Learning a Second Language Through Interaction by Rod Ellis

Originally posted to the CETEFL list.  (Dennis Newson)

--------------

Learning a Second Language Through Interaction by Rod Ellis (and research assistants), an acknowledged expert on language acquisition,   (Studies in Bilingualism 17, John Benjamins, Amsterdam and Philadelphia, 1999 ISBN (Europe 90 272 4125 2 and 1 55619 737 3 (US) 285 pp), being a book concerned with theoretical issues and reports of theoretically-orientated experiments is not an easy book to read and, frankly, the busy, practicing teacher with lessons to prepare and written work to correct is not likely to give it priority. But issues of central relevance to classroom practice are discussed here and the main purpose of this review , written by someone relieved of the duties of daily classroom teaching, is to pass on a summary of its most important statements and findings.

The book consists of five sections: Section 1 (written by Ellis) is entitled Theoretical Perspectives on Interaction and Language Learning. The other sections are: Input, Interaction and Vocabulary Learning, Interaction and Grammar Learning, and two final sections by Ellis, Pedagogical Perspectives (Making the Classroom Acquisition Rich) and a conclusion (Retrospect and Prospect).

In all 8 new research studies are reported on.

Unlike the teacher the researcher isn't satisfied to say, anecdotally: This method seemed to work well with my pupils. (S )he needs to produce statistics demonstrating that procedure A produced measurably better results than procedure B. This narrows, of course, what can be tested, and the teacher suspicious of research may well think, reading the results of a particular study: "I could have told him that.".

Ellis himself states that the purpose of the book is to give answers to three key questions concerning the role of interaction in second language (L2)acquisition. These are:

  1. In what ways does interaction/input contribute to L2 acquisition?
  2. Which types of interaction/input promote L2 acquisition?
  3. What kind of language pedagogy is needed to ensure that classroom learners experience acquisition-rich interaction?
  4. I would summarise these carefully worded questions as amounting to: Which classroom methods  are most effective ?

What Elllis means by "acquisition" is when the learners have really got it, when the item or feature of language under discussion has been well and truly internalised and entered the learner's interlanguage and long-term memory - acquired it   - so that it can be used for communicative purposes.

What are Ellis's answers to his own three questions?

Interaction is, indeed, very important, but some of the research reported here shows that it is not essential, at least for some learners. In one experiment reported  learners who merely listened to classmates doing specific tasks learned just as much as their more active friends.

Most of the experiments carried out, as interpreted by Ellis, demonstrated one of his main theses: the best kind of interaction i.e. the most productive, the kind most likely to result in acquisition is when topics are selected by the learner and not the teacher. This is not simply a question of the learners deciding what they want to talk about, but a matter of the learners being able to guide the direction of the discourse between themselves and the teacher so that they ensure it answers their needs and does not simply follow a plan worked out by the teacher in advance. There is even evidence to suggest that improvised, non-planned discourse is likely to be more productive precisely because it is responding to learners' needs on the spot as they are revealed. And working in pairs or groups ensures that the teacher is not calling the tune: ""With the teacher out of the way learners are free to manage the discourse for themselves." (p226).

The kind of pedagogy that makes acquisition more likely to take place is one which aims at producing an acquisition rich  environment.And what is especially important is encouraging intrapersonal "talk", i.e.. the sort of talk that it is assumed (it is impossible to observe directly) is going on in a learner's head as a lesson is taking place. One piece of research ("Reacting with a text") has recommendations for a strategy that learners can be taught to train them in developing the habit of productive "talk" i.e. silent questioning that will lead them to a fuller and deeper understanding of the text being studied.

And what does research show about the effectiveness of teaching or not teaching grammar?

Ellis is no supporter of the teaching of prescriptive grammar rules. What he and his co-workers have investigated and demonstrated is the effectiveness of consciousness-raising by using task-based activities to get students noticing the structure of language and talking about grammar in the hope that this raised-awareness will bring long-lasting rewards in a general fashion.

Here is a list of the main  statements that are the claims made as a result of the studies reported on or quoted  in this book.

  1. 6 or 7 exposures seem to be the minimum necessary before the meaning of a word can be "known". (p47)
  2. Giving learners training in a general strategy aids in comprehending written texts and learning new vocabulary. (p61)
  3. Active participation is  neither necessary nor detrimental to comprehension/acquisition. Those learners who prefer the quieter route can benefit from the activity of their more public comrades. (p82/83)
  4. Interaction facilitates comprehension, but is not crucial. (p111/112)
  5. Learners vary in their preparedness to negotiate meaning by asking questions. (p111)
  6. The most important factor contributing positively to the acquisition of word meanings   is the number of different contexts in which a target word appears. (112)
  7. Listen-and-do tasks lead to a high level of comprehension, even when this involves listening to or using unfamiliar words. (p127)
  8. Providing "scaffolding", supportive, aim-orientated conversation, creates ideal conditions for language acquisition. (p129).
  9. Ideal conditions for language acquisition  occur when participants can respond to a task by constructing an activity that matches their own goals. (p129)
  10. When learners are trained in how to interrogate a text they are more likely to learn and remember new words. (p165).
  11. Improvements in the use of past tenses under specific test conditions "...failed to show any effect on learners' subsequent ability to use past tense forms accurately." (p184/5)
  12. In making the classroom acquisition rich, learner discourse control was found to be central in creating opportunities for language learning. (p 211)
  13. Cognitive development is facilitated when learners have the opportunity to interact with a more experienced person who can guide, support and shape their actions (i.e. "scaffold" their attempts to perform new functions). (p224)
  14. Learners' control of discourse occurs more in task-based language teaching where learners attend primarily to what is talked about rather than how language is used. (p225)
  15. Learner control of discourse is fostered when 'framework' communication in the classroom is conducted in the target language. (p227)
  16. Pair and small group work ensures that learners have control of the discourse. (p226).
  17. Cultural expectations may make it very difficult to create the conditions necessary for learner discourse control. (p228)
  18. Some teachers may be happier with the 'monastery tradition'.... "with containerization of life, clearly-defined norms and strict discipline" rather than the 'marketplace tradition'.... "the rough and tumble of life where anything that works is welcome." ...."Yet it may be something akin to the market-place tradition that is needed if learners are to have the chance to control the discourse." (p229).
  19. "Future research on interaction needs to complement the current learnING-centred approach with a learnER-centred one." (p238)
  20. Repeatedly, what emerges in the studies reported above is that it is "exposure to input" that is most productive, though Ellis cautions that for acquisition to take place it needs to be exposure to input that the learner understands.

Should teachers be concerned with research? I would have thought that they should, if the research provides insights like those reported above that can make teaching more effective. Ellis himself is much more reticent.

 "....research findings cannot be used to prescribe, proscribe or even advise teachers what  or how to teach. Research findings are best treated as 'provisional' specifications...which teachers can choose to act on or ignore, in accordance with the exigencies of their own teaching situation. In cases where they decide to act, the specifications should be treated as hypotheses about teaching and thus subjected to critical scrutiny through teaching accompanied ideally by some form of action research." (p248)

Perhaps, after all, it is intuition, classroom experience, reflection on that experience and the discussion of it   with other practitioners that is most important.

Dennis Newson