Grammar and how much it matters

Almost every adult learner has spent more time on grammar than the evidence supports, and generally at the wrong point in the process.

What the evidence suggests

The picture, with appropriate caution.

Explicit grammar instruction produces measurable benefits for adults, particularly for rules that are simple, reliable and not easily noticed in input.

It produces less benefit for complex or probabilistic patterns, which appear to be acquired through exposure rather than instruction.

Explicit knowledge does not automatically become fluent production, which is why learners can state a rule and still break it in speech.

And input remains the primary driver: no amount of rule study substitutes for encountering the language.

Which suggests grammar as a supplement to input rather than a replacement for it.

What to study explicitly

The rules worth the time.

Basic word order, which is high-frequency and quickly learned.

Verb conjugation patterns for the most common tenses.

Anything that carries meaning your own language expresses differently — aspect, case, evidentiality — since these will not be noticed without being pointed out.

Rules with few exceptions.

And anything causing errors that impede comprehension rather than merely marking you as a learner.

What to leave to exposure

The other category.

Article usage, which in many languages resists rule-based description.

Preposition choice, which is largely idiomatic.

Collocation.

Register and formality.

Rare constructions.

These are learned by meeting them repeatedly, and studying rules for them produces frustration and little gain.

Timing

When the rule helps.

A rule encountered after you have met the pattern in input is far more useful than one learned in advance, since there is something to attach it to.

Which suggests reading and listening first, noticing a pattern, and then looking up the rule that explains it.

This is the reverse of the textbook order and is more effective for most adults.

And it means grammar reference material is more useful than grammar coursework.

Errors

How to think about them.

Errors are a normal part of acquisition rather than evidence of failure, and learners pass through predictable stages.

Some errors resist correction until the underlying system is ready, which is why the same mistake persists despite being corrected repeatedly.

Which means correction is useful but not immediately effective, and that persistent errors are not a sign of not trying.

And errors that impede comprehension should be prioritised over those that merely sound non-native.

Fossilisation

The genuine risk.

Errors that become permanent, typically in speakers who reached functional fluency and stopped receiving correction.

Which is an argument for periodic formal instruction even at advanced levels, and for asking people to correct you rather than assuming they will.

Most native speakers will not correct a fluent learner unprompted, out of politeness, which is why the errors persist.

And explicit attention to a specific fossilised error can shift it, though it takes deliberate effort.

Correction

What kind works.

Recasting — repeating the utterance correctly without comment — is gentle and is frequently not noticed by the learner.

Explicit correction is noticed and is more disruptive to the conversation.

Prompting the learner to self-correct appears to be effective and is less discouraging than being told.

Which means asking a tutor for a specific kind of correction, and for a list at the end rather than interruption throughout, generally works better than either extreme.

Grammar and speaking

The practical tension.

Monitoring your own grammar while speaking slows production and reduces fluency, and there is only so much attention available.

Which means speaking practice should prioritise fluency, and accuracy work should happen separately — in writing, where there is time to monitor.

Writing is the natural place for grammar practice for exactly this reason.

And accepting errors while speaking is not laziness; it is the allocation that produces progress.

A reasonable allocation

Putting it together.

The large majority of time on input and use.

A modest, regular amount on explicit grammar, focused on the rules that are worth it.

Writing as the place to apply the rules deliberately.

Periodic correction from someone competent.

And a reference grammar consulted when a pattern is noticed, rather than worked through from the beginning.