What do you think about using PAO for Chinese characters? Normally in PAO, each of the digits 0-99 are represented by unique PAO combinations. What about representing each Chinese character component as its own PAO combination, and then combining them in writing order to represent a Chinese character (or part of a character if a character has more than 3 components)
1 Like
I think it would be most useful for handwriting Chinese characters, however I don’t think it’s a good method for reading. When you read, you want to go from character → meaning as quickly as possible, so going from character → person + action + object → meaning would be too slow.
1 Like
Sorry, I read in your other post that you were looking for Chinese character memory palace help. I know Chinese and Japanese use the characters differently, but I wanted to show you the method that I use to see if it would be helpful.
Create a memory palace with 52 stations. (I picked this number because this commons 75% of the most commonly used kanji in Japanese). I associate 48 stations radicals, reserve one station for katakana, one for hiragana, and one for characters that use uncommon radicals. (Since you are learning Mandarin, please modify this number to your desire.)
When you come across a new character, you will “file” it under one of the radicals in your palace. If it has two or more places it could fit just choose wherever you want.
When making assocations with the character, imagine that the character itself is doing the action.
For example, when I see the word 理論, I do not want to think “Okay, the 王radical is king, so that’s King Charles, then the 田 radical is rice field, so King Charles is rice farming…”
I just imagine the word 理論 fervently explaining its mad theory. That way I can go “理論? That’s in this location…oh! theory”
You can also use other associations like auditory, olfactory, etc as long as you keep the character itself in your mind.
I hope you find this helpful!! Please try it out and see if it works for you.
2 Likes
I disagree that character > meaning is the connection you want, character > pronunciation is better
Is your original suggestion to just not use mnemonics at all for chinese characters?
When you were a beginner, didn’t you find the problem of not recognising that this is a character you’ve actually seen before? How can you imagine the character itself doing the action, before you are familiar enough with characters to have “face memory” of them?
In my opinion, charactern > pronunciation is important but I don’t think it’s necessarily better. It’s possible to read and understand without knowing the correct pronounciation.
For example, in English, there are many words that I’ve gone my entire life mispronouncing because I’ve only seen them in writing and never heard them, yet I’m still able to understand and use them correctly.
I don’t mean to say you shouldn’t use mnemonics at all, I just don’t think PAO is the best one to use for your goals (assuming reading is your primary goal rn)
Yes, I did. That’s actually why I began using memory techniques because I got so frustrated by it 
That is why you start off with memorizing radicals. When you can recognize the radicals that make up the character it is simple to have a “face memory” of them. Instead of seeing a bunch of lines your brain sees the character as a bunch of radicals which is easier to have a “face memory” of.
I think reading should encode pronunciation for sure, at least for beginners of a language, because we as humans are so talented at spoken language, we learn it instinctively and quickly, and our passive vocabularies are better than our active ones. If I hear a word, I am extremely likely to know it’s meaning, but I may not be able to go the other way, from meaning to pronunciation. I’m trying to leverage our natural talent for spoken language by making a reading pneumonic system that only encodes the pronunciation
Do whatever you want. Nobody’s going to arrest you if break a mnemonic phonetic law. With 2 digit or cards you will be used to it
Yeah, you can do what you want. I personally focus on recognizing the words but I also add the pronunciation if I want to, I just don’t make it the main focus of my study. You can for sure adapt any system you want to your learning
Another idea for Chinese is to choose a modifier, action, etc for each tone. For example, 1st tone is explode, 2nd tone is melt, etc. Then you can add these to your mnemonic images to help remember the pronunciation better.
Whatever you choose to do, don’t forget to tell us what you tried and how it went!
I’m always interested about what people use for languages (have a PAO myself but for Japanese), but most ask questions, without afterwards sharing with us what they did.
Which is a pity because I always find new ideas interesting (even if I don’t use them myself, or have to adapt to use them myself)
1 Like
I have a list of O (objects) for over 400 Chinese radicals. The 214 you see in other places are only the “dictionary” radicals. I have a P & O (person, object) system using initials and finals for pronunciation. I also have 5 A (actions ) for the tones. I combine these together with other actions to encode meaning+pronunciation+writing.
For instance: 好, hao3, (good, ok, fine)
I imagine a Dolly Parton (女). Putting tacks (子) in a Mr. Goodbar while a guy in a hazmat suit (h) scoops (ice cream scoop = ao) out flying bats (3rd tone) out ot the Mr. Goodbar.
I use SRS flashcards mostly though. I find that I mostly use this system when first learning a character. Later as I become more familiar with the character I find that the review using SRS software is more important. I use Pleco as it is special for learning Chinese.
The order of the actions is only important for the radicals. As radicals are not placed linearly like in English. So I might make sure the story helps me remember where the radicals go. The order for pronunciation or writing doesn’t matter in the story.
It takes some work to start but I have found it quite useful.
did you find a list of over 400 character components somewhere?
I made the list. I found a chinese website that allowed me to draw radicals and had obscure characters/radicals.
Here is one: 㔾 jie2.
Or: 𠃌 gun3
You won’t find them in the list for looking up in a dictionary but they are common enough radicals.
There are many others.
I can’t remember the website as I haven’t dabbled with it in years.
I’m not sure where or if I should post my list Pleco? Anki? Art of Memory?
Anywhere you post it, please share it with me!
@Hari-P here’s exactly what I do, I’m 400+ characters in, at ~5 Hanzi / day.
Perhaps some parts of the below-described system will resonate and will work well for folks:
For every location, I store full pronunciation, meaning, and the full Hanzi character (@gravity yes, I use component PAO for the Hanzi). I call this 1:1 “compression.” (all that in just one place). Here’s how:
SETUP
-
Decide on some full list of characters (I’m doing Taiwan Ministry of Education 4808 characters) that you want to learn ahead of time, and get them into a spreadsheet or list. I’d say go big, at least 5,00 or 1,000, so that the groups can be bigger (see next step).
-
Sort them by pronunciation, then by tone, in some linear order, for example, alphabetical, and then sub-sorted by tone number. This way you save all the space per locci for the Hanzi components + meaning. Grouping the homophones like this is also useful for reasoning through the meaning of spoken words you hear!
-
Start a “component” list in a spreadsheet or similar that grows beyond 214 “Radicals”. This is to drill down to every subcomponent, every time. No “face recognition”-- full itemization. (hopefully face memory also works! I’m just making a bet here
) Give each component a full PAO set. Ignore component/radical “meaning” and pronunciation - just go by what it looks like.Yes the component dictionary will be large (I’m at ~250 components). But, they all already kinda look like something, and most are Hanzi in their own right. And you can build it as you go instead of upfront, taking advantage of the frequency effect that says anything you encounter on the way is likely to be a very common component. Also a site like Hanzicraft can tell you exactly how common components are, to help you decide what warrants its own PAO, and what to just build out of smaller components.
-
Set up a pronunciation Person-Action set where the person is the initial, and the action the final. This is relatively straightforward and useful on its own.
EXECUTION
-
Going in order, in the beginning of each group, set up a “signpost”. This consists of the Person-Action for the pronunciation, and then the Object is a major-system word deliniating how many locci are dedicated to each tone. In one locci, you should be able to get something like "ㄉㄨㄣ: [3, 0 ,2, 9] " Note: ironically this image might have to be created and placed last.
-
Sort the characters in the section even further to chain them by shared components. Some sections have ridiculous chains you can take advantage of. For example: 丁 仃 叮 盯 釘 酊 訂, and 抵 牴 砥 詆 底 邸 嘀 鏑 嫡 敵 . I think you can trust yourself to place in the image only the things that have changed since the last character and save a ton of space for the meaning image.
-
Create a PAO image for the Hanzi (or PA-, or P-, as the case may be.) The order can help you know which components go in what order. I generally proceed from left to right, or top to bottom, or outside to instide for components that sort of swallow other components. It’s subtle suggestion rather than totally foolproof, but it seems to be enough! In this image, somehow creatively leave space for the meaning component (next step)
-
Using Pleco or similar strong dictionary, think hard and make one image that represents a lot of the flavors and meanings of the word, can even sprinkle in bonus meanings but make sure they are marked as bonus.
It takes some work in the beginning, but you can generally fit Hanzi, Pronunciation, and Meaning in one Locci this way!
Maybe someone finds this useful. Let me know!
Best of luck on the memory quests,
Hopper
1 Like
Here’s my current list, though it’s always growing:
⼀ ㇒⼁ 亅 ⼃(乀 / ⺄) ⼂丶 ㇀ ⼄ ( 乛㇖) 乚 乚 丁丆 ⼆ ⼔匕 ⼗ 十 [十 丶] ⼚ 广 ⼕ ⺉ ⼘/⺊ ⼌冂 ⺅ ⼋(丷) 分 人 入 ⼓勹 勺 勿 ㇆㇇ ㇖ ㇗ ⺇ 几 ⼉ 儿ㄦル ㇈ ⻈/ 言 ⼙卩 ⻖阝 ⻏ ⼐ ⼑ ⼒力 ⼛厶(龴) ⼜ 又 ⼵ 廴 ⼮ ⼯ ⼟ 土 ⼠ 千 干 手(⺘龵) ⺾ ⼨寸 ⼶ 开 大 犬 矢 ⼪ 尤 ⼩(⺌) ⺍小 龸 ⼝ 吅 ⼞ 巾 ⼭ 山 出 ⼻ ⼺ 夕 多 夊 夂 攴 / 攵⺙ 飠 ⺦/ 爿 門 ⼇ ⼍ ⼧ ⻌ 辶 ⼹(⼹) ⼰⼰己 包 巳 巴 ⼸ 弔 ⼦ 子 ⼥ ⺰ 馬 烏 ⼳幺 王(⺩) 主(丰龶) ⺺ 毋 母 ⽎殳 韋 木 林 朩 犭 歹 車 ⽐ 瓦 止 ⻊ 日 水 氺 貝 見 牛(牜,⺧) ⺧ 午 ⽑ 气 片 斤 爪爫 爫 父 乂 月 那 欠 風 文 方 万 斗 ⺀ 火 灬 尸⼫ 戶 礻 心 (⺖忄) 肀(聿) 弋 戈 用 示 石 龍 龙 目 田/⽥ 由 网罒 皿 𦉫 而 禾 白 瓜 鳥 --島烏 疒 立 辛 穴 衤 礻 疋(⺪) 皮 矛 臣 自 老(耂) 耳 西 襾 覀 頁 虍 虫 缶 舌 竹(⺮) ケ 𠂉 臼 𦥯 血 舟 艮 米 糸 糹 行 辰 ⾚ ⾛ ⾖豆 酉 酋 鹵 里 足 身 釆 豸 角 言 辛 ⻮(⿒) 金 ⻘(!⾭) ⾬ 雨 ⾫ 魚 長 音 革 骨 ⿁ 食(飠) 髟 鬥 鹿 麻 麥 ⿊黑 ⿏ ⿐ 龍 龜 且 ⺈ 龰 𤴓 走 从 氵 冫 冫 五 六 三 豕 彑 彐 ⾯ 面 麥 肉(⺼) 反 生 至 屮 艸(⺿艹) 非 羊(⺶⺷) 禸 ㄐ 𠂇 了 古 兌 癶 乍 ⺺ 巠 巛 衣 𧘇 甘 廿 䒑 羽 犮 咅 北 甫 共 合 台 虍 氏 氐 真 占 者 ㇎ 乃 (段-殳) 垂 屯
Note the many duplicates exist because the computer thinks they are different characters (maybe they have different unicode mappings?) And having both can help find the set using cmmd/ctrl + F “find” commands in a spreadsheet or text document.
I use hanzicraft.com , you can search any character and get a bull breakdown tree, it’s awesome. Then, per component it will show “Component in x characters”. So as you build up your personal component dictionary, you can use its frequency to determine what is worthwhile to have its own entry, and what is not.
It is hard to find unicode characters (computer-available, copy/paste-able units) for everything though, I also draw my own.
2 Likes