I interviewed a guy who learned to speak utterly phenomenal Chinese within 1 year while living in the UK. Here's how he did it

I recently watched a video about a guy who reached unreal levels of Chinese within a year while in UK. I reached out to interview him. His story is among the most astonishing achievements in language learning I’ve come across. Here’s what I learned: Interviewing This Master of Oral Chinese Made Me Rethink Everything I Believed About Language Learning – I'm Learning Mandarin

UPDATE: The full podcast interview will be published this Sunday 12:00 UK time. STAY. TUNED. You can subscribe to the I’m Learning Mandarin podcast on Apple or Spotify or on imlearningmandarin.com to have new episodes pinged straight to your email.


That is unreal spoken fluency which is achieved within 1.5 year.
Hopefully, in your podcast you delved into his language learning methods.
And what was like listening vs speaking vs reading vs SRS Anki ratio, How much time was dedicated to each activity. Matt vs Japan has a serious competition.

1 Like

Yup, that’s what we talked about in the podcast. Coming soon to a podcast provider near you…


TLDR: This is highly motivating because it confirms a bunch of what I’ve cobbled together myself.

Long version:

This guys method is similar to my method:
anki plus tons of audio and youtube.
He augmented it with talking to tons of people afterwards (which is different than my method).

It’s interesting, though, that he took an intense focus on something that I know about and used a little bit but didn’t focus on entirely, which is the IPA system.

My hypothesis comparing my own results to his (I’m at just under a year in Russian) is that my Russian was slowed down because I did NOT focus enough on IPA. The phonology of Russian is sufficiently different than English that it killed my comprehension of the individual vowels and consonants. It took me nearly 7-8 months before the sounds started to feel natural to me.

I haven’t yet talked much (only 5 times) to native speakers so I can’t compare there.

The best comparison I have is what I did with Spanish: almost identical but at about a year I immersed myself in Latin culture and talked for at least 5 hours a week. I speak pretty good Spanish with a reasonable Mexican accent.

So I have no doubt at all that this guy’s method works.

Specific commentary on his method follows:

EDIT: one thing he does that is interesting is that he “speaks to himself”. I definitely don’t do that so maybe that’s why my speaking is nowhere as good as my listening but I can say that the Universe of memory guy says that speaking words out loud helps massively. The Glossika guy’s entire method is just based on repetition of phrases out loud. And there’s also stuff like Pimsleur and Dave Thomas so maybe there’s something to it.

EDIT 2: There is one thing the guy has right but his reasoning is wrong. He is right that you need to keep a habit up (i.e. do some every day - in my case 1-2 hours). He’s wrong about it merely being to do with lack of motivation.
IMHO the lack of motivation is caused by lack of belief in success due to a very long feedback situation as well as a high threshold before you feel like you can understand anything. The difficult part is to know that it works and keep on going. In my case I had no problem “knowing” that my method would work for French because I had done something similar with Spanish (albeit I didn’t have audio for Spanish in the anki equivalent I used back in the day). The second thing I did was give myself a time limit so that my target was do it HARD for “just” six months and then quit. So both of those things over-rode my lack of belief, which again IMHO is the total killer of motivation. With Russian, however, the lack of belief crept in because I had never had success at a “hard” language. I had to come up with a proxy belief by the fact that Ikenna and Steve Kaufman’s methods both overlap with my own. At times I felt like giving up because the progress on Russian felt painfully slow. But at six months I had definitely made acceptable progress. And right now at 11 months I can’t quite say I can “speak” Russian (although I absolutely can) because my grammar is horrible. But I can definitely have a conversation and I can understand several youtubers in Russian perfectly. Although I haven’t quite achieved my target of being able to perfectly understand TV shows in netflix yet.

EDIT 3: I’m delighted to hear he shares my belief in “don’t study grammar” as I suspect this means that the Mandarin Grammar is not very complicated to master.
In Russian, however, the grammar is pretty complicated because it has cases, declensions and conjugations. So I agree AND disagree with him. What I would say is this: don’t worry about grammar in the beginning in any language. If, however, your language is grammar heavy (e.g. Russian) you are definitely going to have to dig into it at some point but I would leave it until at least six months in. If it’s not grammar heavy then yeah, ignoring grammar you’ll pick up what grammar it does have just by listening/reading/being exposed to it.


He augmented it with talking to tons of people afterwards.

He spoke from the start and all the way through. Speaking, being corrected by native speakers, recording those corrections and drilling them as whole sentence Anki cards wa absolutely central to his method.

This is a story about how early output - if combined with phonetics training and immersion - can lead to spectacular results.


xxdb, his anki collection contains sentences mined from different sources such as speaking. What about yours? Only individual words with sounds?
My concern about his method is that. is it possible to acquire sophisticated grammar knowledge just from audio input alone to an extent that he speaks like a native speaker and with that kind of fluency in 1.5 year, if 3 years was a timeline, i would have believed him but again who knows…he speaks with his Chinese girlfriend all day long…

1 Like

is it possible to acquire sophisticated grammar knowledge just from audio input alone to an extent that he speaks like a native speaker and with that kind of fluency in 1.5 year

Probably not and that isn’t what he did. He didn’t only ‘rely on audio input alone.’ In addition to audio immersion he drilled whole sentence flashcards (including characters). He memorised hundreds/ thousands of sentences this way and practiced using them at the same time eventually developing a feel for the structures so he could produce them flexibly himself. Relying on immersing in audio input alone is far less efficient.


Interesting. His method overlaps with glossika if he’s doing whole sentences.

So my opinion on grammar is that if the language doesn’t have much grammar you will just kind of pick it up. Also what he’s doing different than me (according to michilini) is whole sentences instead of vocab words. (I’m not changing my method to do whole sentences because I find it painful).

He says “don’t do grammar” but I think in fact he is doing grammar.

His method differs from mine (if what michilini says is correct) in that he drills sentences.

I can hypothesize that you’ll get grammatical “forms” from drilling model sentences. That’s the basis of glossika.

Also: yeah Chinese girlfriend will massively help his active speech.

I did something similar with Spanish: I was immersed in Spanish all weekend every weekend for more than a year after I had already memorized 7000 words plus watched a bunch of telenovelas.

So yeah I believe him when he describes his method.

Interesting. Nonetheless. Looking forward to hearing the whole episode. Really appreciate your efforts for conducting such interviews.


Right. Which is what the glossika method is; memorizing thousands of sentences which are carefully selected to model particular grammar patterns.

In fact I did something similar with Spanish to get my head around particular conjugations.

Interesting. Maybe I’ll reactive my glossika subscription.

1 Like

Interesting. Speaking from the start is Benny Lewis’s method.

It is possible that this guy has developed the optimal method.

It’s a hybrid of my method(lkenna’s method) plus Benny Lewis plus Glossika.

Yes. Mine is just individual words with sounds.

I’m re-evaluating my position after seeing this.

My method definitely works for understanding.

My output isn’t great in the case of Russian. I have a very good pronunciation and I can definitely string together grammatically incorrect (but understandable) sentences.

But… I cannot produce cases at all even though I can understand when I hear them.

What he is doing appears to be very similar to glossika combined with Benny Lewis’s method.

1 Like

Not quite. My understanding is Glossika feeds you sentences. That’s far less effective than mining sentences yourself, selecting based on what you want to talk about and say. Also mining sentences from corrections from native speakers during conversations. The whole thing is entirely tailered to his own needs.


Sure. We are splitting hairs though. Both Glossika and the guy are using sentences as models and both Glossika and the guy are memorizing those sentences. The guy is using anki to memorize them and glossika is using glossika to memorize them.

That one is specific to him and the other is generalized is true, but it’s un-neccessary IMO to draw a distinction between them when the methodology is they are both memorizing sentences: the grammar problem is solved by either of the two methods.

Where I would argue that the “tailor to your own needs” part would be superior to glossika is that he is essentially learning how to speak his own stories. In essence conversations are story telling and often when meeting other people you are going to tell them your specific stories from your own life. Which is obviously better than knowing disconnected sentences that don’t match your stories.

For my own purposes, however, I only need to improve my grammar. I can figure out how to tell my stories later.

1 Like

Nope, it’s not splitting hairs. It’s an utterly crucial distinction. In one case your accumulating sentences which may or may not be relevant to your life and which you therefore may not have much opportunity to use in conversation with native speakers. In the other case your’e only memorising sentences which you know you will then be reinforcing through use in daily conversations.


“Nope, it’s not splitting hairs. It’s an utterly crucial distinction. In one case your accumulating sentences which may or may not be relevant to your life and which you therefore may not have much opportunity to use in conversation with native speakers. In the other case your’e only memorising sentences which you know you will then be reinforcing through use in daily conversations.”

OK you can’t get past it and are in confirmation bias mode instead of discussion mode. I’m stepping off here.

Thanks for the initial post.


Maybe he has a point when you curate sentences from your reading and listening, you already accumulate a lot of background information while reading or listening. You can relate to the content emotionally so sentences will stick better. Everyone has different needs. For example, someone who is learning German in Canada does not have an immediate need to order a cup of coffee
at a local bakery in the German language. Since I am living in Germany, I have this immediate need.

Just to illustrate this point or reinforce his position, I came across the word “das nicht”, while watching a movie. I added it in Anki and memorized it because I need this sentence for my needs. Nearby I go to a Turkish shop for buying doner kebab. It consists of many individual ingredients, sauce, yogurt, onions, etc
The vendor asks me which ingredient I do not want to have. I use this phrase and he understands it perfectly well. Similarly, I came across a sentence in a German grammar book “Ich will einen Kaffee”. I prefer Cappucino. However, this sentence is relevant to my needs. I simply replace Kaffee with Cappucino. And communication flows and the vendor understands it.

There are hundreds of individual needs that may vary from person to person and going through premade cards is not completely useless you always pick up on something but are they a good value for your time and end results?


Kudos to Will Hart!
It wouldn’t be surprising to achieve a high level of fluency in a short amount of time in a not too distant L2, but it’s awesome to achieve this in a more distant L2.

@Michilini: But why do you have to rethink everything you believed about language learning?
The strategies Will mentions (e.g. in this video 6 tips I used to learn chinese in a year! 学好中文只要一年,揭秘语言学习捷径 - YouTube) are all well-established in language learning.
There’s absolutely nothing new or revolutionary here: his approach just seems to be a mix of “output first” à la Benny and input orientation / mass immersion while
focusing mainly on the oral dimension.

This may come as a surprise only to the “input purist” faction of language learners…

Reg. the “time frame” (here: 1 year / 1.5 years):
Years, months. etc. are completely useless time units when it comes to learning a practical skill.
So the real question is: How many hours / minutes of quality time per day, week, etc. did Will spent on learning Chinese? I would be really surprised if it was less than 1000-1500 hours in a year.


BTW, this seems to be a typical day of “Khatzumoto” (the founder of AJATT) when he was learning Japanese in an 18-month period (between 2004 and 2006) so that he was able to work for a Japanese company:

So we know that it is possible to learn distant L2s in a relatively short amount of time.
However, the question is: How many hours per day / week can learners consistently invest in their language studies?


“There’s absolutely nothing new or revolutionary here: his approach just seems to be a mix of “output first” à la Benny and input orientation / mass immersion while focusing mainly on the oral dimension.”

Taken in isolation perhaps none of his methods are particularly novel. But I have not come across a case quite like it and I’ve interviewed many learners for my podcast. I was especially interested to learn how he effectively married mass input with early output & phonetics training.

He is the perfect antidote to the purism you allude to which is unfortunately rife in online language learning circles.

“his approach just seems to be a mix of “output first” à la Benny and input orientation”

I don’t think that’s accurate. His early output method differs to Benny in two significant respects.

  1. He combined early output with phonetics training. As far as I know Benny didn’t begin by studying IPA nor did he nail the tones. Benny believes in coming back to nailing pronunciation later which may work well for some languages but for mandarin tends to be a disaster.
  1. Will didn’t just engage in conversations and expect to improve naturally. He noted down every correction and drilled it as an Anki flashcard to make sure he didn’t make the same mistake again.