job without knowing the language

As I recently posted Google Translate has added three indigenous Latin American languages, Guaraní, Aymara and Quechua. Obviously there are still problems, for example Quechua is not a language but a language family, and some translations seem to be a bit awkward but the point is: It’s out there and everyone can use it.

So this brought me to the question how we can use this tool and make it possible to eventually add these languages to LingQ. The main requirement are the 60 Mini Stories and to LET them translate by a translator was by now too laborous and expensive.

But I might have a solution: We can create Google Docs with the 60 Mini Stories (In Latin American Spanish) with one column for the stories in Spanish, one with the translation from Google Translate and then we leave two columns for native speakers, where they can correct the GT version and write some remarks if they want. This would be super easy to use and you can share the link with many native speakers.

What do you think? Here is an example:

1 Like

why spanish?

Other than that, when I see these posts on indigenous languages, I always think “is there so much to read in a langauge like aymara?”

Lingq is mostly about reading, so just how much interesting content is there to import? Worthwile literature? netflix? Youtube channels? (It’s an open question, I don’t really know anything about these languages)

1 Like

Spanish because the speaker of these languages are mostly bilingual in their language and Spanish. You could do the same thing for another language with English as a base. Guaraní for instance is an official language in Paraguay with (let me not lie) more than 60% of the population speaking the language. Obviously there exists less material than for Spanish or French but still there is material and hopefully there will be soon more.

I’ve noticed that for catalan the google translate is better between spanish and catalan than between english and catalan. So kind of supports what’s said above

1 Like

Para todos los interesados, pueden echar un vistazo al documento que ya ha avanzado mucho. A mí me gusta mucho cómo es la dinámica del proceso.

Google Translate + editing is definitely the faster way of translating a document. If you have a little bit of money to pay for an (doesn’t need to be a high-class, maybe just a uni student or something) editor, it’s definitely the cheaper option. If you think others may be interested in the language, you could always try and crowdsource some money.

I come back to this thread because some users constantly demand certain new languages and there is even a discussion about the necessity of MiniStories going on. In my opinion they provide some good beginner material but apart from that it’s just logical that LingQ needs some basic requirements for adding a new language. They can’t just add X-language because 5 people in the forum want so. This costs time and money and although LingQ is quite different, they still compete with Apps like Busuu, Babbel or other language learning apps that by far don’t have this variety of languages. Here, people cry for having Tagalog but at the end how many people really learn Tagalog on LingQ? I mean for the few Tagalog or Armenian learners that’s great but you also should consider that it is a generous user service of LingQ. For marketing issues it would be better to promote the “big” languages.

So you can’t expect that LingQ creates the Mini Stories and finding natives who volunteer in translating 60 stories will be really rare and even then there might be quality issues.

So I repeat my idea: If you want to have for instance Hindi on LingQ, why don’t you translate the 60 stories with Google Translate and then you can give them a native for correction. Maybe you know some basics and can correct bigger mistakes on your own, so that there is less work for the native. You can make the corrections also with your italki tutor, you can post the texts on Journaly or italki or you find a tandem.

Google translate obviously isn’t perfect but with the simplicity of the MS it won’t fail to much. Let’s take for example this sentence from the first MS:

I make breakfast and drink a coffee.

I mean this is so easy. If there is any Hindi speaker here, maybe you can tell me if this sentence is correct:

मैं नाश्ता बनाता हूं और कॉफी पीता हूं।


Agreed. As much as I’d like to have every small language on LingQ, at the end of the day, LingQ is a business and needs to priorities its money-making languages, which in turn provides excess funds that can be used for smaller languages. Already as it is, LingQ offers a lot of niche languages that I’ve never seen on any other language learning service - it’s not as if LingQ isn’t already doing a good job of providing resources for lesser learned languages. But, you’ll never please everyone, especially on a site like this where even the smallest minority has the ability to publicly express their displeasure on a public forum.

Then there’s also the fact that adding a new language isn’t as simple as just adding it to the list and letting people use it. There’s a lot of work that their dev team will have to do for each new language. Yeah, once a language is added, it’s pretty much self-sustaining - but until then, they have to add dictionaries, hook up TTS, add the new forums for that language, add the language to the importer, make different websites in that language usable with the importer, etc… It’s a lot of work, no doubt.