learning about japanese 2019

links and tips

the japanese language education center once posted an average of 2000 study hours needed to pass the jlpt n2 language proficiency test. by this estimate, if studying two hours every day it would take about three years

favorite links

study guide

level 1

  • read human japanese basic for grammar, basic vocabulary and general fundamentals
  • use the kana stroke order tables from above to learn the kana and write from memory all kana characters in gojūon order correctly. you are able to do that in your native language with ease and you should be able to do it in japanese, too
  • use human japanese anki deck to test and consolidate the knowledge
  • use kanji-deck with stroke count order to learn to distinguish the first 1000 kanji. stroke count is a natural order where parts come before compounds
  • regularly write down all words you know. you could use a paper notebook and a fudepen for that

level 2

  • read human japanese intermediate
  • use kanji-deck to learn to distinguish the remaining common kanji
  • use kanji tree to improve word reading proficiency
  • use the core 2000 anki deck to consolidate advanced vocabulary and example sentence with kanji reading knowledge
  • satori reader
  • jisho search plugin for mark and right-click dictionary lookups
  • create a bookmark with the address https://jisho.org/search/%s and keyword ji. then you can enter "ji something" into the address bar for dictionary lookups

level 3

  • find new vocabulary in wikipedia section headlines, vocabulary lists, news and other media
  • install a japanese input method (linux - i use ibus for now because it is works simpler for me than fcitx, android)

more links


  • kana are about 142 characters in two sets that can represent the pronounciation of all words in the japanese language. kana are used as suffixes to kanji for conjugation and particles
  • both hiragana and katakana are in common use. some words use the one and some the other character set. loan words are usually written in katakana (and there are many) and in dictionaries, some types of kanji readings are given usually in katakana
  • small character version: small tsu is silent and stresses the next character, other small characters are combined with the previous one to form a new one
  • romaji is japanese written with latin characters. there are mulitple systems: hepburn (most common, corresponds more to pronounciation) and kunrei (only two latin characters per kana). both have ambiguities
  • each character has an associated stroke order that describes in what order the parts of a character is to be written which is important for handwriting to make it look right and recognisable
  • character similarities

    • hiragana: さき, さちら, けはほに, たにこい, たな, ほま, あおぬめ, われね, ろるそ, とっう
    • katakana: ユコ, ソンツシ, モチテ, マム, リル, ワウフヲラ, クケヌタ, メナ
    • hiragana/katakana: へヘ, かカ, やヤ, サせセ, うラ, とヒ
  • stroke direction: ソ first stroke low then top to bottom, ン first stroke high then bottom to top, シ first two strokes in downward order then bottom to top, ツ first two strokes in upward order then top to bottom
  • every syllable is roughly to be spoken with the same duration or beat
  • the duration that characters are spoken with is increased by appending vowels. vowel duration is doubled by repeating the same vowel (あ -> ああ). consonant syllables are lengthened with the vowel they are ending with, except for o and e, which are lengthened with u and i to produce combinations like kou/こう and kei/けい
  • the characters that i found the most difficult to write: ふそあかやれい


the most common character order is called gojūon (五十音) and goes like this: aiueokstnhmyrwn, were the consonants are to be substitude with their syllables in vowel order, for example kakikukeko. to remember it, i pronounce ordered syllables as a fictional name: kasatana hamaya rawan. characters modified with dakuten and handakuten, for example べ and ぺ, directly follow their unmodified variant in this order. the consonant order with modified characters is kgsztdnhbpmyrw

the 142 commonly used kana

あいうえおかがきぎくぐけげこごさざしじすずせぜそぞただちぢつづてでとどなにぬねのはばぱひびぴふぶぷへべぺほぼぽまみむめもやゆよらりるれろわをん アイウエオカガキギクグケゲコゴサザシジスズセゼソゾタダチヂツヅテデトドナニヌネノハバパヒビピフブプヘベペホボポマミムメモヤユヨラリルレロワヲン

without modifiers

あいうえおかきくけこさしすせそたちつてとなにぬねのはひふへほまみむめもやゆよらりるれろわをん アイウエオカキクケコサシスセソタチツテトナニヌネノハヒフヘホマミムメモヤユヨラリルレロワヲン

kanji (漢字)

  • jouyou kanji (常用漢字) is an official list of about 2130 kanji that are commonly used. in some contexts like official documents and newspapers, texts are limited to these characters
  • most japanese words use kanji
  • kanji are not each completely separate drawings but share 200-300 components. knowing the components makes most kanji recognisable as a combination. this should make it easier to recognise, write (you know how to write each radical), learn (grass, field, beast -> cat) and look up kanji in the dictionary. it is not always easy to recognise the components used and there are variations in how exactly they are integrated into a kanji
  • the reading or actual pronounciation of kanji appears in the context of words. kanji can have different readings in different words. dictionaries usually show a list of readings of kanji but i have found that the readings often arent actually used at all in the words im interested in, so i ignored the standard readings and focused on the reading of words, which is probably the best starting strategy anyway
  • kanji have an associated stroke order
  • kanji can be looked up in dictionaries by component, stroke count, hiragana pronounciation or by topic
  • some kanji that i had difficulty with: 又夂攵夕久具首頁真壬禾夫未矢牛午几兀元示
  • knowing the components is useful for recognition but wont help you for meaning




  • ichidan: "one base", have suffix iru or eru, with a few exceptions that have the suffix but arent ichidan
  • godan: "five bases", suffix u
  • irregular: small number (< 10) of words that conjugate differently

verb stem

formal present: remove masu


the "->" stands for "replaced with". replaced are suffixes.

  • present

    • godan: u -> imasu
    • ichidan: ru -> imasu
  • present negative: present: masu -> masen
  • past

    • godan: u -> imashita
    • ichidan: ru -> mashita
    • irregular: kuru kimashita, suru shimashita, and others
  • past negative: present negative: masen -> masen deshita


  • present: dictionary form
  • present negative

    • godan

      • u -> anai
      • {vowel}u -> wanai
    • ichidan: ru -> nai
    • irregular: kuru konai, suru shinai
  • past

    • godan

      • ku -> ita
      • gu -> ida
      • u/tsu/ru -> tta
      • nu/bu/mu -> nda
      • su -> shita
    • ichidan: ru -> ta
    • irregular: kuru kita, suru shita
  • past negative: present negative: nai -> nakatta
  • te-form: informal past: ta -> te, da -> de

want-to form with tai

  • verb-stem + tai


both classes can precede noun to modify the noun


  • i-adjective: i suffix, never ei

    • can conjugate and function like verb, also called verbial adjective
  • na-adjective: na suffix

i-adjective declension


  • negative: i -> kunai
  • past: i -> katta
  • past negative: i -> kunakatta

how to make your own anki deck

import from csv

  • prepare them in a spreadsheet. one column per field
  • export to comma separated values / csv file
  • ensure that in anki a card with the right number/type of fields exists
  • in anki go to file -> import, choose the target deck and map fields
  • first field is the question, other fields are the answer
  • edit the card template to have the fields and layout you want

export apkg deck file

  • go to browse
  • select the deck in the tree left
  • this file can be imported in anki on desktop or android and shared

add/remove available fields

  • browse -> select deck in tree -> fields... button
  • fields can be deleted from all cards if no card or all cards are selected

edit the card template

  • browse -> select deck in tree -> cards... button
  • templates use html and css

the 2136 jouyou kanji as of 2010