On-site links :
Comments or questions to tmason@timothyjpmason.com
Thunk 3 - The Input hypothesis
Date: Sat, 25 Sep 1999 06:59:59 +0200
From: Timothy Mason To: Foreign Language Teaching Forum FLTEACH@LISTSERV.ACSU.BUFFALO.EDU
Although Stephen Krashen's general theory is usually referred to as the Monitor Theory, the bit that puts grit under people's collars is this one. You can imagine that it goes something like this :
Babies do not speak. However, we know that they listen. They are already listening while they are in the womb, for they recognize their mothers' voices from the moment they are born. (Françoise Dolto, the French child psychoanalyst used to claim that they actually understand what their mothers are saying even before they are born, but that sounds a little far-fetched to me). After a few weeks, they begin to babble - and quite soon, the babbling begins to echo the sounds of the mother tongue. So they are listening. They keep on listening for quite some time before they actually say anything. Yet during this time, they are processing language, and getting ready to talk.
In areas without a modern monolinguistic state there may be a high density of languages, and a man or woman who moves around a lot will have to learn several tongues. It has been observed in a number of cases that the learner spends some time simply listening to what is being said, without attempting to join in. The time that he or she spends in this way may be quite long - anything up to a few months.
These two sets of observations lead some theorists to believe that there is a necessary 'Silent Period' - TPR is based on this idea - during which the learner listens to, absorbs and processes the language. How the processing is carried out is something of a mystery.
Stephen Krashen takes a further radical step : he says that listening - input - is all that is necessary to language acquisition. So long as the learner receives comprehensible input, he or she will automatically (all things being equal - see Thunk 5) become competent in the new language. In such societies as our own - literate ones - the input can be both oral and written. What is important is that the learner should want to listen or to read the material, on the one hand, and that it should be comprehensible on the other. It should not be too easy - in that case there is nothing new to learn - or too difficult, but should be pitched at such a level that the learner must work a little in order to understand.
Part of the teacher's job is to provide the scaffolding which enables the learner to understand it. Gestures, pictures, realia can all help towards this end. The material should also contain enough language that is just a little ahead of the learner's present level of competence. This is the i+1 that Stephen put me right over a couple of days ago. As we can never be absolutely sure where 'i' is with any one learner, and as we may want several learners to work from or on the same material, we cannot choose texts that are fine-tuned to the learners' needs. But that's all right ; we don't have to do so. If the texts are rough-tuned, if they are rich enough to contain a little i+1, and if they are sufficiently contextualized so that the i+2 and the i+3 do not trip the learners up, acquisition will take place.
Teachers, then, should not be challenging learners to speak all the time : a learner will talk when he or she is good and ready, and if s/he remains silent, s/he has good reasons for doing so. In any case speaking and writing do NOT lead to acquisition. Output is not a necessary part of the learning process. Of course, if the learner *wants* to speak, we are not going to stop her. But if she doesn't, that's fine.
Now, this particular thunk cuts across the grain of classroom practice ; even today, I believe that most teachers in most schools spend a great deal of their time trying to entice their pupils into production. This often begins on the first day of the first year in the FL, and term reports by language-teachers are full of remarks about how much or how little their charges participate. If Stephen Krashen is right, then this energy is mis-spent and teaching is poorly focussed.
As you may imagine, the input hypothesis has come in for some stick. (By the way, one point I'm not clear on is whether the language provided as input should be specially produced for the learner or not. Stephen Krashen has pointed out that people tend to talk in a peculiar way when addressing non-native speakers - foreigner talk - that simplifies the grammar and the vocabulary, and that this is, in some ways, similar to 'motherese' (baby-talk) - and there is some indication that special texts need to be produced, particularly in the early stages of learning. However, Stephen's present position in favour of 'whole-language' suggests to me that he may now balk at the idea of modified input).
How does input-only stand up to the beating? Well, I think that in part, it stands up fairly well : nearly all SLA experts who are bothered about what happens in classrooms - and not all of them are - agree that input is extremely important, and that learners should be spending far more time than they used to on listening to or reading in the foreign language. But criticism there has been.
First, the critics have focused on the concept of 'comprehensible input' itself, and have indicated that Krashen is not altogether clear about what he means by this term. Sometimes it appears that it is the language itself that should be comprehensible, and sometimes it is to be rendered comprehensible through external contextual clues.
Second, critics have challenged the idea that output is of no value in language acquisition. As I think the arguments here are of central interest to practicing teachers, (and as I have to go off and teach) I shall hold them back for two later posts which will be labelled 'Thunk 3 - Comprehensible?' and 'Thunk 3 - Input and Output'
Regards
Timothy Mason
P.S. Stephen Krashen kindly takes the time to comment on Thunk 2. He notes that the criticism of morpheme order studies also applies to First Language Acquisition - Roger Brown's work. I agree. He also appears to dismiss my suggestion that he believes that grammar lessons in the FL may be of some use because they provide CI. I seem to recall reading something from his pen to this effect, but as my library is still in turmoil, I can't check this. Stephen, perhaps you have changed your mind? If so, how does this fit with your remark to Ron Sheen that you do not recommend "any program for high school students which deprives students of grammar instruction"?
Université de Paris 8
This work is licensed under a Creative Commons License.