Bunpro should not use N+1 vocabulary in N grammar

10ten/rikaichamp works perfectly for this in case that helps

1 Like

I’ll look into that, thanks!

1 Like

Be happy to! Admittedly, I haven’t read all the studies myself before - when I was studying to become an english teacher, it was taught as ā€œbest teaching practiceā€. It also happens to anecdotally line up with my own experience of language learning.

That said, the studies are definitely out there.

This study is definitely one you’ll want to check out. It’s about incidental learning, which is exactly the topic of this discussion. It makes the claim that when participants were asked to learn a word, only two previous exposures were necessary to cause a significant performance boost in recall and understanding.
https://onlinelibrary.wiley.com/doi/10.1111/lang.12085

This study reports on adult learners learning new meanings for words, and found that exposure to unfamiliar usage linerally increases familiarity. They find significant evidence to imply that exposure to the word helps long-term recall. https://onlinelibrary.wiley.com/doi/10.1111/lang.12313

I skimmed a bunch more, but I think I’ll stop there for now. Hopefully you will find this helpful.

10 Likes

I remember reading similar works when I did my English teacher trainings, and during my linguistics degree. Additionally, being exposed to vocabulary and grammatical constructions you’re unfamiliar with as you’re reading, for example, and being able to grasp their meanings without relying on dictionaries or other forms of translation the minute you perceive you don’t recognize something, proves to be beneficial for long term recall as well.

Personally, I really like that Bunpro’s example sentences use a diverse range of vocab even at early levels - not only in that it helped with being exposed to words I learned just after seeing them over and over again in said sentences, but that it pointed out exactly where there were holes in my learning (i.e. I don’t know basic medical terms, political terms, basic terms for ideas I like talking about etc.) It’s not that I didn’t get enough exposure elsewhere, but I found it nice to have natural language in the grammar exercises. There’s only so many sentences you can write with a constrained lexicon to demonstrate grammatical points.

Even though ā€œnaturalā€ language is an ideologically sticky term filled with implicit biases, I still think there’s something to be said about what it represents conceptually. That is, shouldn’t we be striving more and more in our studies to learn a wide range of phrases and words not only to improve our own ability to communicate the finer points of our feelings and thoughts, but also the ability of others to understand us? And as we communicate more, we learn what does and doesn’t work in a wide range of social contexts (i.e. when something sounds too formal, stilted, childlike, slightly off in one place versus another). And importantly, since even N1 only gets you to the level of a high schooler, if I’m personally trying to communicate as an adult and a professional, even in casual situations, I’m using very different lexical items than what the JLPT ā€œN rankingsā€ might be imagining in terms of the ā€œimagined language learnerā€ (again, another category filled with sticky ideological hangups and implications) it was designed for. For me, I found learning vocab aggressively and grammar little by little worked well to make me a more effective communicator in all my languages, since the concept of a what a word even is can be quite complicated, and the ā€œaverageā€ person knows a LOT of them, regardless of language.

Idk, I feel that a lot of this is reflective of my personal pedagogy as an educator and as someone deeply invested in language education as both a learner and a teacher. If my opinions and perspectives don’t vibe with anyone, that’s always chill - at the end of the day you do what works for you through tools like Migaku, or curating your own custom decks for vocab and grammar that are tailored to your own, more specific needs and desires. I just wanted to put in my two cents, whatever that’s worth, and hopefully add another perspective to this discourse.

10 Likes

I’m saving this one for the next [N1 speedrun in one year] post

1 Like

No one ever mentioned N1 speed runs :wink:

Maybe there are people out there who can truly pass N1 in a year. N1 is a great accomplishment, but there is a great chasm between passing N1 and true language acquisition :nerd_face:

2 Likes

I like getting words I don’t know in my grammar lessons. I’ve learned many new words this way!

4 Likes

I mean… why not? You gotta learn those too anyway. The order doesnt really matter. You dont have to wait 3 years and master n2 before learning a word thats tagged as n1. Those categories are just for convenience. If you see a word just learn it.

5 Likes

Just to give my two cents here, I love the way Bunpro does it.
Please don’t change it ^^

Sentences have increasing complexity in length and meaning and also exposure to new vocab the higher you push them in the srs system and the higher you go in the JLPT rating.
We get somewhat natural sentences this way and it allows a more holistic study experience.

Trying to steelman the other position, if your primary goal is for example passing the N4 and not Fluency or something along that lines, then yes, this approach is slowing you down towards your goal.
But I like that Bunpro is not a ā€žcram for the JLPT testsā€œ website. As it was said before in the thread, even the N1 is not at all comparable to language proficiency (or at least doesn’t guarantee it), which is presumably the goal of most here.

So please don’t compromise too much Bunpro team ^^

4 Likes

Natural Japanese is one thing and I’m grateful for the fact that Bunpro is using them. But I also agree with other users complaining about habitually glossing over sentences because they are extremely complex for the level they’re supposed to be on. Take a look on these from ćØć•ć‚Œć¦ć„ć‚‹:


And sure, I don’t really need to know the content of X in a sentence ā€žX is considered Yā€ to type in ā€žis consideredā€. However, it is unfair to suggest that I will learn anything from a sentence I don’t know half the words of. The sentence becomes gibberish to me.

4 Likes

I actually came across this specific sentence just a week ago or so and found it was a great sentence for learning. It is also very smartly designed.

The reason being that it is considered common knowledge that Nero burned Rome and both his Name and Rome are written in Katakana, so you just reading katakana + the fire kanji will probably give you a pretty strong idea of the context.

From there you can go on an try to quickly deduce some of the worst, like ēš‡åøćƒćƒ­ probably meaning Emperor Nero. And it very closely resembles the way we acquire more vocabulary once we are more proficient with the language as well.

Great, now you will have a very strong context for this word and this alone is probably way better than any Wanikani or personal mnemonic, bc you spent a couple seconds thinking about it yourself.

At least I still remembered both the words for evidence and emperor when I came across the sentence now in your answer.

So I completely disagree, sentences like these are great! Yes, they are hard, the context is a more academical one as seen in the beginning (essay), but why should bunpro only cater to everyday situations?

2 Likes

That’s how I work myself through the sentences, too. Trying to get everything from context first, then looking at the translation and guessing some more words until the sentence clicks. It’s does not only give practice and context for extra words, every sentence deciphered that way is great reading practice for native content.

Sure, it’s more work in the beginning, but it pays off quite fast.

This way of learning just doesn’t click with everybody. Some don’t do well with it or some just hate it, that’s all fair and understandable. But if it doesn’t work for some, then looking for another ressource would be the way to go. It’s not like one way is inherently better for everybody. There is no one superior way to learn japanese.

3 Likes

Not to be all elitist here, but there are definitely inherently better ways to study a language :grinning_face_with_smiling_eyes:

We all agree probably that you will not effectively learn japanese ONLY by listening to japanese radio for example. Or maybe you are because you are genius but it will still take forever and you would be a lot faster even if you would just add a kana dictionary and transcripts of the radio into the mix.

2 Likes

Oh sure, there definitely are ways that are terrible for learning, but a lot of strategies are just mediocre or ok-ish and still get lots of critic. If an ok-ish way is fun or easy to follow longterm, it’s usually a better choice than an optimal learning strategy that you hate or just don’t really vibe with.

A lot of people in this thread made clear they don’t vibe with bunpros way of bonus vocab practice, while others totally thrive on it. Personally I would be totally bored with an n+1 approach, so bunpro is a great fit for me. But both ways are great approaches and supported by multiple studies as ā€œeffectiveā€. Just like any form of SRS is supported, but lots of people learn languages without the use of a dedicated system.

I’m wondering if there are any good n+1 ressources for those people who don’t like to get bonus vocab and reading practice with every review ^^

2 Likes

I’m wondering if there are any good n+1 ressources for those people who don’t like to get bonus vocab and reading practice with every review ^^

I think anyone that’s complaining WANTS the bonus reading practice. It’s the bonus vocab and grammar they don’t like. Because it’s quite difficult to practice reading if you don’t know the words or grammar. The unknownunknownunknown that unknownunknownunknown, is horrible reading practice for me. When I see a clause like that one, I just settle for using the srs system as word recall only.

The best N+1 resources I’ve found:
For listening/speaking Pimsleur. 150, 30 minute lessons, that all build on the prior lessons. If a new word is introduced it’s usually explained in advance, or immediately following a short listening exercise, and then the listening exercise repeats. It incorporates spaced repetition by bringing up older words in future lessons.

For textbook learning Genki is structured N+1. Explains all the vocab, then explains all the grammar. Then you practice with only that vocab, grammar, and the previous lessons vocab/grammar.

A standard SRS tool, like anki, learning vocab is usually 0+1, because you’re just learning one bite of information with no other connections to anything. When Bunpro has a sentence where I have learned all the words and grammar it levels up to N+1.
Bunpro is the best option I’ve found for N+1 SRS learning. Even if I can’t always benefit from the reading practice.

Hopefully once my foundation is better established, I can leave the safe easily digestible structure of N+1 learning, and join the rest of you in happily learning in random order.

1 Like