Stephen Krashen argued in the 1980s that people acquire language fastest when they receive comprehensible input, meaning language they can mostly understand but that still contains some new words and structures. That idea became one of the main explanations for why reading is such a powerful way to build vocabulary. Instead of learning isolated words first and using them later, learners often meet new words inside meaningful sentences, connect them to known words, and gradually absorb how they are used.
Krashen’s input hypothesis developed from research on second-language acquisition in the United States, where scholars were trying to explain why some learners improved rapidly outside formal grammar study. Krashen claimed that progress depended less on memorizing rules and more on receiving large amounts of understandable language. Reading fit that model especially well. A book or article lets learners control speed, reread difficult parts, and infer meaning from context. Spoken language moves on immediately, but written language stays in front of the learner long enough for analysis.
As this theory spread, researchers began comparing contextual learning with direct vocabulary study such as flashcards. Flashcards can help with intentional memorization, especially for beginners or for technical terms that must be learned quickly. Studies by Paul Nation and others showed that explicit vocabulary learning has a real effect, particularly when learners use spaced repetition. But researchers also found a limit: a flashcard usually teaches a basic meaning, while reading shows grammar, collocations, register, and repeated use across situations. A learner may memorize the word run on a card, yet only repeated reading reveals differences between run a company, run out of time, and run fast.
This led to an important quantitative question: how much of a text must a learner already know in order to learn from context instead of becoming lost? Research by Batia Laufer and later by Norbert Schmitt, Xi Chen, and Paul Nation focused on lexical coverage, the percentage of running words in a text that a reader already knows. Their findings suggested that 95 percent coverage is a minimum threshold for reasonable comprehension, while 98 percent gives much better conditions for comfortable reading and vocabulary growth. In practical terms, 95 percent coverage means about 1 unknown word in 20. That is still demanding, but it usually leaves enough context for the reader to guess meaning and continue.
Once this threshold became clearer, the case for extensive reading became stronger. Extensive reading means reading large amounts of relatively easy material for general understanding rather than stopping to analyze every sentence. Researchers such as Day and Bamford argued that this method works because it supplies exactly the kind of repeated, comprehensible exposure Krashen described. When learners read texts near the right difficulty level, they do not meet a word once. They meet it again and again, in slightly different contexts, until form, meaning, and usage start to connect.
Later studies tracked what repeated encounters actually do. Research on vocabulary acquisition from reading found that one meeting with a new word rarely creates full knowledge. Learners usually need multiple exposures, and each exposure may teach a different part of the word: spelling, meaning, grammatical pattern, or common partner words. Extensive reading increases the number of these encounters far more efficiently than short lists can. A novel, graded reader, or series of articles naturally recycles high-frequency vocabulary while adding a manageable number of new items.
By the 2000s, the strongest position was no longer that flashcards or reading should exist separately. Researchers such as Nation recommended a balanced approach, with deliberate study supporting broad reading. But the sequence matters. Flashcards can introduce or reinforce words, while reading provides the volume of meaningful input that turns partial knowledge into usable vocabulary. When learners choose texts at about the 95 to 98 percent coverage range and keep reading extensively, they place new words inside a system of context, repetition, and comprehension that isolated memorization cannot supply.