AI Parallel Texts for Learning Two Similar Languages

I’ve seen a fair few social media posts recently about linguist Michael Petrunin’s series of Comparative Grammars for polyglots. They seem to have gone down a storm, not least because of the popularity of triangulation as a polyglot strategy.

They’re a great addition to the language learning bookshelf, since there’s still so little formal course material that uses this principle. Of course, you can triangulate by selecting course books in your base language, as many do with Assimil and other series like the Éditions Ellipse.

Parallel Texts à la LLM

But LLMs like ChatGPT, which already do a great job of the parallel text learning style, are pretty handy for creative comparative texts, too. Taking a story format, here’s a sample parallel text prompt for learners of German and Dutch. It treats each sentence as a mini lesson in highlighting differences between the languages.

I’m learning Dutch and German, two closely related languages. To help me learn them in parallel and distinguish them from each other, create a short story for me in Dutch, German and English in parallel text style. Each sentence should be given in Dutch, German and English. Purposefully use grammatical elements which highlight the differences between the languages, which a student of both does need to work hard to distinguish, in order to make the text more effective.

The language level should be lower intermediate, or B1 on the CEFR scale. Make the story engaging, with an interesting twist. Format the text so it is easy to read, grouping the story lines together with each separate sentence on a new line, and the English in italics.

You can tweak the formatting, as well as the premise – specify that the learner already speaks one of the languages more proficiently than the other, for example. You could also offer a scenario for the story to start with, so you don’t end up with “once upon a time” every run. But the result is quite a compact, step-by-step learning resource that builds on a comparative approach.

ChatGPT creating parallel texts in German and Dutch with an English translation.

ChatGPT creating parallel texts in German and Dutch with an English translation.

Variations and Limitations

I also tried prompting for explanatory notes:

Where the languages differ significantly in grammar / syntax, add an explanatory note (in English) to the sentences, giving details.

This was very hit and miss, with quite unhelpful notes in most runs. In fact, this exposes the biggest current limitation of LLMs: they’re excellent content creators, but still far off the mark in terms of logically appraising the language they create.

It is, however, pretty good at embellishing the format of its output. The following variation is especially impressive in an LLM platform that shows a preview of its code:

I’m learning Spanish and Portuguese, two closely related languages. To help me learn them in parallel and distinguish them from each other, create a short story for me in Spanish, Portuguese and English in parallel text style. Each sentence should be given in Spanish, Portuguese and English. Purposefully use grammatical elements which highlight the differences between the languages, which a student of both does need to work hard to distinguish, in order to make the text more effective.

The language level should be lower intermediate, or B1 on the CEFR scale. Make the story engaging, with an interesting twist.

The output should be an attractively formatted HTML page, using a professional layout. Format the sentences so they are easy to read, grouping the story lines together with each separate sentence on a new line, and the English in italics. Hide the English sentences first – include a “toggle translation” button for the user.

Claude by Anthropic creating an HTML-formatted parallel story in Spanish and Portuguese.

Claude by Anthropic creating an HTML-formatted parallel story in Spanish and Portuguese.

It’s another use case that highlights LLMs’ greatest strength: the creation of humanlike texts. For linguists, it matters not a jot how much (or little) deep understanding there is beneath that. With the language quality now almost indistinguishable from real people-speak, AI texts serve as brilliant ‘fake authentic’ language models.

e-Stories as parallel texts are yet another fun, useful flavour of that!

A picture of foam pyramids to illustrate triangulation, a language learning technique. Image from FreeImages.com

Triangulation : Familiar Stepping Stones To New Languages

It bears repeating: triangulation is one of my favourite language learning methods, and one I recommend to anyone who has at least one other foreign language already and is looking to add a new dimension to their polyglot journey.

For newcomers to the technique, it’s certainly worth going back over what triangulation is, how it works, and why it can be more effective than learning through English. I’ll also share some of my personal favourite resources for triangulation, covering various language pairs and levels. (You knew it would come to back to books eventually!)

What is triangulation?

Triangulation is a language learning method that uses one of your stronger foreign languages to learn a new one, bypassing English. For example, if you already know French, you can use French as your base language to learn German through French materials like books, podcasts, and courses.

How does triangulation work?

Triangulation works by leveraging your existing knowledge of a foreign language to acquire a new one. By using a foreign language as your base, you activate both languages in your brain, creating connections between them. This can help you improve your vocabulary, grammar, pronunciation, and comprehension skills in both languages simultaneously. You can also benefit from the exposure to different cultures and perspectives that each language offers, without looking through the anglophone lens.

Why is triangulation more effective than learning through English?

Triangulation can be more effective than learning through English for several reasons:

  • It can reduce the cognitive load and interference that English may cause when learning a new language. By avoiding English, you can eliminate the need to translate everything into your native language and instead focus on the meaning and context of the words and sentences in the target languages.
  • It reduces the ‘thinking in English’ aspect of language use, reducing the need to translate from your native language as you speak in real time.
  • It can increase your motivation and confidence as a language learner. By using a foreign language as your base language, you challenge yourself and prove that you are capable of learning multiple languages without working solely in English. You can also enjoy the satisfaction of seeing your progress in both languages at the same time.
  • It can enhance your linguistic awareness and sensitivity. By comparing and contrasting two foreign languages, you can notice the similarities and differences in their structures, sounds, expressions, and cultures. You can also discover new aspects of each language that you may have overlooked or taken for granted when learning through English.

What are some of the best resources for triangulation?

If you are interested in trying triangulation, here are some of the best resources I’ve found for various language pairs and levels in my own sights:

    • Petit vocabulaire actuel allemand. This is a French-German vocabulary book that covers over 3000 words and phrases in various topics, such as politics, economy, society, culture, etc. It also includes exercises and tests to help you practice and review your knowledge.
    • Assimil : Le Turc Sans Peine. This is a French-Turkish course that follows the Assimil method of natural assimilation. It consists of 100 lessons that introduce you to the basics of Turkish through dialogues, exercises, notes, and audio recordings.
    • Langenscheidt Komplett-Grammatik Italienisch. This is a German-Italian reference book that covers the most important words and grammar rules for learners of Italian. It also provides tips and explanations on how to use them correctly and effectively. Langenscheidt, like Assimil, has a long an proud heritage of producing respected foreign language guides.
    • Stein på stein: Norsk-tysk ordliste. This is a Norwegian-German word list that accompanies a textbook for intermediate Norwegian learners. It contains about 2500 words and expressions used in the textbook, along with translations in German. Like Finnish and Swedish vocabulary guides in similar second language series, these Norwegian guides are available in many of the languages of immigration to Norway, including Polish, Ukrainian and Urdu.

For other language pairings, it’s hard to beat the Assimil  Sans Peine and Ohne Mühe ranges. These feature many titles using French and German as the base languages respectively. They’re available in other languages too, although less comprehensively. That said, it’s definitely worth hunting them down in Italian and Spanish if the language pairs suit.

In any case, I hope this post has given you some inspiration to give triangulation a try. If you have any experiences or resource tips to share, please let us know in the comments.

And happy triangulating!

Triangulation - a polyglot approach to language learning. Image by Nils Thingvall, FreeImages.com

Everyday Triangulation : Three Sides to Every Language Story

A study colleague popped up in our group forum this week, sharing an interesting resource. It was a set of quiz flashcards for the current term’s Swahili vocabulary. But it came with a triangulation twist. It was a Swahili-Spanish set, rather than Swahili-English.

Triangulation – learning one of your foreign languages through another, rather than your first – is nothing new, of course. A beloved technique of polyglotters, it can be an easy, quick-win strategy to learn and maintain / strengthen skills at the same time. Many readily available resources support it, too. Both Duolingo and Glossika have options for learning via a different base language.

The assumption is often that it works best with quite different language pairs, like my colleague’s Swahili-Spanish set. There is certainly a logic to this, as some might expect possible counter-interference with closely related languages. I’ve certainly got some good use out of Langenscheidt’s Polish course for German speakers (a slightly more updated version of my ancient copy is available here!).

Close Triangulation

That said, triangulating with close language pairs does come with a unique advantage. Namely, it shines a bright light on false friends and misleading pairs, which might otherwise remain invisible if English is the medium to learn both.

Take Norwegian and Icelandic, for example. There is an apparent cognate in Icelandic líka and Norwegian like. However, they mean different things: also and alike respectively. If you learn both languages via English, the two will never come into contact with other (at least in your mind), and that discrepancy will remain in the dark. Well, at least until you confuse them in conversation with a native speaker (yes: guilty!).

However, if you create a set of learning resources in Icelandic and Norwegian that makes explicit this (dis)connection, you have a head start.

The same happens with words that are cognates, but slightly overlap in usage. For instance, Icelandic and Norwegian have the cognates sem and som. These can both be used as relative pronouns (the dog that I saw, the doctor that treated me and so on). However, Norwegian som can also be used for the comparing like, as in noen som ham (someone like him). In Icelandic, that doesn’t work at all. Instead, you have to use the term eins og, giving us einhvern eins og hann for the same phrase. It’s exceptionally tricky to learn that distinction if you learn Icelandic and Norwegian through English, but separately from each other.

Triangulating Existing Resources

Great, if you are just starting out, you might say. But what if you are already halfway down the road? By the time I realised the benefit of triangulating Iceland and Norwegian, I already had a ton of English-based Anki flashcards in separate decks for each one. Starting a third set for Icelandic-Norwegian was a less than fun prospect. It felt like treading the same ground all over again.

Tech tools to the rescue, though. There are some clever tricks you can play with your existing data sets to create triangulated versions without starting over. This export / collation technique using Anki and Excel, for example, produces a merged list than can then, in turn, be used to create a fresh Anki deck.

Aside from that, auditing via Excel is a great way to check what you know in one of your languages but not the other.