Interactional Hypothesis (Michael Long)
Explanation of Main Features
L2 students are able to achieve the target language fluency and reach a better comprehension by interacting with native speakers and / or non-native speakers with a higher language level. The interactions are able to help the L2 students modify their speech and improve vocabulary structure by understanding the meanings.

The example of conversation modifications:
1. Comprehension checks-
2. Clarification requests-
3. Self-repetition or paraphrase-


Classroom Implications
1. Comprehension checks-
Teacher: The train leaves at 5:00. Do you know what I am saying?
2. Clarification requests-
Non Native Speaker: Sorry, can you say that again? I don’t understand it.
3. Self-repetition or paraphrase-
Teacher: Right now it’s 4:50. The train will leave in 10 minutes. You should take the train right now.

Reference:

Ellis, R. (1999). Learning a second language through interaction. Amsterdam: John Benjamins Publishing Company.
Lightbown, Pasty M, & Nine , Spada. (2006). How language are learned. Oxford University Press.



Critique of the Interactional Hypothesis (Michael Long)

Despite the fact that the interactional hypothesis makes a very compelling argument as a theory of second language acquisition, its benefits have yet to be proven. Many studies have been conducted to try and prove that the use of negotiation will increase English language proficiency, but the data is not strong enough to demonstrate its value. Long took a native English speaker and a language learner and observed their interactions to see what assisted them in their conversations and found that negotiating for meaning was key. However, the subjects in the study were volunteers that were willing to work with each other for the sake of the study.
It appears that Long argues for the use of comprehensible input as a means of communicating with language learners in the classroom, and then for the use of typical language and then negotiate or communicate after the message has been conveyed to develop more meaning. However, what Long assumes is that native speakers in the classroom will want to speak with the language learners and assist the learners in their comprehension and learning. In "real" (non-testing situations), native speakers often may feel frustrated with constantly trying to understand the language learners, and unwilling to help them out.

Ellis, R. (1991) “The Interaction Hypothesis: A Critical Evaluation.”