Char's Study Log & Hanzi Hero Journey

My current status:
characters: 1,949
character/day: 13
words: 479
words/day: 2

Too Long; Didn’t Read: Hanzi Hero is really great and fun. It is super easy to get started with and continue. There are issues sometimes, but the developers listen to feedback. I would like zhuyin pronunciation and characters as components. I would like Hanzi Hero to be free software. I am focused on learning characters.

I’ll just tell my whole Chinese story so far here.

Chinese Before Hanzi Hero

I did some Pimsleur which is audio only. After knowing some things, I learned about pinyin and tried typing using an input method (IME). I didn’t study too much pinyin, so I was surprised as to why stuff like 会说 is hui shuo instead of hway shwo. Eventually, I got to the end of my Pimsleur lessons, and I kinda forgot about Chinese for a while.

Enter Hanzi Hero

I was waiting at the bus stop with a friend. He couldn’t stop telling me about his Chinese studies using Hello Chinese and Hanzi Hero. It seemed fun, so I got both. I made sure not to push myself too hard, so I kept the default 10 cards/day, and I didn’t do more than one section in Hello Chinese per day. I really liked that Hanzi hero took the time to explain the components of each character as well as the pronunciation all wrapped up in a nice mnemonic. I actually liked Hello Chinese better because I felt like I was learning more useful things, but I learned that they leave out a lot of characters. I don’t have google play, so paying for Hello Chinese is basically impossible for me. I continued doing only Hanzi Hero.

Continuing with Hanzi Hero

When my free trial ran out, I enjoyed the break of only learning new words for a week. I enjoyed it so much that I decided to subscribe and continue learning. Doing my daily study, some things started to annoy me.

The #1 annoyance was having to study each character 2 times because the pronunciation and meaning were far apart. Thankfully, this has been solved in a way that I find very satisfying. It has unlocked new study abilities. I can start typing the meaning as soon as I enter the pronunciation; so fast! Also, while driving somewhere, I can have a friend describe a character to me, and I can tell them the pronunciation and then the meaning right after without having to describe the same character again later.

The #2 annoyance is having to study the meaning a character twice as much when there is a component with the same meaning as a character. My work around is to skip the component, and prioritize the character with the matching meaning. For example 舟 is a component in many HSK2 characters, so the component is needed early on, but the character doesn’t show up until HSK7! This lead me to learn many characters “out of order”, but I do not care because I am not interested in taking any kind of test. I like this because when I see 舟 as a component, I can think zhōu as well as boat; the pronunciation helps me. This is especially useful when the component is phonetic like 羊 (component HSK1, character HSK3). At the minimum, I would like to see Hanzi Hero make it easier to do this. With the current lesson flow, it doesn’t work well with skipping things and prioritizing things in the middle of a lesson, so I have to back out and start the lesson over.

At the maximum, I would like to see Hanzi Hero fully embrace this “characters as components” idea. As an example, Hanzi hero has the 鸟 component (HSK1) and character (HSK2) , and it also has the completely unrelated 乌 component and character (both HSK 6). I would like to see the 乌 and 鸟 components go away. There is now only a 乌 character which is a prerequisite for the 鸟 character (along with the 丶 component) which is a prerequisite for the 鸡 character (along with the 又 component). I know this is kind of radical and not everyone would want this. The way I see it, I would have learned 乌 character instead of spending time thinking about a bird component.

I like words in Hanzi Hero. I think they do a great job of their purpose of re-enforcing characters. That said, I like learning characters in Hanzi Hero a lot more. That is why I settled on only doing 2 new words/day. I have tried going up to 20 cards/day, but it is not sustainable, so I settled on 15 cards/day.

I have found that doing Hanzi Hero reviews on the computer is just way faster than on mobile. I can type much faster on the computer with less mistakes. I still use mobile quite a bit when I do not have access to a computer. I really like the “wrap up review” feature so that I can switch between devices.

I really appreciate Hanzi Hero’s pinyin validation.

The developers have been super responsive to inquiries about mnemonics, components, and feature requests.

Supplemental Study

I have tried some Du Chinese, and I like it, but I am not doing it regularly for 2 reasons. Reason #1 is that they are still on HSK 2.0, so even reading stories on my level leaves me having to look up a lot of characters. Reason 2 is that I just do not have enough time to focus on both learning characters and reading and listening. This is why I am focusing on mass character learning. Then I will focus in on reading and listening. At my current rate, I will learn all the characters Hanzi Hero has some time in December. At that point, I will shift my focus to reading and listening. I will of course continue my daily review and maybe learn a few words with Hanzi Hero.

The more I did Hanzi Hero, the more I noticed that I consistently miss the characters that I just learned yesterday. I thought something that might help with this is to write the character down on paper and think about the mnemonic as I am writing the character. This has helped quite a bit and it is a lot of fun. Of course I make sure to use the correct stroke order. I use omgchinese.com or strokeorder.com for this. I am only writing simplified characters currently. I like that omgchinese shows corresponding traditional version of the character. I would like to see Hanzi Hero have support for learning both traditional and simplified at the same time. Writing the characters also makes me see certain components differently. For example, 隹 is really a slenderman with a lid and a bean stalk.

ㄅㄆㄇㄈ

Several times throughout my journey, I have come across different input methods. The other main phonetic one for Mandarin being zhuyin. I was interested, but I dismissed it because it looked too foreign. After learning quite a few Hanzi, and returning, it looked a lot more familiar! For example, ㄖ is basically a quick version of 日; ㄡ is 又; ㄦ is 儿; ㄒ is just 下 without the extra stroke; ㄊ is basically elbow pad, wait, that isn’t a character! You can see explanations of the other symbols on the wikipedia page. I think it is pretty cool that some of them come from ancient versions of characters. In fact the more I learn about zhuyin, the more I become obsessed. The way all the similar sounding symbols are aligned vertically on the keyboard is so satisfying and practical. All the initials on the left. All the finals on the right. It is so nice. I have already switched all my input methods to zhuyin. I am slow to type currently, but I think it will be worth it in the long run (especially since chewing IME supports stenographic input). The presence of tones also narrows down the selections list quite a bit. I have also been writing the zhuyin annotations next to the characters when I learn them.

Despite Hanzi Hero’s great pronunciation guides and text to voice pronunciation, zhuyin helped me realize that I was still pronouncing some things incorrectly. For example, ㄘ (pinyin c), has always been a hard one for me; learning that it is just aspirated ㄗ (pinyin z), has helped me a lot. Having pinyin e be separated into ㄜ and ㄝ has helped me understand the “few cases where it does in fact sound like the ‘-e’ in ‘electricity’”.

I would absolutely love it if I could input zhuyin for the pronunciations in Hanzi Hero. I imagine that it would not be too hard since Hanzi Hero already uses zhuyin for its text to voice internally. The main issue would be that there is no convenient way to input actual ㄅㄆㄇㄈ letters. Even if there was an input method for it, it would be extremely inconvenient to constantly switch between it and English for the pronunciation and meaning. I think the best way to do this would be different for desktop and mobile. For mobile, just make your own zhuyin keyboard that appears for pronunciation prompt, when I press ㄅ on this Hanzi Hero keyboard, it immediately inputs a ㄅ in the pronunciation text box. No IME features; no conversion to Hanzi. For desktop, I think the best thing to do would be sense key strokes and input the appropriate symbol. When I press 1, immediately input a ㄅ into the pronunciation prompt.

Of course the bigger issue with integrating zhuyin into Hanzi Hero is the mnemonics. For some sounds there is a one to one match. ㄇㄨ is Mulan plain and simple. For some, it is not so straight forward though. One might think that ㄒㄧ could be Xena. It works for some syllables like xian & ㄒㄧㄢ, but xiong is equivalent to ㄒㄩㄥ. It might make sense to have character sound mnemonics for zhuyin medials. ㄧ is already Yoshi. ㄨ is already Walter White. ㄩ could be Quasimodo or some other original character. This means that ㄉㄨㄥ (pinyin dong) would become Dracula and Walter White in the English Manor. Going along with with this, ㄇㄩ would become Mario and Quasimodo in the house. Yes, this means that the Ong Temple and Mulan would go away completely (at least for zhuyin mode. I am not trying to take away anyone’s pinyin). Pinyin is also still great for romanization.

I realize that integrating zhuyin into the very pinyin centric Hanzi Hero is a huge ask, but I think it would be truly awesome.

Licensing

I really like Hanzi Hero. I also really like free software. I like not having to fiddle with Anki decks or make up my own mnemonics. I really like paying Hanzi Hero developers to work on mnemonics, fix errors, and build features for Hanzi Hero. I would like it even more to pay them to do that if Hanzi Hero was free software. I would love it if Hanzi Hero was made free software with an AGPLv3 license. I think it is cool that Hanzi Hero is written in Elixir (not as cool as Common Lisp :stuck_out_tongue:). It might be nice to propose mnemonic or component corrections with a pull request or a patch email instead of bothering Phil all the time. I might even work on the zhuyin pronunciation input (no promises).

Conclusion for Now

I like using Hanzi Hero. The developers did a great job at making a polished application. I would like zhuyin pronunciation and characters as components. I would like Hanzi Hero to be free software. I am focused on learning characters. Writing Hanzi is fun. I have already learned a bunch of Hanzi, and I can’t wait to learn more.

4 Likes

Thanks for your review (and for your help in fixing up errors/issues you run across)! It helps a lot :slight_smile:

I can see an argument for this. We initially leaned toward the side of less components as that’ll mean less to learn, but I have a suspicion that by not embracing the “characters as components” model – or, in other words, componentizing as much as possible – it can sometimes make it more difficult to remember.

It’s also possible that it’s the same difficulty. The question is whether having more smaller components helps or hinders :thinking:

I would be curious how much this has boosted your recall – writing is another great association to make when it comes to remembering :thinking:

Yeah, the main barrier-of-entry to providing zhuyin as an answer method is the keyboard question. Your suggestion of providing our own virtual zhuyin keyboard is an interesting one!

If we were to go down this route, to reduce confusion/keyboard issues I was thinking instead of providing a pinyin → zhuyin real-time convertor as you type.

Thus you don’t have to learn a new layout, but you get to see the zhuyin associated, and it helps differentiate the nuances of pronunciation since zhuyin makes it explicit.

This also makes it so we don’t have to do anything to the mnemonics, leaving them pinyin centric but letting the zhuyin highlight how things are pronounced when you type it in.

I think zhuyin is pretty cool and would love to support it in the future!

Here’s a thread previously discussing this.

It’s amusing you mentioned this, as I had a stint of using Common Lisp and at one point we even discussed using Common Lisp!

Common Lisp is cool for sure. The main argument for Elixir in the end is that everything is structured, so you don’t spend too much time thinking about the code and instead think about the end results :thinking:

No worries, it’s never a bother :wink:


Thanks for the valuable feedback/insights on your experience :sunglasses:

Yes! Now whenever I come across 鸟, I will think crow and drop even though that is not the Hanzi Hero mnemonic. I can’t unseen it.

That is definitely an interesting idea. The main reason I was thinking about virual keyboard is so that I can build the muscle memory for the different keyboard layout.

1 Like

Thanks for your detailed feedback and thoughts! :slight_smile:

For Zhuyin, I think the first steps is showing the Zhuyin as secondary text alongside the pinyin. And after that, as you mentioned, at least on desktop, having some sort of input remapper that maps latin letters to the Zhuyin equivalent. The automatic live translate thing Phil mentioned seems cool, but may be more difficult in practice. I also see ourselves, separate from the app, adding some supplementary Zhuyin tools/quizzes at some point when we find the time for it.

The components and characters thing is definitely something we may be able to improve upon eventually. At the very least, once we have better word coverage of the 3k HSK characters, I see ourselves going and revisiting our componentization. I don’t see ourselves being able to make huge changes there, but at least we can fix the cases I see where we could combine a common group of components (which almost always maps to a character) into a new component. This would also reduce the number of components in many mnemonics that use that group of components, making it easier to remember.

I assume you mean something like 暴 and 高.

1 Like

It is of course hard to measure. It has a noticeable improvement, but I still get some wrong that I just was introduced to yesterday.

Something I have noticed is that some characters are mainly or purely phonetic or mainly used as a surname. Currently the “meaning” part of these characters is just the pronunciation without the tone. I have started adding “phonetic” and “surname” respectively as meanings for these types of characters. If I got the pronunciation, then I know what the surname is or what the sound is. I have really liked it. The only problem is that there is a bunch that I have already learned that do not have these alternate meanings. What do you think about adding them to all these types of characters? Another advantage is that it clears up the meaning; currently the only way to know if it is a surname or a phonetic is to see Mr. Zhao in the mnemonic.

It would also be nice if characters that are mainly used as classifiers had some distinction.

1 Like

For classifiers, I can see eventually showing some additional information there, or grouping them into a “list” (once we add that). On a separate note, I’ve also thought about making some sort of mini quiz or something, probably separate from HanziHero that has some combination of:

  1. Show a picture of something and one can type (or select from a list) the appropriate measure words
  2. Show an English word (probably with Chinese equivalent below it) and select from a list the appropriate measure word(s).
  3. Show a measure word and select the appropriate item(s) from a list (probably in English with secondary Chinese) that would be appropriate to use.

For the phonetic/surname thing, I can see the utility of. I’ll think about it some more. One downside (or upside) is that all phonetic or surname alternate meaning words will now be in the same “similar meaning” group.

2 Likes

I had not considered that. I wouldn’t be opposed to it. Alternatively, maybe there could be an exception for these meanings.

1 Like

I just remembered add “dynasty” to this list.

Which list are you referring to?

The list of characters where the meaning is basically the pronounciation:

  • phonetic
  • surname
  • dynasty