Correcting or nagging?

Studies have actually failed to show that explicit error correction causes a permanent improvement in production for learners of a second language.  Lightbown (1985) and others have then come to the conclusion that isolated error correction is usually ineffective in changing behaviour.

If you think about it, direct error correction usually relies on the student having an understanding of the correct form, they just mispoke because they are in the habit of mispeaking having never been successfully corrected before. 

If they do not have this underlying understanding of the correct form, they must be exposed to it in order for the language to make sense to them before they will reliably produce it correctly.  Surely, this will take multiple exposures and in the few hours that most ESL students spend in the classroom with many other students, a reoccurrance of the same error will take some time.

Point of this being that if you really want to correct behaviour, the best way to do this is to correct the belief that underlies the behaviour.

Lightbown, P. (1985).  Great expectations: Second language acquisition research and classroom teaching.  Applied Linguistics, 6, 173-189.

A couple of steps that should not be forgotten

Okay, so in the 50’s we had Contrastive analysis.  This is where a bunch of geniuses sat around and said, “If we compare L1 and L2, we can see where the students are gonna go off track and we can focus on these areas”.

These guys were debunked by the younger, hipper crowd in the 70’s who said, “yeah, but a lot of errors that we expect to occur under contrastive analysis don’t eventuate.  Maybe what we should be looking at are the errors that the students actually make.” 

This was the birth of error analysis.  Now, what kind of teacher doesn’t try to help students by looking at the mistakes that they are making?  We would be remiss to ignore such vital indications of our individual student’s understanding.  But to base your entire approach on errors now seems to be a little negative doesn’t it?  Don’t forget the positive reinforcement.

But that was a step that our scientist friends needed to take in order to get to the theory of interlanguage.  “What if these errors are not just random occurrences?” they mused as they poured over their sample of errors.  “What if there was some linear progression from one language to another language that we could plot our learners on?  We already know that L1 is learnt in roughly the same order by all children who learn under comparable circumstances, perhaps there is a similar correlation here.”

The part that I find interesting is that sometimes, a progression appears to be a regression.  Children learn words like played, did, went and use them perfectly fine as individual lexical units.  Then, they learn the rule; add -ed if you are talking about the past.  Suddenly, they start using words like doed and goed even though they were using the correct form last week.  They progress from using individual lexical parts to understanding the rule but applying it blindly to applying the rule and understanding the exceptions.

I don’t like to think of the old theories as debunk.  I prefer to think that they are a step that we had to go through to get where we are now and each step contributed something to a fuller understanding of how people learn to communicate.  So look back at your contrastive analysis and error analysis and see what they can offer to what you do in the classroom today.