How to Remember Grammar Rules When You Already Know Them
There is a specific kind of frustration in explaining a rule correctly to someone else and then breaking it two minutes later in your own sentence. It feels like a memory failure. It is usually not one, because the knowledge is plainly still there.
Most learners who complain about forgetting grammar have not forgotten anything. Ask them to explain the rule cold and they will do it accurately, sometimes better than a native speaker could. The rule is intact. What is missing is any route from the rule to the sentence they are producing in real time.
That distinction changes what you should do about it. Rereading the grammar chapter strengthens something you already have, which is why it feels productive and changes nothing.
Knowing a rule and using it are different jobs
Stating a rule is a piece of knowledge you retrieve consciously, with as much time as you need. Applying it mid-sentence has to happen in under a second, while you are also choosing words, tracking what the other person said and worrying about your accent. Under that load, anything that requires deliberate lookup gets skipped.
Stephen Krashen made a version of this argument decades ago, separating language you have acquired from language you have consciously learned and can only monitor with when there is time. The strong form of his claim is disputed and has been for years, but the everyday observation behind it survives all of the criticism: rules you can only state do not show up when you are speaking quickly.
So the target is not remembering the rule better. It is building a version of it that fires without being called.
Store the pattern, not the statement
A rule written as a statement can only ever be recited. A rule stored as a sentence you can say gives you something to reach for directly, which is why proficient speakers rely far more on ready-made chunks and collocations than on live rule application. They are not computing the preposition. They already have the phrase.
In practice that means keeping one short, real sentence per rule, ideally one you would plausibly say. "I have been waiting since Tuesday" does more for your present perfect than the paragraph explaining when to use since and for, because it is retrievable at speed and the rule is baked into its shape.
Keep the sentence in the first person and in a situation you actually find yourself in. A sentence about a train timetable is harder to reuse than a sentence about your own week.
Card formats that transfer, and one that does not
Most grammar cards fail for the same reason: they ask you to recognise or recite, and then people are surprised that recognition is what they get. A card should ask you to produce the form under mild pressure.
| Card format | What it asks | What it builds |
|---|---|---|
| "When do we use the present perfect?" | Recite the rule | Almost nothing you can use in speech |
| Fill the gap in a sentence | Choose the right form in context | A start, though recognition still helps you |
| Translate a sentence from your own language | Produce the whole structure from scratch | The closest thing to speaking |
| Correct a wrong sentence you actually wrote | Notice the error and repair it | Attacks the specific mistake you make |
| Say the sentence out loud from a prompt | Produce it at speed, aloud | The version that survives conversation |
Mine your own mistakes instead of studying rule lists
A grammar textbook covers what learners in general get wrong. Your errors are a much shorter and far more useful list, and they repeat with remarkable consistency: most people have somewhere between five and fifteen structures they keep breaking, and everything else is fine.
Collecting them is the whole trick. Every time someone corrects you, or you catch yourself, keep the sentence you got wrong next to the corrected version. Ten of those are worth more than a grammar course, because each one is proven to be a live problem rather than a hypothetical one. The common mistakes at B1 and B2 are a decent place to check whether yours are the usual suspects.
Richard Schmidt argued that noticing a gap between what you said and what you should have said is what lets the correct form get in. Writing the pair down forces the noticing to happen rather than hoping it does.
Which rules deserve explicit study at all
Not every rule is worth a card. Some are better left to exposure, and spending review time on them is a quiet waste.
- Worth studying: rules where your first language pushes you the wrong way. These do not fix themselves with exposure, because you are not hearing the error you make.
- Worth studying: rules that change meaning rather than just form, since a mistake there gets you misunderstood rather than merely marked.
- Not worth studying: anything you already produce correctly most of the time. Review it and you are polishing what works.
- Not worth studying: rare exceptions and edge cases. They are memorable precisely because they are rare, and they cost you nothing when you get them wrong.
Spacing works on grammar too, with one adjustment
Structures fade like words do, and they respond to the same treatment: repeated attempts spread over days rather than crammed into one sitting, with real retrieval at each attempt. What changes is the unit. A vocabulary card has one right answer, while a grammar pattern has infinitely many, so repeating the same sentence eventually teaches you that sentence and nothing more.
The adjustment is to vary the sentence while keeping the structure. If the pattern is a conditional, produce it about a different situation each time you meet the card. It is slower per repetition and it is the only version that generalises.
Where MindDory fits
Grammar mistakes are hardest to collect exactly when they happen, which is mid-conversation. If you are practising with an AI assistant, MindDory can capture the correction over MCP as you go: your sentence, the fixed version and the structure behind it, without breaking the conversation to open an app.
From there it is the same machinery as vocabulary. Cards carry the sentence and a cue rather than a rule statement, and sessions are built from what is closest to slipping rather than what you added last. If your goal is the jump where grammar stops being the bottleneck, the guide on going from B1 to C1 covers where that work sits in the wider picture.
The rule you keep breaking is not one you have forgotten. It is one you have only ever stored as a sentence about the language, rather than as a sentence in it.