Read a Lot or Read Closely? Maximising Acquisition Through Reading

Two people can spend the same hour reading the same language and get entirely different things out of it. One reads forty pages of a story, understands most of it, looks nothing up and remembers the plot. The other reads a page and a half, takes apart every clause, writes down eleven words and knows exactly why the verb was in the subjunctive. Both are reading. They are not doing the same activity, and the question of which one you should be doing has a real answer.

The two modes

Extensive reading means large quantities of text you find easy, read for meaning, at speed, without stopping. The material is below the edge of your ability rather than at it; the unit is the book, not the sentence; and success is measured in pages. In the research literature it comes with a set of conditions — learners choose their own material, the material is easy, reading is its own reward, speed is fast rather than careful — set out most influentially by Richard Day and Julian Bamford.

Intensive reading means short passages studied closely. The material may be harder than you could read fluently; the unit is the sentence; you stop, look things up, notice how the grammar works, and come away with a page of notes rather than a chapter of story. This is what most classroom reading has always been.

The argument between the two is old and mostly false, because they are answers to different questions.

What extensive reading gives you

It is the only practical way to get input in the volume that acquisition needs. Words are picked up from context in small increments — a few per cent of what there is to know about a word per encounter is a common estimate — so the process depends entirely on the number of encounters. Ten thousand words of text a week produces those encounters. A page of close analysis a day does not.

It also produces things that cannot be studied at all:

• Reading speed. Fluency is a function of practice at speed, and speed only develops on material that is easy for you.
• Collocation and idiom. Which words go with which is learned by exposure, not by rule. Nobody has ever memorised their way to sounding natural.
• Grammar you were not taught. Structures met dozens of times in meaningful contexts become expectations. This is the mechanism the input hypothesis is about.
• The habit itself. Extensive reading is pleasant, which is not a small point: it is the only mode anyone sustains for years.

The evidence for it is good by the standards of educational research. Programmes that flooded classrooms with accessible books — Warwick Elley’s studies in Fiji are the best known — produced gains in reading and writing over conventional syllabuses, and extensive reading programmes since have generally reported advantages in reading rate, vocabulary and attitude.

What extensive reading does not give you

Precision. A reader who only ever reads for gist can go a long way with a private approximation of the grammar and never be contradicted, because the text does not answer back. Merrill Swain’s observation of French immersion students — years of rich input, persistent errors in production — is the standard citation here, and anyone who has read comfortably in a language for a year and then tried to write a paragraph in it recognises the pattern.

Extensive reading is also inefficient for anything you need on a deadline. If a particular tense or a particular set of two hundred words is required by an exam in April, waiting to meet them incidentally is a poor plan.

And there is a quieter failure: the word you have met forty times and still cannot define. Incidental acquisition works well for words whose meaning the context makes obvious and badly for words the context leaves vague — abstract nouns, discourse markers, near-synonyms. Those need a deliberate look, once.

What intensive reading is for

Done properly, intensive reading is not slow extensive reading. It is a different job with a different unit of work: taking a short passage and using it to notice things you would otherwise read past.

• Confirming what you inferred. A guess held for three chapters is worth five minutes of checking. Wrong guesses are sticky.
• Grammar in context. One paragraph containing four instances of a construction teaches more than a grammar page about it, because you already know what the sentences mean.
• Register and style. Why this word and not the neighbouring one; what the writer’s sentence does that yours would not.
• Feeding your review. The words worth putting into spaced repetition come out of close reading, in the sentences you met them in.

The failure mode is doing it to a whole book. Intensive reading of three hundred pages is not a method, it is an ordeal, and it produces the reader who has studied a language for four years and never finished anything in it.

The ratio

Paul Nation’s four strands is the most useful frame here: a balanced course gives roughly equal time to meaning-focused input (reading and listening you understand), meaning-focused output (speaking and writing), language-focused learning (deliberate study of vocabulary and grammar) and fluency development (doing what you can already do, faster). Three quarters of that is not deliberate study, and only one quarter is.

For someone whose access to the language is mostly a book, the practical version is simpler:

Read four pages fast for every page you read slowly.

A ratio of about four to one by time, with the extensive share as the default and the intensive share deliberately bounded, keeps the volume high enough for acquisition while leaving room to fix what volume alone would leave crooked. If you have to err, err towards the extensive side: the commonest mistake among serious learners is not too little study, it is too little reading.

One qualification. Below about A2 there is not enough language in place for extensive reading to work at all — the coverage figures make that arithmetic unforgiving — and the balance necessarily tilts towards study until there is a vocabulary to read with. The ratio above is for readers who can already get through a graded text.

Doing both with one text

The tidiest arrangement uses the same story for both, in passes, which costs less time than two texts and exploits the fact that rereading is nearly free.

1. Read the story fast, once, for the plot. Footnotes only, no dictionary. This is the extensive pass and it should feel like reading.
2. Go back to one page. Not the chapter — one page, chosen because something in it resisted you. Take it apart properly.
3. Reread the whole story a day or two later. Fast again. The page you studied will now be transparent, and so, oddly, will several you did not study.

Listening to the same text while reading it adds a fourth pass at almost no cognitive cost and fixes the pronunciation of words you have only ever met on paper. Fluency practice, in Nation’s sense, is exactly this: material you already understand, taken faster.

How much is a lot

Some numbers to calibrate against, all approximate. A short graded volume is fifteen to thirty thousand words. Half an hour a day at a modest hundred words a minute is three thousand words a day, or around a million a year — enough for the incidental effect to be large. Recommendations in the extensive reading literature commonly land near a book a week for a class, which is the same order of magnitude.

Whatever the target, count something. Pages, stories or minutes: the quantity of input is the variable that predicts progress, and it is the one most learners never measure while measuring their vocabulary app streak to the day.

The next article puts the two passes, the review and the listening into a single routine you can run on one story a week: a complete reading method, step by step.