Krashen's input hypothesis and the Monitor Model

The five hypotheses concern acquisition and learning, the monitor, natural order, input, and the affective filter. This page explains the model from Principles and Practice in Second Language Acquisition (1982), especially chapter II. It also describes its limits. Start with the short definition if the term is new to you.

Who Krashen is and when he wrote

Stephen Krashen is an American linguist and education researcher. He worked for many years at the University of Southern California. His theory grew from research in the 1970s, including studies of the order in which learners acquire English grammar. The three key books are:

• Second Language Acquisition and Second Language Learning (Pergamon, 1981)
• Principles and Practice in Second Language Acquisition (Pergamon, 1982), which Krashen later made free to read online
• The Input Hypothesis: Issues and Implications (Longman, 1985)

With the teacher Tracy Terrell he also wrote The Natural Approach (1983), which turned the theory into a classroom method. Krashen later wrote widely about reading, for example The Power of Reading.

1. The acquisition–learning hypothesis

Krashen said that adults have two separate ways to develop ability in a second language.

• Acquisition is subconscious. It is similar to the way children pick up their first language. You are not aware that it is happening. You only notice that you are using the language to communicate. The result is a "feel" for what sounds right.
• Learning is conscious. It means knowing the rules, being aware of them, and being able to talk about them. In everyday language, it is "knowing about" a language: grammar.

Krashen argued that adults can still acquire. The ability does not disappear at puberty. His most controversial claim was that learning does not turn into acquisition. In his view, practising a rule many times does not make it part of your subconscious system. Researchers call this the "non-interface" position.

2. The monitor hypothesis

If acquisition and learning are separate, how do they work together? Krashen's answer: acquisition starts your sentences and gives you fluency. Learned knowledge has one job. It acts as a monitor, an editor that can change the form of a sentence before or after you say or write it.

Krashen (1982) said the monitor can only work when three conditions are met:

1. Time. You need time to think about the rule. Normal conversation usually does not allow it.
2. Focus on form. You must be thinking about correctness, not only about what you want to say.
3. Know the rule. You must actually know the relevant rule, and nobody knows them all.

This explains a common experience. Many learners write more accurately than they speak, because writing gives them time to check. Krashen also described "over-users" of the monitor, who speak hesitantly because they check everything, and "under-users", who rely only on feel.

3. The natural order hypothesis

Studies in the 1970s found that learners of English tend to acquire certain grammar endings in a similar order. Dulay and Burt (1974) studied Chinese-speaking and Spanish-speaking children learning English. Both groups showed approximately the same sequence. Bailey, Madden and Krashen (1974) reported a similar order in adults. For example, the -ing ending ("he is playing") and the plural -s tend to come early. The third-person -s ("she lives") and the possessive 's tend to come late, even though they are easy to explain.

Krashen did not recommend teaching grammar in this order. His point was the opposite: the order is not controlled by teaching, so a grammar syllabus cannot force it. This is why a learner can "know" the third-person -s rule for years and still drop it in fast speech.

4. The input hypothesis

This is the heart of the theory, and the source of the term comprehensible input. In Principles and Practice, Krashen stated it in four parts. In plain English:

1. The input hypothesis is about acquisition, not learning.
2. We acquire by understanding language that contains structures a little beyond our current level. Krashen wrote this as i+1. We can understand it because of context, our knowledge of the world, and other non-linguistic information.
3. If the input is understood and there is enough of it, i+1 will be present automatically. So teachers do not need to aim at the next grammar point on purpose.
4. Speaking ability emerges on its own as a result of acquisition. It is not taught directly. Early speech is not accurate, and accuracy improves with more input.

Krashen described understanding input as a necessary condition for acquisition, but not a sufficient one. The affective filter (below) is the other condition. He supported the hypothesis with several kinds of evidence: the simplified way adults talk to young children, the silent period some children show when they move to a new country, and studies in which input-based classroom methods did well.

5. The affective filter hypothesis

"Affective" means related to feelings. Krashen borrowed the idea of a filter from Dulay and Burt and used it to explain why two people with the same input can make different progress. He grouped the relevant feelings into three types: motivation, self-confidence, and anxiety. If a learner is anxious, unmotivated or insecure, a mental "filter" goes up, and input does not reach the part of the mind that acquires language, even when it is understood.

The model supports relaxed, interesting input with little pressure. But the filter is a proposed explanation, not a measured barrier inside the brain. Anxiety is not a learner's fault. This metaphor must not become a way to blame a learner when a method fails.

The main criticisms

The Monitor Model was one of the most discussed theories in second language acquisition, and also one of the most criticised. The main objections were:

The key terms are hard to test

Barry McLaughlin's Theories of Second-Language Learning (1987) argued that "acquisition" and "learning" are not defined in a way that allows a clear test. If a result fits the theory, it counts as acquisition; if not, it can be called learning. A theory that cannot be shown wrong is weak as science. Kevin Gregg (1984) made a detailed critique of the same kind and rejected the strict separation of acquisition and learning.

i+1 is vague

Nobody can measure a learner's current level i, or say exactly what +1 is. Lydia White (1987) also pointed out that some errors are hard to fix from input alone, because input shows what is possible but not always what is impossible. White did not reject input. She argued that the hypothesis needed to be made more precise.

Input may not be enough

Merrill Swain (1985) studied French immersion students in Canada. They had years of comprehensible input and understood French very well, but their speaking and writing were still less accurate than native speakers'. Swain argued for a role for producing language, or "comprehensible output", in developing more accurate language. Later output research expanded this account. Richard Schmidt (1990) argued that learners must consciously notice features of the input for them to be learned. Both ideas challenge the claim that understanding is the only route.

Grammar teaching does help

Later research reviews, such as Norris and Ortega (2000), found that focused grammar instruction produces clear gains, and that explicit instruction tended to work better than implicit instruction in the studies they reviewed. This sits uncomfortably with the claim that learned knowledge only serves as a monitor. Researchers still debate how long these gains last and whether they appear in spontaneous speech.

The natural order and the filter are less firm than claimed

The morpheme-order studies looked at a small set of English grammar endings, and later researchers questioned their methods and how far the findings generalise. The affective filter is a useful metaphor, but it is difficult to measure directly.

What still holds up

Many of Krashen's specific claims are disputed. His core idea is widely accepted: learners need large amounts of meaningful input that they can understand. Few researchers today think you can learn a language without it. The disagreement is about whether input is enough, and what else helps.

The next page looks at that question with the evidence: does comprehensible input work? If you want a practical summary of how input and study fit together, Jibber Jabber's guide to acquisition and deliberate study takes a combined view, and its page on why comprehensible input works gives the practical case. Definitions of every term on this page are in the glossary.