Thirty three letters, and only about seven of them are the work. The alphabet divides into three groups and they cost wildly different amounts. Letters that look Latin and sound the same are free. Letters that look new are cheap, because nothing interferes with them. Letters that look Latin and sound like something else are the entire difficulty, and there are six or seven of them.
So the first afternoon should not be spent on the alphabet as a list. It should be spent on В, Н, Р, С, У and Х, because those are the ones your eye will keep reading as English.
The letters that cost the most, first
Half the alphabet is recognisable and about six letters do all the damage.
- В is v, Н is n, Р is r, С is s. Four letters that look like Latin ones and say something else, and they are among the commonest in the language.
- Ж is the s in measure, a sound English has and never spells with a single letter.
- Щ is a long soft sh, held and pushed forward, distinct from Ш.
- Ы has no English equivalent, made by saying ee with the tongue pulled back, and it separates real word pairs from И.
- Ь makes the consonant before it soft. It is not a sound of its own; it changes the one in front of it, and it changes meaning.
- Я, Ю, Ё, Е carry a y before the vowel at the start of a word, so Я is ya and Ёлка is yolka.
Interference, not memory
Reading is automatic and that is the problem. You do not decode a familiar shape, you recognise it, and recognition fires before any decision can be made about it. So a learner who knows perfectly well that Р is an r will still read ресторан as something starting with p for weeks, and being told again does not help, because being told is not what the eye is doing.
What breaks it is forcing the decode. Read words made almost entirely of the dangerous letters and read them aloud: ресторан, сувенир, Россия, универсам. They are mostly international words, so the payoff is immediate and the meaning confirms whether the decode was right, which is the feedback a letter chart cannot give.
The second surprise arrives later and catches almost everybody. Russians write in cursive, and Russian cursive is not the print letters joined up: several of them change shape entirely, so a handwritten т looks like an m and a handwritten д looks like a g. Italic type does the same thing. A learner who has only ever met the print forms can read a book and not a note, and finding that out from a handwritten menu is worse than spending an hour on it now.
Reading without ever reaching speed
Here is the stall and it is easy to miss because it looks like success. The letters are learned, the words can be decoded, and the learner declares the alphabet finished. What they are actually doing is solving each word, at perhaps a fifth of the speed they read English, and at that speed reading is not input: nothing is left over to notice a pattern with, and every page is work.
The threshold that matters is not accuracy, it is automaticity, and the only route to it is volume of very easy text read fast, accepting mistakes. Signs, labels, transliterated brand names, the first page of a children's book, ten minutes a day. Somewhere in the second or third week the recognition flips and stops feeling like decoding, and until it does, everything built on top of reading is being built on a slow foundation. That is the general argument in why you forget, and the schedule that beats forgetting applied to a skill rather than to a word.
Two alphabets, one language
The letters that change shape when somebody writes them
Russians write in cursive, so a reader who only knows the printed forms can read a book and not a note.
The right hand column is why a handwritten Russian word can look like a row of identical humps. Native readers use context and the underline; a learner needs an hour of copying and then stops noticing.
Two things that force the decode
Put a Russian keyboard on your phone and use it. Typing requires you to find each letter rather than recognise it, which is exactly the operation the eye is refusing to do, and the mapping between the Latin key and the Cyrillic one is precisely the interference you are trying to break. It replaced writing the alphabet out twenty times by hand, which trains the hand and leaves the eye where it was. Free, and it takes two minutes to add.
For handwriting there is no substitute for a cursive chart with a page of real Russian handwriting beside it. It replaced discovering the difference from somebody's shopping list, which is how most learners find out. An hour with the two of them side by side, copying a paragraph, covers the letters that change shape and there is no second hour needed.
The mistake: learning the letters by their names
Every alphabet chart gives the letters names, and the names are not the sounds. Learning that the letter is called эр does not help you read ресторан any faster, and it adds thirty three items to memorise that will never be used again outside a spelling bee.
Learn each letter as a sound inside a word instead, and skip the names entirely until they turn up naturally. The exception is worth noting: the two letters that are not sounds at all, the hard and soft signs, cannot be learned this way because they modify the consonant in front of them rather than making a noise. Those two are a grammar fact wearing a letter's clothes, and what they actually do belongs to stress that moves, and the vowels it reduces.
What the letters do not give you
An alphabet that can be read is not a word that can be said, because Russian does not print its stress and the stress decides the vowels, which is why reading aloud correctly is a separate skill from reading. Once reading is automatic the next constraint is the ending on every word, and what each one does is what each ending does, and which preposition demands it. The words themselves are built in layers that make them decodable once the joins are visible, in roots and prefixes that build whole families of words. And the gap between reading a word and catching it in speech is the widest in this subject, in real speed, and the sounds that disappear in it.