The Input Vs Output Fallacy English Language Essay

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

Dr Krashen's article on the relative values of the comprehensible output and the comprehensible input hypotheses in last week's issue of Share Magazine, contains some interesting discussion points.

The comprehensible output hypothesis holds that language acquisition takes place when learners fail to transmit messages due to poor command of the language being learnt, but re-formulate them as many times as necessary and finally succeed in producing an accurate, appropriate message. By doing this, learners acquire the items thus produced. (Swain, 1985). The comprehensible input hypothesis states the value of input as opposed to that of output (Krashen, 1985) in the process of language "acquisition", defined as a sub-conscious process, different from formal learning. The value of comprehensible input far exceeds that of comprehensible output, to the point that "...humans acquire language in only one way - by understanding messages..." (Krashen,

The discussion of one point of view versus the other seems slightly pointless, as we know that attributing the success of learning or acquisition primarily to only one factor might exclude many others which are equally relevant and play important roles in the process. However, we would like to discuss Dr Krashen's comparison of both hypotheses and his conclusions, namely:

"The comprehensible output hypothesis has numerous difficulties.

- Output and especially comprehensible output is too scarce to make a real contribution to linguistic competence.

- High levels of linguistic competence are possible without output.

- There is no direct evidence that comprehensible output leads to language acquisition.

In addition, there is some evidence that suggests that students do not enjoy being "pushed" to speak."

To support his assertion that there is scarce comprehensible output among EFL learners, he cites quantitative studies which demonstrate that instances in which learners re-formulate their messages in the way described in the CO hypothesis are extremely rare. In these studies, out of thousands of instances of erroneous utterances produced by learners, only a few were readily repaired by the learners themselves, when their teachers or interlocutors asked for clarification of the inaccurate or otherwise faulty utterances. This is supposed to demonstrate the scarcity of CO and therefore minimise its value, when actually it is only demonstrating that the very concept of CO (error-request for clarification-repair-acquisition of the new form) is false, because it does not take into consideration the three stages of error described by Pit Corder (1973): pre-systematic, systematic and post-systematic. At pre-systematic stage, learners are not even aware that they have made an error; at systematic stage, they can self-correct the error only if it is pointed out to them and at post-systematic stage, learners can readily correct their errors with very little prompting from the interlocutor - maybe a raised eyebrow - or no prompting at all. What the studies cited by Dr Krashen demonstrate is that there are always very few errors at post-systematic stage and therefore the whole concept of CO is reduced to this domain.

Why should there be very few examples of post-systematic errors, and consequently CO, at experiments or lesson observations where learners interact with native speakers or their teachers for the researchers to count these instances of self-repair? The answer to this question depends largely on the level of the conversation relative to the learners' level of command of the foreign language. To self-repair an utterance, the learner needs to have recourse to the correct form, which is impossible to fetch from the thin air or devise creatively. In other words, the learner needs to "know" what to produce. When made to perform above this level, he/she will find self-repair impossible.

In our own study of errors (1978 - 2001) we were forced to make a distinction between compensatory errors, those made by learners in their attempt to fill a communication gap with language elements they did not possess and non-compensatory errors, those involving language they had been taught, but had not yet learnt. The distinction was not easy at upper-intermediate or advanced levels, but it could very often be clarified through authoritative explanation (Pit Corder, of the error, that is, by discussing it with the learner, who normally provided the necessary information. Whereas compensatory errors tell the researcher a lot about the learners' capability and style for coping with language limitations, non-compensatory errors add information as to what has been learnt and about the stage of the learners' errors. Within this paradigm, nobody could expect a learner to repair a compensatory error, because compensation is just that and will never produce CO as defined by Swain. In fact, in Tarone and Liu (1995), Liu admits that one of the subjects in their experiments "may have been pushed beyond the limits of his language competence" by the researchers themselves.

A quantitative method does not seem to be fully reliable here, as we are told that the studies cited took place in different countries, with learners of different ages and backgrounds and there does not seem to have been any attempt at controlling important variables of the experiments, such as methods used for teaching these subjects, environmental and cultural aspects, and the level of demand placed on the learners' production, to mention but a few. Scarcity of CO, in these studies, may have been caused by a number of factors which, if modified, might produce a completely different type of production. This does not expose the scarcity of CO but it may show that teachers and researchers fail to create the necessary conditions for CO to emerge or that they are relatively unaware of the processes behind error production and their implications.

To support his claim that language acquisition is possible without output, Dr Krashen mentions a laboratory experiment concerning the learning of a few lexical items, a study on second (not foreign) language acquisition and several other experiments, all conducted under different circumstances and by an assortment of methods, which resulted in learners showing "a modest vocabulary gain" as measured in terms of words per minute, versus those who had been exposed to conversational interaction with a native speaker. Again, we may question the methods used, the focus of the experiments, the lack of control of loose variables and even the advisability of offering varied instances of research as an exploration of such a comprehensive concept as "acquisition without output". The quality and level of this acquisition should also be measured with reference to appropriacy and register, not merely by counting words or instances of correct usage of grammatical items. Otherwise, we might conclude that input is more important than output only in terms of achieving accuracy.

To further support the value of the comprehensible input hypothesis, Dr Krashen claims that there is evidence suggesting that students do not enjoy being "pushed to speak". This seems to indicate that we should not push students to speak, not that output is not important. Methods should respect the learners' pace, learning styles and personality and never push them to do anything, let alone talk. We might also state there is some evidence that adult learners do not appreciate not being asked to speak, as they are usually eager to acquire some conversation skills in as short a time as possible and that pushing learners to do something is just as inadvisable and authoritarian as preventing them from doing it.. Whether they are inhibited or not in class, or ashamed of making mistakes, or embarrassed to speak in a foreign language, or ready for a particular type of production, depends largely on their personality, cognitive hypotheses and the teacher's ability to create a supportive, relaxed classroom atmosphere.

The claim that "providing more comprehensible input seems to be a more reasonable strategy than increasing output", seems a rather unnecessary remark to make, since it is to be demonstrated how we could possibly increase output with less input, as input necessarily provides the source of elements for language production. For years, we have known that learners should be exposed to listening and reading materials slightly above their level of production, so that they can have a whole wealth of input as a valuable resource.

The input-versus-output discussion appears to atomise the problem of growing into the command of a language, where no element can be singled out as dominant. Input and output interact and this goes hand in hand with theories of learning which place great emphasis on the role of social interaction, the creation of meaning and the formation of systems during language learning. Meaning, on the other hand, is created with reference to the language user's previous knowledge, view of the world and personal needs and as nothing succeeds like success, learners find confirmation of their hypotheses about the foreign language by trying them out and confirming or correcting them, a process by which they also build up their self-esteem. This goes beyond trial and error, as Piaget (1960) demonstrated, with each trial producing a re-accommodation of the whole system, not just of the item being tried out. Although periods of intake of information without overt performance are necessary, language is ultimately a social tool and it cannot be consolidated until it has been tried and tested in social interaction. An exploration of the nature of the necessary interplay of input and output and how it takes place in different environments and with different individuals would shed more light on the process of language learning and acquisition.

Krashen, S.D. (1985) The Input Hypothesis: Issues and Implications. New York: Longman Inc.

Piaget, J. (1960) The Psychology of Intelligence. New York: Humanities Press Inc.

Pit Corder, S.(1973) Error Analysis. In Allen, J.P.B. and Pit Corder, S. (Eds.) The Edinburgh Course in Applied Linguistics, Vol.4, pp 122-154. Oxford: Oxford University Press

Swain, M. (1985) Communicative competence: Some roles of comprehensible input and comprehensible output in its development. In Gass, S. and Madden, C. (Eds.), Input in Second Language Acquisition, pp. 235-256. New York: Newbury House.