Stephen van Vlack

Sookmyung Women`s University

Graduate School of TESOL

Second Language Learning Theories

Spring 2007

Week 7 - Answers

M&M, Chapter 6: Input and interaction in second language learning

1. How can we define the term input? What kinds of English input are Koreans exposed? What are the sources of this input?

Input is anything that the students are exposed to and maybe expected to form some reaction to whether conscious or subconscious. From a neurological view, input is anything that will or can activate the sensory system. Following this very broad definition, we need to make a distinction between linguistic input and extra-linguistic input or non-linguistic. Obviously linguistic input is going to be input which is coming in the form of some acceptable language. From an SLA point of view this can be spoken (auditory) input in the target language and/or (as well as) written (visual) input in the target language. In addition, we have extra-linguistic input which is input which often accompanies the situation in which the linguistic input is delivered or perceived. From a pedagogical point of view extra-linguistic input is there to heighten or enhance the linguistic input being given. We do not, however, always need to have linguistic input. Non-linguistic input is primary input given that lacks overt linguistic forms, such as photos or smells (anything that activates any one of five perceptual systems but do not contain linguistic forms). We can generate linguistic output from non-linguistic input of all kinds.

An interesting aspect of this basic input idea is the basic idea that we are receiving all sources of input simultaneously. It is like being able to watch many different TV channels at the same time. In the interactional model as we will be discussing this week, it seems to be of the utmost importance that different types of sensory input are produced and perceived at the same time. Based on the simple concept of co-occurrence and association the richer the context (the more elaborate the input is in respect to different perceptual information being processed simultaneously) the better one will be able to encode and retrieve the target linguistic input. Such `rich` input would seem to be a necessary prerequisite (or at the very least a big help) for, as we talked about last week, form to function mapping. So, really, input is pretty much anything that people are exposed to which is perceivable based on our physical/cognitive mechanisms. It bears repeating that linguistic input is only a small part of the possible input which students can be supplied with (exposed to). And to sum this up we can say that extra-linguistic input and/or non-linguistic input is a necessary accompaniment for linguistic input. It is important because without this extra-linguistic input output will be virtually impossible.

Now, if we look at the Korean situation as related to English teaching we find that Koreans tend to be exposed almost exclusively to purposely tiny amounts of linguistic input and this predominately in the written form. Most classrooms do not provide any extra-linguistic input and one thing about written linguistic input is that it also often ignores or fails to integrate extra-linguistic input Integration of extra-linguistic input in written forms needs to be overt. This means that in general Korean learners are getting a very small aspect of the entire input spectrum. In the Korean situation sources for input tend to be textbooks, sometimes teachers, and rarely other students. This again shows an overall lack of diversity or richness in the input situation in Korea as regards English. It also probably not be unfair to say that the total amount of input which students are exposed to in and surrounding the target language is minimal. Thus, we have a problem.

This problem becomes more serious when we begin to consider how it is that the brain actually works. Input is like fuel for learning in the brain. The sensory systems, as one part of the human cognitive chain, are essential in driving that chain. None of the other systems work without a prerequisite amount, or threshold level, of input coming in. This, put in the simplest possible way, could indicate that many Korean learners as related to English could best be described as `cognitively deprived`. The overall lack of input combined with the one-sidedness of the input probably simply does not allow students to actually use their cognitive systems effectively related to English to the point where no other systems will be built in relation to English. What we find that is that English again never actually does attain the status of a fully functioning cognitive system, but is relegated to some sort of side system which will never function effectively.

2. What is the status of input and interaction in first language acquisition?

Despite the conflict between pro and anti-Chomsky researchers in the area of first language acquisition it seems that input in the form of interaction does play an important role in the first language acquisition process. The most important finding that researches have come up with is that children will only learn their first language in an interactional setting. Therefore, it is impossible for children to learn their first language by watching TV or by listening to the radio. Simply put, children need interaction because interaction is meaningful and again it provides all sorts of extra-linguistic as well as non-linguistic input to accompany the linguistic input. Once more, output is a must. Any parent will tell you how they long for a give immediate and grossly over-exaggerated reactions to any output produced by their child. This in fact, though not at all discussed in our textbook, could actually be the key to the process of language acquisition. Meaningful situations to accompany meaningful language accompanied by expected output! In fact, one things that is often over looked in the linguist`s over zeal for structure is that linguistic structure (forms) are only made meaningful by the contexts in which they are embedded. Meaning does not come from language as such but from the association between linguistic and extra-linguistic input. Interestingly, non-linguistic input can be meaningful without accompanying linguistic input but linguistic input without accompanying extra-linguistic input is virtually meaningless.

The studies in this area have been controversial primarily because it has been found that not all adults engage in the same type of interactional behavior with their children. Having said that, all adult communities do in some way, shape, or form interactions with their children and we know from unfortunate cases where interaction has not occurred that learning simply doesn't occur.

3. What are the strengths and weakness of Krashen`s input hypothesis?

The main strength of Krashen's input hypothesis is that for the first time it stated that input was necessary and not just any old input. Krashen claimed that input needed to be of a certain type; namely comprehensible (and, presumably, directed) input. This was an interesting claim to make at the time because Chomsky was simultaneously downplaying the role of input in first language acquisition as regarded quality or quantity. Yet, as we have already said, one of the main weaknesses of Krashen's input hypothesis is its vagueness. How does one make input comprehensible? This is unclear especially when there is no specific definition for what comprehensible actually means. One other problematic aspect of Krashen's input hypothesis is the idea that it severely underrates the role of output. In fact, it basically states that input alone is enough, provided it is of the right type. Output, then, is virtually unnecessary for the acquisition process. In this view, output occurs or is made possible simply as a byproduct of internal processes. As time and a plethora of research has shown this is definitely not true.

4. How is Long`s interaction hypothesis different from the input hypothesis?

The underlying foundation of Long's interaction hypothesis is the idea of interaction as promoting a specific type of input that learners seem to need in order to naturally acquire the target language. In designing his hypothesis Long has, therefore, taken the general idea of input as initially put forth by Krashen and extended it and defined it in and through the interaction process. This claim is based on observations made between native speakers and non-native speakers during the interaction process. While the specific strategies employed by native speakers in dealing with non-native speakers vary from the strategies employed by caretakers or adults in dealing with young children, native speakers seem to engage in specific behaviors designed to help the non-native speakers to understand the language they were producing. The specific strategies of the native speakers in such interactions did, upon testing, improve the comprehension of the non-natives trying to learn the language. Thus, interaction has positive effects. This is, in effect, what the interactional hypothesis claims.

In looking at this from another angle, we can see that the interaction hypothesis follows in similar lines, possibly by default, certain aspects of socio-cultural models of language acquisition in that both envision the mentor-apprentice type of interaction as being pivotal, to learning. This leaves open the all important question for us about what happens when two students, neither of whom is an expert, interact in the foreign language classroom? Is this interaction useful or are they simply going to lead each other astray? My personal feeling is that interactions do not need to be between expert and novice or mentor and apprentice to be useful, but they do need to be set up and managed more carefully under such a situation.

5. Why are the concepts of negotiation and interaction so important in the interaction hypothesis?

First it should be made clear that negotiation can only occur with some sort of interaction. This is not merely limited to the realm of spoken interaction between two humans. Similar interactional effects have been shown in reading as well as listening under certain conditions. The basic idea is that readers and listeners need to somehow interact with a text even if there is no actual person with who a further interaction can take place. In interactional speaking, negotiation is not only possible but simply occurs as a necessary part of the communicative process. The beauty of negotiation is that learners can receive the input they require from the text again and in a different form which allows them to process the same basic idea in a different way. This seems to be an extremely important aspect of language acquisition. To simplify, it is basically the idea that a learner is able to receive the same basic semantic or functional code but with a variety of forms provided immediately by an interactional partner. It is through this negotiation of meaning of the idea, the reformulations and recast provided by the interactional partner, that a learner is able to deeply process and better understand the language in the input and the linguistic system as a whole. This is a particularly compelling idea because the learner her/himself is able to chose the areas which they are not sure about for negotiation.

6. What can we conclude from the vast array of studies conducted on interaction?

From the studies conducted for the most part in the 1990s we can see that interaction does indeed have some sort of effect on language acquisition, but it is hard to really see how it affects acquisition specifically. As with Krashen's original input hypothesis the interaction hypothesis of the late '80s and early '90s was simply too vague and ill-defined. This could possibly be because it was being formulated outside of a state/trait theory of language. As was evident in the beginning of the chapter, a lot of these input/interaction hypotheses were designed, like the processing models we looked at last week, to enhance current models of language acquisition not by focusing on the language part but on actual usage side of language/acquisition. A problem arises when such enhancement is undertaken without being closely linked to the linguistic model they're supposed to enhance and this is what seems to be evident in the earlier versions of Long's interaction hypothesis. Again, effects of interaction seem to be evident in the acquisition process. There is no doubt about that, but the specific effects are not clear again because a the effects of interaction are not well defined or explained within an overall model of acquisition or language for that matter.

7. Why is negative evidence perceived as being so tremendously important to language acquisition and SLA in particular?

Negative evidence became the most important aspect of Long's modified version of the interaction hypothesis which he proposed in 1996. The general idea behind the importance of negative evidence for Long was that it acted as a trigger for attentional focus. In this way Long was able to link his new interaction hypothesis with many of the current processing models which were being developed at the time. It is interesting to note, however, that Chomsky flat out rejects the notion of acquisition from negative evidence. He claims that it is more positive evidence which triggers the acquisition process. Thus, we have a strong conflict in the views of Long and Chomsky. This may be in part due to the fact that they are describing different acquisitions, Long focusing on second language acquisition and Chomsky focusing on first language acquisition. Their conflict may also be simply due to the fact that they value the role of interaction and input itself very differently. All the same, Long's model is still suffering from the same vagueness which plagued Krashen a decade earlier. Yes, he has linked the interaction hypothesis with some information processing models, but there is still no clear linguistic model which it can actually be paired with. It should be obvious that his conflicts with Chomsky are too strong to be over come. Thus, we wind up with a separate and distinct model for second-language acquisition following this interaction model than UG researchers have prosed for first language acquisition.

8. What is Braidi`s main criticism of the interaction research and what is her suggestion?

Braidi, a UG-based researcher, criticizes this interaction research due to its overall vagueness, which we already discussed above. As she is a theoretical UG linguist focusing on linguistic elements of the second language acquisition it would seem obvious that she rejects, or at least is severely disappointed by, interaction approaches in their almost total absence of linguistic considerations. To date interactional research has focused almost exclusively on semantic or functional aspects of second language acquisition. She would like to see more of a focus on the forms acquired and how their use develops overtime. From her comments we get an interesting viewpoint. It seems that these interactional approaches would benefit greatly from a link not necessarily with information processing models but with form to function mapping models where the concern is on linguistic forms. This in effect is what Braidi is claiming.

9. How important is focus on form in SLA?

It should be clear from our discussions last week that a focus on form is basically referring to output. The difference between output and focus on form as terms should by the same, also be clear. Focus on form is one of the components of output. It is largely related to noticing or monitoring (in a Levelt (1989) sense) forms as they are produced and using this noticing to reevaluate, and check hypotheses in the internal system (competence).

Once we are focusing attention on output and focus on form, we immediately need to try to study what it is that can affect such forms and SLA has been coming up with many studies doing just this. In this way studies of feedback are important and have become quite common in the SLA literature, but again with no clear findings yet. There seems to be too much variability in the effects of feedback and even in the various types of feedback given. Thus, we know that feedback, as one important aspect of interaction, does work to aid in the second language acquisition process but we are still not sure how and what some of the other variables are.