1. Introduction: Input versus Output. A general overview
In order to assess how compatible Krashen’s and Swain’s views are, it is essential to first outline the basics of each view, that is, the main tenets of their hypotheses.
As part of his Monitor Model, Krashen (1981,1982, 1985) formulated the Input Hypothesis, which claims that language input (listening and reading comprehension) constitutes the main communicative process through which we acquire a second language. Krashen believes that fluency in speaking or writing in a second language will naturally come about after learners have built up sufficient competence through comprehending input. However, it is not just any kind of input that is appropriate or effective, or as Krashen puts it, not all input will produce intake. The term “intake” is closely linked to how affective factors affect second language acquisition (SLA from now on), and this is how this author refers to the amount of input that is effectively assimilated by the learner. In such direction, he stated that it was only “comprehensible input” which would be effective for SLA. Such input is the one which is only slightly above the current level of the learner’s competence, which he represented with the simple formula I + 1, where I = input. This input is made comprehensible because of the help provided by the context. Thus, if the learner receives understandable input, language structures will be naturally acquired, according to Krashen. Therefore, the ability to communicate in a second language will emerge as a consequence of comprehensible input. Moreover, as part of his Affective Filter Hypothesis, previously put forward by Dulay and Burt (1977), Krashen argues that learners are not to be forced to produce language, as this would bring about a considerable amount of anxiety, which would cause them to develop a high affective filter that would prevent them from acquiring the target language smoothly.
In opposition to Krashen’s Input Hypothesis lies the Output Hypothesis, issued by Swain (1985). In contrast to the former, Swain’s hypothesis proposes that it is through language production (written or spoken) that SLA may be more likely to occur. This is so because, as claimed by its author, it is during language production stages that learners realise what they know and what they don’t. This may happen when a learner is trying to convey a message but his or her linguistic knowledge of the second language is insufficient to do so. It is then that the learner realises that s/he ignores some useful language structures and/or words needed to express a desired message. This issue is what Swain refers to as the “gap” between what one can say and what one would like to be able to say. And it would be on realizing this gap, that learners are motivated towards modifying their output in order to learn something new about the target language. Besides, this hypothesis asserts that language production aids learners in four different ways (Swain, 1993). The first derives from the fact that language production provides opportunities for meaningful practice, allowing the development of automatic linguistic behaviours. The second is related to that which forces the learner to switch from semantic mental processes to syntactic ones. As Krashen (1982) suggested: “In many cases, we do not utilize syntax in understanding, we often get the message with a combination of vocabulary, or lexical information plus extra-linguistic information”. Whereas in an understanding process the use of syntax may not be essential, it is in the production stages that learners are forced to consider syntactic aspects of the target language.
The third way in which language production helps learners in acquiring a L2 is through testing hypotheses, since output provides students with the opportunity to test their own hypotheses, and withdraw their own conclusions. This third aspect is closely related to the fourth one, which deals with the responses of other speakers of the language, especially native ones, which can give learners information on how comprehensible or well-formed their utterances are.
It must be said that, despite all emphasis being laid on output, Swain admits that output is not solely responsible for SLA.
To sum up, where Krashen sees input hugely responsible for language acquisition, Swain considers output; where the latter claims language production to be of utter importance, the former regards it as not necessary, as something that should not be forced, since it will appear naturally after a certain amount of comprehensible input.
Before continuing with this article, it must be noted that no distinction between the terms “learning” and “acquisition” is being made, as most authors do not consider it amongst their theories of SLA.
2. Input and Output: rejecting or complementing each other?
In this section we will be looking at how the terms input and output have been dealt with by other authors, and whether these support either Krashen’s or Swain’s views of SLA, and in what ways they do so. We will also consider if these two concepts are opposites or simply two sides of the same coin.
Originated by the work of Chomsky (1957), the Generative Paradigm arose as a clear opposition to the structural approach to linguistics. And, although this paradigm did not deal with how languages were learned, it did however consider the term output within one of its main features, given the importance of the creative nature of language use within this paradigm. It is here where output is first remotely considered, as creativity calls for production and this may be understood as the very core of output. Moreover, according to Chomsky, creativity has to come hand in hand with compliance to rules, as any type of creation ought to take part within a framework governed by a set of rules. It is here where Swain’s hypothesis may receive support, since she believes that production leads learners to consider syntax as such, which can be considered as that set of rules which governs a particular communicative framework.
Moving now towards the field of SLA specifically, we find three different theories that aim at explaining how language is acquired, and these are the behaviourist, nativist and interactionist theories. We will focus firstly on behaviourist and nativist views.
As far as behaviourism is concerned, a language is learned by the creation of a series of habits which are acquired by imitation. Thus, we can find both input and output in this theory, since learners imitate (output) something that has previously been assimilated (input). As regards nativist theories, while learning a language, learners are constantly forming hypotheses based on the information received (input). However, they also test these hypotheses through speech (output) and comprehension (input).
So we can see how, within behaviourist theories, output is considered as imitation, which accounts for Swain’s argument related to the creation of automatic linguistic behaviours. From a nativist point of view, the Output Hypothesis is also backed, since it would be through speech that learners test what they know and what they don’t. In the same way, both behaviourist and nativist theories stand beside Krashen’s Input Hypothesis, as they both explicitly consider output to be a natural consequence of input. So it is at this point that we can see how these two seemingly opposite hypotheses start complementing rather than denying each other’s validity.
Insofar as interactionist theories are concerned, they regard the acquisition of a language as the result of the interaction between the learner’s mental process and the linguistic environment (Arzamendi, Palacios and Ball, 2012, p.39). It is here where we can also appreciate a combination of both input and output, working as one. Interactionist theories believe in interaction as the main reason of language acquisition. It is therefore a clear example of the validity of both input and output hypotheses.
The importance of interaction as the cause of language learning is supported by a study carried out by Pica, Young and Doughty (1987), which proved up to a certain point that Krashen’s comprehensible input was less effective than interaction, which implies not only input but also output.
In the same direction, Ellis (1985), defined an “optimal learning environment”, to which he bestowed several features related to output as well as input. He talked about the importance of exposure to a great deal of input, which comes hand in hand with Krashen’s Input Hypothesis, but he also stressed the significance of output. He does so by highlighting the need for learners to perceive L2 communication as something useful (meaningful communication, as Swain puts it). Besides, the opportunity for uninhibited practice in order to experiment is also stressed by this author. In this last statement we can see not only Swain’s view of output as a means of language hypothesis testing, but also Krashen’s importance of a low affective filter, since inhibition would clearly restrain a learner’s linguistic performance. In this way, not only Swain’s and Krashen’s hypothesis look more alike, but they start needing each other in order to exist flawlessly.
Within sociolinguistic models of SLA, input is clearly dealt with, especially within the Nativisation Model (Andersen, 1979). This model emphasises the importance of input and how learners internalise the L2 system. According to this model, learners interact with input in two ways, they adapt input to their view of the L2 and they adjust their internal linguistic system to suit that particular input, in order to acquire L2 form features. This theory clearly matches the importance Krashen gives to input as the means of acquiring a language.
If we move onto linguistic models of SLA, we will find that Hatch (1978) deals with the importance of both input and output in his Discourse Theory. Hatch places meaning negotiation at the core of his theory. In this way, input gains importance, as L2 advanced or native speakers adjust their speech when addressing an L2 learner. Thus, input becomes comprehensible for the learner, which is a key factor in Krashen’s hypothesis. However, this theory also states that the natural way of acquiring a language is a consequence of learning how to hold conversations. And it is in this sense that output becomes important too, since in order to engage in conversation, which involves language production, it is as essential as understanding. Also, and according to this SLA theory, the learner uses vertical structures to construct sentences, which implies borrowing chunks of language from preceding discourse to which s/he adds elements of his or her own. In this way, learners are experimenting and testing their hypotheses on the language, which is one of the ways in which output leads to SLA, according to Swain (1985, 1993).
And this is how we arrive at Swain’s Output Hypothesis, which is a linguistic model, and Krashen’s Input Hypothesis, which constitutes a cognitive model for SLA. Although the main tenets of one seem to reject those of the other, we have seen how, far from opposing, they complement each other.
3. Reconciling Krashen’s input and Swain’s output views
It is time now to tackle the main purpose of this assignment, reconciling Swain’s and Krashen’s views. In order to do so, we will see how both hypotheses are right but incomplete at the same time.
The Input Hypothesis claims that fluency in speaking or writing in the L2 will naturally emerge after learners have achieved sufficient competence through comprehensible input (Wang and Castro, 2010). However, the studies of Tanaka (1991) and Yamakazi (1991), in Wang and Castro (2010), reveal that although input facilitates greatly the acquisition of vocabulary in the target language, it does not cater for the acquisition of many syntactic structures. Therefore, comprehensible input is essential but not sufficient in achieving SLA. It is the Output Hypothesis that takes care of this flaw. According to Swain (1993), producing language would force learners to recognise what they do not know or know only partially, which she calls the “gap” between what learners can say and what they want to be able to say. In her opinion, when encountered with such gap, learners can react in three different ways. One would be to ignore it. Another to search in their own linguistic knowledge to find or construct the answer; and the last one is to identify what the gap is about and then pay attention to relevant input which may cater for this lack of knowledge. This third response establishes a relationship between input and output that benefits SLA. As a result of this, learners are more likely to enhance their input processing capability because their output has focused their attention on the need to do so. (Swain, 1993)
We can see now how Swain’s Output Hypothesis accepts input as an important part of SLA, whereas Krashen’s view is slightly more slanted. In his work Comprehensible Output (1998), in which he assesses the effectiveness of comprehensible output (CO), Krashen criticizes CO as a means of acquiring a L2. Amongst other issues or flaws in Swain’s hypothesis, he argues that being forced to speak, as part of CO, leads to discomfort, that is to say, to anxiety on the part of the learner. According to Young (1990) and Laughrin-Sacco (1992), in Krashen (1998), foreign language students find speaking to be the highest anxiety-causing activity. Moreover, he puts forward what Price (1991) stated, that not being able to communicate effectively leads to a great deal of frustration.
These two arguments clearly support Krashen’s Affective Filter Hypothesis. Anxiety and frustration may cause low motivation and little self-confidence, which may provoke high affective filters on the part of the student and, hence, little intake may take place.
Although Krashen has made a good point on how CO may have less advantages than it seems to, he also grants it a place in his Monitor Model, as part of his Monitor Hypothesis. According to Krashen (1985) the “monitor” is an internal editing device that may work before or after output taking place. In order to do so, the learner has to know the appropriate rules of speech. Despite the lack of supportive research evidence for this hypothesis, if we take Krashen word by word, we understand that we edit or correct what we utter before or after we do so. In this way, if we do it before, we are using inner knowledge in order to edit something we are about to produce; if we do it after, we are correcting a mistake, which is basically testing a hypothesis that has proven to be wrong. After doing so, we can re-arrange it in our head to correct it or simply focus our attention on the knowledge we need to acquire to be able to produce a hypothesis which turns out to be right. It is here where we see two of the advantages of output mentioned by Swain: testing a hypothesis and recognising what one does not know but needs to.
It is clear by now that both hypotheses are neither wrong nor complete. In any case, they can complement each other in order to produce a more integral hypothesis.
As a final conclusion, one might propose certain guidelines so as to put an end to this unsettling disagreement.
Firstly, a certain amount of comprehensible input is necessary before producing any kind of output whatsoever. This might be more important with young learners than with adults, since the latter have a better control over affective issues. Young learners however, apart from not having enough linguistic knowledge so as to reflect on their own output, they might become more anxious by being forced to speak, if it is not done in a careful way.
Secondly, the use of either input or output may vary according to the type of language acquisition we are trying to achieve. If the focus is on syntax, we shall use output strategies, which allow for a greater amount of reflection and self-correction. However, if we are working on vocabulary acquisition, an input approach will probably prove to be more effective.
Finally, learners ought to make use feedback that they can obtain from other speakers of the language, and this is achieved only through language production. Other speakers’ responses will provide learners with informative feedback on the comprehensibility and/or accuracy of their utterances. In a language learning environment, this feedback may come from the teacher or from other learners.
If we follow these guidelines, drawn from both Krashen’s and Swain’s arguments, the ability to produce the language will not only be the result of language acquisition, as the former argues, but also the cause, as Swain believes.
- Arzamendi, J., Palacios, I. and Ball, P. (Eds.) (2012). Second Language Acquisition. FUNIBER.
- Krashen, S.D. (1981). Second Language Acquisition and Second Language Learning. Oxford: Pergamon.
- Krashen, S.D. (1985). The Input Hypothesis. Issues and Implications. New York: Longman.
- Ellis, R. (1985). Classroom Second Language Development. A Study of Classroom Interaction and Language Acquisition. Oxford: Pergamon Press.
- Krashen, S.D. (1982). Principles and Practice in Second Language Acquisition. Oxford: Pergamon.
- Krashen, S.D. (1998, June). Comprehensible Output. System, 26(2), 175-182. Obtained on 11th February 2013, from http://www.sdkrashen.com
- Swain, M. (1993, October). The Output Hypothesis: Just Speaking and Writing Aren’t Enough. The Canadian Modern Language Review, 50(1), 158-164.
- Wang, Q. and Castro, C.D. (2010, June). Classroom Interaction and Language Output. English Language Teaching, 3(2), 175-186.