BrushPath · Blog

Why I built BrushPath

8 November 2026 · Khay

I grew up speaking Cantonese and never really learned how to read or write it, which for most of my life didn't matter, because I lived in the UK and a younger me firmly believed I would never need it. That changed when I met my girlfriend, who is Shanghainese, and we started taking trips to China together. I loved it there, and I felt like I was reconnecting with a part of myself I'd ignored for years.

Travelling with her was easy, because she did all the talking and all the navigating, but it left me feeling rather useless, and somewhere on one of those trips I caught myself wishing I'd spent more time studying the language. I wanted to be able to get around on my own, and in China that means something specific. You don't get around with signs and a map any more; you get around with Alipay, WeChat, Didi, Meituan and the city's own apps, for the metro, the bike, the taxi, the restaurant and the bill, and every one of them is in characters. I wanted to be able to open those apps and use them, and at the very least to read the signs.

Why write, and not just read

I decided the only way for me to really learn a character was to write it. Reading is passive, in that you can look at a character a dozen times and still not know it the next time you see it without the pinyin, whereas writing involves more engagement. You have to pay attention to every stroke, where it starts and which way it goes, and that attention is what makes it stick.

There was a less noble reason too. I'm a bit of an introvert, and by that I mean I would totally avoid talking to someone if possible, so if there's a version of a task that doesn't involve another person, that's the version I'll do. Speaking practice means a tutor or a stranger and getting it wrong in front of them, whereas reading and writing need nobody. I could sit with a character for as long as I liked, get it wrong fifty times, and no one would ever know.

Learning with TOFU

I'd tried tuition and YouTube, and they helped a bit, but the app that clicked with me was TOFU Learn. I went through HSK 1 to 3 in about a month and slowly progressed to HSK 6 over the next three to six months. I practised several times a day, short but frequent, and for the first time I felt like I was making progress, not just in writing but in reading and speaking too. Once I could read a word I recognised it when I heard it, and once I recognised it I could say it.

It worked for me because I'm lazy, in that I can't commit to hours of studying at a time, but I can manage a few minutes several times a day, and TOFU never asked for more than that, so it was easy to come back.

Then TOFU died

Then, unfortunately, TOFU died. Its server kept going down, and in February 2025 it was pulled from the app stores, and when it went I stopped learning, because nothing else I'd tried made things stick the way writing had.

So I thought, why not build another TOFU, one without the issues TOFU had, and with graded reading in the same place. That's how BrushPath came about, and what it does differently is a post of its own: What TOFU Learn got wrong, and what BrushPath does about it.

Who it's for

It's not only aimed at complete beginners, though beginners are welcome and the first level is free. It keeps heritage speakers in mind, anyone who already knows how a word sounds, in Cantonese or Mandarin, and has never been shown how it's written. Being Cantonese, I already knew most of the words; I'd just never seen them written, and that turned out to be true of most heritage speakers. It's for people who really want to learn not just to write but to read, and to know how each stroke is made, in order.

Writing came first for me, then reading, and once I could read, listening and speaking followed, so that's the order BrushPath is built in. Hopefully it will help its learners read and write, and from that, help them with listening and speaking too.