Comprehensible input,
and where flashcards fit

What Krashen claimed

In the late 1970s and 1980s the linguist Stephen Krashen set out the input hypothesis: we acquire a language in one way, by understanding messages in it (Krashen, Principles and Practice in Second Language Acquisition, 1982). He wrote the ideal input as i+1: language a little beyond your current level (i), made understandable by context, pictures and what you already know.

• Acquisition and learning are different. In Krashen’s view, rules and word lists studied on purpose (“learning”) mostly help you check what you say; the language you can use without thinking comes from input (“acquisition”).
• Speaking comes later, by itself. Output is the result of acquisition, not its cause.
• A relaxed learner takes in more. Anxiety and boredom act as a filter that keeps input out.

The idea changed teaching: more reading and listening, graded readers, and classes run in the language itself. Many learners today follow it as a method of their own, with series, podcasts and books in place of textbooks.

What later research says

Few researchers doubt that input matters most. What they question is the strong claim that understanding is enough.

• Hard to test. Barry McLaughlin pointed out that nobody can say exactly what a learner’s “i” is, or which input was comprehensible, so the claim is hard to prove wrong (Theories of Second-Language Learning, 1987).
• Understanding is not producing. Merrill Swain studied children in Canadian French immersion. After years of understanding French all day, their comprehension was close to native speakers’, but their speaking and writing were not. She argued that having to produce language makes you notice what you cannot yet say: the output hypothesis (1985).
• You learn what you notice. Richard Schmidt argued that input becomes learning only when you pay attention to the form in it (Schmidt, Applied Linguistics 11(2), 1990). A word you hear but never register stays noise.
• A balance of four strands. Paul Nation proposes a course of four roughly equal parts: meaning-focused input (reading and listening), meaning-focused output (speaking and writing), deliberate study of words and grammar, and fluency practice (Nation, “The four strands”, 2007). Deliberate vocabulary study has a place, about a quarter of the time, not all of it.

The fair reading: input is necessary and does most of the work, and a little deliberate attention to words makes it faster.

How much you need to understand

“Mostly understand” has been measured. Researchers count the share of running words in a text a reader knows. In a study of learners reading in English, about 95% known words was a minimum for reading with support, and about 98% for reading on your own; the authors linked those to about 4,000 to 5,000 and about 8,000 word families (Laufer and Ravenhorst-Kalovski, Reading in a Foreign Language 22(1), 2010). Another study found that comprehension grows steadily with every extra percent, with no sudden threshold (Schmitt, Jiang and Grabe, The Modern Language Journal 95(1), 2011).

Television is easier than books. Across 88 English programmes, the 3,000 most common word families, with proper nouns and words like um and oh, covered about 95% of the words (Webb and Rodgers, Language Learning 59, 2009).

In practice: at 95%, one word in twenty is new, about one in every two lines of a book. At 98%, one in fifty. If you are looking things up in every sentence, the input is too hard to be comprehensible; pick something easier and come back later. How many words each level is estimated to need: CEFR levels for vocabulary.

Why words met in input slip away

Words picked up from reading and listening, without trying, are called incidental learning, and it works, slowly. A meta-analysis of studies on it found that the number of times you meet a word is one of the strongest predictors of whether you learn it, though not the only one (Uchihara, Webb and Yanagisawa, Language Learning 69(3), 2019). A common word meets you again and again; a word you met once, in one episode, usually does not.

That is where a few flashcards fit: they give the words you care about the repetitions the input would not. Bring a word back from memory a few times, at growing gaps, and it stays (active recall, spaced repetition). Then every time you meet it in input again, you recognise it, which is the point.

Saving words without breaking the flow

• Watch or read first, collect second. Stopping on every unknown word turns a story into a study session. Note a word and move on; decide later which to keep.
• Keep the sentence. The line you met a word in carries its meaning, its grammar and the moment you met it. It is your example, ready made.
• Use the tools of the medium. In a video, the subtitle line; on a Kindle, the words you looked up, kept in Vocabulary Builder; on a web page, a selection; on paper, a pencil mark in the margin.
• Prefer words that come back. A word you met twice this week is a better card than a rare one you met once.

The routines for each medium: learn vocabulary from videos, remember new words from books, and sentence mining for keeping one new word per sentence.

How many words to keep

Fewer than you think. Every word you keep brings reviews over the next weeks, so five to ten words from an hour of input is plenty, and on a busy day none at all is fine: the input still counts. If your daily reviews grow longer than the time you spend watching or reading, keep fewer. How many new words a day works out the load, and the spaced repetition calculator shows yours.

How Lexpad fits in

Sources, read 25 October 2026

• Krashen, S. (1982). Principles and Practice in Second Language Acquisition. Pergamon (free PDF from the author).
• McLaughlin, B. (1987). Theories of Second-Language Learning. Edward Arnold.
• Swain, M. (1985). Communicative competence: some roles of comprehensible input and comprehensible output in its development. In S. Gass and C. Madden (eds.), Input in Second Language Acquisition. Newbury House.
• Schmidt, R. (1990). The role of consciousness in second language learning. Applied Linguistics 11(2), 129–158.
• Nation, P. (2007). The four strands. Innovation in Language Learning and Teaching 1(1), 2–13.
• Laufer, B. and Ravenhorst-Kalovski, G. C. (2010). Lexical threshold revisited. Reading in a Foreign Language 22(1), 15–30.
• Schmitt, N., Jiang, X. and Grabe, W. (2011). The percentage of words known in a text and reading comprehension. The Modern Language Journal 95(1), 26–43.
• Webb, S. and Rodgers, M. P. H. (2009). Vocabulary demands of television programs. Language Learning 59, 335–366.
• Uchihara, T., Webb, S. and Yanagisawa, A. (2019). The effects of repetition on incidental vocabulary learning: a meta-analysis of correlational studies. Language Learning 69(3), 559–599.