After months of hard work, we’ve just added our 144th language to the uTalk app—Ladino! In this blog post, we share how we go about adding new languages, as well as the unique issues that arise when it comes to adding languages that are endangered.
Way back at the beginning of 2019, two individuals met and had a conversation that would eventually lead to the development and release of Ladino on the uTalk app. Specifically, Carlos Yebra López, from the organisation Ladino 21, met Kevin Sun, a man who has famously topped the leaderboard at many of our Kahoot quizzes—and they got to talking about the ways in which Ladino, an endangered language that Carlos has been trying to help preserve and promote, could be further helped.
So when uTalk was then approached by Carlos in March, it was the first time many of the staff had even heard of the language. Carlos offered the assistance of Ladino 21 in translating our script and finding native speakers and it was clear from the start that uTalk’s values lined up with what Ladino 21 want to achieve—record and preserve Ladino via digital means while reaching a wider and younger audience.
How do we go about adding a new language?
At 144 languages, it’s clear that we have a system going that works for us—although, of course, problems can still arise. In fact, every language comes complete with its own set of issues, and this is especially true when it comes to minority or endangered languages such as Ladino.
The first step is always choosing a language and finding people from that language community who would like to get involved and help. If we find no speakers, then that language doesn’t go ahead—but with Ladino, we were lucky, as Ladino 21 approached us first.
So, with the language chosen, it’s time to look for the first bit of help we’ll need: translators. Since the script we have is the same for every language, we get hold of two translators, who work independently of one another. With around 2,500 words and phrases, this is about a week’s worth of work, though it can take more time than that, depending on what other work our busy translators have to do!
Once they’ve both translated the entire script, they then swap files and check each others’ work. This is an easy way to spot any common errors or hash out initial discrepancies. After this, the translated scripts are returned to our Languages Manager, who checks them and then jumps on a call with the translators to go over any lingering issues. There are always a few, but by the end of this stage, everything is more-or-less ready to go.
Next, it’s time for our voice actors, aka, the people you hear when you use the app. For every language, we source one female and one male native speaker (very few languages are exceptions to our native speaker rule: Latin and Ancient Greek being the obvious ones!) and have them come and record at our studio.
The advantage to our office being based in London is that London is one of the most multicultural cities in the world, with approximately 200-300 languages spoken here. This means that when we need native speakers for the app, we usually only have to put the word out and we can find our voice actors locally.
However, with Ladino, much like with Manx, we had to search a little further afield. Fortunately, Carlos and his deputy director, Alejandro Acero, already had people in mind. Our voice actors were sent the script, as they usually are, to also get two new sets of eyes on it, and then flew over to record.
There are always more checks on the script while recording—as any translator knows, sometimes what is written down does not always lend itself well to speech!—and this recording takes at least a couple of days with both speakers sitting in and listening out for any mistakes.
After the recording is all wrapped up, we go into post-production: editing the audio for things like background noise and mouth noise (a super common thing!), and then into app building, lots and lots and lots of testing, and finally the release.
What about Ladino?
Working with Ladino 21 was really a great partnership for us because they already had speakers in mind for the app and on top of that, they offered to share their connections in the Ladino community, which is important for us when trying to drum up support. By October, we were all ready to go with the recording, our voice actors were ready to fly over… but then, disaster struck!
At the last minute, one of our speakers couldn’t make it, so another Ladino speaker, Benni Aguado, was roped in, dropping everything in New York last minute so he could fly over and help out. This did mean, however, that he didn’t really have a chance to check out the script before his arrival in London.
Between this and the fact that Ladino is now spoken by a diaspora that is geographically scattered, there was a lot of discussion throughout the week between our voice actors, Rachel Bortnick and Benni Aguado, as to the word or phrase they would use in a specific situation, as sometimes the two had very differing opinions. Sometimes one of them would use a word the other had never heard of, too!
One of these examples is the word for ‘kite.’ In the app, we have chosen to use the word ‘la filandra,’ which is used among Ladino speakers from Turkey or Greece. However, Rachel’s family—and she herself—are from Izmir, where they instead say ‘el hacho.’ This was something that had to be discussed, especially as the word ‘hacho’ actually comes up in a different context: jellyfish! The Ladino word for jellyfish is ‘el hacho de mar,’ or, ‘the kite of the sea,’ but ‘la filandra’ was chosen for ‘kite,’ not to lose out on this imagery, but for the fact that it is likely to be understood by a higher number of speakers.
So, there was a lot to consider and this meant that recording took about a week! After the recording was finished—and after Rachel and Benni had returned home to Texas and New York, respectively—the post-production fun began, with Emily, our Languages Manager, working hard on audio editing before everything was handed off to be built into the app.
Checks after checks and tests after tests were performed, including a round of everyone in the office doing their best to update their app after Ladino first went live, and what exists now in the app is a labour of love that has taken our company around nine months to achieve.
And that nine months? While it’s longer than it usually takes for us to get a language finished and in the app (we estimated that our absolute quickest time to add a language would be around a month), considering the scope of this project—a severely endangered language, speakers scattered around the globe, a very fixed time limit on how long the recording could take—nine months is not all that bad.
We don’t think so, anyway!
Of course, if you’d like to get your hands on Ladino, you can get a subscription (with 40% off) here. You might not find a native speaker local to you (though you might be surprised, especially if you’re in the US!) but there’s a strong online community who will surely all be happy to speak with you.
From all of us at uTalk, mazal bueno (good luck) with your Ladino journey—and drop us a comment if you start learning; we’d love to hear all about it!