Do You Really Need to Study Grammar?

Yes, a little, late, and for different reasons than the ones you were given at school. That is the whole answer; the rest of this article is why, and what “a little” means in hours.

One boundary first. This is about explicit grammar study — sitting down with rules — and what it does for a reader. The larger argument between whole teaching systems, where grammar is only one of the differences, is a separate article.

Two things called grammar

The confusion in this debate is almost entirely a confusion between two things that share a name.

The first is the system you use in real time: whatever it is that makes a wrong word order sound wrong before you can say why. Every competent speaker has it and almost nobody can state it. The second is the description of that system — the tables, the rules, the terminology. It is a map, drawn afterwards, by people looking at the territory from outside.

Nobody disputes that the first is necessary; it is knowing the language. The question is what studying the second does for acquiring the first. And the honest answer is that it does something real and narrow, which is neither what its enemies nor its defenders usually claim.

Why rules cannot drive reading

Krashen’s account of this is the monitor: explicit knowledge is available not as a producer of language but as an editor of it, and using it requires three conditions at once — enough time, attention directed at form, and knowing the relevant rule.

Reading satisfies none of them. A page moves at two or three hundred words a minute, your attention is on what happens next, and the relevant rule is one of several hundred you would have to have available. Nobody parses a novel by rule. Whatever is happening when you read a Russian sentence and understand it, it is not the recall of a case table — and the proof is that readers who cannot name a single case ending understand the sentence anyway, having met the pattern four hundred times.

Writing is different, because the clock stops. Speaking is in between, and mostly behaves like reading. So the first practical conclusion is already clear: explicit rules pay off most in the slow, monitorable activities and least in the fast ones.

What explicit grammar actually does

Its real job is noticing. Adults do not register everything on the page; features that carry little meaning, or that their own language does not mark, slide past unseen for years. A brief explanation makes a feature visible, and once it is visible the reading starts feeding it.

The sequence matters, and it is the reverse of the classroom’s. You do not learn the rule and then apply it. You learn the rule and then start seeing the thing the rule is about, in text, dozens of times a week, until it stops needing the rule. Five minutes on Russian verbal aspect will not let you choose an aspect correctly. It will make you notice, from that day on, that the perfective keeps turning up where the action completes — and three months of noticing is what eventually lets you choose.

Read that way, a grammar book is a lens rather than a rulebook. It changes what you can see in the input you were going to read anyway.

Where it earns its place for a reader

Four areas repay explicit attention, and they have something in common: in each, the form carries meaning you cannot recover from context.

• Morphology that changes who did what. In Russian the case endings are the sentence’s wiring; a reader who ignores them can misidentify subject and object in a sentence whose words they all know. Fifteen minutes on the accusative and the instrumental is worth more than fifteen hours of guessing.
• Distinctions your own language does not make. Slavic aspect, Romance imperfect against perfect, the subjunctive, articles for speakers of languages without them. These are invisible until pointed out, and they are everywhere.
• The small closed classes. Pronouns, prepositions, conjunctions, negation, the handful of irregular verbs that carry the plot. They are a few dozen items, they occur constantly, and a table learned in an afternoon removes a permanent source of confusion.
• Sentence architecture. Relative clauses, subordination, what German does with its verbs, what a participial construction is doing in a long Russian sentence. When a sentence collapses, this is usually why — and taking one apart deliberately is the close pass, not the reading pass.

Notice what is not on the list: the exhaustive paradigm, the exceptions, the terminology, and the ability to name what you are doing. Those are the parts of grammar study that consume the most time and return the least.

What it cannot do

The evidence here is consistent enough to state plainly. Learners acquire certain features in a broadly stable order, and teaching them in a different order does not change it: a structure the learner is not ready for stays unlearned however well it is explained. Explicit instruction reliably produces explicit knowledge — the ability to complete an exercise, pass a test, correct a sentence with time to think. Converting that into the fast, automatic system takes exposure, and there is no shortcut through the rule.

Hence the familiar figure who can recite the conditional and cannot read a page. Nothing went wrong; the wrong thing was bought.

A working prescription

For a reader, the following holds up:

• Keep it under a tenth of your time. Half an hour a week against five hours of reading is about right. If grammar is taking more than reading, the ratio is upside down.
• Study in arrears. Look up the point after the text has raised it, not before. A rule that answers a question you actually have is remembered; a rule delivered in advance is a fact about a language you have not met.
• Own a reference grammar, not a course. A course dictates a sequence. A reference lets you take the one page you need, which is the only way the point above works in practice.
• One point at a time, and let it run. Take a single feature, spend twenty minutes on it, then read for a fortnight noticing it. Feature by feature, this is faster than it sounds and it sticks.
• Reread the same twenty pages of the grammar every few months. The same explanation means something different once you have met the pattern two hundred times. This is the cheapest good habit in the whole business.
• Do not shop for rules while reading. Mark it, keep going, deal with it later — the same discipline as unknown words.

Two exceptions. If an examination is coming, the examination tests explicit knowledge and you should study explicitly for it, without mistaking that for language learning. And in the first weeks of a heavily inflected language, a slightly larger dose is justified — enough morphology to make the page parseable at all, because otherwise there is no comprehensible input to be had.

How this weighs against a whole grammar-first course is the comparison of methods; where the close, deliberate pass over a difficult sentence belongs is in the method itself.