Comprehensible input
Comprehensible input is language, heard or read, that a learner can understand. The term is central to the input hypothesis of the American linguist Stephen Krashen, who argued in the late 1970s and early 1980s that people acquire a second language in only one way: by understanding messages in it [1]. In this view, grammar study and drills play a small role, and speaking ability "emerges" on its own after enough understood input [1].
Krashen set out the idea as five linked hypotheses: the acquisition-learning distinction, the natural order hypothesis, the Monitor hypothesis, the input hypothesis and the affective filter hypothesis [1]. The theory was influential in language teaching and widely criticized by researchers, who argued that its key terms could not be tested [2][3] and that output and conversation also drive learning [4][5]. Its practical core, large amounts of understandable reading and listening, is supported by studies of extensive reading and by research on how many words a learner must know to follow a text [6][7].
Contents
Key facts
- Krashen's input hypothesis states that a learner moves from one stage to the next by understanding input that contains structures slightly beyond the current level, written i + 1 [1].
- In a Fiji "book flood" study, 380 primary pupils given 250 high-interest English story books per class progressed in reading and listening comprehension at about twice the normal rate after eight months [8].
- A 2014 meta-analysis of 34 extensive reading studies with 3,942 participants found a medium effect (d = 0.46) for group comparisons and a larger one (d = 0.71) for before-and-after comparisons [6].
- Hu and Nation (2000) estimated that readers need to know about 98 percent of the words in a fiction text to understand it without help [7][9].
- If 98 percent coverage is needed, reading English requires about 8,000 to 9,000 word families and understanding speech about 6,000 to 7,000 [10].
- Critics such as Barry McLaughlin and Kevin Gregg argued that the theory's central claims could not be tested [2][11][3].
Krashen's theory
Origins and history
The Monitor Model
Krashen first proposed the Monitor Model in the mid-1970s as a general account of adult second language performance [2]. He drew on studies showing that learners acquire grammatical features of English in a similar order regardless of their first language. Heidi Dulay and Marina Burt, for example, compared Spanish-speaking and Chinese-speaking children learning English and found approximately the same acquisition sequence for 11 English function words and endings [12].
Principles and Practice
The theory reached its best-known form in *Principles and Practice in Second Language Acquisition* (1982), which presented the five hypotheses and argued that classroom time should be spent supplying comprehensible input in low-anxiety situations [1]. Krashen described methods that do this as ones in which "class time is devoted to providing comprehensible input", with the focus on the message and not the form [1].
The five hypotheses
Five linked hypotheses
Krashen's five hypotheses build on one another [1].
- The acquisition-learning distinction: adults have two separate ways of developing a second language. Acquisition is a subconscious "picking up" of language, similar to how children learn their first language; learning is conscious knowledge of rules, "knowing about" a language [1].
- The natural order hypothesis: grammatical structures are acquired in a predictable order. For English, the -ing ending and plural -s come early and third person -s (as in "he goes") comes late [1][12].
- The Monitor hypothesis: acquired knowledge produces fluent speech, and learned rules act only as an editor, or "Monitor". The Monitor works only when three conditions are met: enough time, attention on form, and knowledge of the rule [1].
- The input hypothesis: learners progress by understanding input slightly beyond their level, helped by context and knowledge of the world. When communication succeeds and there is enough input, the next structures are supplied automatically, so lessons need not target a "structure of the day" [1].
- The affective filter hypothesis: motivation, self-confidence and low anxiety let input reach the acquisition process; high anxiety acts as a mental block that keeps input out. Krashen credited the idea of the filter to Dulay and Burt [1].
The silent period
Krashen also used the input hypothesis to explain the "silent period", in which children in a new language environment may say very little for several months while building competence by listening [1]. Krashen pointed to methods such as James Asher's Total Physical Response and Tracy Terrell's Natural Approach, in which students are not expected to speak until they decide they are ready [1].
Research evidence
Evidence from reading programs
The Fiji book flood study
The strongest classroom support for input-based teaching comes from reading programs. Warwick Elley and Francis Mangubhai gave each Class 4 and Class 5 class in eight rural Fijian schools 250 illustrated story books in English. After eight months the 380 pupils had progressed in reading and listening comprehension at twice the normal rate compared with 234 matched pupils in the standard structured program, and after 20 months the gains had grown and spread to related skills [8]. Elley later reviewed several such "book flood" studies and reported that children exposed to many high-interest books consistently learned the target language faster than children in structured, audiolingual programs [13].
Combined extensive reading studies
Extensive reading means reading large amounts of easy, interesting text for meaning. Takanori Nakanishi's meta-analysis (a statistical summary of many studies) combined 34 studies with 43 effect sizes and 3,942 participants. It found a medium effect for students who read extensively compared with those who did not, and concluded that extensive reading improves reading proficiency [6].
Vocabulary coverage
The 98 percent estimate
Input is only comprehensible if the learner knows most of its words. Lexical coverage is the percentage of words in a text that a reader already knows. Hu Hsueh-chao and Paul Nation's 2000 study, based on a regression analysis of 66 university students in New Zealand, concluded that about 98 percent coverage is needed for unassisted reading of fiction [7][9]. A 2023 replication with 104 adult learners in Sri Lanka could not fully reproduce the original results [9].
Later coverage studies
Other studies refine the figure. Schmitt, Jiang and Grabe tested 661 readers in 8 countries and found a roughly linear link between words known and comprehension, with no sudden threshold [14]. For listening, van Zeeland and Schmitt found that most participants could follow informal spoken stories at 90 percent coverage, and that non-native listeners understood well, with much less variation, at 95 percent [15]. The figures and their sources are compared in Knowledge frontier.
Word families
A word family is a base word with its inflected and derived forms (for example *hablar*, *habla*, *hablamos*, *hablante*). The estimates in the table were calculated for English; comparable counts for Spanish have not been established by the same studies.
Criticism and competing accounts
Early critiques
Barry McLaughlin argued in 1978 that the research behind the Monitor Model did not support it and proposed a more economical account based on general information processing [2]. In his 1987 book *Theories of Second-Language Learning* he reviewed the theory in detail, criticizing the reliance on conscious and subconscious processes that cannot be observed [11]. Kevin Gregg made a similar case in "Krashen's Monitor and Occam's Razor" (1984), arguing that the theory's claims were too vague to test [3].
Interaction and output accounts
Two influential alternatives give a larger role to the learner's own speech and to conversation. Michael Long showed that native speakers talking with learners change the structure of conversation, using checks, repetitions and clarification requests, and that these interactional changes are what most often make input comprehensible [5]. Merrill Swain argued that producing language makes learners notice gaps in what they can say, which pushes them to process grammar more deeply than comprehension alone does [4]. These views are covered in Output and corrective feedback.
Use in teaching
Forms in teaching today
Common practices
Several current practices apply the idea of comprehensible input.
- Graded readers and extensive reading: books written or simplified for a set vocabulary level, read in quantity for meaning [6][13].
- Narrow listening: repeated listening to several short recordings by different speakers on one familiar topic, so that vocabulary and context recur, proposed by Krashen in 1996 [16].
- Extensive listening and viewing: large amounts of audio and video at a level the learner can follow; the coverage studies suggest spoken material needs fewer known words than written text [15][10].
- Teaching Proficiency through Reading and Storytelling (TPRS): a classroom method in which teacher and students co-create stories, using personalized questions and repeated structures. Research on it remains limited [17].
- Study abroad and immersion, which supply input in daily life, are discussed in Immersion and study abroad.
A Spanish listening task
For Spanish learners, a typical input-based task is listening to a short story in Mexican Spanish about ordering food, such as a customer asking *¿Me da dos tacos al pastor, por favor?*, then hearing several other speakers tell versions of the same scene.
Relation to other methods
Comprehensible input defines the target of the Knowledge frontier: material just above the current level that can still be understood. Input supplies the encounters with words, and Spaced repetition and Retrieval practice are often used alongside it to keep new words from being forgotten. Long-term models of skill growth, such as the Stages of talent development, and level scales such as the CEFR levels describe the larger path along which this input is meant to move a learner.
See also
Pages that link here
References
- Krashen, S. D. (1982). Principles and Practice in Second Language Acquisition. Pergamon Press. www.sdkrashen.com/content/books/principles_and_practice.pdf
- McLaughlin, B. (1978). The Monitor Model: Some methodological considerations. Language Learning, 28(2), 309-332. doi:10.1111/j.1467-1770.1978.tb00137.x
- Gregg, K. R. (1984). Krashen's Monitor and Occam's Razor. Applied Linguistics, 5(2), 79-100. doi:10.1093/applin/5.2.79
- Swain, M., Lapkin, S. (1995). Problems in output and the cognitive processes they generate: A step towards second language learning. Applied Linguistics, 16(3), 371-391. doi:10.1093/applin/16.3.371
- Long, M. H. (1983). Native speaker/non-native speaker conversation and the negotiation of comprehensible input. Applied Linguistics, 4(2), 126-141. doi:10.1093/applin/4.2.126
- Nakanishi, T. (2014). A meta-analysis of extensive reading research. TESOL Quarterly, 49(1), 6-37. doi:10.1002/tesq.157
- Hu, M. H.-c., Nation, I. S. P. (2000). Unknown vocabulary density and reading comprehension. Reading in a Foreign Language, 13(1), 403-430. doi:10.64152/10125/66973
- Elley, W. B., Mangubhai, F. (1983). The impact of reading on second language learning. Reading Research Quarterly, 19(1), 53-67. doi:10.2307/747337
- Kremmel, B., Indrarathne, B., Kormos, J., Suzuki, S. (2023). Unknown vocabulary density and reading comprehension: Replicating Hu and Nation (2000). Language Learning, 73(4), 1127-1163. doi:10.1111/lang.12622
- Nation, I. S. P. (2006). How large a vocabulary is needed for reading and listening?. The Canadian Modern Language Review, 63(1), 59-82. doi:10.3138/cmlr.63.1.59
- McLaughlin, B. (1987). Theories of Second-Language Learning. Edward Arnold. archive.org/details/theoriesofsecond0000mcla
- Dulay, H. C., Burt, M. K. (1974). Natural sequences in child second language acquisition. Language Learning, 24(1), 37-53. doi:10.1111/j.1467-1770.1974.tb00234.x
- Elley, W. B. (1991). Acquiring literacy in a second language: The effect of book-based programs. Language Learning, 41(3), 375-411. doi:10.1111/j.1467-1770.1991.tb00611.x
- Schmitt, N., Jiang, X., Grabe, W. (2011). The percentage of words known in a text and reading comprehension. The Modern Language Journal, 95(1), 26-43. doi:10.1111/j.1540-4781.2011.01146.x
- van Zeeland, H., Schmitt, N. (2012). Lexical coverage in L1 and L2 listening comprehension: The same or different from reading comprehension?. Applied Linguistics, 34(4), 457-479. doi:10.1093/applin/ams074
- Krashen, S. (1996). The case for narrow listening. System, 24(1), 97-100. doi:10.1016/0346-251X(95)00054-N
- Printer, L. (2019). Student perceptions on the motivational pull of Teaching Proficiency through Reading and Storytelling (TPRS): A self-determination theory perspective. The Language Learning Journal, 49(3), 288-301. doi:10.1080/09571736.2019.1566397