Names and identifying details of every student here are changed, as they are everywhere on the blog.
The ones that feel efficient are usually the ones that are not
I’ve finally found something that works, a man who had been at it eighteen months told me, and described an evening of flashcards that sounded pleasant enough to do forever.
He could recognize about nine hundred words. He could produce maybe forty of them in a sentence without stopping. Names and identifying details of every student here are changed, as they are everywhere on the blog.
Nothing on this page reduces the work. The good ones move it somewhere it pays.
That gap is the whole subject. Every method below either moves effort from recognizing to producing, or it does not, and the ones that do not are the ones people stick with, because they are the comfortable half of the bargain.Seven, in the order I would introduce them
Spaced repetition
The only one a machine can run for you
Meeting a word again just as it starts to slip.
Comprehensible input
Recognition work
Russian you can mostly follow, hard enough to stretch and easy enough to enjoy.
Shadowing
Production work
Speaking along with a recording a half-second behind, so your mouth learns the shape before your head catches up.
Conversation from day one
Production work
Badly, with somebody who will let you be bad at it for as long as it takes.
Phrases over words
The cheapest change on offer
A phrase arrives whole and a word has to be built into something.
Recording yourself
Four minutes a week
What you meant against what actually came out.
Narrow reading
The fastest thing here
The vocabulary repeats and the subject carries you over the gaps.
Order matters more than count here. Somebody who starts with the last three and adds the first four later tends to arrive. Somebody who starts with the two a machine can run and promises to add the rest later usually does not, because readiness of that kind never announces itself. I’ve watched both sequences enough times to stop treating it as a matter of temperament.
Why the first four are not equal
Two of them a machine can do with you and two of them it cannot, and this is where most of the wasted time goes.
Spaced repetition and comprehensible input are recognition work. They build the half you already build fastest, and left alone they produce a person who reads well and says nothing, which is the most common shape of failure I see. It is also the most invisible. Everything about it looks like progress from the inside: the word count climbs, the articles get easier, and nothing in the evening tells you the other half has not moved since spring.
I can read the news and I can’t order lunch, one student said, and he was describing eighteen months of doing only the comfortable two.A method you enjoy from the first week is usually training the skill you already have.
Shadowing and conversation are production work. Both are unpleasant in the specific way that matters: your mouth is doing something it cannot yet do, in front of either a recording or a person, and there is nowhere to hide.Where the effort goes
The three nobody puts on the list
Phrases over words is the cheapest change on offer and the one people resist hardest, because a word count feels like progress in a way a phrase count never quite manages to.
Recording yourself takes four minutes and tells you what no amount of listening will, which is the gap between what you meant and what came out, and the gap is never where you expected it. The transcript was brutal, a student wrote to me after the first attempt, and that is the reaction of somebody who has just learned something.Narrow reading is the one I push hardest. An engineer reading engineering Russian meets the same three hundred words until they stop being words and start being furniture, and the subject knowledge fills in what the language does not, so the whole thing moves at a pace general reading never manages.
Three courses I did not finish
I’ve abandoned more language courses than most of my students have started, and they had one thing in common that took me years to see.

Every one of them was chosen because it looked sustainable. Twenty minutes a day, gentle progression, nothing that would make me feel stupid in front of anybody. I kept each for about six weeks, made real progress in the first three, and drifted off when the interesting part ran out, which is exactly when the useful part would have begun.
The course I did finish was the one I disliked most, where a teacher made me speak for twenty minutes before I was ready and I came out of it feeling flattened.Every method I abandoned was one I could do alone. The one that worked had somebody in the room.
That is not a coincidence and it is not a sales pitch. It’s the same thing the seven methods are trying to say from different angles: the part that speeds you up is the part you can’t do comfortably by yourself, which is also why people stop somewhere around month seven.Related reading at Palme School
Building a home study routine
Why practicing alone stops producing results somewhere around month seven
How to think in Russian instead of translating
What the pause before you answer actually is, and what shortens it
Speaking Russian with confidence
If the obstacle is not the method but being heard getting it wrong
Frequently asked questions
Some combination of production work and narrow reading in a field you already know, and it is not what most people start with. Recognition methods feel faster and build the half that was never the bottleneck.
Yes, for recognition, and it is the only method on the list a machine can run for you. It will not teach you to produce case endings under time pressure, and no flashcard schedule has ever managed that.
Speaking along with a recording about half a second behind, so your mouth learns the rhythm before your head has finished the sentence. It is worth it, it feels ridiculous, and those two facts are related.
From production methods, about six weeks, and you will usually notice it as somebody else interrupting you less rather than as a feeling of progress.
Palme School: Conversation school · Russian for bilingual children · Free materials for learners
Lily Rozberg has been teaching for sixteen years, first English in Russia and now Russian to bilingual children across the US and Canada. She writes for Palme School about heritage language, reading, and the things report cards do not measure.





