How I hacked the language learning process to jump-start my Japanese listening skills

It’s a known fact that language learning becomes difficult with age. Little kids are somehow able to pick it up very quickly, yet adults struggle through the process.

Wanting to get my mind off of engineering, I decided to test this assertion as a quick side project and see if I’d be capable of picking up a new language.

Picking a target language

The first question I needed to figure out is what language I wanted to learn. Since, I’ve always been a fan of Japanese anime, I decided to go with Japanese. This reasoning is critical because to improve any language skills, you need exposure, and the more, the better.

Therefore, since I enjoyed watching anime in its native tongue (with English subtitles), I’d also be able to use that time to get more exposure to Japanese!

What to learn

So now that I know what language I wanted to learn, the next question I had to figure out was how I should focus my efforts. Should I try learning how to read, write, speak, and listen at once? Or should I pick a subset to focus on?

Ultimately, I decided to focus my efforts on just the listening comprehension. The hope was that if I can pull this off, it’d be easier for me to learn how to speak. Afterwards, I can focus on learning to read and write, if the motivation is still there.

What I tried

Alright, so at this point, I’ve decided on a language, and decided to focus on developing my listening skills. Now, I had to figure out how I was going to go about improving my Japanese listening comprehension.

Here are two things I tried that didn’t work very well for me:

Just watching anime and hoping that I magically pick things up.

I did this for ~20 years and all I picked up were a few phrases and a bunch of curse words, so it clearly wasn’t an effective technique.

Taking a Japanese class.

I took a short YouTube class, and while it helped with things like how to count, how to talk about date/time, and basic sentence structure, I found that it wasn’t really helping me improve my listening skills.

A lot of that had to do with what I was hearing in the anime episodes wasn’t aligning with what I was learning in class, so I felt like I was always in square one. Furthermore, it seemed like the classes are structured to teach formal Japanese, whereas in the episodes they spoke more informally. Thus, from a practical perspective of understanding day-to-day dialogues, I didn’t feel like I was learning the right thing.

Developing a plan

Given the experience above, I decided to change up my strategy. Instead of focusing my learning around classes, I instead decided to focus my learning around the anime episodes.

One observation I made was that an episode generally focused on a specific part of the overall story. As a result, the themes were generally consistent across the entire episode. The reason why this was important was because words related to the themes would be repeated over and over throughout the episode. Thus, from a listening perspective, it gives you repeated exposure to a set of words, thereby maximizing the odds of you remembering them!

The process

The strategy I followed relied on the key observation that when we watch an episode with subtitles, we already have a mapping from Japanese audio to English words. And if the translation is of high quality, we can be fairly confident that there’s a one-to-one mapping between the audio and the translation.

So then the key question becomes: what is the audio snippet that maps to the translated word in the subtitle?

Unfortunately, there’s no easy way to answer this question, so I had to use Google Translate on each of the English words, and then identify the right audio snippet that was spoken in the dialog.

So here’s a concrete example. Let’s say I’m watching an episode and the following dialog occurs:

What I would do first is identify the English words that are likely to remain consistent regardless of translation quality. In this case, maybe I’d start with ‘fly.’

Next, the goal is to identify the snippet of audio that translates to ‘fly.’ Since I don’t know in advance what that audio snippet is likely to be, I go to Google Translate and look up ‘to fly’ in Japanese.

Looking at the results, we can see a list of Japanese words that means ‘to fly’ along with their audio clips. The goal is then to go through each of these audio clips and identify the one that is spoken in the episode.

This is a bit time consuming because you have to listen to each of these audio clips and then compare it to the Japanese dialog in the episode. Sometimes, you might need to repeat the Japanese dialog multiple times to identify the various words being spoken before successfully finding the match.

However, I think this exercise has a ton of value because it forces you to listen actively. At first, the dialog would sound like a continuous stream of gibberish, but after forcing myself to listen actively, I was able to detect pauses between the different words. Additionally, it forces your brain to process the words more quickly. At first, I felt like the dialog was too fast, and I couldn’t keep up with what was being spoken. But with repeated exposure, I realized my brain was more able to process the dialog with less effort.

After I found the right Japanese audio snippet for the English word, I kept track of it in a spreadsheet:

The above now gives me a mapping from English to Japanese. And because I know the Japanese word, I can quickly look it up on Google Translate to hear the audio clip whenever I want.

With this process in place, I basically repeated it over-and-over for as many words as possible across 45 episodes, each of which were ~25 minutes in length. I couldn’t catch all the words in the episode, likely due to translation quality, but I tried to catch as many as possible.

The results

All in all, I invested about 1–2 hours per night for each of the 45 episodes. This exercise spanned for about a month.

As for results, I’m happy to say that I started seeing results very quickly. Not only was my brain able to process the dialog more rapidly, but I realized that it started catching words I learned in previous episodes.

Additionally, there were three interesting side effects that I hadn’t anticipated. First, I noticed my brain started catching words from previous episodes being used in different contexts to mean different things. Secondly, for unfamiliar words, I found myself making educated guesses on the mapping between the Japanese audio and English subtitle, and being right many times! And finally, after learning a bit of vocabulary, I sometimes found myself detecting inconsistencies between the audio and the subtitles. The subtitles might be a good approximation of what’s being spoken, but it’s certainly not a one-to-one mapping.

Conclusion

It might be true that language learning is difficult for adults, but this experiment has proven to me that it’s not impossible. With a little bit of effort, all it takes is to develop a process, being disciplined about following it, and getting repeated exposure.

The key observation here was that each anime episode has a central theme that’s consistent across the whole episode. Additionally, there are a set of words that are linked to this theme. And because the theme is consistent across the entire episode, those words will get repeated many times throughout the episode. Therefore, this approach will not only help you learn this set of words, but also give you repeated exposure to them throughout the episode. Consequently, this exposure will help cement the English to Japanese mapping as you hear it over and over.

And while this method isn’t perfect, I do believe it’s a good first step for getting comfortable listening to Japanese dialog. Once you’ve hit a certain level of proficiency and can’t extract additional value from this approach, you can always apply other methods to fill in the gaps, and I’d expect that to take a lot less time going forward!

I’m planning to refine this methodology over time and explore other strategies for maximizing my listening comprehension. If there’s interest, I’m happy to share my findings.

Let me know if you guys have any feedback, thanks!

Upal Hasan