Wow, there is a lot of negative gatekeeping on learning Japanese in this thread. I recommend people chill out.
Dan, this is cool, and I can tell you enjoyed writing this up and thinking it through as you wrote. Don’t let the haters get you down.
I think you’d like Orbit (withorbit.com), a tool that got built out of some experimental learning work I think originally done at quantum.country — it’s intended as a sort of toolkit to help turn blog content into spaced repitition inline memorization tools.
Check it out! I think the post would benefit from something like this - it’s fun for most to play with seeing how they are understanding this, and you spend a lot of time trying to create a bit of interactive reading/thinking in the blog.
Don’t want to sound negative, just want to share my observations as someone who’s been actively learning Japanese for a couple of years. I much prefer how Japanese people actually teach verbs: 一段 (ichidan), literally "1-step"; 五段 (godan), literally "5 steps", plus a few exceptions sometimes called irregular verbs. It's not hard; it makes sense, just some textbooks (e.g. Genki) teach it really bad.
Learning Japanese is about learning a new way of thinking and structuring your thoughts. The more you learn, the more you realize it just doesn't fit into the English world. You can't really translate Japanese into English without losing nuance — and sometimes that nuance is important. So start early and start training your brain to think in the language, instead of trying to translate it and force it into English or some other language brackets. It won’t work; it won’t make sense; you will get stressed and confused.
> Learning Japanese is about learning a new way of thinking and structuring your thoughts.
I suspect this is true with most languages that are not in the same language family. The indoeuropean languages are all pretty similarly structured (don't @ me for not knowing the eighteenth tense of lithuanian), but it still takes time and effort for an english speaker to think in french, even if for the most part concepts translate directly. But bantu languages, eg kirundi or zulu, lean heavily on verbs with an entirely different conception of sentence structure and morphology, where you can stick entire clauses into one word, and you realize that your brain is picking up on patterns to decipher grammar that wouldn't make much sense in indoeuropean languages like dahl's law (except apparently in greek and sanskrit, where it's called grassman's maw? Huh... now I know). Hausa is different still where you need to think about tense in an entirely different way—the pronoun is conjugated and the verb remains unchanged, and sometimes it feels like there are more irregular words than regular. Mandarin is beautiful, and actually quite simple to speak a little conversationally, if probably as difficult as english (or maybe more so) to master.
Learning foreign languages really makes you realize how central language is to basic cognition. You see the world in a different way, with different values and relations, depending on which language you speak/think in.
Ironically, I think this makes people extra susceptible to thinking that chatbots are intelligent (or even conscious, mr dawkins), even though it fails basic tests of memory and learning necessary to convincingly mimic understanding of, idk, time passing. Or motivation. Or emotion. But you can tell when it's awkwardly translating from another language in a way that a human would likely catch quite quickly if exposed to billions of documents.
As an unrelated tangent, I love seeing it when the “thinking” in a native language leaks into the writing in another, I find it so endearing and intriguing. My French colleagues sometimes produce the weirdest English sentences (perfectly understandable just unusually phrased), that alone occasionally sparks my interest on picking up a bit of French — I just want to learn how different the grammar and sentence construction is!
Obviously, it applies to other languages, I’ve just been working a lot with French. Well, and my own language often leaks into English, of course.
> I much prefer how Japanese people actually teach verbs: 一段 (ichidan), literally "1-step"; 五段 (godan), literally "5 steps", plus a few exceptions sometimes called irregular verbs
This is literally what I teach in the article, including these translations. Quoting it:
> in ichidan ("one-row") verbs like taberu, the last syllable of the stem is fixed. it's always going to be be, no matter the suffix:
>
> (table)
>
> it stays on a single row in the hiragana table, hence "one-row".
>
>on the other hand, in godan ("five-row") verbs like nomu, the final syllable of the stem alternates between ma, mi, mu, me, and mo:
> (table)
> it spans all the five rows, which is why it's godan ("five-row"). the m* "wildcard" represents the entire ma/mi/mu/me/mo column.
You’re also mischaracterising my approach. I am not teaching to “think in English”. Quoting from the article:
> i'm using romaji as a convenient way to refer to phonetics in text. however, your "mental algebra" should match the hiragana table. so this is a reminder to not think in romaji when you do calculations. when we conjugate godan verbs, we literally go up and down the column. (maybe all these textbooks that used hiragana had a point!)
If you have objections, please engage with the article’s actual content, not with what you assume it is based on a glance (“oh he’s using romaji, this is thinking in English”). I’m using romaji for specific reasons that are motivated and explained in text, and I show every single pitfall of that choice as well.
There’s nothing wrong with trying to make sense of something new using terms and concepts you already know and that click for you. We all do it, me included.
I was speaking more broadly about learning Japanese and what I see on HN. Every six months or so, somebody discovers a clever pattern in the Japanese language. It’s almost always related to something taught in the beginning, the N5 level. And that pattern seems to have been eluding rest of us.
Sure. I like Cure Dolly. I’m actually surprised you bring her up because she gets a lot of purist haters too.
I’ve found this lesson difficult to digest when I tried her approach. It’s actually where I fell off and lost the interest the first time I tried to learn the language.
The grandparent says it’s easier to teach by saying ichidan and godan actually mean “one-step” and “five-step”. That’s exactly what I’m teaching so it can’t be an argument against my article. This is something I not only say, but explicitly show visually with a table — as I have just quoted. From this I conclude the grandparent hasn’t read the article, and is commenting on vibes.
Then the grandparent says something I agree with (don’t force the language into another language) but I don’t think it’s a fair description of what I’m doing. It sure looks like that’s what I’m doing, but I strongly believe that learning conjugation is primarily phonetical (it’s about how it sounds, not kana itself) and therefore romaji is just a better pedagogical choice for someone not already fluent with kana. And no, I don’t buy that you have to put being fluent in kana as a prerequisite. The whole conceit of the article is basically that you can learn almost the entire conjugation system in one evening with zero prior knowledge of the language. That alone justifies the small shortcut I took to get there.
I figured it was alluding to more conventional approaches to learning them. Reads like you’re making assumptions and defending instead of engaging with criticism in good faith.
“Don’t want to sound negative […] I much prefer” reads to me like a comment phrased against the article’s approach. Otherwise why would it “sound negative”? And what I’m saying is that it matches the article’s approach. Thus, the assumption. Where am I wrong or uncharitable?
There are lots of people in the comments somehow offended by the author's genuine excitement over the method that worked well enough for them that they wanted to share it.
As someone who's never tried learning Japanese, I thoroughly enjoyed reading the deep dive and am now less afraid to check out some more serious tutorials (though I wish everyone put as much effort into explaining the system behind something so often dismissed as "just memorise it").
As someone who is fluent in Japanese, the thing is that this is not an example of a method that works well, it’s an example of a beginner coming up with an overwrought headcanon (while calling it “simple”, nonetheless!) that ultimately will hinder their understanding, as I will detail elsewhere.
If this was a similar post about programming, it would not get upvoted here because more people would recognise it for what it is: a “Monads are Like Burritos” post from a well-meaning but misguided beginner.
I used to get these terms mixed up all the time because some textbooks use "Group 1" and "Group 2" to refer to these verbs, but Group 1 doesn't mean ichidan.
For all that I'm not totally sold on this article's idea of "stems" and "suffixes", I think it does a good job of avoiding this pitfall and correctly explaining the groups.
This kind of content is usually shared by people who are learning a language out of genuine interest for the first time. Most of the time it's Japanese because of the popularity manga and anime enjoy, and the challenge of learning a language that's so different from one's mother tongue. There are countless blog posts written by people to whom it "clicks" for the first time, and they excitedly run to their computer to share their novel experience. I don't see anything wrong with this, let them enjoy the moment. But if they stick to their learning routine, they will inevitably learn the truth that is that the only way to learn a language is to interact with it a lot through all possible channels - speaking, listening, reading, writing. Mnemonics may work for some time, but in the end one won't be able to actually use the language if they don't learn it intuitively. And there is also the fact that most of those people live in places where Japanese is not in everyday use, limiting their opportunities to practice it.
> Mnemonics may work for some time, but in the end one won't be able to actually use the language if they don't learn it intuitively.
A crutch can be cast away gradually.
Finnish has two verbs, ottaa "take" and ostaa "buy". As a n00b I confused them all the time. So I decided that the "s" in ostaa was a dollar sign, and I used that quickly in conversation to select the right verb. And as I got quicker at it, it became "intuitive". I threw down my crutch.
Yes, this is a valid point, mnemonics are totally valid when you are new to a concept.
I'm learning Turkish right now for example, and the dative forms for me/you (bana/sana) perfectly fit the accusative conjugation pattern in my native Polish. So my mnemonic is explicitly "the opposite of what you feel it is".
Senior management at a Japanese tech company, 2 kids, most of my daily life in Japanese here. Can confidently talk circles around most N1s.
I agree. I had plenty of strange abstractions and crutches when starting out learning Japanese, and starting in romaji was absolutely fine. While the post wouldn't benefit me now, starting out any useful mental shortcuts to producing Japanese would have been most welcome.
This is a quite weird article. It opens with several paragraphs about romaji, implying the target audience are not familiar with romaji/kana already. But it's way, way too early for such a reader to worry about verb conjugation.
I wanted to write a weird article! My whole shtick here is that learning conjugation is completely orthogonal to learning most of the language, and you can frontload it into a single evening like this. I don’t respect the whole “it’s too early to do X” dogmas in learning and teaching, and I trust some of my readers will find useful what I find useful.
> There are lots of people in the comments somehow offended by the author's genuine excitement over the method that worked well enough for them that they wanted to share it.
Did it work for them though? They apparently never got past the basics. So IMO it's more likely the opposite; they've distracted themselves from getting on with learning.
I started learning a couple of months ago (with a couple of false starts before). What level am I supposed to be at by this point? This is so incredibly condescending I don’t even know what to say.
Thanks! I’d say that a big part of the motivation for me was to show that this piece is orthogonal to the rest of the language. You can sit down and understand how to conjugate almost any word to almost every suffix in the time it takes to read this. Yes — not fluently — and not “understanding” the language — but it’s wild to me that you can do that at all! It wasn’t obvious to me that this knowledge is orthogonal to everything else.
That’s what motivated me to write about it, really. In language courses all of this is often spread out over weeks or months. I thought it would be fun to write something that you can read in one evening and map out almost the entire system. With no prerequisites.
Because the whole article is built upon a straw man (it's not usually taught the way he claims) and the "method" is just a normal explanation (see this section on wikipedia).
I’m just describing my experience learning. I understand that some compressed explanations match it, and I agree this one is good! This is not the explanation I was exposed to when trying to learn it the first time.
So you agree my explanation is correct (but unnecessarily longwinded and builds a strawman)? I’ll take that.
I had to stare at this for a while to figure out why the author thought it was wrong. "si" is rendered as し on every IME keyboard I've ever used, but the author wants it to be written as "shi".
I don't think this article is really simpler than just learning the table and letting your pattern recognition neural wetware kick in and do its thing. Or better yet, go read some books. After a while, incorrectly conjugated verbs just look/sound wrong.
Yes, this is native speakers of Japanese applying a sound law pervasive in their native language (that the consonant /s/ cannot occur before the vowel /i/ and must turn into a sh sound), and applying it to English, which does not have that sound law and that happily allows /s/ to occur before our /i/-like vowels.
An equivalent phenomenon going the other way (at least for American English speakers) is clearly distinguishing また mata "again" and まだ mada "yet" - American English speakers tend to merge /t/ and /d/ in that kind of intervocalic position (think about pronouncing "latter" and "ladder" identically), and it takes deliberate effort to pronounce distinct /t/ and /d/ sounds there in Japanese, where the American English sound law that merges them does not apply.
I wasn't really going for kunreishiki, more ワープロ, aka "whatever is the least typing to make the desired characters appear".
I think typing in IME is how most people actually interact day-to-day with romaji, and I lex "si" and "shi" identically because the difference is usually not important.
I feel like you’re going out of your way to misinterpret the article. As the article says below:
> this is why it's important that you don't actually "think in" romaji. i'm using romaji as a convenient way to refer to phonetics in text. however, your "mental algebra" should match the hiragana table.
Then the article includes an exercise that verifies the reader’s understanding.
I also included a note:
> (note i could also have used a different romanization that renders し as "si", つ as "tu", and ち as "ti" for this article. i decided to not because everyone else uses romaji, and once you understand this point once, you shouldn't have a difficulty doing this in your head.)
Where is the factual mistake here? “si” is invalid romaji, my article uses romaji, therefore it’s invalid.
"Factual mistake" is a bit harsh, but the missing piece is that there are multiple ways to romanise Japanese; all of them produce "valid Romaji" but only in the particular system being used. Si is how you write し in romaji using the kunrei-shiki romanisation. In the Hepburn romanisation it's shi. See https://en.wikipedia.org/wiki/Romanization_of_Japanese#Diffe... .
Kunrei has been deprecated, Hepburn is the official one even the government of Japan recommends these days and matches better the pronunciation of the language.
But nobody cares about formal romanization rules in the first place. People have informal bastardizations of Hepburn that's consistent enough for typing, and Kunrei deprecation is more or less just admission of the status quo.
I see, maybe my use of “romaji” is sloppy because I implied “Hepburn romaji” specifically because that’s what I chose in the article. I do explicitly mention that other romanizations exist and that I chose not to use them — so I’m not worried about misleading someone who reads the article. But on pedantic level I see why “si doesn’t exist” sounds overly broad.
I think also that anyone who's spoken Japanese for a while already has internalized that "si" === "shi" because there literally isn't the sound "si" in modern Japanese, as the other commenters mentioned it's often romanized both as "si" and "shi" in daily life, if you typed "si" into a keyboard it renders し, it goes on. The original comment on this thread includes one such person who literally didn't follow why "si" is wrong, and I felt the same way too as a long-time Japanese learner. It's a very "copy paste Western language concepts onto Japanese" way of conceptualizing the language, which is IMO a great way to set oneself up for great struggles when trying to learn a language that is structurally different, because it's not the right mental model.
I’d normally be with you and I call this out as a concern purists would have my article. I’m doing something differently here and maybe more daring — the conceit of this article is that you can learn the entire system in one evening with no prior knowledge of Japanese at all. In that case I think the “s_ + i = shi” is as fine a pedagogical moment as any to introduce why the hiragana table matters. And for readers who already know it, I assume they won’t actually have a problem following the content, setting aside the pedantism and the annoyance at someone seemingly “teaching in English” and “copypasting concepts”. Like I get what you’re saying, I just think that it’s okay to break the rules for what I’m going for here.
This is pedantic but you're thinking of "romanization", the act of transliterating Japanese with the Roman alphabet (romaji). There are different systems of romanization, most notably Hepburn and Kunrei. In Hepburn shi is correct, in Kunrei si is correct.
It sounds like you're saying "si" are not valid characters in the Roman alphabet.
Ok yea I see what you’re getting at — my terminology was overly narrow. By “romaji” I specially meant “romanization I’ve chosen for this article”. I do offer an explicit note that other romanizations exist, but outside of that note, I tried to stay within a single self-consistent system, so this nuance was lost.
> I feel like you’re going out of your way to misinterpret the article.
Nope. You correct yourself after, sure, but what I wrote is how it came across at the time when reading.
> Where is the factual mistake here? “si” is invalid romaji, my article uses romaji
No, that's not what "romaji" means. If you mean Hepburn, say Hepburn. And if you don't know the difference, that's a sign to learn more before presuming to teach others.
I’ve already conceded in sibling comments that saying “romaji” to mean “Hepburn romaji” specifically is technically imprecise. Note I still do explicitly mention other romanization systems, want to stay consistent within the article, and don’t want to overload the reader with terminology. My motivation here is that realistically that’s the romaji you’re going to be exposed to the most as a learner. At least that’s been my experience.
As for “correcting” myself, yes, I expect the reader invested in arguing online to also be able to follow more than a single paragraph of text. I think it’s fine that you stumbled there as that paragraph wasn’t for you. I don’t think I’ve hurt your understanding of Japanese this way.
The only people confused by that paragraph are people who already learned kana or correct pronunciation (or both). That paragraph isn’t for them.
> there is no "si" in the hiragana table, so s_ + (i) = shi. […] this is why it's important that you don't actually "think in" romaji. […] i'm using romaji as a convenient way to refer to phonetics in text. however, your "mental algebra" should match the hiragana table.
That’s just false, “si” is in the hiragana table as し. The romanization “si” is /si/ which is pronounced [ɕi] (or [ɕi̥] or some other possibility). This is basic Japanese phonetics.
If you fix all the errors that are in the article, at best there is an argument buried here that Hepburn romanization should not be used to teach Japanese to English speakers—but I think that point is really my own argument that I’m making with the fragments of the article that make sense.
Romanization can be more consistent with Japanese phonetics or it can be more consistent with English phonetics, and the Hepburn romanization is more consistent with English phonetics, which is why it’s a good choice for English speakers that don’t know Japanese, but a bad choice for English speakers who are trying to learn Japanese.
Okay, we’re fighting over definitions here. There is no “si” in Hepburn romanization. I am intentionally using Hepburn romanization in the article. Therefore, in my article “si” is a compile error.
You may argue with my choice, or maybe you can argue that referring to cells in Hiragana table solely by my chosen romanization is somehow bad, and I should instead be inconsistent and give the same mora two different romanizations within a single article. Is that what you’re suggesting?
First, As a basic part of Japanese language education, students are expected to be familiar with different romanization systems. If you ask a student where “si” is in the table, they should be able to find it. If a student says “it’s not in the table” then they’ve failed the lesson or there is something wrong with the teaching material.
Second: am I arguing that the choice of using Hepburn here is somehow bad? Yes, that’s correct. I think Hepburn is a bad choice here. A good choice is Nihon-shiki. JSL romanization is also fine.
We’re talking about the same thing but you insist that there is only one angle under which things aren’t confusing. I disagree. That’s fine. The two systems are isomorphic, and I genuinely believe that, given I’ve described every single caveat of Hepburn in the article, I’ve paid my dues for using it. YMMV. I even include the “finding in the table” part.
I think I agree that Nihon-shiki and explaining it upfront would’ve made the article more elegant. One constraint I wanted to hit is that a person should be able to read this article with zero knowledge of Japanese, and walk away with being able to conjugate almost every verb to every suffix correctly. This is more of a challenge to myself as a writer than any practical need but hope it shed some light on the choices and the framing. I liked Hepburn because it’s closer to how it sounds. You can imagine I’m using IPA instead if you want.
The problem is that your explanation confuses phonemes with letters.
A spoken language is described by decomposing the spoken words into phonemes, where phonemes are sounds that distinguish words, in the sense that replacing one phoneme in a word with another phoneme will produce a different word.
While ideally each phoneme should be recognized by a distinct pronunciation, in the majority of the languages of the world a phoneme does not have a single pronunciation, but it is pronounced in different ways, depending on the context.
It does not matter at all how one chooses to write a Japanese word, with hiragana or with one of the various methods of romanization. For any writing system, you must know the correspondence between phonemes and how they are written. For very few writing systems there is a one-to-one mapping between phonemes and letters.
The Hepburn romanization does not attempt to be a phonemic writing system, but it attempts to be close to a phonetic writing system from the point of view of an English speaker. The Kunrei-shiki romanization attempts to be closer to a phonemic writing system than to a phonetic writing system. I my opinion a phonemic writing system is superior to a phonetic writing system, but it appears that for most English speakers it has been too difficult to understand the difference between such writing systems, so the Japanese government eventually gave up and they switched to Hepburn, to please the less sharp-witted English-speaking visitors.
Japanese has an "s" phoneme, which happens to be pronounced differently before the vowel "i" than before the other vowels, and before "i" it is pronounced similarly to an English "sh".
In the same way, the Japanese phoneme "t" is pronounced before "i" similarly to an English "ch".
Once you know these two rules, and the few other rules about the other Japanese phonemes whose pronunciation depends on the context, like "n" becoming "m" before "b", there is no point in mentioning them again.
In your discussion about conjugation there is nothing exceptional about the variations in pronunciation that are reflected in the Hepburn Romanization. They are just the general rules of Japanese pronunciation, like for any other words.
So any discussion about these spelling variations is misplaced in the discussion about conjugation, where it occupies a space without contributing anything to the understanding of the conjugation rules.
I get that. I’ve compressed this as an aside in an article about something else. It’s a choice; like an article about React could spend a bit of time on arrow functions vs function declarations as a choice. Or even let/var. It all depends on audience’s prerequisite knowledge and what you choose to assume.
I strongly suspect that if I were using Kunrei-shiki, there would be just as many comments here saying my article is wrong because “si” is pronounced closer to English “shi”, but my article makes it seem like it doesn’t — so this is why you should learn kana bla bla bla.
I assume my reader (1) has zero prerequisites and (2) wants words to sound correctly while seeing them the first time. Those are the constraints that motivated my approach. You could argue that it’s a strange set of constraints to pick when teaching but I wanted it to be fun.
The systems are obviously not isomorphic—Japanese kana are not entirely phonetic (they are just mostly so) and the different romanization systems choose differently whether to follow orthography or phonetics more closely.
> hanas* + (i)masu = hanasimasu (wrong!)
I cannot wrap my head around how this line in the article could be defensible. Like, if I don’t understand how Japanese is pronounced or written, and I just rely on Hepburn, I guess pasting these fragments of Hepburn together don’t produce the right Hepburn in the end?
YMMV indeed, but I think the lesson here is “this is why you don’t use Hepburn when you’re writing an article about Japanese verb conjugations”.
Hepburn does make sense for somebody with zero knowledge of Japanese but it just gets in the way when you are trying to explain how Japanese works. So lesson zero is “don’t rely on Hepburn” and IMO if you are interested in pronunciation and listening you should be using audio as your primary source.
I’m saying that Hepburn is isomorphic to Nihon-shiki since each is an encoding of kana. Each of them is a bijection to kana (actually that’s wrong; see EDIT below), therefore there’s a bijection between them. Obviously I’m not saying that arbitrary latin characters are isomorphic to kana, that would make zero sense.
I sympathise with your point about the benefits of Nihon-shiki romanization here. It might’ve been a better choice for this article.
> I cannot wrap my head around how this line in the article could be defensible
I think the reader would just read the next section where I use your argument to critique my own approach? And then make up their own mind whether it’s defensible to do something in the article, to raise pros/cons for why I did it, and then to keep on with the choice.
I wanted to illustrate this confusing point, and that’s how I chose to illustrate it. I think it’s confusing either way. I trust that a reader who actually wants to learn, and isn’t just being a pedant, would carry away the right set of conclusions, and would understand the isomorphism (again — see EDIT below) after those two sections.
> Like, if I don’t understand how Japanese is pronounced or written, and I just rely on Hepburn, I guess pasting these fragments of Hepburn together don’t produce the right Hepburn in the end?
Yeah. So that’s a learning opportunity that kana row shifting doesn’t quite follow rules you might expect from many other languages. Maybe that’s a clunky way to introduce it. I personally like this framing. As I noted somewhere else, you could imagine that I’ve chosen IPA notation instead.
—
EDIT: Actually wait, Hepburn is not bijective for zu and ji. I haven’t thought about that. It’s not relevant to any of the conjugations so it doesn’t break the article, but that may be a good argument that it’s not worth the effort rescuing Hepburn.
> I think the reader would just read the next section where I use your argument to critique my own approach? And then make up their own mind whether it’s defensible to do something in the article, to raise pros/cons for why I did it, and then to keep on with the choice.
I think that’s a long wait; I don’t want to rely too heavily on analogies but it is like teaching somebody arithmetic roman numerals and then explaining in a parenthetical that there are other ways to do arithmetic (but not naming them). Maybe the reader can make up their own mind—but I don’t think the pros and cons are raised in the article, or if the are raised, I couldn’t find it.
I don’t want to pile on here but it sounds like you are, in this conversation, learning about why the different romanizations exist and what the pros and cons are. Or if you already knew, you are getting what they call an object lesson. (Like you noted—in Hepburn, ji and zu correspond to two different kana each.)
> As I noted somewhere else, you could imagine that I’ve chosen IPA notation instead.
This just resurfaces a similar problem with different symbols—if you put your IPA notation in slashes // you get phonemes, which will get you something mostly equivalent to Kunrei-shiki romanization. If you put your IPA in brackets [] then you get something sort of equivalent to Hepburn (in that it’s designed to show pronunciation). Both choices will on some level obscure a regular pattern that could be revealed with kana or romaji. Orthography is funny like that; in both Japanese and English it can show the origin of words even when the pronunciation changes.
I think the other lesson here is that students will mostly learn morphophonology intuitively by absorbing examples with some light explanations of the rules, and if you overexplain the rules you end up with too much “scaffolding” which gets in the way. Like when people use mnemonics or try to memorize kanji by thinking pictorially.
I genuinely haven’t thought about zu/ji here (conceded!) It’s not relevant to conjugation though.
In general, I find your attitude a bit condescending. This is what I wrote about my choice:
> note i could also have used a different romanization that renders し as "si", つ as "tu", and ち as "ti" for this article. i decided to not because everyone else uses romaji, and once you understand this point once, you shouldn't have a difficulty doing this in your head
My main mistake seems to be meaning “[Hepburn] romaji” by writing “romaji”. I was obviously aware of other systems because that is what the sentence says but I thought it’s acceptable to refer to Hepburn as just “romaji” as a sort of the default one. Maybe that’s wrong.
Other than this terminology nit, I think I’ve made myself quite clear there. I genuinely don’t think it’s a big deal. Maybe I overestimate my readers’ intelligence but I don’t find this difficult to live with at all once you get it.
Roman numerals is a funny parallel but it doesn’t hold very well. The difficulty of using Hepburn is O(1) shortcut: for conjugation, you only have to “remember” three special cases and they’re always applied just-in-time. It’s just substitutions — and are arguably inherent phonetically. Arithmetic with Roman numerals requires many stacked adjustments where you have to match pairs of things. And lack of orders really screws with ability to do multiplication. This just isn’t an intellectually honest comparison.
Re: your last point I actually kind of agree. I’m that annoying student who likes to un-extrapolate backwards from examples to the rules, knowing which gives me a warm fuzzy feeling, after which I can go back to examples. My article is for people like me. Maybe there’s a few more of them.
> In general, I find your attitude a bit condescending.
Yeah—I can understand why I’d come across as condescending. There’s a balance here—I want to be clear when I say that I have problems with the article, but I don’t want to be hurtful and I don’t want to make criticisms that are not supported by the text.
Rather than defend my comments as “correct” let’s say that I failed in my goals of not coming across as condescending. The reason I want to frame it this way is that similarly, I think the article failed in its goals as coming across (to me) as “look at this neat thing about Japanese”.
It is just kind of the nature of written communication that it takes a lot of editing and polish to make it clear, correct, and concise. I had the good fortune to sign up for Japanese 101 when my professor was in the middle of writing a new Japanese textbook—it was pretty exciting, with the changing lesson plans, the flock of master’s students hanging around, revisions and drafts to teaching materials, and those endless hours of classroom observation. The teachers occasionally gave us a “peek behind the curtain” and explained why they chose to teach things a certain way or another. I’ve rarely gotten that kind of explanation in any class that I’ve taken so I thought it was pretty special.
I don’t expect you to put in the textbook-level of polish into your article but there is a kind of verbosity (the article is long, which makes it kind of hard to respond to because there is just so much to sift through), there are some problems with clarity (the issue of romanization and orthography is mixed in with the conjugation, and maybe it would be better to separate those issues) some problems with correctness (various) and some problems with completeness (the patterns omit some conjugations that I think you don’t know, and I don’t think they follow the pattern).
I have certainly put effort into articles that have gotten brutal negative feedback; I think it was right for me to write the article, and then feel like shit from the feedback, and then maybe retract and revise it. If there is one actual error here, a true error, I think the error is fighting out criticism in the HN comments.
Actually let me just try to explain my pedagogical approach and philosophy here. Maybe that makes it clearer.
I assume no prerequisites at first. So my reader has never seen a kana table and doesn’t know which syllables exist.
I choose to teach conjugation first. That’s an unorthodox choice but I like it! That’s what I set out to do. So we get far enough until it breaks down. And it breaks down when a rule (which worked so far) doesn’t help with “s” because saying “si” would sound wrong.
That’s the moment I use to teach kana table and its importance. This “you made a mistake” is a pedagogical vehicle for introducing kana rows. And we go over the exact ones that you’d make a mistake with. So each special case is walked through.
At this point we could discard Hepburn but I choose to keep going because if you know special cases, there’s no issue. And at some point you’ll learn kana anyway.
So that’s how I chose to layer it. Maybe it’s a bit unholy but I like it. It is definitely self-consistent.
I understand why you wrote the article this way, I think the lesson here is “we have learned why Japanese textbooks do not teach the content in this order” and there are a couple reasons why this order is not good:
1. It relies on people not understanding certain things. In general, you cannot expect people to have exactly the right misunderstanding necessary for a lesson.
2. Spending extra time with Hepburn reinforces it, and it shouldn’t be reinforced.
I am in general extremely skeptical of lessons which try to engineer a way for the students to make mistakes. What I have seen in real classrooms and in informal teaching is that the mistakes are habit-forming and the outcomes of this kind of engineering are unpredictable.
Mistakes are appealing to the developers on HN because we understand things more by seeing them fail. But this does not mean that you can engineer somebody to experience the same moment of enlightenment that you did, because it requires constructing the same (incorrect) mental model that you had when you made that mistake that led to useful insight, and it both difficult and counterproductive to try and make that happen to students. Give people the best chances to learn by giving them the best chances to avoid mistakes, and the mistakes and insight will happen organically on their own, in unique ways for each student.
Okay I see where you’re coming from, sure. We assume different things about the reader and the context of the article. This is an article for people like me who like this approach — maybe engineers or people with an engineering mindset.
I also genuinely think it’s not that deep and that there’s no complex mistake being engineered here. I don’t believe you that the “mistake” of “sa with a replaced by i must be si” is an an unusual one for someone who hasn’t yet internalized kana. If we test this on random people on the street, I’m highly confident an overwhelming majority will make this exact mistake.
I agree with your broader point that “teaching via mistakes” is a risky path not worth it when the mistakes start getting combinatorial. I also think it’s absolutely fine when everyone does the same exact mistake, and there’s exactly one way to avoid it.
> My main mistake seems to be meaning “[Hepburn] romaji” by writing “romaji”. I was obviously aware of other systems because that is what the sentence says but I thought it’s acceptable to refer to Hepburn as just “romaji” as a sort of the default one.
"Romaji" does not (in English) mean "romanisation", as most people who've studied Japanese to at least beginner level know.
Hepburn is the official romanization chosen by the Japanese government (it's a relatively recent change), kunrei-shiki has been deprecated and all the signs etc are in the process of being converted to Hepburn.
I've looked at over a dozen hiragana tables and they all use Hepburn romanization.
Obsessing over romanization, something that a student ought to outgrow, is a sure fire way for a student to get overwhelmed by irrelevant details that discourage learning. The hard part is putting in the work, not learning less than a dozen exceptions.
I think it’s probably a mistake to use Hepburn if you’re learning Japanese, it kinda gets in the way. Either learn kana (which takes what, a week?) or use one of the other romanization systems which maps more cleanly to Japanese orthography
It’s a deliberate choice in the article. I cover every single caveat with it explicitly. I also mention this:
> (note i could also have used a different romanization that renders し as "si", つ as "tu", and ち as "ti" for this article. i decided to not because everyone else uses romaji, and once you understand this point once, you shouldn't have a difficulty doing this in your head.)
I think the choice is not a good one, whether it is deliberate or by accident, it is not a good choice either way. The main caveat to Hepburn is that it’s unsuitable for explaining how Japanese works and it’s unsuitable for learning Japanese—so before you start working on verb conjugations, you pick up kana or one of the romanizations which is more aligned with Japanese.
The idea that you “shouldn’t think in romaji” is really “you shouldn’t think in Hepburn”. This is an important distinction! Japanese has a relatively small inventory of phonemes, somewhere around 20 or 22 of them, and they map very neatly to the latin alphabet.
But the article doesn’t make this distinction, and seems to rely on confusion induced by the Hepburn romanization in order to make its points.
IMO, this is kind of like seeing an article about how monads are burritos. Thinking that a monad is a burrito does not help me understand monads.
Nomu -> noma-nai / nomi-masu / nome-ru / nomo-u
Miru -> mi-nai / mi-masu / mire-ru / miyo-u
The ichidan and godan verbs are not assigned different categories because existing scholars of Japanese are just bad at explaining how they work, and you can still understand them just fine in romaji. I put the hyphens above to mark a place where you could think that the verb ends and the common conjugation forms end, and you can see that the part on the left has somewhat different rules for ichidan and godan verbs, even when you apply the “tricks”—but some of these forms may be unfamiliar if you are are starting out (are you familiar with miru -> miyou conjugation, or miru -> mirareru?)
I concede that using Nihon-shiki maybe would’ve been more elegant for what I tried to do in the article.
> But the article doesn’t make this distinction, and seems to rely on confusion induced by the Hepburn romanization in order to make its points.
Not at all. I give it two sections and then we move on. It doesn’t affect literally anything else on the page. You just learn to shift rows and move on. To make what points?
> you can see that the part on the left has somewhat different rules for ichidan and godan verbs, even when you apply the “tricks”—but some of these forms may be unfamiliar if you are are starting out
I’m not quite sure what you mean to say in this part. I do cover -[r]eba and -[y]ou in the final section (“one more thing”) which extends the model to clearly handle that disappearing consonant. I think -[r]eru fits in there the same way, just as -[r]u itself.
I think explaining it as mi + [y]ou = miyou, but nom_ + [y]ou = nomou is a clearer way to think about this. The rule is that the hole burns down the leading consonant (but takes the vowel).
I've never understood how people can claim that learning kana takes a week. It clearly takes more time than that, considering how similar some of the symbols are and a lot of them only differ by double dashes or a stroke (think nu vs me, ne vs re, ro vs ru, chi vs sa, and so on). Then there are the combinations and even if you managed to learn hiragana, you still have to learn katakana.
Oh and I forgot, you have to actually learn how to listen, pronounce and speak them, not just learn a useless romanization mapping. I've heard way too many English speakers just say the romanization with English pronunciation. At that point their learning efforts turn into self sabotage.
In total that's definitively a month of effort, albeit spread out over the first year of learning.
I think "a week" is slightly optimistic, but I also think "a month" is slightly pessimistic. When I learned hiragana, I spent my free time drilling on RealKana [0]. I'd focus on a new column of the kana table, then bring in columns I'd already practiced, until eventually I could flash reliably on cards drawn from the entire table. This legitimately didn't take much longer than a week, because learning a single column in isolation is very quick, and the real difficulty comes in distinguishing similar kana (as you say). But I was able to drill similar kana by selecting two or three kana that would force me to see them often. (I still struggle slightly with wa, re, and ne, but I definitely know them.)
I also drilled on a drag-n-drop kana table [1] in a few ways -- sometimes I'd start from the kana and try to figure out where they should go in the table, and sometimes I'd go along rows or columns in the table and try to find the kana that belong there. These two directions drill both recognition and recollection.
Proper pronunciation is a cross-cutting concern. As a whole, it's not something you can reasonably learn solely from kana, but the aspects that are relevant are not difficult to pick up. Every kana breaks into one (vowels and N) or two (the rest) phonemes, and for the most part, the way you pronounce those phonemes is consistent across rows and columns of the table (admitting exceptions like "shi" and "tsu"). If you are taught those basics, learning how to pronounce kana is not hard. Training your ear to "hear" distinctions among English allophones, and to distinguish pitch accent from the more familiar stress accent, is much harder, and really has to come from experience, not just kana.
The Japanese phonological system doesn't allow a /s/ sound to occur before the vowel /i/, the consonant must undergo palatalization and become /ɕ/ (the IPA symbol for the Japanese sh-like sound). Because this is a regular sound rule, the native writing system doesn't have a way to distinguish the nonexistent */si/ sequence from the /ɕi/ sequence that actually occurs, and this is the syllable that hiragana し or katakana シ indicate.
In the Hepburn romanization system, which generally tries to be transparent to speakers of English or other European languages, し is romanized as _shi_, because this indicates to English speakers that the /s/ -> /ɕ/ sound change happens. In the Kunrei-siki romanization system, which tries to be more faithful to the distinctions made in the Japanese phonological system, し is romanized as _si_ to be consistent with the other possible syllables _sa_ _su_ _se_ _so_ that begin with the consonant /s/.
And yeah the fact that the article-writer hasn't internalized this sound change yet is a sign that their command of Japanese isn't all that good yet.
> And yeah the fact that the article-writer hasn't internalized this sound change yet
I don’t understand where this misunderstanding about my article comes from. I am saying that the sound at the intersection of “s” column and “i” row is “shi”. My article uses romaji so this is self-consistent. I am also mentioning that there is an alternative system that would romanize it as “si” but that’s not the one I’m using in my article.
> I am saying that the sound at the intersection of “s” column and “i” row is “shi”.
That is exactly the problem. Japanese doesn't distinguish between 'shi' and 'si', so all you're really gaining by pointing that out is learning how to correctly romanize Japanese, in a single system. Instead of learning the language you're learning how to represent the language in a foreign way.
The rule 0 of learning a language is to get rid of the crutches as soon as possible. Use their native writing system (or if they're one of the latin alphabet users, use their pronunciation rule), learn words of the target language using said language, and learn how to formulate concepts with the language rather than translating it from what you already know. Crutches should only be used to get to this point and no more. If you do that, details like 'si' and 'shi' are not even worth mentioning. Romanization methods have their own goals, and rarely is it about facilitating language learning.
> this is actually completely intentional. some purists may dislike that, but [..] i think it provides a much clearer intuition for Japanese verb conjugation than writing it in hiragana.
You are that purist that I’m talking about. I get your argument. I just think it’s a non-issue in my chosen pedagogical approach to convey what I wanted to move on. People unaware of how romanization works will nod at it and move on. People aware of it will not have issue with it. People with opinions about proper way to teach the language are the ones that will get stuck on this.
The article is engineer-brained. It’s for people like me. It’s not for the people you’re teaching, assuming these strong opinions come from you being a teacher.
As someone that recently went through an introductory Japanese course in Japan, I don't find this much different than how it's taught. Or maybe I'm missing something?
It seems like the article is trying to make the case that in romaji, you can split the letters and isolate the vowel (e.g. the asterix in the article's conjugation).
But we were simply taught to change from the う- row to the い- row (u- row to i- row). I switched to Japanese to illustrate that you can make that statement even without romaji. In that case, it seems like basically the same thing?
As an anecdotal point, my class was mostly non-english speakers and I didn't find the above to be a sticking point for my classmates. The real sticking points were messing up the ichidan verb exceptions (ichidan verbs that look like godan) and conjugating the correct form for the different grammar points. Te and ta form were also a bit tricky. But the article doesn't seem to offer anything new to help there.
If you’re taught to shift rows, and you already comfortably think in kana — yes, I think that’s equally good. Some of the materials I’ve tried learning from when I got started with the language didn’t do that, and instead described each case as special. That’s what I didn’t like.
That said, part of the challenge to myself with this article was to allow someone to learn Japanese conjugation even if they have zero knowledge of the language (even no kana). So that’s another constraint influencing my choice. I also wanted to have the visual “gluing” throughout the article as an aid for intuition, so that’s another reason I used romaji.
I've found that any resource relying on any romaji after the first chapter or two is often a complete waste of time.
It slows down beginners needing to make the hard jump, since romaji is never used except for signs in real life, and it just becomes a distraction to the material for anyone who is not a complete beginner. Furigana is helpful to the intermediate learner, romaji just becomes harder to read at that point.
Agree. Especially how easy it is honestly to learn hiragana. You can practically learn it in a day and keep a table next to you to look up every time you forget one.
If you're at the point you're learning verbs you'd be mad not to know how to read some kana.
The benefit you give (be able to "cut" a kana in the middle) is really weak, I've never seen anyone being confused by that when learning in kana.
This is completely nullified by all the drawbacks of using romaji while learning and they're well known already.
The only reason to use romaji for Japanese grammar is to explain the concepts to someone who has no interest in learning the language, just for their general knowledge.
I’ve explicitly addressed this in the article. Cmd+F for a section called “why romaji is actually good” and then “why romaji is actually bad”. You may disagree with the approach, but I outlined my reasons for choosing it (as well as its downsides).
You glossed over my point. I’m not using it as a “crutch” for reading. I’m using it to have notation for the stem — the thing before -u. I could choose alternative notation with kana (e.g. just always using the -u ending, or the idea of variable stems like i-stem and a-stem) but then the visual “gluing” wouldn’t work. Which is the whole point of mental model I’m communicating. It’s fine if you don’t find this mental model helpful but it’s the point of the article.
I’ll be honest that I also wanted (as a challenge) to write this article so that a person with zero Japanese knowledge would be able to correctly conjugate almost every word to every ending. This is more of a teaching drill for myself though but it’s another reason for the romaji choice.
The explanation made sense to me: romaji works well for vowel shifts (as the vowels aren't glued to consonants) while kana works well for consonant shifts (because the vowels are glued to consonants).
Latin text's smaller tokens/phonemes have advantages and disadvantages, but they are a convenient notation for getting the author's point across.
The difference in phonemes reminds me of how game designer Naramura came up with the (Spanish-sounding) name "La Mulana" for his game by spelling his name backwards in kana. In romaji it would have been "Arumaran" which is completely different (while in kanji it would have been "Muranara".)
Not quite. If you change the order of some kanji, the general case is that the resulting text has no definite pronunciation. You definitely would not expect that the sounds assigned to the kanji in one ordering would be the same ones assigned in a new ordering.
This is a phenomenon the Japanese sometimes play with. In the novel Musashi, Musashi comes up with that name by reinterpreting the characters of his actual name (which, in the novel, is Takezō).
This is true, the only practical place where romaji might be used daily is for IME input.
This is why I think all the commenters obsessing over romaji just add fuel to the fire by being elitist over something that should only matter to an absolute beginner.
It's especially vain if the primary reason to choose one romanization strategy over another is to save keystrokes on a keyboard.
I wonder what the author’s native language is as they seem to struggle with the concept of alternation.
For me, the first half of the article could be removed and the learning could simply start with “there are two types of verbs in Japanese (+ some irregular verbs), one type conjugates without alternation in the root, the other with”. That’s enough to get the mental model but my native languages have alternation to begin with so it’s an intuitive concept.
The comments to this article are another example of something I see so often in Japanese language learning dicussions I see online. It's always filled with debate, disagreement, arguements over incredible subtle things, and everyone trying to optimize the best method. It can be really discouraging space for early learners.
Theorycrafting efficient ways of learning Japanese while being barely conversational yourself is a completely different hobby from actually learning Japanese (or actually studying the teaching of Japanese), and sadly in online spaces the former often swamps out discussion of the latter.
The online spaces can be really discouraging, true—but it can also be really discouraging to be in a classroom or in a foreign country, struggling to use a language you barely know. Meanwhile, there are also a lot of ways to spend effort trying to learn a language without gaining mastery. Truly frustrating.
If you want some decent level of fluency then there is something you have to do, which is to communicate with other people in context and with specific goals in mind (get information, give information, make a request, etc). Whatever you can do to arrange for that to happen is probably more valuable than anything you can do online or with books. I personally like to recommend finding classes at a local college.
If you can’t get that, then I think the next best thing is reading and listening.
Drills are also necessary but you can easily fill your time with drills without advancing your ability to communicate or understand people.
There is plenty of research about what is / is not effective when it comes to learning languages so I encourage people to at least take a look at the results of that research rather than just go with whatever people recommend online (I’m just some random person online, I may be no better than the next). AI tools reportedly have a positive effect but they are not nearly as good as human interaction.
It's not discouraging at all. Japanese concepts do not have a 1-1 mapping with English concepts, so there is a lot of debate about how it can be taught. I find it fascinating.
Author here — it’s a really fun, elegant, and beautiful language. I highly recommend it if you’re interested. Maybe controversial but I think you can largely ignore the Japanese-learning community if/when the vibe isn’t right.
Mostly it helps to find learning resources that gel with your style of learning, and if possible, a tutor so you have a roadmap and more motivation. I found mine on italki. I also find Claude very helpful for sentence drills based on words and grammar I know.
Japanese language learning environments really are ridiculously hostile and I've stopped engaging with them unless im searching for a specific resource.
I've dabbled a bit in Mandarin and while eventually i ended up liking Japanese more as a language, the Mandarin language learner community felt like such a warm bath in comparison. The people were friendly and welcoming and willing to help and genuinely excited to find more people wanting to learn the language.
And, since the English equivalent of those sounds doesn't exist, there's no confusion the way there would be between "she" and "see" in english. Complaining that there's no english equivalent of the russian (взгляд / vzglyad)'s initial cluster would be similar in feel - no english words use it, so the romanization can be whatever you like, really.
> And, since the English equivalent of those sounds doesn't exist, there's no confusion the way there would be between "she" and "see" in english.
Erm, wtf? The English "si" sound does exist and sounds different from し. There is a reason people don't want to write Sinzyuku, and while I think they're making the wrong tradeoff, it is a tradeoff and should be acknowledged as such.
"As a result, the sequences [ti si di (d)zi] do not occur in native or Sino-Japanese vocabulary."[1] Unless I'm missing something, Japanese phonology doesn't include the sound english has as "si", only "shi"? I'm not a native speaker though, it's entirely possible that I am missing something.
It isn't. It falls slightly apart in the `s` column, and completely in the `t` column which contains both "chi" and "tsu". It also breaks for godan words that end in "u" which become "wa" in the negative form.
Mu, bu and nu also all obey the -nda transformation due to phonetics, and not due to how "if we just shuffle the letters around and presto! Nomu becomes nonda".
Japanese already has plenty of its own reading inconsistencies, so adding another layer on top isn't going to help you.
Finally, there's going to be so much kana in your every day life that learning conjugation in romaji is guaranteed to cripple your reading, because instead of recognizing kana (e.g. you see a billboard that says お茶を飲んだ方がいい！ as you frantically try to back-translate everything into romaji, but also removing excess w's and converting nda's as you go) you've spent the first n hours on trying to "hack" the language instead of just learning it.
> It isn't. It falls slightly apart in the `s` column, and completely in the `t` column which contains both "chi" and "tsu".
All of this is described just below, in the section called “why romaji is actually bad”, with these specific examples being tested.
> It also breaks for godan words that end in "u" which become "wa" in the negative form.
This is also described further below in the post.
> as you frantically try to back-translate everything into romaji, but also removing excess w's and converting nda's as you go
I see “understanding the system” and “applying it fluently” as two separate activities. I find romaji more illustrative for the former because the system is phonetical on a sub-mora level. Whereas kana is more helpful for the latter. I don’t assume my reader is an idiot, and so I assume that they would be able to pick the right tool for the job on their learning journey. The fluency always layers on separately anyway, understanding doesn’t “help” there. It’s just that personally I find understanding a great motivator (and fallback) for developing fluency.
-u godan verbs historically ended in ふ /fu/ (and earlier /pu/), and were written that way until the 20th century kana reform. So the historical conjugation pattern of a verb like tamau was tamafa-/tamafi-/tamafu-/tamafe- + additional inflectional endings. The /f/ became a /w/ before /a/ but weakened and was lost entirely in other positions, leading to the modern pattern of tamau, tamaimasu, etc. but tamawanai.
As someone who is learning Japanese and is struggling through learning conjugations.... this is not explained simply to me.
Maybe it's because I'm at a different part of my learning journey, but this seems to go on and on and on about godan v ichidan verbs but doesn't touch on the actual conjugation (ie. how to turn "drink" into "didn't want to drink"), nor on plain form (everything is in polite form), and te-form gets a very very short mention at the end x_x
Also, the romaji did my head in but I understand that was done for a reason (I don't agree with the reason, but at least there was one)
A plug for my favourite book here. Richard Webb's 80/20 Japanese covers these topics extremely well, and in a very logical / productive order for a foreign speaker.
(and it comes with very enjoyable audio tapes and anki cards to boot)
I started to learn Japanese 30 years ago, and in my experience the people who try to be smart and build systems almost never get decent. It’s procrastination while thinking they’re actually productive.
To add insult to injury this article hasn’t discovered anything new, makes it sound way more complicated than it is, and in the end still requires you to just remember which verbs are of the eru/iru group, and which are not (which was posed as a problem to solve in the intro).
Just make cards and mark the stem, learn it along with the verb. No need for heuristics. If you ever forget, you’re bound to remember the masu-form and can reverse engineer the stem from that 100%.
Agree. Experience shows that fluency arises when you don't have to think about rules anymore. My advice is to not spend too much time learning grammar rules (actually, no time, like native learners). Leave the rule discovery to your unconscious brain and get going with rote repetition.
Your brains "language module" is not a slow computer, computing rules, it's a fast lookup-table.
Similarly, when complaining about how you have to memorize a big table of verb conjugations in the intro, the author links to a table of... -ta forms, a verb form for which the author later concludes you just have to memorize a big table.
The te/ta-form is genuinely a separate system that doesn’t reduce further. I think there’s still value in having a solid model for everything else. At least I personally found it valuable, which is why I thought to share it with people.
I'm not totally sure this "stems and suffixes" mental model really works well for everything else. Forms like the imperative (食べろ), volitional (食べよう), provisional (食べれば), potential (食べられる), and causative (食べさせる) aren't cleanly handled either -- they work similarly for godan verbs, but you have to add a different suffix for ichidan verbs.
It's definitely useful to understand how "chi" and "tsu" fit into the hiragana chart, and if your asterisk notation helps you remember which verbs are ichidan vs godan then that's great, but I'm not sure it's worth trying to unify -masu and -nai into one model.
> Forms like the imperative (食べろ), volitional (食べよう), provisional (食べれば), potential (食べられる), and causative (食べさせる) aren't cleanly handled either -- they work similarly for godan verbs, but you have to add a different suffix for ichidan verbs.
They are cleanly handled in the final section (“one more thing”) that introduces a notion of disappearing consonant like -[r]u, -[r]eba and so on, and gives a rule for it. This is a perfect inversion of what happens with -(i)masu and friends. The hole in the stem accepts the leading vowel but burns down the leading consonant.
"Disappearing consonant" doesn't work for the potential form, unless you expand the representation to allow writing -[rar]eru. (Edit: And I think imperative would require like "-[ro](e)".)
Which, like, is clean in the sense that Redux is technically Turing-complete (you can encompass _any_ difference between two strings by saying that one string uses the stuff in brackets and the other string uses the stuff in parentheses), but that doesn't make it a good idea.
Okay, that one’s fair! I remember there was also some other one that had it split into two completely different suffixes.
My answer to this is that by the time you’re learning those, you’re already so fluent in conjugation that a couple of special cases will layer on fine. It’s way better then you get in most languages. (And pedantically I’d still say [] works for the cases above, as you’ve shown.)
I honestly don’t understand the cynicism here. If I could read this article when I started learning, it would’ve saved me a ton of time. That’s why I wrote it. I hope it’ll be useful to someone else but it’s fine if not. As an educator I’m proud of how much it crams in that’s usually spread over many weeks, and how the simple model almost perfectly generalizes. But yeah sure it’s making some unorthodox choices and leaves a couple of advanced cases within one indirection. I’m still very happy with it.
Why would you expect the article where I’m describing what worked for me to “discover something new”? I’m literally sharing the mental model that I personally found helpful. There’s nothing “new” in learning or teaching a language. But this is the most minimal model I’ve found useful, compared to others, and I wanted to share it with other people.
I think you’re taking a lot of stuff for granted. “Just” do cards etc. You’re using the word “stem” but what’s a stem? Why do we sometimes inject -i or -a (or -wa) there and sometimes we don’t? You still have to learn that and understand that. That’s what I’m describing in the article. If you already know stems and how they compose with suffixes, congratulations, you won’t find my article useful.
Honestly, you don't need to "learn" or "understand" much grammar explicitly. I think it definitely helps get you off your feet, as you can "decode" sentences if you remember grammar rules, but eventually the grammar has to be internalized anyway. This happens when you are repeatedly exposed to the same patterns in context. I don't know how English or Norwegian grammar works, and I'm fluent in those. I skipped grammar in Japanese and focused on reading, yet I can understand most things and I can tell when something sounds wrong.
One big change I had when learning Japanese was that someone introduced to me Cure Dolly videos on YouTube, and it has been an eye-opener: All these verb conjugations are actually attaching another verb to extend its meaning
First, we learnt verbs in the -masu form. Nomimasu, tabemasu and so on.
Then we learnt this song (to the tune of Clementine)
chi ri i tte
mi ni bi nde
kiite
giite
It's a quick mneumonic to help you go from the polite verb to the "te-form" ending. I hummed it in my head while working out the conjugation before it became natural and "obvious".
For me the biggest problem was never memorizing the suffixes but determining whether the verb belongs to group 1 or 2. While you can tell immediately that some verbs belong to group 1, the others (e.g. kaeru mentioned in the post) are not so easy (as far as I know there's no algorithm for that and you just need to memorize with every verb to which group it belongs).
Funny how some conjugated forms of verbs collide with dictionary forms of other verbs (esp. if we ignore pitch accent differences):
E.g. potential of 買う (kau, to buy) is 買える (kaeru), which is spelled like 帰る (kaeru, to return home).
It reminds of you how "lay" is a verb (to put something into a flat resting position, but is also the past tense of "to lie" (take on a supine position).
Today, I lay bricks; yesterday I lay in bed all day.
Plus lie and lie are examples of how English verbs can be homonyms in dictionary form, but conjugate differently, something we see in Japanese (either actual homonyms or near homonyms modulo pitch accent).
Romaji are great, and in some ways more instructive because they reveal patterns which are otherwise a little hidden. You just have to realize that S+I is shi, T+U is tsu, etc. I don’t want to get too deep into it but there is a regularity to the language, and rules, and different choices of writing system reveals different pieces of the puzzle.
Next, the conjugation itself. There are massive categories of conjugations missing! Like, how do you get from taberu / nomu in this system to tabereru / nomereru? It turns out that these ichidan and godan verbs actually do have some differences in conjugation. Who’d have thought? (There is the -i stem, but there are other forms.)
Both of these things are described in the article. The first one is in “why romaji is actually bad” section, the second one is in “one more thing” part at the very end.
Looks good. It essentially separates phonotactics out of conjugation patterns instead of conflating them, which works also very well with how I learn languages. How to conjugate then breaks down to finding the stem (yes, some learning is always involved), and then there are not even different conjugations, but there are simply vocalic and consonantal stems -- all the morphology is in the suffix (the 'secret vowel') and the rest is phonotactics. I am sure there are irregular verbs, but probably this explains 90% or more of the verbs with simple rules. For me, this creates a perfectly logical system. Thanks for the overview!
"hanasimasu" is not exactly wrong; there is a romanization system in which "si" is how you write "shi".
If you want to invent scheme for understanding conjugation which works by cracking the romanized versions of words to create a pseudo-stem that could not actually exist in spoken language, it behooves you to adopt "si" and "ti", because they bring in a consistency needed by such a system to be complete.
> If you want to invent scheme for understanding conjugation which works by cracking the romanized versions of words to create a pseudo-stem that could not actually exist in spoken language
That's how all conjugation schemes work. There's nothing weird about this. Stems aren't supposed to exist in the spoken language. But they are observable in the spoken language.
Compare how a modern dictionary will give you ποιέω, a full and fully-inflected word which doesn't actually exist in ancient Greek, as the first principle part of that verb. This is done because the stem of the verb is ποιε-, and the epsilon ending the stem can be easily observed by its effect on most of the conjugational endings. It doesn't happen to affect the first-person singular ending -ω (to be precise, the contraction of ε- with -ω is -ω), so the dictionary form is synthetic, chosen to be informative.
Japanese is known as an agglutinative language [0], and how verbs are conjugated also has a lot to do with politeness, as well as local dialects. That's why you can turn on an anime and hardly understand it, even after a couple years of study.
I got to the third year college level in my own Japanese studies, and at that point, memorizing kanji was starting to compete with my computer science studies, so I had to drop it. I got to travel to Japan and live with host families (we kind of settled on a Japanese/English pidgin), so I don't regret the experience.
Fun, and a programmatic perspective. However, it can be too easy and fun to get super caught up in these details, if your goals are some level of fluency and ability to communicate/read. The majority of people that I know who have gained any level of fluency in Japanese as an adult mostly avoided stuff like this because (for many people; of course everybody is different) doing all of this mental math to dive down to the last detail was nowhere near as effective as some speaking and reading drills.
By all means, it is fun to play with a language. And every person's brain works differently.
I like to use this metaphor, though. You're hiking a mountain, this journey to the "peak" is reaching some goal of fluency.
It's fun to stop and look at rocks, examining, comparing and whatnot. But it doesn't necessarily get you closer to the peak. I mean, it might, because you'll better understand your footing every-so-slightly. Not a perfect metaphor.
I'm a bit dismayed by all the negativity you're getting. Thanks for all your level-headed and detailed responses here.
I have a hard time understanding why people object to this kind of attempt to systematize something that conventionally relies on rote learning. I've always found that figuring out the underlying rules helps with the rote learning. When you see some new word or conjugation, it's super helpful to have a method for explaining why it's that way; it helps fix it in your mind.
I wonder if native speakers object to this approach because it isn't really how you learn as a kid. You don't learn any abstract rules, you just absorb a huge amount of training material and construct your own model. But that doesn't mean adults need to learn in the same way; we can and should leverage our painfully-acquired knowledge and skills from other domains.
Maybe it’s not very clear but I don’t suggest studying from articles like this alone. Obviously you need to do sentence drills and talk to a tutor or native speakers to have any chance of success. I still find documenting my mental model helpful because most articles I’ve seen before were not teaching it clearly enough for my taste.
If it's taking you this much effort to do the trivial conjugations (seriously, the whole page barely mentions the interesting ones 80% of the way down, and falls back on "yeah, you just have to memorize the patterns" for た/て forms), yeah, just memorise them. Language learning and exercise are the two things where I've found the programmer's instinct to "work smarter, not harder" works against you; you actually just have to put the time and effort in.
I prefer having a system to simply memorizing. I don’t know what you mean by “so much effort”. I am literally just describing the system as it is brick by brick. If you see an opportunity to simplify, you’re welcome to provide a specific suggestion. I find this system rather elegant, and I tried to build it piece by piece because that’s my preferred way both to learn and to teach.
>the whole page barely mentions the interesting ones 80% of the way down
The te/ta-form is genuinely a separate system linguistically with its own heritage. So it makes sense to look at it separately. I don’t consider it more “interesting” and I’d argue getting the details right with other forms is much more useful coverage-wise. So I didn’t spend much time on te/ta-form. (That said, even for -te/ta form, I find it calming to think of -nda as a contraction of -nita, and so on, which AFAIK is in the ballpark of what historically happened.)
> Language learning and exercise are the two things where I've found the programmer's instinct to "work smarter, not harder" works against you
I agree you need to put time to practice and all that. But if there’s a genuinely simple system underneath, I always prefer to see it. Even if there’s a layer of memorization and repetition to achieve actual fluency. Japanese conjugation is a rare case where the system actually is very clear and methodical. The article is written for people like me who also prefer to know it. There’s literally thousands of resources that teach it your preferred way, so I don’t understand the impulse to complain about someone teaching it differently for a change.
This is a pretty long blog post covering really not very much.
> The te/ta-form is genuinely a separate system linguistically with its own heritage. So it makes sense to look at it separately. I don’t consider it more “interesting” and I’d argue getting the details right with other forms is much more useful coverage-wise.
It's not just te/ta, you don't mention anything other than the basic polite/casual, positive/negative, and desiderative. At the very end you point to conditional and causative but say you haven't studied them, and no mention at all of passive, imperative, causative passive, or volitional.
> I agree you need to put time to practice and all that. But if there’s a genuinely simple system underneath, I always prefer to see it.
And how's that working out for you?
> There’s literally thousands of resources that teach it your preferred way, so I don’t understand the impulse to complain about someone teaching it differently for a change.
I find it very presumptive to propose to "teach" what you haven't really learnt. How many people have successfully become remotely close to fluent following this approach? It's 0, right? What makes you think you're "teaching" rather than leading people astray?
Indeed, especially for a language with forms of verb as regular as in Japanese. The whole language has two and a half irregular verbs. Compare that to Spanish and realize how fortunate you are to study Japanese verbs.
meh, language learning has an inconvenient truth: sometimes it’s just rote memorization. it's the reflexive belief that every human endeavor must have a hidden optimization waiting to be discovered. Language learning is one of those domains that stubbornly replies, "Cool flowchart. Now memorize 500 words and spend 200 hours listening."
There’s no clever engineer hack that replaces time spent with the language. and with regard to japanese, please stay away from romaji, unless you're still in beginner stage and typing things out to communicate words you already know the phonetics to.
I mean I think it’s both! As an author, I wrote this to settle my mental model. This doesn’t mean I could use it during fast speech, but I find it calming to understand the actual linguistic system behind the tables. Especially when it’s so elegant.
The choice of romaji is deliberate for multiple reasons and is defended in the article (with counter-arguments for why it’s bad too).
this is quite intriguing, as a native speaker and someone with friends trying to learn Japanese, I always had a hard time explaining all the different patterns and just defaulted to "it just is". Will use this in the future!
Categorizing Japanese verbs as -ru or -u requires more context.
I prefer the term "group 2 verbs" to "-ru verbs." Group 2 verbs are verbs that end in -eru or -iru, not just -ru. Of course there are some exceptions, like kaeru, which ends in -eru but is actually a -u verb. Conjugation is easy: remove the final -ru and append -masu, -mashita, etc.
"Group 1 verbs" (again, -u verbs) are verbs that are not group 2 verbs. Conjugation is a bit more difficult because the -nu, -bu, -mu, and -u verbs have many suffixes. However, after memorizing these two (-nbmu and -u, because -nu, -bu, and -mu are almost the same), the rest are easy.
There are only two irregular verbs: kuru and suru. Just memorize them.
I learned Japanese by just memorizing. Once you have memorized enough verbs and their conjugations, you can figure out the conjugation of a new verb even if you don't understand how it works.
there are more irregular verbs than just kuru and suru. iku and aru are also irregular, for example.
Irregulars notwithstanding, the conjugation pattern is actually completely lossless if you just remember the imperative form (e.g. 着ろ kiro, 切れ kire) instead of the infinitive, which is lossy (e.g. 着る kiru, 切る kiru). Then there's no need to have to remember, "oh... is this -iru verb group 1 or group 2?"
I learned this 35 years ago to the tune of Clementine, youre using black sheep, my spouse uses another tune, but what funny is I learned it not with the 'i' endings, but with the dictionary form (sometimes called base-3) 'u' forms.
u tsu ru tte
bu mu nu nde
ku ite, gu ite
without the shi shite as that had been learned well ahead of the lesson adding ta/te forms.
I just think it's interesting how readily a little ditty tune helps people with recall, regardless of the actual tune.
I've generally found this kind of thing annoying in my various attempts to learn languages - in one of my high school spanish classes the teacher thought it was important to have us learn a little song to memorize the six present tense forms of the to-be verb ser: (yo) soy, (tu) eres, (el/ella) es, (nosotros) somos, (vosotros) sois, (ellos/as) son. My thinking was, there's only six forms to memorize, and the verb is an incredibly common one ("to be"!) so they'll get constantly reinforced anyway. The song is silly and isn't really helping anyway.
I feel similarly about the transformations for the Japanese -te forms and -ta past tense marker. The entire system is:
So basically ten patterns which group into 5 subpatterns. There's some logic behind them - the -te and -ta morphemes originally got added to the -i/pre-masu/ren'youkei stem and then underwent some idiosyncratic consonant reductions in godan verbs. But, really, it's only ten patterns, you can just memorize them; and these are incredibly common verb forms that get used all the time so you'll have them reinforced frequently if you are at all engaging with the language. It's a lot less to memorize than if you were learning Ancient Greek or Sanskrit or something.
Wow, there is a lot of negative gatekeeping on learning Japanese in this thread. I recommend people chill out.
Dan, this is cool, and I can tell you enjoyed writing this up and thinking it through as you wrote. Don’t let the haters get you down.
I think you’d like Orbit (withorbit.com), a tool that got built out of some experimental learning work I think originally done at quantum.country — it’s intended as a sort of toolkit to help turn blog content into spaced repitition inline memorization tools.
Check it out! I think the post would benefit from something like this - it’s fun for most to play with seeing how they are understanding this, and you spend a lot of time trying to create a bit of interactive reading/thinking in the blog.
Keep writing, please, it’s good.
Don’t want to sound negative, just want to share my observations as someone who’s been actively learning Japanese for a couple of years. I much prefer how Japanese people actually teach verbs: 一段 (ichidan), literally "1-step"; 五段 (godan), literally "5 steps", plus a few exceptions sometimes called irregular verbs. It's not hard; it makes sense, just some textbooks (e.g. Genki) teach it really bad.
Learning Japanese is about learning a new way of thinking and structuring your thoughts. The more you learn, the more you realize it just doesn't fit into the English world. You can't really translate Japanese into English without losing nuance — and sometimes that nuance is important. So start early and start training your brain to think in the language, instead of trying to translate it and force it into English or some other language brackets. It won’t work; it won’t make sense; you will get stressed and confused.
> Learning Japanese is about learning a new way of thinking and structuring your thoughts.
I suspect this is true with most languages that are not in the same language family. The indoeuropean languages are all pretty similarly structured (don't @ me for not knowing the eighteenth tense of lithuanian), but it still takes time and effort for an english speaker to think in french, even if for the most part concepts translate directly. But bantu languages, eg kirundi or zulu, lean heavily on verbs with an entirely different conception of sentence structure and morphology, where you can stick entire clauses into one word, and you realize that your brain is picking up on patterns to decipher grammar that wouldn't make much sense in indoeuropean languages like dahl's law (except apparently in greek and sanskrit, where it's called grassman's maw? Huh... now I know). Hausa is different still where you need to think about tense in an entirely different way—the pronoun is conjugated and the verb remains unchanged, and sometimes it feels like there are more irregular words than regular. Mandarin is beautiful, and actually quite simple to speak a little conversationally, if probably as difficult as english (or maybe more so) to master.
Learning foreign languages really makes you realize how central language is to basic cognition. You see the world in a different way, with different values and relations, depending on which language you speak/think in.
Ironically, I think this makes people extra susceptible to thinking that chatbots are intelligent (or even conscious, mr dawkins), even though it fails basic tests of memory and learning necessary to convincingly mimic understanding of, idk, time passing. Or motivation. Or emotion. But you can tell when it's awkwardly translating from another language in a way that a human would likely catch quite quickly if exposed to billions of documents.
As an unrelated tangent, I love seeing it when the “thinking” in a native language leaks into the writing in another, I find it so endearing and intriguing. My French colleagues sometimes produce the weirdest English sentences (perfectly understandable just unusually phrased), that alone occasionally sparks my interest on picking up a bit of French — I just want to learn how different the grammar and sentence construction is!
Obviously, it applies to other languages, I’ve just been working a lot with French. Well, and my own language often leaks into English, of course.
> I much prefer how Japanese people actually teach verbs: 一段 (ichidan), literally "1-step"; 五段 (godan), literally "5 steps", plus a few exceptions sometimes called irregular verbs
This is literally what I teach in the article, including these translations. Quoting it:
> in ichidan ("one-row") verbs like taberu, the last syllable of the stem is fixed. it's always going to be be, no matter the suffix:
>
> (table)
>
> it stays on a single row in the hiragana table, hence "one-row".
>
>on the other hand, in godan ("five-row") verbs like nomu, the final syllable of the stem alternates between ma, mi, mu, me, and mo:
> (table)
> it spans all the five rows, which is why it's godan ("five-row"). the m* "wildcard" represents the entire ma/mi/mu/me/mo column.
You’re also mischaracterising my approach. I am not teaching to “think in English”. Quoting from the article:
> i'm using romaji as a convenient way to refer to phonetics in text. however, your "mental algebra" should match the hiragana table. so this is a reminder to not think in romaji when you do calculations. when we conjugate godan verbs, we literally go up and down the column. (maybe all these textbooks that used hiragana had a point!)
If you have objections, please engage with the article’s actual content, not with what you assume it is based on a glance (“oh he’s using romaji, this is thinking in English”). I’m using romaji for specific reasons that are motivated and explained in text, and I show every single pitfall of that choice as well.
There’s nothing wrong with trying to make sense of something new using terms and concepts you already know and that click for you. We all do it, me included.
I was speaking more broadly about learning Japanese and what I see on HN. Every six months or so, somebody discovers a clever pattern in the Japanese language. It’s almost always related to something taught in the beginning, the N5 level. And that pattern seems to have been eluding rest of us.
Specifically about your post, I think there's a shorter and simpler explanation of the verbs. One good example: https://kellenok.github.io/cure-script/5-verb-groups-and-the...
Sure. I like Cure Dolly. I’m actually surprised you bring her up because she gets a lot of purist haters too.
I’ve found this lesson difficult to digest when I tried her approach. It’s actually where I fell off and lost the interest the first time I tried to learn the language.
The comparable lesson is primarily https://kellenok.github.io/cure-script/7-5-conjugation.html, not the one you quoted. I prefer teaching it before the -te form because it’s more orderly.
If we reorder these two lessons and rely on romaji over kana, you basically get my approach. I found it helpful so I wrote it down.
Where does the commenter say it was only a glance? I share the feeling.
The grandparent says it’s easier to teach by saying ichidan and godan actually mean “one-step” and “five-step”. That’s exactly what I’m teaching so it can’t be an argument against my article. This is something I not only say, but explicitly show visually with a table — as I have just quoted. From this I conclude the grandparent hasn’t read the article, and is commenting on vibes.
Then the grandparent says something I agree with (don’t force the language into another language) but I don’t think it’s a fair description of what I’m doing. It sure looks like that’s what I’m doing, but I strongly believe that learning conjugation is primarily phonetical (it’s about how it sounds, not kana itself) and therefore romaji is just a better pedagogical choice for someone not already fluent with kana. And no, I don’t buy that you have to put being fluent in kana as a prerequisite. The whole conceit of the article is basically that you can learn almost the entire conjugation system in one evening with zero prior knowledge of the language. That alone justifies the small shortcut I took to get there.
I figured it was alluding to more conventional approaches to learning them. Reads like you’re making assumptions and defending instead of engaging with criticism in good faith.
“Don’t want to sound negative […] I much prefer” reads to me like a comment phrased against the article’s approach. Otherwise why would it “sound negative”? And what I’m saying is that it matches the article’s approach. Thus, the assumption. Where am I wrong or uncharitable?
Are there any books you'd recommend that get it right?
How about kellenok.github.io/cure-script/
But actually, I prefer the app "human japanese"
There are lots of people in the comments somehow offended by the author's genuine excitement over the method that worked well enough for them that they wanted to share it.
As someone who's never tried learning Japanese, I thoroughly enjoyed reading the deep dive and am now less afraid to check out some more serious tutorials (though I wish everyone put as much effort into explaining the system behind something so often dismissed as "just memorise it").
As someone who is fluent in Japanese, the thing is that this is not an example of a method that works well, it’s an example of a beginner coming up with an overwrought headcanon (while calling it “simple”, nonetheless!) that ultimately will hinder their understanding, as I will detail elsewhere.
If this was a similar post about programming, it would not get upvoted here because more people would recognise it for what it is: a “Monads are Like Burritos” post from a well-meaning but misguided beginner.
How is this my headcanon when this is literally how it works linguistically? Please provide a counterexample.
E.g:
> not so lucky and got a "ru" ending? check what vowel before the "ru". it it's one of -aru, -oru, -uru, then it's also a godan verb.
Wrong: 煽る (aoru) is ichidan: 煽った、煽りたい、煽らない、煽ろう、煽られる 。。。
煽る is a godan verb, consistent with your example conjugations (if it was ichidan it would be 煽た、煽たい, etc)
I used to get these terms mixed up all the time because some textbooks use "Group 1" and "Group 2" to refer to these verbs, but Group 1 doesn't mean ichidan.
For all that I'm not totally sold on this article's idea of "stems" and "suffixes", I think it does a good job of avoiding this pitfall and correctly explaining the groups.
This is so obviously incorrect that I have to wonder if it was an AI response.
This kind of content is usually shared by people who are learning a language out of genuine interest for the first time. Most of the time it's Japanese because of the popularity manga and anime enjoy, and the challenge of learning a language that's so different from one's mother tongue. There are countless blog posts written by people to whom it "clicks" for the first time, and they excitedly run to their computer to share their novel experience. I don't see anything wrong with this, let them enjoy the moment. But if they stick to their learning routine, they will inevitably learn the truth that is that the only way to learn a language is to interact with it a lot through all possible channels - speaking, listening, reading, writing. Mnemonics may work for some time, but in the end one won't be able to actually use the language if they don't learn it intuitively. And there is also the fact that most of those people live in places where Japanese is not in everyday use, limiting their opportunities to practice it.
> Mnemonics may work for some time, but in the end one won't be able to actually use the language if they don't learn it intuitively.
A crutch can be cast away gradually.
Finnish has two verbs, ottaa "take" and ostaa "buy". As a n00b I confused them all the time. So I decided that the "s" in ostaa was a dollar sign, and I used that quickly in conversation to select the right verb. And as I got quicker at it, it became "intuitive". I threw down my crutch.
Yes, this is a valid point, mnemonics are totally valid when you are new to a concept.
I'm learning Turkish right now for example, and the dative forms for me/you (bana/sana) perfectly fit the accusative conjugation pattern in my native Polish. So my mnemonic is explicitly "the opposite of what you feel it is".
Please provide a counterexample if you think the article has a mistake or builds a system that doesn’t work.
I'm not claiming any of those.
Senior management at a Japanese tech company, 2 kids, most of my daily life in Japanese here. Can confidently talk circles around most N1s.
I agree. I had plenty of strange abstractions and crutches when starting out learning Japanese, and starting in romaji was absolutely fine. While the post wouldn't benefit me now, starting out any useful mental shortcuts to producing Japanese would have been most welcome.
This is a quite weird article. It opens with several paragraphs about romaji, implying the target audience are not familiar with romaji/kana already. But it's way, way too early for such a reader to worry about verb conjugation.
I wanted to write a weird article! My whole shtick here is that learning conjugation is completely orthogonal to learning most of the language, and you can frontload it into a single evening like this. I don’t respect the whole “it’s too early to do X” dogmas in learning and teaching, and I trust some of my readers will find useful what I find useful.
> There are lots of people in the comments somehow offended by the author's genuine excitement over the method that worked well enough for them that they wanted to share it.
Did it work for them though? They apparently never got past the basics. So IMO it's more likely the opposite; they've distracted themselves from getting on with learning.
I started learning a couple of months ago (with a couple of false starts before). What level am I supposed to be at by this point? This is so incredibly condescending I don’t even know what to say.
Thanks! I’d say that a big part of the motivation for me was to show that this piece is orthogonal to the rest of the language. You can sit down and understand how to conjugate almost any word to almost every suffix in the time it takes to read this. Yes — not fluently — and not “understanding” the language — but it’s wild to me that you can do that at all! It wasn’t obvious to me that this knowledge is orthogonal to everything else.
That’s what motivated me to write about it, really. In language courses all of this is often spread out over weeks or months. I thought it would be fun to write something that you can read in one evening and map out almost the entire system. With no prerequisites.
Because the whole article is built upon a straw man (it's not usually taught the way he claims) and the "method" is just a normal explanation (see this section on wikipedia).
https://en.wikipedia.org/wiki/Japanese_conjugation#Verb_grou...
I’m just describing my experience learning. I understand that some compressed explanations match it, and I agree this one is good! This is not the explanation I was exposed to when trying to learn it the first time.
So you agree my explanation is correct (but unnecessarily longwinded and builds a strawman)? I’ll take that.
> now let's try to apply the rules:
> hanas* + (i)masu = hanasimasu (wrong!)
I had to stare at this for a while to figure out why the author thought it was wrong. "si" is rendered as し on every IME keyboard I've ever used, but the author wants it to be written as "shi".
I don't think this article is really simpler than just learning the table and letting your pattern recognition neural wetware kick in and do its thing. Or better yet, go read some books. After a while, incorrectly conjugated verbs just look/sound wrong.
Many Japanese English speaksers instinctively pronounce si as shi. They say shistem for system, bashic for basic, shix for six, etc
Yes, this is native speakers of Japanese applying a sound law pervasive in their native language (that the consonant /s/ cannot occur before the vowel /i/ and must turn into a sh sound), and applying it to English, which does not have that sound law and that happily allows /s/ to occur before our /i/-like vowels.
An equivalent phenomenon going the other way (at least for American English speakers) is clearly distinguishing また mata "again" and まだ mada "yet" - American English speakers tend to merge /t/ and /d/ in that kind of intervocalic position (think about pronouncing "latter" and "ladder" identically), and it takes deliberate effort to pronounce distinct /t/ and /d/ sounds there in Japanese, where the American English sound law that merges them does not apply.
Kunrei shiki was abandoned a few months ago.
Thanks, this was news to me. TIL.
I wasn't really going for kunreishiki, more ワープロ, aka "whatever is the least typing to make the desired characters appear".
I think typing in IME is how most people actually interact day-to-day with romaji, and I lex "si" and "shi" identically because the difference is usually not important.
Um... 話します is the correct conjugation for 話す, what am I missing here?
The author is using an anglicised romaji system and evidently thinking in English, so they think writing 話します as "hanasimasu" is "wrong".
I feel like you’re going out of your way to misinterpret the article. As the article says below:
> this is why it's important that you don't actually "think in" romaji. i'm using romaji as a convenient way to refer to phonetics in text. however, your "mental algebra" should match the hiragana table.
Then the article includes an exercise that verifies the reader’s understanding.
I also included a note:
> (note i could also have used a different romanization that renders し as "si", つ as "tu", and ち as "ti" for this article. i decided to not because everyone else uses romaji, and once you understand this point once, you shouldn't have a difficulty doing this in your head.)
Where is the factual mistake here? “si” is invalid romaji, my article uses romaji, therefore it’s invalid.
"Factual mistake" is a bit harsh, but the missing piece is that there are multiple ways to romanise Japanese; all of them produce "valid Romaji" but only in the particular system being used. Si is how you write し in romaji using the kunrei-shiki romanisation. In the Hepburn romanisation it's shi. See https://en.wikipedia.org/wiki/Romanization_of_Japanese#Diffe... .
Kunrei has been deprecated, Hepburn is the official one even the government of Japan recommends these days and matches better the pronunciation of the language.
But nobody cares about formal romanization rules in the first place. People have informal bastardizations of Hepburn that's consistent enough for typing, and Kunrei deprecation is more or less just admission of the status quo.
Most people use hepburn these days, it's people who learned japanese 20 years ago or so who are still stuck with kunrei.
I see, maybe my use of “romaji” is sloppy because I implied “Hepburn romaji” specifically because that’s what I chose in the article. I do explicitly mention that other romanizations exist and that I chose not to use them — so I’m not worried about misleading someone who reads the article. But on pedantic level I see why “si doesn’t exist” sounds overly broad.
I think also that anyone who's spoken Japanese for a while already has internalized that "si" === "shi" because there literally isn't the sound "si" in modern Japanese, as the other commenters mentioned it's often romanized both as "si" and "shi" in daily life, if you typed "si" into a keyboard it renders し, it goes on. The original comment on this thread includes one such person who literally didn't follow why "si" is wrong, and I felt the same way too as a long-time Japanese learner. It's a very "copy paste Western language concepts onto Japanese" way of conceptualizing the language, which is IMO a great way to set oneself up for great struggles when trying to learn a language that is structurally different, because it's not the right mental model.
I’d normally be with you and I call this out as a concern purists would have my article. I’m doing something differently here and maybe more daring — the conceit of this article is that you can learn the entire system in one evening with no prior knowledge of Japanese at all. In that case I think the “s_ + i = shi” is as fine a pedagogical moment as any to introduce why the hiragana table matters. And for readers who already know it, I assume they won’t actually have a problem following the content, setting aside the pedantism and the annoyance at someone seemingly “teaching in English” and “copypasting concepts”. Like I get what you’re saying, I just think that it’s okay to break the rules for what I’m going for here.
This is pedantic but you're thinking of "romanization", the act of transliterating Japanese with the Roman alphabet (romaji). There are different systems of romanization, most notably Hepburn and Kunrei. In Hepburn shi is correct, in Kunrei si is correct.
It sounds like you're saying "si" are not valid characters in the Roman alphabet.
Ok yea I see what you’re getting at — my terminology was overly narrow. By “romaji” I specially meant “romanization I’ve chosen for this article”. I do offer an explicit note that other romanizations exist, but outside of that note, I tried to stay within a single self-consistent system, so this nuance was lost.
> I feel like you’re going out of your way to misinterpret the article.
Nope. You correct yourself after, sure, but what I wrote is how it came across at the time when reading.
> Where is the factual mistake here? “si” is invalid romaji, my article uses romaji
No, that's not what "romaji" means. If you mean Hepburn, say Hepburn. And if you don't know the difference, that's a sign to learn more before presuming to teach others.
I’ve already conceded in sibling comments that saying “romaji” to mean “Hepburn romaji” specifically is technically imprecise. Note I still do explicitly mention other romanization systems, want to stay consistent within the article, and don’t want to overload the reader with terminology. My motivation here is that realistically that’s the romaji you’re going to be exposed to the most as a learner. At least that’s been my experience.
As for “correcting” myself, yes, I expect the reader invested in arguing online to also be able to follow more than a single paragraph of text. I think it’s fine that you stumbled there as that paragraph wasn’t for you. I don’t think I’ve hurt your understanding of Japanese this way.
The only people confused by that paragraph are people who already learned kana or correct pronunciation (or both). That paragraph isn’t for them.
hanasimasu = 話します
I also still don't understand why the author thought this was wrong?
Because the author of the article hasn’t internalized that si is pronounced “shi”, is my guess.
The article literally says:
> there is no "si" in the hiragana table, so s_ + (i) = shi. […] this is why it's important that you don't actually "think in" romaji. […] i'm using romaji as a convenient way to refer to phonetics in text. however, your "mental algebra" should match the hiragana table.
That’s just false, “si” is in the hiragana table as し. The romanization “si” is /si/ which is pronounced [ɕi] (or [ɕi̥] or some other possibility). This is basic Japanese phonetics.
If you fix all the errors that are in the article, at best there is an argument buried here that Hepburn romanization should not be used to teach Japanese to English speakers—but I think that point is really my own argument that I’m making with the fragments of the article that make sense.
Romanization can be more consistent with Japanese phonetics or it can be more consistent with English phonetics, and the Hepburn romanization is more consistent with English phonetics, which is why it’s a good choice for English speakers that don’t know Japanese, but a bad choice for English speakers who are trying to learn Japanese.
Okay, we’re fighting over definitions here. There is no “si” in Hepburn romanization. I am intentionally using Hepburn romanization in the article. Therefore, in my article “si” is a compile error.
You may argue with my choice, or maybe you can argue that referring to cells in Hiragana table solely by my chosen romanization is somehow bad, and I should instead be inconsistent and give the same mora two different romanizations within a single article. Is that what you’re suggesting?
First, As a basic part of Japanese language education, students are expected to be familiar with different romanization systems. If you ask a student where “si” is in the table, they should be able to find it. If a student says “it’s not in the table” then they’ve failed the lesson or there is something wrong with the teaching material.
Second: am I arguing that the choice of using Hepburn here is somehow bad? Yes, that’s correct. I think Hepburn is a bad choice here. A good choice is Nihon-shiki. JSL romanization is also fine.
We’re talking about the same thing but you insist that there is only one angle under which things aren’t confusing. I disagree. That’s fine. The two systems are isomorphic, and I genuinely believe that, given I’ve described every single caveat of Hepburn in the article, I’ve paid my dues for using it. YMMV. I even include the “finding in the table” part.
I think I agree that Nihon-shiki and explaining it upfront would’ve made the article more elegant. One constraint I wanted to hit is that a person should be able to read this article with zero knowledge of Japanese, and walk away with being able to conjugate almost every verb to every suffix correctly. This is more of a challenge to myself as a writer than any practical need but hope it shed some light on the choices and the framing. I liked Hepburn because it’s closer to how it sounds. You can imagine I’m using IPA instead if you want.
The problem is that your explanation confuses phonemes with letters.
A spoken language is described by decomposing the spoken words into phonemes, where phonemes are sounds that distinguish words, in the sense that replacing one phoneme in a word with another phoneme will produce a different word.
While ideally each phoneme should be recognized by a distinct pronunciation, in the majority of the languages of the world a phoneme does not have a single pronunciation, but it is pronounced in different ways, depending on the context.
It does not matter at all how one chooses to write a Japanese word, with hiragana or with one of the various methods of romanization. For any writing system, you must know the correspondence between phonemes and how they are written. For very few writing systems there is a one-to-one mapping between phonemes and letters.
The Hepburn romanization does not attempt to be a phonemic writing system, but it attempts to be close to a phonetic writing system from the point of view of an English speaker. The Kunrei-shiki romanization attempts to be closer to a phonemic writing system than to a phonetic writing system. I my opinion a phonemic writing system is superior to a phonetic writing system, but it appears that for most English speakers it has been too difficult to understand the difference between such writing systems, so the Japanese government eventually gave up and they switched to Hepburn, to please the less sharp-witted English-speaking visitors.
Japanese has an "s" phoneme, which happens to be pronounced differently before the vowel "i" than before the other vowels, and before "i" it is pronounced similarly to an English "sh".
In the same way, the Japanese phoneme "t" is pronounced before "i" similarly to an English "ch".
Once you know these two rules, and the few other rules about the other Japanese phonemes whose pronunciation depends on the context, like "n" becoming "m" before "b", there is no point in mentioning them again.
In your discussion about conjugation there is nothing exceptional about the variations in pronunciation that are reflected in the Hepburn Romanization. They are just the general rules of Japanese pronunciation, like for any other words.
So any discussion about these spelling variations is misplaced in the discussion about conjugation, where it occupies a space without contributing anything to the understanding of the conjugation rules.
Otherwise, I think that your article is fine.
I get that. I’ve compressed this as an aside in an article about something else. It’s a choice; like an article about React could spend a bit of time on arrow functions vs function declarations as a choice. Or even let/var. It all depends on audience’s prerequisite knowledge and what you choose to assume.
I strongly suspect that if I were using Kunrei-shiki, there would be just as many comments here saying my article is wrong because “si” is pronounced closer to English “shi”, but my article makes it seem like it doesn’t — so this is why you should learn kana bla bla bla.
I assume my reader (1) has zero prerequisites and (2) wants words to sound correctly while seeing them the first time. Those are the constraints that motivated my approach. You could argue that it’s a strange set of constraints to pick when teaching but I wanted it to be fun.
The systems are obviously not isomorphic—Japanese kana are not entirely phonetic (they are just mostly so) and the different romanization systems choose differently whether to follow orthography or phonetics more closely.
> hanas* + (i)masu = hanasimasu (wrong!)
I cannot wrap my head around how this line in the article could be defensible. Like, if I don’t understand how Japanese is pronounced or written, and I just rely on Hepburn, I guess pasting these fragments of Hepburn together don’t produce the right Hepburn in the end?
YMMV indeed, but I think the lesson here is “this is why you don’t use Hepburn when you’re writing an article about Japanese verb conjugations”.
Hepburn does make sense for somebody with zero knowledge of Japanese but it just gets in the way when you are trying to explain how Japanese works. So lesson zero is “don’t rely on Hepburn” and IMO if you are interested in pronunciation and listening you should be using audio as your primary source.
I’m saying that Hepburn is isomorphic to Nihon-shiki since each is an encoding of kana. Each of them is a bijection to kana (actually that’s wrong; see EDIT below), therefore there’s a bijection between them. Obviously I’m not saying that arbitrary latin characters are isomorphic to kana, that would make zero sense.
I sympathise with your point about the benefits of Nihon-shiki romanization here. It might’ve been a better choice for this article.
> I cannot wrap my head around how this line in the article could be defensible
I think the reader would just read the next section where I use your argument to critique my own approach? And then make up their own mind whether it’s defensible to do something in the article, to raise pros/cons for why I did it, and then to keep on with the choice.
I wanted to illustrate this confusing point, and that’s how I chose to illustrate it. I think it’s confusing either way. I trust that a reader who actually wants to learn, and isn’t just being a pedant, would carry away the right set of conclusions, and would understand the isomorphism (again — see EDIT below) after those two sections.
> Like, if I don’t understand how Japanese is pronounced or written, and I just rely on Hepburn, I guess pasting these fragments of Hepburn together don’t produce the right Hepburn in the end?
Yeah. So that’s a learning opportunity that kana row shifting doesn’t quite follow rules you might expect from many other languages. Maybe that’s a clunky way to introduce it. I personally like this framing. As I noted somewhere else, you could imagine that I’ve chosen IPA notation instead.
—
EDIT: Actually wait, Hepburn is not bijective for zu and ji. I haven’t thought about that. It’s not relevant to any of the conjugations so it doesn’t break the article, but that may be a good argument that it’s not worth the effort rescuing Hepburn.
> I think the reader would just read the next section where I use your argument to critique my own approach? And then make up their own mind whether it’s defensible to do something in the article, to raise pros/cons for why I did it, and then to keep on with the choice.
I think that’s a long wait; I don’t want to rely too heavily on analogies but it is like teaching somebody arithmetic roman numerals and then explaining in a parenthetical that there are other ways to do arithmetic (but not naming them). Maybe the reader can make up their own mind—but I don’t think the pros and cons are raised in the article, or if the are raised, I couldn’t find it.
I don’t want to pile on here but it sounds like you are, in this conversation, learning about why the different romanizations exist and what the pros and cons are. Or if you already knew, you are getting what they call an object lesson. (Like you noted—in Hepburn, ji and zu correspond to two different kana each.)
> As I noted somewhere else, you could imagine that I’ve chosen IPA notation instead.
This just resurfaces a similar problem with different symbols—if you put your IPA notation in slashes // you get phonemes, which will get you something mostly equivalent to Kunrei-shiki romanization. If you put your IPA in brackets [] then you get something sort of equivalent to Hepburn (in that it’s designed to show pronunciation). Both choices will on some level obscure a regular pattern that could be revealed with kana or romaji. Orthography is funny like that; in both Japanese and English it can show the origin of words even when the pronunciation changes.
I think the other lesson here is that students will mostly learn morphophonology intuitively by absorbing examples with some light explanations of the rules, and if you overexplain the rules you end up with too much “scaffolding” which gets in the way. Like when people use mnemonics or try to memorize kanji by thinking pictorially.
I genuinely haven’t thought about zu/ji here (conceded!) It’s not relevant to conjugation though.
In general, I find your attitude a bit condescending. This is what I wrote about my choice:
> note i could also have used a different romanization that renders し as "si", つ as "tu", and ち as "ti" for this article. i decided to not because everyone else uses romaji, and once you understand this point once, you shouldn't have a difficulty doing this in your head
My main mistake seems to be meaning “[Hepburn] romaji” by writing “romaji”. I was obviously aware of other systems because that is what the sentence says but I thought it’s acceptable to refer to Hepburn as just “romaji” as a sort of the default one. Maybe that’s wrong.
Other than this terminology nit, I think I’ve made myself quite clear there. I genuinely don’t think it’s a big deal. Maybe I overestimate my readers’ intelligence but I don’t find this difficult to live with at all once you get it.
Roman numerals is a funny parallel but it doesn’t hold very well. The difficulty of using Hepburn is O(1) shortcut: for conjugation, you only have to “remember” three special cases and they’re always applied just-in-time. It’s just substitutions — and are arguably inherent phonetically. Arithmetic with Roman numerals requires many stacked adjustments where you have to match pairs of things. And lack of orders really screws with ability to do multiplication. This just isn’t an intellectually honest comparison.
Re: your last point I actually kind of agree. I’m that annoying student who likes to un-extrapolate backwards from examples to the rules, knowing which gives me a warm fuzzy feeling, after which I can go back to examples. My article is for people like me. Maybe there’s a few more of them.
> In general, I find your attitude a bit condescending.
Yeah—I can understand why I’d come across as condescending. There’s a balance here—I want to be clear when I say that I have problems with the article, but I don’t want to be hurtful and I don’t want to make criticisms that are not supported by the text.
Rather than defend my comments as “correct” let’s say that I failed in my goals of not coming across as condescending. The reason I want to frame it this way is that similarly, I think the article failed in its goals as coming across (to me) as “look at this neat thing about Japanese”.
It is just kind of the nature of written communication that it takes a lot of editing and polish to make it clear, correct, and concise. I had the good fortune to sign up for Japanese 101 when my professor was in the middle of writing a new Japanese textbook—it was pretty exciting, with the changing lesson plans, the flock of master’s students hanging around, revisions and drafts to teaching materials, and those endless hours of classroom observation. The teachers occasionally gave us a “peek behind the curtain” and explained why they chose to teach things a certain way or another. I’ve rarely gotten that kind of explanation in any class that I’ve taken so I thought it was pretty special.
I don’t expect you to put in the textbook-level of polish into your article but there is a kind of verbosity (the article is long, which makes it kind of hard to respond to because there is just so much to sift through), there are some problems with clarity (the issue of romanization and orthography is mixed in with the conjugation, and maybe it would be better to separate those issues) some problems with correctness (various) and some problems with completeness (the patterns omit some conjugations that I think you don’t know, and I don’t think they follow the pattern).
I have certainly put effort into articles that have gotten brutal negative feedback; I think it was right for me to write the article, and then feel like shit from the feedback, and then maybe retract and revise it. If there is one actual error here, a true error, I think the error is fighting out criticism in the HN comments.
Actually let me just try to explain my pedagogical approach and philosophy here. Maybe that makes it clearer.
I assume no prerequisites at first. So my reader has never seen a kana table and doesn’t know which syllables exist.
I choose to teach conjugation first. That’s an unorthodox choice but I like it! That’s what I set out to do. So we get far enough until it breaks down. And it breaks down when a rule (which worked so far) doesn’t help with “s” because saying “si” would sound wrong.
That’s the moment I use to teach kana table and its importance. This “you made a mistake” is a pedagogical vehicle for introducing kana rows. And we go over the exact ones that you’d make a mistake with. So each special case is walked through.
At this point we could discard Hepburn but I choose to keep going because if you know special cases, there’s no issue. And at some point you’ll learn kana anyway.
So that’s how I chose to layer it. Maybe it’s a bit unholy but I like it. It is definitely self-consistent.
I understand why you wrote the article this way, I think the lesson here is “we have learned why Japanese textbooks do not teach the content in this order” and there are a couple reasons why this order is not good:
1. It relies on people not understanding certain things. In general, you cannot expect people to have exactly the right misunderstanding necessary for a lesson.
2. Spending extra time with Hepburn reinforces it, and it shouldn’t be reinforced.
I am in general extremely skeptical of lessons which try to engineer a way for the students to make mistakes. What I have seen in real classrooms and in informal teaching is that the mistakes are habit-forming and the outcomes of this kind of engineering are unpredictable.
Mistakes are appealing to the developers on HN because we understand things more by seeing them fail. But this does not mean that you can engineer somebody to experience the same moment of enlightenment that you did, because it requires constructing the same (incorrect) mental model that you had when you made that mistake that led to useful insight, and it both difficult and counterproductive to try and make that happen to students. Give people the best chances to learn by giving them the best chances to avoid mistakes, and the mistakes and insight will happen organically on their own, in unique ways for each student.
Okay I see where you’re coming from, sure. We assume different things about the reader and the context of the article. This is an article for people like me who like this approach — maybe engineers or people with an engineering mindset.
I also genuinely think it’s not that deep and that there’s no complex mistake being engineered here. I don’t believe you that the “mistake” of “sa with a replaced by i must be si” is an an unusual one for someone who hasn’t yet internalized kana. If we test this on random people on the street, I’m highly confident an overwhelming majority will make this exact mistake.
I agree with your broader point that “teaching via mistakes” is a risky path not worth it when the mistakes start getting combinatorial. I also think it’s absolutely fine when everyone does the same exact mistake, and there’s exactly one way to avoid it.
> My main mistake seems to be meaning “[Hepburn] romaji” by writing “romaji”. I was obviously aware of other systems because that is what the sentence says but I thought it’s acceptable to refer to Hepburn as just “romaji” as a sort of the default one.
"Romaji" does not (in English) mean "romanisation", as most people who've studied Japanese to at least beginner level know.
Hepburn is the official romanization chosen by the Japanese government (it's a relatively recent change), kunrei-shiki has been deprecated and all the signs etc are in the process of being converted to Hepburn.
I've looked at over a dozen hiragana tables and they all use Hepburn romanization.
Obsessing over romanization, something that a student ought to outgrow, is a sure fire way for a student to get overwhelmed by irrelevant details that discourage learning. The hard part is putting in the work, not learning less than a dozen exceptions.
There's different romanization systems and English learners usually use Hepburn
I think it’s probably a mistake to use Hepburn if you’re learning Japanese, it kinda gets in the way. Either learn kana (which takes what, a week?) or use one of the other romanization systems which maps more cleanly to Japanese orthography
It’s a deliberate choice in the article. I cover every single caveat with it explicitly. I also mention this:
> (note i could also have used a different romanization that renders し as "si", つ as "tu", and ち as "ti" for this article. i decided to not because everyone else uses romaji, and once you understand this point once, you shouldn't have a difficulty doing this in your head.)
That note isn’t much to go on.
I think the choice is not a good one, whether it is deliberate or by accident, it is not a good choice either way. The main caveat to Hepburn is that it’s unsuitable for explaining how Japanese works and it’s unsuitable for learning Japanese—so before you start working on verb conjugations, you pick up kana or one of the romanizations which is more aligned with Japanese.
The idea that you “shouldn’t think in romaji” is really “you shouldn’t think in Hepburn”. This is an important distinction! Japanese has a relatively small inventory of phonemes, somewhere around 20 or 22 of them, and they map very neatly to the latin alphabet.
But the article doesn’t make this distinction, and seems to rely on confusion induced by the Hepburn romanization in order to make its points.
IMO, this is kind of like seeing an article about how monads are burritos. Thinking that a monad is a burrito does not help me understand monads.
Nomu -> noma-nai / nomi-masu / nome-ru / nomo-u
Miru -> mi-nai / mi-masu / mire-ru / miyo-u
The ichidan and godan verbs are not assigned different categories because existing scholars of Japanese are just bad at explaining how they work, and you can still understand them just fine in romaji. I put the hyphens above to mark a place where you could think that the verb ends and the common conjugation forms end, and you can see that the part on the left has somewhat different rules for ichidan and godan verbs, even when you apply the “tricks”—but some of these forms may be unfamiliar if you are are starting out (are you familiar with miru -> miyou conjugation, or miru -> mirareru?)
I concede that using Nihon-shiki maybe would’ve been more elegant for what I tried to do in the article.
> But the article doesn’t make this distinction, and seems to rely on confusion induced by the Hepburn romanization in order to make its points.
Not at all. I give it two sections and then we move on. It doesn’t affect literally anything else on the page. You just learn to shift rows and move on. To make what points?
> you can see that the part on the left has somewhat different rules for ichidan and godan verbs, even when you apply the “tricks”—but some of these forms may be unfamiliar if you are are starting out
I’m not quite sure what you mean to say in this part. I do cover -[r]eba and -[y]ou in the final section (“one more thing”) which extends the model to clearly handle that disappearing consonant. I think -[r]eru fits in there the same way, just as -[r]u itself.
I think explaining it as mi + [y]ou = miyou, but nom_ + [y]ou = nomou is a clearer way to think about this. The rule is that the hole burns down the leading consonant (but takes the vowel).
I've never understood how people can claim that learning kana takes a week. It clearly takes more time than that, considering how similar some of the symbols are and a lot of them only differ by double dashes or a stroke (think nu vs me, ne vs re, ro vs ru, chi vs sa, and so on). Then there are the combinations and even if you managed to learn hiragana, you still have to learn katakana.
Oh and I forgot, you have to actually learn how to listen, pronounce and speak them, not just learn a useless romanization mapping. I've heard way too many English speakers just say the romanization with English pronunciation. At that point their learning efforts turn into self sabotage.
In total that's definitively a month of effort, albeit spread out over the first year of learning.
I think "a week" is slightly optimistic, but I also think "a month" is slightly pessimistic. When I learned hiragana, I spent my free time drilling on RealKana [0]. I'd focus on a new column of the kana table, then bring in columns I'd already practiced, until eventually I could flash reliably on cards drawn from the entire table. This legitimately didn't take much longer than a week, because learning a single column in isolation is very quick, and the real difficulty comes in distinguishing similar kana (as you say). But I was able to drill similar kana by selecting two or three kana that would force me to see them often. (I still struggle slightly with wa, re, and ne, but I definitely know them.)
I also drilled on a drag-n-drop kana table [1] in a few ways -- sometimes I'd start from the kana and try to figure out where they should go in the table, and sometimes I'd go along rows or columns in the table and try to find the kana that belong there. These two directions drill both recognition and recollection.
Proper pronunciation is a cross-cutting concern. As a whole, it's not something you can reasonably learn solely from kana, but the aspects that are relevant are not difficult to pick up. Every kana breaks into one (vowels and N) or two (the rest) phonemes, and for the most part, the way you pronounce those phonemes is consistent across rows and columns of the table (admitting exceptions like "shi" and "tsu"). If you are taught those basics, learning how to pronounce kana is not hard. Training your ear to "hear" distinctions among English allophones, and to distinguish pitch accent from the more familiar stress accent, is much harder, and really has to come from experience, not just kana.
[0]: https://realkana.com/hiragana, wow it's improved since I last used it
[1]: https://ohelo.github.io/usagi-chan/hiragana/
The Japanese phonological system doesn't allow a /s/ sound to occur before the vowel /i/, the consonant must undergo palatalization and become /ɕ/ (the IPA symbol for the Japanese sh-like sound). Because this is a regular sound rule, the native writing system doesn't have a way to distinguish the nonexistent */si/ sequence from the /ɕi/ sequence that actually occurs, and this is the syllable that hiragana し or katakana シ indicate.
In the Hepburn romanization system, which generally tries to be transparent to speakers of English or other European languages, し is romanized as _shi_, because this indicates to English speakers that the /s/ -> /ɕ/ sound change happens. In the Kunrei-siki romanization system, which tries to be more faithful to the distinctions made in the Japanese phonological system, し is romanized as _si_ to be consistent with the other possible syllables _sa_ _su_ _se_ _so_ that begin with the consonant /s/.
And yeah the fact that the article-writer hasn't internalized this sound change yet is a sign that their command of Japanese isn't all that good yet.
> And yeah the fact that the article-writer hasn't internalized this sound change yet
I don’t understand where this misunderstanding about my article comes from. I am saying that the sound at the intersection of “s” column and “i” row is “shi”. My article uses romaji so this is self-consistent. I am also mentioning that there is an alternative system that would romanize it as “si” but that’s not the one I’m using in my article.
> I am saying that the sound at the intersection of “s” column and “i” row is “shi”.
That is exactly the problem. Japanese doesn't distinguish between 'shi' and 'si', so all you're really gaining by pointing that out is learning how to correctly romanize Japanese, in a single system. Instead of learning the language you're learning how to represent the language in a foreign way.
The rule 0 of learning a language is to get rid of the crutches as soon as possible. Use their native writing system (or if they're one of the latin alphabet users, use their pronunciation rule), learn words of the target language using said language, and learn how to formulate concepts with the language rather than translating it from what you already know. Crutches should only be used to get to this point and no more. If you do that, details like 'si' and 'shi' are not even worth mentioning. Romanization methods have their own goals, and rarely is it about facilitating language learning.
I am literally saying in the article:
> this is actually completely intentional. some purists may dislike that, but [..] i think it provides a much clearer intuition for Japanese verb conjugation than writing it in hiragana.
You are that purist that I’m talking about. I get your argument. I just think it’s a non-issue in my chosen pedagogical approach to convey what I wanted to move on. People unaware of how romanization works will nod at it and move on. People aware of it will not have issue with it. People with opinions about proper way to teach the language are the ones that will get stuck on this.
The article is engineer-brained. It’s for people like me. It’s not for the people you’re teaching, assuming these strong opinions come from you being a teacher.
As someone that recently went through an introductory Japanese course in Japan, I don't find this much different than how it's taught. Or maybe I'm missing something?
It seems like the article is trying to make the case that in romaji, you can split the letters and isolate the vowel (e.g. the asterix in the article's conjugation).
But we were simply taught to change from the う- row to the い- row (u- row to i- row). I switched to Japanese to illustrate that you can make that statement even without romaji. In that case, it seems like basically the same thing?
As an anecdotal point, my class was mostly non-english speakers and I didn't find the above to be a sticking point for my classmates. The real sticking points were messing up the ichidan verb exceptions (ichidan verbs that look like godan) and conjugating the correct form for the different grammar points. Te and ta form were also a bit tricky. But the article doesn't seem to offer anything new to help there.
If you’re taught to shift rows, and you already comfortably think in kana — yes, I think that’s equally good. Some of the materials I’ve tried learning from when I got started with the language didn’t do that, and instead described each case as special. That’s what I didn’t like.
That said, part of the challenge to myself with this article was to allow someone to learn Japanese conjugation even if they have zero knowledge of the language (even no kana). So that’s another constraint influencing my choice. I also wanted to have the visual “gluing” throughout the article as an aid for intuition, so that’s another reason I used romaji.
I've found that any resource relying on any romaji after the first chapter or two is often a complete waste of time.
It slows down beginners needing to make the hard jump, since romaji is never used except for signs in real life, and it just becomes a distraction to the material for anyone who is not a complete beginner. Furigana is helpful to the intermediate learner, romaji just becomes harder to read at that point.
Agree. Especially how easy it is honestly to learn hiragana. You can practically learn it in a day and keep a table next to you to look up every time you forget one.
If you're at the point you're learning verbs you'd be mad not to know how to read some kana.
I explicitly address my reason for choosing it over kana in the article. If you disagree, please engage with the argument that’s already there.
The benefit you give (be able to "cut" a kana in the middle) is really weak, I've never seen anyone being confused by that when learning in kana.
This is completely nullified by all the drawbacks of using romaji while learning and they're well known already.
The only reason to use romaji for Japanese grammar is to explain the concepts to someone who has no interest in learning the language, just for their general knowledge.
I’ve explicitly addressed this in the article. Cmd+F for a section called “why romaji is actually good” and then “why romaji is actually bad”. You may disagree with the approach, but I outlined my reasons for choosing it (as well as its downsides).
I'm saying I disagree.
It is always a crutch for the first year student or the barely passed second year student that never helps with real Japanese.
Writing out romaji in Japan is likely to confuse more than help someone else understand.
You glossed over my point. I’m not using it as a “crutch” for reading. I’m using it to have notation for the stem — the thing before -u. I could choose alternative notation with kana (e.g. just always using the -u ending, or the idea of variable stems like i-stem and a-stem) but then the visual “gluing” wouldn’t work. Which is the whole point of mental model I’m communicating. It’s fine if you don’t find this mental model helpful but it’s the point of the article.
I’ll be honest that I also wanted (as a challenge) to write this article so that a person with zero Japanese knowledge would be able to correctly conjugate almost every word to every ending. This is more of a teaching drill for myself though but it’s another reason for the romaji choice.
The explanation made sense to me: romaji works well for vowel shifts (as the vowels aren't glued to consonants) while kana works well for consonant shifts (because the vowels are glued to consonants).
Latin text's smaller tokens/phonemes have advantages and disadvantages, but they are a convenient notation for getting the author's point across.
The difference in phonemes reminds me of how game designer Naramura came up with the (Spanish-sounding) name "La Mulana" for his game by spelling his name backwards in kana. In romaji it would have been "Arumaran" which is completely different (while in kanji it would have been "Muranara".)
> while in kanji it would have been "Muranara"
Not quite. If you change the order of some kanji, the general case is that the resulting text has no definite pronunciation. You definitely would not expect that the sounds assigned to the kanji in one ordering would be the same ones assigned in a new ordering.
This is a phenomenon the Japanese sometimes play with. In the novel Musashi, Musashi comes up with that name by reinterpreting the characters of his actual name (which, in the novel, is Takezō).
This is true, the only practical place where romaji might be used daily is for IME input.
This is why I think all the commenters obsessing over romaji just add fuel to the fire by being elitist over something that should only matter to an absolute beginner.
It's especially vain if the primary reason to choose one romanization strategy over another is to save keystrokes on a keyboard.
I wonder what the author’s native language is as they seem to struggle with the concept of alternation.
For me, the first half of the article could be removed and the learning could simply start with “there are two types of verbs in Japanese (+ some irregular verbs), one type conjugates without alternation in the root, the other with”. That’s enough to get the mental model but my native languages have alternation to begin with so it’s an intuitive concept.
The comments to this article are another example of something I see so often in Japanese language learning dicussions I see online. It's always filled with debate, disagreement, arguements over incredible subtle things, and everyone trying to optimize the best method. It can be really discouraging space for early learners.
Theorycrafting efficient ways of learning Japanese while being barely conversational yourself is a completely different hobby from actually learning Japanese (or actually studying the teaching of Japanese), and sadly in online spaces the former often swamps out discussion of the latter.
The online spaces can be really discouraging, true—but it can also be really discouraging to be in a classroom or in a foreign country, struggling to use a language you barely know. Meanwhile, there are also a lot of ways to spend effort trying to learn a language without gaining mastery. Truly frustrating.
If you want some decent level of fluency then there is something you have to do, which is to communicate with other people in context and with specific goals in mind (get information, give information, make a request, etc). Whatever you can do to arrange for that to happen is probably more valuable than anything you can do online or with books. I personally like to recommend finding classes at a local college.
If you can’t get that, then I think the next best thing is reading and listening.
Drills are also necessary but you can easily fill your time with drills without advancing your ability to communicate or understand people.
There is plenty of research about what is / is not effective when it comes to learning languages so I encourage people to at least take a look at the results of that research rather than just go with whatever people recommend online (I’m just some random person online, I may be no better than the next). AI tools reportedly have a positive effect but they are not nearly as good as human interaction.
It's not discouraging at all. Japanese concepts do not have a 1-1 mapping with English concepts, so there is a lot of debate about how it can be taught. I find it fascinating.
> It can be really discouraging space for early learners.
And it's probably for their (our) good since interacting with other non-native speakers online is counterproductive for language learners.
Author here — it’s a really fun, elegant, and beautiful language. I highly recommend it if you’re interested. Maybe controversial but I think you can largely ignore the Japanese-learning community if/when the vibe isn’t right.
Mostly it helps to find learning resources that gel with your style of learning, and if possible, a tutor so you have a roadmap and more motivation. I found mine on italki. I also find Claude very helpful for sentence drills based on words and grammar I know.
Japanese language learning environments really are ridiculously hostile and I've stopped engaging with them unless im searching for a specific resource.
I've dabbled a bit in Mandarin and while eventually i ended up liking Japanese more as a language, the Mandarin language learner community felt like such a warm bath in comparison. The people were friendly and welcoming and willing to help and genuinely excited to find more people wanting to learn the language.
Up until just last year "si" "ti' and "tu" were the proper official way to romanize "shi" "chi" and "tsu": https://en.wikipedia.org/wiki/Kunrei-shiki
"Official" in only the strictest sense. Everyone has used Hepburn since forever, the government just got around to acknowledging that.
And thank the lord that the romanization of Japanese (with its few little quirks) is one of the simplest transliterations there is.
And, since the English equivalent of those sounds doesn't exist, there's no confusion the way there would be between "she" and "see" in english. Complaining that there's no english equivalent of the russian (взгляд / vzglyad)'s initial cluster would be similar in feel - no english words use it, so the romanization can be whatever you like, really.
> And, since the English equivalent of those sounds doesn't exist, there's no confusion the way there would be between "she" and "see" in english.
Erm, wtf? The English "si" sound does exist and sounds different from し. There is a reason people don't want to write Sinzyuku, and while I think they're making the wrong tradeoff, it is a tradeoff and should be acknowledged as such.
"As a result, the sequences [ti si di (d)zi] do not occur in native or Sino-Japanese vocabulary."[1] Unless I'm missing something, Japanese phonology doesn't include the sound english has as "si", only "shi"? I'm not a native speaker though, it's entirely possible that I am missing something.
[1] https://en.wikipedia.org/wiki/Japanese_phonology
Ah, you mean those sounds don't exist in (standard) Japanese? I got confused by you talking about "the English equivalent".
> why romaji is actually good
It isn't. It falls slightly apart in the `s` column, and completely in the `t` column which contains both "chi" and "tsu". It also breaks for godan words that end in "u" which become "wa" in the negative form.
Mu, bu and nu also all obey the -nda transformation due to phonetics, and not due to how "if we just shuffle the letters around and presto! Nomu becomes nonda".
Japanese already has plenty of its own reading inconsistencies, so adding another layer on top isn't going to help you.
Finally, there's going to be so much kana in your every day life that learning conjugation in romaji is guaranteed to cripple your reading, because instead of recognizing kana (e.g. you see a billboard that says お茶を飲んだ方がいい！ as you frantically try to back-translate everything into romaji, but also removing excess w's and converting nda's as you go) you've spent the first n hours on trying to "hack" the language instead of just learning it.
> It isn't. It falls slightly apart in the `s` column, and completely in the `t` column which contains both "chi" and "tsu".
All of this is described just below, in the section called “why romaji is actually bad”, with these specific examples being tested.
> It also breaks for godan words that end in "u" which become "wa" in the negative form.
This is also described further below in the post.
> as you frantically try to back-translate everything into romaji, but also removing excess w's and converting nda's as you go
I see “understanding the system” and “applying it fluently” as two separate activities. I find romaji more illustrative for the former because the system is phonetical on a sub-mora level. Whereas kana is more helpful for the latter. I don’t assume my reader is an idiot, and so I assume that they would be able to pick the right tool for the job on their learning journey. The fluency always layers on separately anyway, understanding doesn’t “help” there. It’s just that personally I find understanding a great motivator (and fallback) for developing fluency.
-u godan verbs historically ended in ふ /fu/ (and earlier /pu/), and were written that way until the 20th century kana reform. So the historical conjugation pattern of a verb like tamau was tamafa-/tamafi-/tamafu-/tamafe- + additional inflectional endings. The /f/ became a /w/ before /a/ but weakened and was lost entirely in other positions, leading to the modern pattern of tamau, tamaimasu, etc. but tamawanai.
As someone who is learning Japanese and is struggling through learning conjugations.... this is not explained simply to me.
Maybe it's because I'm at a different part of my learning journey, but this seems to go on and on and on about godan v ichidan verbs but doesn't touch on the actual conjugation (ie. how to turn "drink" into "didn't want to drink"), nor on plain form (everything is in polite form), and te-form gets a very very short mention at the end x_x
Also, the romaji did my head in but I understand that was done for a reason (I don't agree with the reason, but at least there was one)
does this help? https://www.youtube.com/watch?v=cGA6Tj9_lSg
A plug for my favourite book here. Richard Webb's 80/20 Japanese covers these topics extremely well, and in a very logical / productive order for a foreign speaker.
(and it comes with very enjoyable audio tapes and anki cards to boot)
Highly recommended.
I started to learn Japanese 30 years ago, and in my experience the people who try to be smart and build systems almost never get decent. It’s procrastination while thinking they’re actually productive.
To add insult to injury this article hasn’t discovered anything new, makes it sound way more complicated than it is, and in the end still requires you to just remember which verbs are of the eru/iru group, and which are not (which was posed as a problem to solve in the intro).
Just make cards and mark the stem, learn it along with the verb. No need for heuristics. If you ever forget, you’re bound to remember the masu-form and can reverse engineer the stem from that 100%.
Agree. Experience shows that fluency arises when you don't have to think about rules anymore. My advice is to not spend too much time learning grammar rules (actually, no time, like native learners). Leave the rule discovery to your unconscious brain and get going with rote repetition.
Your brains "language module" is not a slow computer, computing rules, it's a fast lookup-table.
Similarly, when complaining about how you have to memorize a big table of verb conjugations in the intro, the author links to a table of... -ta forms, a verb form for which the author later concludes you just have to memorize a big table.
The te/ta-form is genuinely a separate system that doesn’t reduce further. I think there’s still value in having a solid model for everything else. At least I personally found it valuable, which is why I thought to share it with people.
I'm not totally sure this "stems and suffixes" mental model really works well for everything else. Forms like the imperative (食べろ), volitional (食べよう), provisional (食べれば), potential (食べられる), and causative (食べさせる) aren't cleanly handled either -- they work similarly for godan verbs, but you have to add a different suffix for ichidan verbs.
It's definitely useful to understand how "chi" and "tsu" fit into the hiragana chart, and if your asterisk notation helps you remember which verbs are ichidan vs godan then that's great, but I'm not sure it's worth trying to unify -masu and -nai into one model.
> Forms like the imperative (食べろ), volitional (食べよう), provisional (食べれば), potential (食べられる), and causative (食べさせる) aren't cleanly handled either -- they work similarly for godan verbs, but you have to add a different suffix for ichidan verbs.
They are cleanly handled in the final section (“one more thing”) that introduces a notion of disappearing consonant like -[r]u, -[r]eba and so on, and gives a rule for it. This is a perfect inversion of what happens with -(i)masu and friends. The hole in the stem accepts the leading vowel but burns down the leading consonant.
It’s quite elegant.
"Disappearing consonant" doesn't work for the potential form, unless you expand the representation to allow writing -[rar]eru. (Edit: And I think imperative would require like "-[ro](e)".)
Which, like, is clean in the sense that Redux is technically Turing-complete (you can encompass _any_ difference between two strings by saying that one string uses the stuff in brackets and the other string uses the stuff in parentheses), but that doesn't make it a good idea.
Okay, that one’s fair! I remember there was also some other one that had it split into two completely different suffixes.
My answer to this is that by the time you’re learning those, you’re already so fluent in conjugation that a couple of special cases will layer on fine. It’s way better then you get in most languages. (And pedantically I’d still say [] works for the cases above, as you’ve shown.)
I honestly don’t understand the cynicism here. If I could read this article when I started learning, it would’ve saved me a ton of time. That’s why I wrote it. I hope it’ll be useful to someone else but it’s fine if not. As an educator I’m proud of how much it crams in that’s usually spread over many weeks, and how the simple model almost perfectly generalizes. But yeah sure it’s making some unorthodox choices and leaves a couple of advanced cases within one indirection. I’m still very happy with it.
Why would you expect the article where I’m describing what worked for me to “discover something new”? I’m literally sharing the mental model that I personally found helpful. There’s nothing “new” in learning or teaching a language. But this is the most minimal model I’ve found useful, compared to others, and I wanted to share it with other people.
I think you’re taking a lot of stuff for granted. “Just” do cards etc. You’re using the word “stem” but what’s a stem? Why do we sometimes inject -i or -a (or -wa) there and sometimes we don’t? You still have to learn that and understand that. That’s what I’m describing in the article. If you already know stems and how they compose with suffixes, congratulations, you won’t find my article useful.
Honestly, you don't need to "learn" or "understand" much grammar explicitly. I think it definitely helps get you off your feet, as you can "decode" sentences if you remember grammar rules, but eventually the grammar has to be internalized anyway. This happens when you are repeatedly exposed to the same patterns in context. I don't know how English or Norwegian grammar works, and I'm fluent in those. I skipped grammar in Japanese and focused on reading, yet I can understand most things and I can tell when something sounds wrong.
The article could just be two words: ichidan, godan. Done.
One big change I had when learning Japanese was that someone introduced to me Cure Dolly videos on YouTube, and it has been an eye-opener: All these verb conjugations are actually attaching another verb to extend its meaning
Here's how I was taught verb conjugation.
First, we learnt verbs in the -masu form. Nomimasu, tabemasu and so on.
Then we learnt this song (to the tune of Clementine)
chi ri i tte mi ni bi nde kiite giite
It's a quick mneumonic to help you go from the polite verb to the "te-form" ending. I hummed it in my head while working out the conjugation before it became natural and "obvious".
For me the biggest problem was never memorizing the suffixes but determining whether the verb belongs to group 1 or 2. While you can tell immediately that some verbs belong to group 1, the others (e.g. kaeru mentioned in the post) are not so easy (as far as I know there's no algorithm for that and you just need to memorize with every verb to which group it belongs).
Funny how some conjugated forms of verbs collide with dictionary forms of other verbs (esp. if we ignore pitch accent differences):
E.g. potential of 買う (kau, to buy) is 買える (kaeru), which is spelled like 帰る (kaeru, to return home).
It reminds of you how "lay" is a verb (to put something into a flat resting position, but is also the past tense of "to lie" (take on a supine position).
Today, I lay bricks; yesterday I lay in bed all day.
Plus lie and lie are examples of how English verbs can be homonyms in dictionary form, but conjugate differently, something we see in Japanese (either actual homonyms or near homonyms modulo pitch accent).
:(
Romaji are great, and in some ways more instructive because they reveal patterns which are otherwise a little hidden. You just have to realize that S+I is shi, T+U is tsu, etc. I don’t want to get too deep into it but there is a regularity to the language, and rules, and different choices of writing system reveals different pieces of the puzzle.
Next, the conjugation itself. There are massive categories of conjugations missing! Like, how do you get from taberu / nomu in this system to tabereru / nomereru? It turns out that these ichidan and godan verbs actually do have some differences in conjugation. Who’d have thought? (There is the -i stem, but there are other forms.)
Both of these things are described in the article. The first one is in “why romaji is actually bad” section, the second one is in “one more thing” part at the very end.
Looks good. It essentially separates phonotactics out of conjugation patterns instead of conflating them, which works also very well with how I learn languages. How to conjugate then breaks down to finding the stem (yes, some learning is always involved), and then there are not even different conjugations, but there are simply vocalic and consonantal stems -- all the morphology is in the suffix (the 'secret vowel') and the rest is phonotactics. I am sure there are irregular verbs, but probably this explains 90% or more of the verbs with simple rules. For me, this creates a perfectly logical system. Thanks for the overview!
For anyone who is interested in learning Japanese, or looking for resources. I've compiled this "Awesome Japanese" repo.
https://github.com/yudataguy/Awesome-Japanese
"hanasimasu" is not exactly wrong; there is a romanization system in which "si" is how you write "shi".
If you want to invent scheme for understanding conjugation which works by cracking the romanized versions of words to create a pseudo-stem that could not actually exist in spoken language, it behooves you to adopt "si" and "ti", because they bring in a consistency needed by such a system to be complete.
> If you want to invent scheme for understanding conjugation which works by cracking the romanized versions of words to create a pseudo-stem that could not actually exist in spoken language
That's how all conjugation schemes work. There's nothing weird about this. Stems aren't supposed to exist in the spoken language. But they are observable in the spoken language.
Compare how a modern dictionary will give you ποιέω, a full and fully-inflected word which doesn't actually exist in ancient Greek, as the first principle part of that verb. This is done because the stem of the verb is ποιε-, and the epsilon ending the stem can be easily observed by its effect on most of the conjugational endings. It doesn't happen to affect the first-person singular ending -ω (to be precise, the contraction of ε- with -ω is -ω), so the dictionary form is synthetic, chosen to be informative.
Calling it concatenation is a little misleading.
Japanese is known as an agglutinative language [0], and how verbs are conjugated also has a lot to do with politeness, as well as local dialects. That's why you can turn on an anime and hardly understand it, even after a couple years of study.
I got to the third year college level in my own Japanese studies, and at that point, memorizing kanji was starting to compete with my computer science studies, so I had to drop it. I got to travel to Japan and live with host families (we kind of settled on a Japanese/English pidgin), so I don't regret the experience.
[0] https://en.wikipedia.org/wiki/Agglutinative_language
It’s concatenation with the asterisk, and the asterisk is what we build throughout the article.
Fascinating to hear non-tech insight from Dan, especially as a fellow (rookie) student of 日本語.
Fun, and a programmatic perspective. However, it can be too easy and fun to get super caught up in these details, if your goals are some level of fluency and ability to communicate/read. The majority of people that I know who have gained any level of fluency in Japanese as an adult mostly avoided stuff like this because (for many people; of course everybody is different) doing all of this mental math to dive down to the last detail was nowhere near as effective as some speaking and reading drills.
It is definitely well written and presented.
I like to do deep dives like this not to memorise but to understand deeper layers, the spirit of the language, the way it moves, the way it unfolds.
By all means, it is fun to play with a language. And every person's brain works differently.
I like to use this metaphor, though. You're hiking a mountain, this journey to the "peak" is reaching some goal of fluency.
It's fun to stop and look at rocks, examining, comparing and whatnot. But it doesn't necessarily get you closer to the peak. I mean, it might, because you'll better understand your footing every-so-slightly. Not a perfect metaphor.
Thanks! Feels like the only person on this page who gets what I was going for.
I'm a bit dismayed by all the negativity you're getting. Thanks for all your level-headed and detailed responses here.
I have a hard time understanding why people object to this kind of attempt to systematize something that conventionally relies on rote learning. I've always found that figuring out the underlying rules helps with the rote learning. When you see some new word or conjugation, it's super helpful to have a method for explaining why it's that way; it helps fix it in your mind.
I wonder if native speakers object to this approach because it isn't really how you learn as a kid. You don't learn any abstract rules, you just absorb a huge amount of training material and construct your own model. But that doesn't mean adults need to learn in the same way; we can and should leverage our painfully-acquired knowledge and skills from other domains.
FWIW I actually haven’t seen negative feedback from the native speakers I’ve shared it with. (Maybe they didn’t want to upset me! I don’t know :)
Well, the native speakers or otherwise self-diagnosed experts who felt obliged to chip in here!
Maybe it’s not very clear but I don’t suggest studying from articles like this alone. Obviously you need to do sentence drills and talk to a tutor or native speakers to have any chance of success. I still find documenting my mental model helpful because most articles I’ve seen before were not teaching it clearly enough for my taste.
I usually just consult my handy sheet of BNF rules while speaking Japanese.
If it's taking you this much effort to do the trivial conjugations (seriously, the whole page barely mentions the interesting ones 80% of the way down, and falls back on "yeah, you just have to memorize the patterns" for た/て forms), yeah, just memorise them. Language learning and exercise are the two things where I've found the programmer's instinct to "work smarter, not harder" works against you; you actually just have to put the time and effort in.
Author here. Strong disagree.
I prefer having a system to simply memorizing. I don’t know what you mean by “so much effort”. I am literally just describing the system as it is brick by brick. If you see an opportunity to simplify, you’re welcome to provide a specific suggestion. I find this system rather elegant, and I tried to build it piece by piece because that’s my preferred way both to learn and to teach.
>the whole page barely mentions the interesting ones 80% of the way down
The te/ta-form is genuinely a separate system linguistically with its own heritage. So it makes sense to look at it separately. I don’t consider it more “interesting” and I’d argue getting the details right with other forms is much more useful coverage-wise. So I didn’t spend much time on te/ta-form. (That said, even for -te/ta form, I find it calming to think of -nda as a contraction of -nita, and so on, which AFAIK is in the ballpark of what historically happened.)
> Language learning and exercise are the two things where I've found the programmer's instinct to "work smarter, not harder" works against you
I agree you need to put time to practice and all that. But if there’s a genuinely simple system underneath, I always prefer to see it. Even if there’s a layer of memorization and repetition to achieve actual fluency. Japanese conjugation is a rare case where the system actually is very clear and methodical. The article is written for people like me who also prefer to know it. There’s literally thousands of resources that teach it your preferred way, so I don’t understand the impulse to complain about someone teaching it differently for a change.
> I don’t know what you mean by “so much effort”.
This is a pretty long blog post covering really not very much.
> The te/ta-form is genuinely a separate system linguistically with its own heritage. So it makes sense to look at it separately. I don’t consider it more “interesting” and I’d argue getting the details right with other forms is much more useful coverage-wise.
It's not just te/ta, you don't mention anything other than the basic polite/casual, positive/negative, and desiderative. At the very end you point to conditional and causative but say you haven't studied them, and no mention at all of passive, imperative, causative passive, or volitional.
> I agree you need to put time to practice and all that. But if there’s a genuinely simple system underneath, I always prefer to see it.
And how's that working out for you?
> There’s literally thousands of resources that teach it your preferred way, so I don’t understand the impulse to complain about someone teaching it differently for a change.
I find it very presumptive to propose to "teach" what you haven't really learnt. How many people have successfully become remotely close to fluent following this approach? It's 0, right? What makes you think you're "teaching" rather than leading people astray?
Indeed, especially for a language with forms of verb as regular as in Japanese. The whole language has two and a half irregular verbs. Compare that to Spanish and realize how fortunate you are to study Japanese verbs.
meh, language learning has an inconvenient truth: sometimes it’s just rote memorization. it's the reflexive belief that every human endeavor must have a hidden optimization waiting to be discovered. Language learning is one of those domains that stubbornly replies, "Cool flowchart. Now memorize 500 words and spend 200 hours listening."
There’s no clever engineer hack that replaces time spent with the language. and with regard to japanese, please stay away from romaji, unless you're still in beginner stage and typing things out to communicate words you already know the phonetics to.
I mean I think it’s both! As an author, I wrote this to settle my mental model. This doesn’t mean I could use it during fast speech, but I find it calming to understand the actual linguistic system behind the tables. Especially when it’s so elegant.
The choice of romaji is deliberate for multiple reasons and is defended in the article (with counter-arguments for why it’s bad too).
this is quite intriguing, as a native speaker and someone with friends trying to learn Japanese, I always had a hard time explaining all the different patterns and just defaulted to "it just is". Will use this in the future!
Categorizing Japanese verbs as -ru or -u requires more context.
I prefer the term "group 2 verbs" to "-ru verbs." Group 2 verbs are verbs that end in -eru or -iru, not just -ru. Of course there are some exceptions, like kaeru, which ends in -eru but is actually a -u verb. Conjugation is easy: remove the final -ru and append -masu, -mashita, etc.
"Group 1 verbs" (again, -u verbs) are verbs that are not group 2 verbs. Conjugation is a bit more difficult because the -nu, -bu, -mu, and -u verbs have many suffixes. However, after memorizing these two (-nbmu and -u, because -nu, -bu, and -mu are almost the same), the rest are easy.
There are only two irregular verbs: kuru and suru. Just memorize them.
I learned Japanese by just memorizing. Once you have memorized enough verbs and their conjugations, you can figure out the conjugation of a new verb even if you don't understand how it works.
there are more irregular verbs than just kuru and suru. iku and aru are also irregular, for example.
Irregulars notwithstanding, the conjugation pattern is actually completely lossless if you just remember the imperative form (e.g. 着ろ kiro, 切れ kire) instead of the infinitive, which is lossy (e.g. 着る kiru, 切る kiru). Then there's no need to have to remember, "oh... is this -iru verb group 1 or group 2?"
They’re sometimes called “semi-irregular” because they are mostly regular with, like, one deviation. The list is not long and it is quick to memorize.
Iku is kinda irregular but it's only itte and itta instead of ite/ita. Also aru/nai is more like antonyms rather than conjugation?
Te-form mnemonic (sung to Ba Ba Black Sheep):
i chi ri tte
bi mi ni nde
ki ite
gi ite
shi shite
I learned this 35 years ago to the tune of Clementine, youre using black sheep, my spouse uses another tune, but what funny is I learned it not with the 'i' endings, but with the dictionary form (sometimes called base-3) 'u' forms.
u tsu ru tte bu mu nu nde ku ite, gu ite
without the shi shite as that had been learned well ahead of the lesson adding ta/te forms.
I just think it's interesting how readily a little ditty tune helps people with recall, regardless of the actual tune.
I've generally found this kind of thing annoying in my various attempts to learn languages - in one of my high school spanish classes the teacher thought it was important to have us learn a little song to memorize the six present tense forms of the to-be verb ser: (yo) soy, (tu) eres, (el/ella) es, (nosotros) somos, (vosotros) sois, (ellos/as) son. My thinking was, there's only six forms to memorize, and the verb is an incredibly common one ("to be"!) so they'll get constantly reinforced anyway. The song is silly and isn't really helping anyway.
I feel similarly about the transformations for the Japanese -te forms and -ta past tense marker. The entire system is:
Ichidan: add -te/ta to stem Godan: -u/ru/tsu -tte/tta -su -shite/shita -mu/bu/nu -nde/nda -ku -ite/ita -gu -ide/ida
So basically ten patterns which group into 5 subpatterns. There's some logic behind them - the -te and -ta morphemes originally got added to the -i/pre-masu/ren'youkei stem and then underwent some idiosyncratic consonant reductions in godan verbs. But, really, it's only ten patterns, you can just memorize them; and these are incredibly common verb forms that get used all the time so you'll have them reinforced frequently if you are at all engaging with the language. It's a lot less to memorize than if you were learning Ancient Greek or Sanskrit or something.