Blog Archive

Tuesday, August 18, 2015

Easily confused Japanese characters

Many kanji have subtle differences that may be confused or not noticed, especially if they are printed or written small, and even more so if you happen to have studied reading kanji but not writing them. It's no secret that recognition is far easier than producing-- and it's not uncommon for someone who's only studied reading and never writing to recognize 2,000 characters but only be able to write a fraction of them. One of the major reasons why it's important to practice handwriting is to pick up on these subtleties.

A (somewhat lame) example could be, everyone can recognize the Mona Lisa, but unless practiced in producing it, you probably can't mimic a copy to the point where someone recognizes it. Fortunately writing one character isn't remotely as difficult or time consuming as reproducing the Mona Lisa, but hopefully it demonstrates the point!

Here are 15 sets of easily confused Japanese characters:
  1. シ vs. ツ
     ('shi' vs. 'tsu')

    The font in the graphic shows the difference more clearly, and the stroke order is significantly different depending on which you write. In シ (left), the order is small dashes first, top to bottom, then the long dash written from bottom to top. In ツ (right), the order is small dashes first, bottom to top, then the long dash written from top to bottom. One trick for most handwriting and fonts would be to imagine a clock, and visualize which hour the third stroke is pointing toward. For instance, シ points more towards 4 o'clock, and ツ points more towards 2 o'clock.

  2. ソ vs. ン
    ('so' vs. 'n')

    Both stroke orders are the same for these, except ソ (left) is written with the second stroke from top to bottom, while ン (right) is written with the second stroke bottom to top. The same trick can be used as the first set of characters above, where ソ is pointing more towards 2 o'clock, and ン points more towards 4 o'clock.

  3. 入 vs. 人 
     ('enter' vs. 'person)

    These are quite similar, with 入 (left) having the second stroke overlapping and longer than the first, and 人 (right) having the first stroke overlapping and the second stroke shorter. With a lot of fonts these characters look almost identical, with the exception of the strong hook shape at the top of 入. 

  4. 犬 vs. 太
    ('dog' vs 'fat')

    These characters could be tricky at first for a 漢字初心者 (kanji beginner). They're both written almost the same way, with the difference being where the last dash is placed. 犬 (left) has it's dash written in the upper right hand corner, while 太 (right) places the dash in the bottom center. If confusing at first, mnemonics can help-- 犬 can be a *large* dog with a *drop* of drool from it's mouth, and 太 can be a *large* (fat) sumo wrestler with a *drop* of loincloth between his legs.  

  5. 捨 vs. 拾
    〔'discard' vs. 'pick up')

    It's no coincidence these two look so similar. 捨 (left) means "discard", while 拾 (right) means "pick up" or "gather". Besides the verb stems when used, which would be 捨てる and 拾う, mnemonics can be used to keep the two straight. With 捨, the ground component is used (albeit within 舎), which could be used to remember discarding something on the ground. With 拾, the 合 component is used which in 合う means to "merge" or "come together", like picking something up. 

  6. 未 vs. 末 
    ('not yet' vs 'end')

    The character 未 (left) has it's first stroke shorter than the second, while 末 (right) has it's second stroke shorter than it's first. Besides this subtle difference the characters are written the same way. One way to keep track of these while reading is to keep in mind that 未 (left) is a prefix, like in 未読 meaning "unread", and 末 (right) is a suffix, like in 週末 meaning "weekend". 

  7. 緑 vs. 縁
    ('green' vs 'fate/link')

    These are the same, except 緑 (left) uses 水 for the last portion of the character, while 縁 (right) uses 豕.

  8.  宇 vs. 字
     ('eaves' vs 'character')

    The character 宇 (left) uses 于, while 字 (right) uses 子. You'll likely run into the character 宇 most often in the word 宇宙, meaning "universe".

  9. 鳥 vs. 烏
    ('bird' vs 'crow')

    These two are purposefully similar, with the difference being that 鳥 (right) has one extra line compared to 烏 (left). With that said, the word for crow nowadays is typically written as カラス, in katakana.

  10. vs. 士
    ('ground' vs 'warrior')

    The proportions are what make these two different, other than that the strokes and order are the same. 土 (left) has a short first stroke, and a longer last stroke. 士 (right) has a long first stroke, and a shorter last stroke. 士 is also used most commonly as a suffix, like in 兵士 meaning "soldier".

  11. 候 vs. 侯
    ('season/weather' vs. 'lord/daimyo')

    These two aren't just visually similar, but also phonetically identical (no surprise as the right phonetic-half is the same). The first character 候 (left) has an extra stroke (third), and 侯 (right) lacks the stroke.

  12. 幸 vs. 辛
    ('happiness' vs 'bitter/spicy')

    It's interesting the difference between 幸 (left) meaning "happiness", and 辛 (right) meaning "bitter" is just one line.

  13. 氷 vs. 永
    ('ice' vs 'eternity')

    氷 (left) has it's dash (third stroke) in the upper left hand corner, while 永 has it's dash (first stroke) at the upper middle. The other difference besides the placement and stroke order of the dash is the hook shape at the top of 永's second stroke.

  14. 日 vs. 曰
    〔'day/sun' vs 'pretext/say')

    The only difference in these are the proportions of how they're drawn. 日 (left) is drawn taller, while 曰 (right) is drawn wider. You'll probably run into the second character most often in the word 曰く, meaning "pretext; history; story", or "says".

  15. 垂 vs. 重
    ('suspend/hang' vs 'heavy')
These two are definetely different, however when printed small could easily be confused.

Are there any not listed here that you find easy to confuse? Let us know in the comments!

Thursday, August 6, 2015

Hacking the most difficult Chinese characters (4-6)

We all have some characters that just refuse to stick and that we keep forgetting over and over. The best way of learning tricky characters in Chinese is to deal with them decisively. For more about how to do that, check the first article in this series.

 The most difficult Chinese characters
In this article, I will go through some of the most difficult characters. This difficulty is not based on my opinion, it's based on statistics fetched from our database. We know which characters Skritter users get wrong most often.

For each character, I will explain:
  • Character frequency and basic definition
  • Pronunciation
  • Character composition and formation
  • The component parts and their functions
  • Common words and/or phrases for context
  • Why the character might be difficult
(For number 1-3, please check the previous article.) 

Note that all characters in this article are meaning-sound compounds. Such compounds are created by one part carrying information about meaning and another about sound. Read more here if you're not sure how this works.

4.  叙/敘 (xù) "talk; narrate" (frequency rank: ~2100)

This character is a left-right composition consisting of 余 (yú), which is the pronoun "I" in classical Chinese and here gives the character its sound. The second component is originally ⺙ (pū), which means "strike", but in simplified Chinese, this has been replaced by 又 (yòu), which is a pictograph of a hand and means "also" in modern Chinese. 

Some very common words including this character are:
  • 叙述/敘述 (xùshù) "narrate"
  • 叙利亚/敘利亞 (xùlìyà) "Syria (transcription)"
  • 叙旧/敘舊 (xùjiù) "reminisce, talk about old times"
My guess is that this character is hard because most students haven't learnt 余 properly and might confuse it with visually similar characters, such as 金, 全 and 于. Another possibility is mixing it up with 徐 or 序, which are pronounced xú and xù respectively (thanks to Simon for pointing this out). My personal mnemonic for this character is a very old me (余) striking (⺙) my glass to call attention to my guests so I can tell them about old times.

Practise writing the character using the scratchpad!

5.  毅 (yì) "firm; resolute" (frequency rank: ~2200)

 This is also a left-right compound. The left part, 豙, is extremely rare, so it's perhaps butter remembered as a combination of 豕 (pig) and 立 (to stand). This is the sound component (豙 is pronounced "yì" and 立 is pronounced "lì"). The right part is 殳, which is a halberd-like weapon. There is no simplified/traditional difference here.

There is only one really common word for this character and that's 毅力 (yìlì) "perseverance; will power". We also have the less common 毅然 (yìrán) "resolutely".

This character was very hard to learn for me before I learnt the components properly. None of the components are really used as individual characters, so if you don't study them specifically, you won't know them. However, they are quite common as components, so learning them makes sense.

My mnemonic for this character is a boy standing (立) on the back of a  hog (豕) swinging a halberd (殳), participating in some kind of contest to see who can persevere the longest.

Practise writing the character using the scratchpad!

6.  抛/拋 (pāo) "abandon; throw" (frequency rank: ~1700)

 The left part of this character is 扌 "hand", which should be one of the first components you learn because it's so common. It's related to the meaning of the character. In simplified Chinese, the right part is a combination of 九 "nine" and 力 "force“. Note that the traditional version is deceptively similar, but uses 尤 instead of 九.

Here are two common words that include this character:
  1. 抛弃/拋棄 (pāoqì) "to discard; to abandon"
  2. 抛开/拋開 (pāokāi) "to throw out; to get rid off"
Again, this character is probably difficult if you don't sort out the component parts. There are many strokes that form no distinguishing features (such as boxes or other recognisable shapes). However, if you know the parts well, it isn't that hard. Try the mnemonic "using your hands (扌)  to throw something away with the force (力) of nine (九) men" (or, if you learn traditional, "with extraordinary force").

Practise writing the character using the scratchpad!


That's it for today! Do you find these characters difficult? Have you developed better mnemonics than those I share here? Or do you have a question? Leave a comment!

Tuesday, July 21, 2015

Why strangers are so hard to understand

Image credit: Bob Smith
Human language is varied. It varies across time and space, but also between individuals. This is true within languages like Chinese and Japanese as well. I think all of us have had the experience where we understand what our teacher says, but when we talk to strangers, we understand little or nothing.

Speaking with teachers and strangers

This is partly because a random person on the street doesn't use the neutral and clear accent your teacher speaks with, but it also because of individual variation, a stranger, even if it's another teacher, simply doesn't speak the same way as your teacher does.

This feeling can be frustrating, but knowing that it happens everyone and that it's a natural part of learning a language should help. It's true that your Chinese or Japanese isn't as great as you thought it was when you only spoke with your teacher, but it's also true that it isn't as abysmal as it can sometimes feel after talking with a complete stranger and feeling completely lost.

Layered communication

My way of explaining what's going on is say that different ways of speaking adds layers on top of the intended meaning. Your teacher has a certain way of speaking and you need to figure out how that works in order to sort your way through the layers to the core meaning. The more you listen to her speaking, the easier it will become. If you have only ever had one teacher, you might even think that there are no layers.

However, as soon as you encounter a different person, you'll figure out that this isn't the case, because this person has a different voice and a different way of speaking. If it's similar to how your teacher speaks, you'll be able to adjust quickly, but if there are major changes, it's harder.

Most of these changes are unimportant for the meaning of what is said. For instance, even when using the same words, male and female speakers have different voices. Children and old people speak differently. This takes time getting used to. Your brain requires practice to figure out what is important and what isn't.

Sorting sounds in your brain

There's substantial research into how this process takes place and what allows you to learn the sounds of a foreign language. One of the most robust findings is that you need variety. If you only ever hear one single voice, you won't be able to tell which characteristics are essential for communication and which are just peculiarities of the specific voice of that person.

In other words, if you only listen to one voice, you will learn to understand what that person says, but not the sounds of the language in general. Your brain has no way of knowing what was an added layer because of individual differences and what was really important for determining what something means. When you are exposed to a new, opaque layer, communication fails and it feels really bad.

How to learn to deal with language diversity

The way to deal with this is to vary the input. Practice makes perfect. You learn to deal with different types of layers by dealing with them. The more strangers you listen to, the easier it will be to understand what the next one says. The real world is a varied place, much different from a controlled classroom environment. But it is the real world you're learning for, not the classroom.

Regional accents

I just came back from a four-week trip to China and since I spent most of that time in areas far from Beijing and travelled around quite a bit, I had the opportunity to consider this question from another angle, namely regional accents.

If someone speaks with an accent, that's yet another layer you have to fight your way through to get to the meaning. If you've never heard that particular accent before, it's going to be really hard. The more you listen to it, though, the easier it becomes.

Most layers contain systematic changes, meaning that they mix sounds or switch them in a predictable way. Let's look at some examples in Mandarin Chinese, but I'm sure there are similar variations in Japanese, albeit not as many since Japan is much smaller, both geographically and in population.

In some places in China, “f” turns into “h”, so “fei” is pronounced “hui”, in others “a” turns into an “e”, so “bai” actually sounds like “bei”, in still other areas, “n” and “l” merge. In some places, tones work differently. There are many, many more examples.

The point is that these are additional layers put on to of the core message and you need to learn how to deal with them. They aren't completely random, though, so the more you study, the better your chance to understand the next stranger becomes. Again, the only way to learn that is to increase the diversity of what you listen to.

Beginners and advanced learners

In general, I suggest that beginners stick with one accent, preferably but not necessarily a standard one. You will have enough problems figuring out the basic sounds and there will still be (should still be) variation between individuals who share the same accent.

For more advanced students, watching TV and listening to the radio will bring you into contact with people from different regions and different walks of life. They speak differently. It's up to you to familiarise yourself with the various accents, it's not their fault that you aren't used to the way they speak. Travelling like I did is also a great way of getting to know regional accents, but an option that isn't open for everyone.

Over to you!

What layers have you struggled with? Where they peculiarities of an individual speaker or general characteristics of a specific region? How did you overcome the problem? Leave a comment!

Tuesday, July 7, 2015

Translating the Skritter dictionary

Students of Chinese and Japanese hail from all corners of the world, not just English-speaking parts. So here at Skritter we want to make our methods of language study as accessible as possible to speakers of all languages.

We're looking for a group of helpful individuals, with a intermediate to high level of Chinese or Japanese, who often create corrections or definitions, to help us work on the Skritter dictionary in exchange for free access. (We call these individuals "ballers").

As long as corrections are made on a consistent basis, you would be able to continue receiving free Skritter access. If you fit into this category and are interested, please let me know at, along with which language pair you would like to work on, and I can fill you in on the details.


Language pairs most needed:
Chinese-Chinese (simp)
(plus any pair not list here which isn't on the 'not needed' list)

Language pairs not needed:

Wednesday, July 1, 2015

Japanese Cultural Post: Onsen (温泉)

Hot springs (onsen: 温泉), or literally "warm spring", are a common part of the average Japanese lifestyle. They're used for relaxation, rejuvenation and fighting off sickness if you think you're about to catch a cold. There are over 2500 hot springs in Japan, many of which are natural springs and others which are man made. Some onsen tap into a natural spring but offer a more modern room style bath, while others have different chemical elements in them, such as chloride, carbon dioxide, sulfate and many more that have therapeutic effects on the body. There are some rules to follow when using onsen, such as before you enter the bath, you must first clean your entire body in either a shower or by sitting on a stool and rinsing with a bucket -- this also applies to most Japanese households where when using the bathtub, there's often a shower built next to the tub with a drain directly on the floor, so you can rinse off before entering the tub.

You must also not dunk your head into the onsen, you can only be in the water from the neck down. This is one rule that most foreigners may not be aware of and probably break often.

In Hakone at Yunessun Spa, there is a popular type of onsen, the coffee-bath, where you bathe in coffee mixed with hot spring water. It's said to be good for the skin along with a boost of energy (it's coffee after all). I know how I feel after drinking too many cups of coffee, so I personally would be hesitant to bathe in an entire vat of it! They also have red wine, green tea and sake onsen there. They really got creative with their onsen.

Another type of more common spa treatment is the sulfur onsen (硫黄泉). These smell of rotten eggs, but are supposed to have amazing benefits for people with lung issues, along with expanding blood vessels. Because it can be toxic, sulfur onsen need proper ventilation. There are so many different types of onsen that help with common diseases and are overall good for your skin. There is the iron onsen (鉄泉) which replenishes the body of lost iron. There is sodium bicarbonate saline onsen (炭酸水素塩泉), which makes your skin very smooth. Salt onsen (塩温泉) helps the body retain body temperature, so you stay warmer for longer once you exit the bath.

While there are many types of onsen that contain minerals, the most abundant onsen in Japan do not contain enough of them to really do much to the body, therefore you can stay in them for longer periods of time without worrying about the effects it might have on your health.

Depending on where you are outside of Japan you may be able to find Japanese style onsen, meaning you wouldn't necessarily have to venture all the way to Japan to experience one, though a lot of tourists travel to visit all of the different onsen that are there.

Wednesday, June 10, 2015

Font differences between Japanese and Chinese

Spawning off from Olle's article on dealing with font issues when learning Chinese, this is about some of the differences you might encounter with various fonts in Japanese versus Chinese. It's luckily not as complicated of an issue as character variances in Chinese, but it's worth going over!

The issue stems from unicode, which is a number assigning system for every character in all languages on computers. Back before there was unicode, languages were encoded independently, so if you were to send a document written in Japanese encoding to someone using English encoding, it would show up as jibberish. Unicode ties all languages together and each character has a unique number assigned to it, so that it will show up in the language it was written in regardless. When it came to Japanese and Chinese, it was decided that kanji/hanzi would be mapped to the same unicode place and not put a distinction of what language it is, since it's technically the same character. The problem with that is there are some differences in the way that Japanese and Chinese characters are typically written. You may have noticed on the iOS app the font for everything besides the handwriting area may not always entirely match the handwriting area, which is due to the way the iOS font for Japanese renders characters in a Chinese style, versus Skritter's font on the canvas.

Here are side by side screenshots of some characters that vary in Japanese (left) and Chinese (right) on Skritter:

These just scratch the surface, but are good examples of some the differences between the typical Japanese and Chinese font. Skritter fortunately pays attention to these differences, and it would be nice if unicode would! With that said, it doesn't necessarily make a character wrong if writing it in one style versus the next. When writing a character in 行書 or 草書 (two kinds of cursive styles), it's still the same character-- just written in a different way.

Tuesday, June 9, 2015

Staying on track with Beeminder

There are many ways to motivate yourself to study more. Some of these are based on cultivating an interest for language or culture, others are instrumental, meaning that you learn because you need to be able to use the language for some reason. These are all great, but sometimes they aren't enough. You know what you should do (learn characters and words on Skritter, for instance), but it's hard to get started or you get derailed easily.

Enter: Beeminder

This is where Beeminder comes in. Beeminder is a tool that plots your progress along a Yellow Brick Road (see picture below) to your goal, sends you reminders to keep studying, and if you go off track, it charges you money. Sounds crazy? It isn't. I actually did something like this many years ago when I couldn't finish a major freelance writing project, I simply gave my dad around $1000 and said that he could keep it if I hadn't reached my goal before a certain date. It worked like a charm. Beeminder does just that, but with more sting and in a more sophisticated way. You can watch a short introduction video on their front page here.

Skritter Beeminder
The reason this article is about Beeminder isn't only because it's a useful tool for staying on track, it's also because there is now integration between Skritter and Beeminder available, meaning that you can connect your Skritter account to your Beeminder account, enabling you to track Skritter progress directly on Beeminder.

How to integrate Skritter and Beeminder

The following is copied from the mirror blog post about Skritter over at Beeminder and tells you how the integration works.Their article also includes much more background information, so if you want more than just the basic instructions, you should check it out.
Using the integration is pretty simple. On the landing page you authorize us to read your Skritter data, pick which language you are studying, and tell us how much time you want to commit to. We’ll set up a graph for you and send reminders if you get too close to the edge. We check for new data before we send a reminder, and at the end of the day, but if you want to update Right Now you can visit your graph page and click the “refresh” arrow above the graph, and that will force a sync.

One thing that’s a bit different about the Skritter integration from our other autodata sources is that we set the goal’s deadline to be 4am. That’s to match Skritter’s end of day, which is set to be 04:00, because it is the least likely time to be awake. Usually for autodata integrations we fix the deadline to be midnight and don’t allow you to change it, but because Skritter is counting work done between 00:00 and 04:00 as still counting toward the previous day’s totals, we need to match that up. Otherwise we’ll potentially miss data done in the twilight hours.
The rest of it is pretty hands off. As long as you keep studying you won’t get charged, but if you fall below the road we call that derailing. We’ll charge you and readjust your road so that you’re back on track. We give you a week off after a derail, so it’s a great time to reassess and adjust how much studying you have to do every day. As a matter of fact, you can adjust your goal any time, it just takes a week to take effect. So you can plan ahead to give yourself a break or adjust how challenging your goal is, but, in the words of another rationality nerd, you can’t change it out of laziness unless you are particularly forward thinking about your laziness (in which case you probably won’t sign up for this).
Good luck, and happy learning!
Indeed! If you feel that you need that little extra push to not fall behind on your reviews, head over to Beeminder and try out the new integration. You need to create an account here and if you're new, don't worry, you'll get a chance to learn how Beeminder works before the system goes live (the first derailment is free of charge). Try it out and let us know what you think!