On-site links :


General Site Index

FLTeach posts

EFL Teaching

Schools & schooling

Comments or questions to tmason@timothyjpmason.com

Thunk 3 - Comprehensible?

Date: Mon, 27 Sep 1999 08:08:48 +0200

From: Timothy Mason Organization: Home To: Foreign Language Teaching Forum FLTEACH@LISTSERV.ACSU.BUFFALO.EDU

Chomsky's model of language acquisition posits a fully autonomous system which needs no more than input ; parents and other caretakers do not actually have to do anything other than providing for basic needs and talking while the child is present. Critics of this position have pointed to the fact that mothers and fathers do not talk the same way to their children as they talk to other adults. It has, in particular, been noted that mothers tend to talk to their children in short sentences, using simple grammatical structures, that they use a higher register, with rising intonation and stick to the 'here and now'. Linguists call this 'care-giver' language.

People often talk to foreigners in a rather similar fashion - 'E ope you no fright' said Mrs. Plornish, thus gaining a great reputation as a linguist*. It might be the case that input of this kind is easier to understand, and that it aids in the acquisition process. So we have two examples of how other people react to learners - children or foreigners - by simplifying their language so that it becomes more comprehensible. It might seem that it is a *natural* reaction to the language-learner.

However, we need to be careful about making such an assumption. First, look at the way mothers speak to their offspring. Some theorists have argued that mothers' language is so easy to understand that there is no reason to believe that an LAD is necessary : Chomsky is wrong. However, there are a number of problems with this position. First of all, care-giver talk is not as simple as it at first appears. Mothers who use it tend to produce a lot of 'Wh-' questions, and linguists agree that these are structurally complex in nature. So people may believe that they are producing simple language, but in fact be using quite difficult structures from an analytical point of view.

Second, it is noted that even if mothers do use this special kind of language, their children do not. There is no correlation between care-giver structures and those that the infants produce.

Third - and this is perhaps the most damning objection - not all mothers use care-giver talk. In our own cultures, upper-middle class mothers tend to speak to their babies as if they were adults. In other cultures, adults do not speak to their children at all - in fact, for an example of Americans who do not address their children directly, see Shirley Brice Heath's 'Ways with Words', a marvellous book that also offers some insights into how best to teach children to read. And yet, these children grow up to speak their mother-tongue with as much ease and grace as anyone else.

Similarly, foreigner talk is not practised by all NS when they come across NNS, nor do all NNS hear it. When they do, they tend to find it irritating and demeaning. (Vivian Cook points out that, in any case, there is no reason to believe that foreigner-talk, if it is practised, will necessarily lead to acquisition. As he says, people have been talking like this to dogs for centuries, and dogs still can't talk ; he argues that Stephen Krashen has not provided any model of the relationship between input and acquisition.)

Nor can the case cited by Krashen of the Amazonian Indian, patiently listening to others speak for weeks before he opens his mouth, be taken as evidence for some 'natural method.' His behaviour is fully culture-bound. Among a number of American Indian cultures, it is considered extremely rude to thrust one's way into a conversation, and a newcomer may be expected to remain silent for some time even if he *does* know the language. All learning, including language learning, is culturally shaped.

Nevertheless, we can agree that most learners (including babies - as anyone knows who has spent some time observing infants) do spend a lot of time listening. Stephen Krashen's belief that input is vital to acquisition may be accepted - even if, as we shall see, a lot of people do not accept that output is unnecessary. But what exactly does he mean by 'comprehensible input', or CI?

I suggested last time that critics had pointed to the fact that the term 'comprehensible' is ambiguous. Stephen Krashen was kind enough to take the trouble to answer this objection - but it is not certain that his reply would satisfy all SLA specialists. We may need to distinguish between Comprehensibility 1 (comprehension provided by linguistic context) and Comprehensibility 2 (comprehension provided by extra-linguistic context). The two may have different effects on the acquisition process : as SK acknowledges, an over-rich C2 may interfere with acquisition. Some members of a class carrying out a TPR exercise may very well perform correctly simply through watching and copying other members. And one of the reasons advanced for the inaccuracies in production of immersion pupils is that both their teachers and other pupils are too quick to understand what is being said, so that they are never obliged to either decode or encode complex utterances.

Which brings us to the third post on Thunk 3 - in which we need to consider the arguments over the (non)necessity of output. Timothy Mason

iufm de versailles


*Little Dorrit

Timothy Mason

Université de Paris 8

Creative Commons License
This work is licensed under a Creative Commons License.

Free Counter