Narrow reading: one topic, many texts

What is narrow reading, and why did Krashen recommend it?

Krashen first made the case in a short 1981 article and restated it in 2004, defining narrow reading as reading several books by one author or about a single topic of interest. His reasons are practical:

• Built-in review. Every writer has favourite expressions and a style of their own, and every topic has its own vocabulary. Stay with them and they repeat.
• Background knowledge. Each text you read on a subject makes the next one easier to follow.
• Fewer cold starts. The first pages of a new author are hard going even in your own language. A diet of short, unrelated texts means starting cold again and again.
• Motivation. He calls this the clearest advantage: people read more of what they are already into.

He also answers the obvious objection, that it narrows what you learn. Much of the vocabulary carries over between topics, he argues, and readers tend to widen their interests over time. Renandya, Krashen and Jacobs later described series reading as a special form of narrow reading, and noted how rarely it is recommended: in a survey of 119 teachers in Asia, 5.9% suggested reading many books by the same author, and 68% were not familiar with narrow reading at all.

Do related texts really bring fewer new words?

Yes, and this part is well measured. In 1989 Kyongho Hwang and Paul Nation compared newspaper stories that followed the same event over four days with sets of four unrelated stories. Assuming a reader who knew the 2,000 most common words, the share of new word families fell across the running stories from 5.8% to 4.6%, 4.3% and 3.7%. In the unrelated stories it stayed between 5.3% and 5.8%.

We ran the same kind of test on serialized fiction. For every series in the Contos library, we counted how much of chapter 2 a reader would already have met after reading chapter 1 of the same series, and compared it with reading chapter 1 of a different series at the same level instead:

Staying with a series gives you a head start of about 15 percentage points at A1, narrowing to about 7 at B2, where the stories draw on a wider vocabulary whatever they are about. Some of the advantage comes from names and the setting, but that is part of the point: they are words you no longer have to stop for.

The same holds across levels. In each language Contos has one series written at all four levels with the same characters. Reading its A1 chapters gives you 59% of the running words in its A2 chapters, against 47% from another A1 series, averaged over the same 11 languages. You can see the recycling in the opening of the Spanish one:

Does narrow reading help you learn more words?

Here the evidence is encouraging but limited, and some of it cuts the other way:

Paul Nation adds a sober note from corpus counts: across novel-length text, treating the topic words a reader learns along the way as known shifts coverage by well under half a percentage point. Narrow reading makes text easier and keeps you reading. It is not a shortcut to a large vocabulary on its own; that still takes volume, and some deliberate review.

How do you read narrowly?

• Finish a series before starting another. The second chapter of a story you know is the cheapest reading you will do all week. Each Contos series runs for six or eight chapters with the same cast.
• Follow one story up the levels. If your language has a series written at several levels, read it at your level, then at the next one, so that the cast and setting carry you through the harder text. Our page on choosing your reading level explains when to step up.
• Stay in one genre for a while. Mysteries share words for clues and suspicion, romances for meetings and misunderstandings.
• Then widen. Chang and Millett's readers didn't carry their speed to unrelated texts after three weeks, so add variety once a topic feels comfortable. Our page on learning with stories covers why repetition across chapters helps memory.