Monday, 9 February 2015

Conscious and unconscious language learning (5)

This blog, the fifth in the series, starts to examine Stephen Krashen's influential hypotheses about language learning and acquisition. Krashen continues to be a significant influence in the world of language teaching and the term "comprehensible input", coined by Krashen, can almost be written without quotation marks, such is its currency. What was Krashen saying?

His hypotheses follow a traditional belief about two types of second language learning, although it makes unusually specific predictions and claims to be based on empirical evidence. It is a boldly expressed set of claims, notable for its attractive elegance. Its very clarity and simplicity arouse suspicion!

Acquisition and Learning

For Krashen these two terms are mutually exclusive and are used in a particularly narrow sense. First, acquisition:

   Language acquisition is a subconscious process; language acquirers are not usually aware of the fact they are acquiring language, but are only aware of the fact that they are using the language for communication (Krashen, 1982, p.10)

Elsewhere Krashen states that acquisition occurs through a process of "creative construction" (Krashen, 1978). It closely resembles the way a young child comes to master its first language. Acquisition is enabled by the existence of the innate Language Acquisition Device (LAD) à la Chomsky. He explains that acquired linguistic competence (his term) is also subconscious and that we have an implicit "feel" for correctness.

Learning is described as:

   ... conscious knowledge of a second language, knowing the rules, being aware of them, and being able to talk about them (Krashen, 1978)

Learning, therefore, relates very specifically to explicit grammatical knowledge and to activities, we presume, such as the memorisation of vocabulary lists and rules of grammar. It also clearly refers not just to a process, but to the product, the knowledge that derives from the process.

Krashen makes the huge claim that learners only become proficient at a language through acquisition, not through learning. Learning can only allow a student to monitor their utterances or utterances they hear as being correct. Put simply, people become proficient at a language by hearing or reading messages they understand. Acquisition is unconscious and will occur naturally if the input is meaningful. This is what is termed comprehensible input.

The Input Hypothesis

Krashen labelled this the Input Hypothesis.  He wrote:

   The Input Hypothesis claims that we acquire language in an amazingly simple way - when we understand messages (Krashen, 1985, p.vii)

More specifically:

   The Input Hypothesis makes the following claim: a necessary (but not sufficient) condition to move from stage i to stage i + 1 is that the acquirer understand input that contains i + 1, where "understand" means that the acquirer is focused on the meaning and not the form of the message (Krashen, 1982, p.21).

The "acquirer" understands the new element (the "+ 1") by using other, non-linguistic information e.g. contextual clues and knowledge of the world. Krashen disagrees with the commonly held view that we acquire structure (i.e. tacit linguistic competence) be learning and practising it.

In sum, the Input Hypothesis relates to acquisition, not learning and we acquire by understanding language that contains language a bit beyond our current level of competence.

So, Krashen claims that learning cannot become acquisition. In other words, what we learn from explicit grammar instruction cannot seep into our naturally acquired competence. He calls this the Non-interface Hypothesis. This means that there may be no point in teaching grammar explicitly as only being exposed to meaningful messages, as a child would be, will cause acquisition to take place.

Krashen's Monitor Hypothesis claims that learning can only help us monitor the accuracy of what we hear or say. It might allow us to correct a mistake we have made, for example.

It is worth adding that by this hypothesis input is much more important than conversation. Conversation is most useful in language acquisition because it is a way of obtaining more input. Talk is output, and as such is of little use for further acquisition. Krashen would therefore recommend lessons to contain as many input activities as possible and relatively few output tasks. This is the thinking of the TPRS (Teaching Proficiency through Reading and Stories) movement, popular with many teachers in North America. This is also why Krashen strongly favours the use of extensive reading.

The Natural Order Hypothesis

Krashen cites Dulay and Burt in claiming that both first and second language acquirers tend to acquire certain grammatical morphemes in a natural, predictable order. The order of acquisition is related to difficulty and is similar in some ways for first and second language acquirers. It follows that:

   Students may be able to learn some structures consciously, but true subconscious acquisition will come only when the students are ready (Dulay, Burt and Krashen, 1982, p.201)

Some writers believed that natural orders could be used to design syllabuses, but Krashen does not take this view. He feels language chosen as input should not be fine-tuned in the traditional way, but rough-tuned.

The Affective Filter Hypothesis

One attractive element of Krashen's hypotheses is that they incorporate affective as well as cognitive domains. Drawing directly from the work of Dulay and Burt (1973, 1974, 1978), he postulates that affective factors relate directly to acquisition, but not learning. Broadly speaking, well-motivated students are predicted by the theory to well at communicative tasks, whereas learners of high aptitude would perform well on grammatical tasks such as cloze tests. Students with poor motivation are said to have strong affective filters (i.e. much of the input will not become intake). So the filter acts as a potential barrier between input and acquired competence.

Other factors which have been held to account for successful acquisition, such as memory, intelligence or language learning aptitude, are thought by Krashen to be good predictors of conscious learning success, but poor predictors of real language acquisition. This claim is supported by the success of first ands second language acquirers in immersion or bilingual environments.


Over the years there has been a fair amount of "Krashen bashin'". Here are some of the criticisms which have been levelled at his hypotheses.

Firstly, the clear-cut distinction between acquisition and learning has been much criticised. Put simply, the distinction between acquisition and learning is impossible to prove. A good model or theory needs to be falsifiable.

Next, critics consider the argument that learning cannot become acquisition questionable. Many language learners feel that language they have learned by conscious application and rote learning can become part of their automised competence.  Learning can become acquisition, one may think. How do we know that what we can say is the result of acquisition and not, in part at least, the product of explicit learning? Krashen's distinction is, as I see it, an attractive but unfalsifiable leap of faith.

Krashen might respond that during explicit practice of patterns and memorisation we are, in fact, being exposed to comprehensible input, but input of a poor quality. Meaning-based acquisition would be more effective. For example, if we do a translation from the target language, we are getting meaningful input, even if it is less efficient than a task where we work with a source text in the target language (getting, in the process, even more input).

The fact is that we cannot be sure that what Krashen says is right and attempts to prove the hypotheses empirically are likely to be unconvincing. How can we design an experiment with two identical, parallel groups of students, with a large enough sample, reproducible, learning by using different methods, with the same teacher, in the same conditions?

As with the acquisition-learning hypothesis, the first criticism of the Input Hypothesis surrounds the lack of a clear definition of comprehensible input; some argue that Krashen never sufficiently explains the values of i or i+1.

More importantly, the input hypothesis focuses on comprehensible input as necessary, although not sufficient, for second language acquisition to the neglect of any possible importance of output. How do we know that output is not also, simultaneously, input? Perhaps through talking students identify gaps in their knowledge and then attend more carefully to relevant input.

As for the Affective Filter Hypothesis, whilst it is attractive to include motivation in a theory of language acquisition (we all know motivation is crucial), how do we know that other, non-affective factors are not just as important? My own experience over many years was that language learning aptitude was a major factor in success and distinguished clearly between pupils of similar motivation in terms of long term proficiency, not just grammar tests. Yes, all students can acquire a new language naturally, but for some it takes much longer.

The Natural Order Hypothesis has come in for criticism. For example, it does not take into account the influence of the first language when students are acquiring a new language. The order of acquisition may be different for the second language acquirer. Others have said that morpheme studies offer no indication that second language learners similarly acquire other linguistic features (phonology, syntax, semantics, pragmatics) in any predictable sequence let alone in any sequence at all.

In the final blog of this series I shall look at the practical implications of the study of conscious and unconscious language learning for the teacher.

To be continued....

H. Dulay and M. Burt (1973) "Should We Teach Children Syntax", Language Learning Vol. 23
H. Dulay and M. Burt (1974) "natural Sequences in Child Second Language Acquisition", Language Learning Vol. 24, p.37-53
H. Dulay and M. Burt (1978) "Some Remarks on Creativity in Language Acquisition", in W.C. Ritchie (ed.)
H. Dulay, M. Burt and S.D. Krashen (1982) Language Two. New York: O.U.P.
S.D. Krashen (1978) "Individual variation in the Use of the Monitor", in W.C. Ritchie (ed.) 175-183
S.D. Krashen (1982) Principles and Practice in Sercond Language Acquisition. Oxford: Pergamon Press (kindly made available online at:
S.D. Krashen (1985) The Input Hypothesis. London: Longman
W.C. Ritchie (1978) Perspectives in Neurolinguistics and Psycholinguistics. New York: Academic Press

Stephen Krashen is making previously published work available online at

1 comment: