ctly based or derived on linguistic or psychological scholarship (Richards and Rodgers, 1986, p. 14).
At least since the 1960s, scholars and educators have incorporated the idea of error analysis or error correction in their models and practices of how to teach a foreign language. This reflects a long-recognized understanding (one that has been challenged with considerably intellectual force in this century but one that still proves to have a certain pragmatic usefulness) that a personÆs first language creates certain heuristics and internal models that will be used to acquire later languages. Such heuristics may in appear to ôinterfereö with secondary language acquisition by causing the learner to err systematically in ways that reflect the structure of the maternal tongue. Because these errors are systematic, they provide a gateway for scholars and teachers into understanding the linguistic mechanisms a person uses in acquiring a second language.
While the basic idea behind error correction is thus a simple one, in practice it is very complicated because the internal model a learner has of a language is highly fluid, changing almost constantly as the learner acquires greater mastery over the new language. (In fact, many scholars believe that, as an individual acquires a new language, she or he creates what is termed an ôinterlanguageö, or a sort of working text that helps the person shift back and forth between the already acquired rules of the first language and the fully developed but as yet undiscovered grammar, syntax and vocabulary of the second language. With each new piece of acquired information about the second language, the interlanguage changes, which may in turn produce changes in the type and/or frequency of error that is made.) As Richards (1985) notes, errors inform both about how well an individual is doing in acquiring a new language and about the process of language acquisition itself:
A primary focus o...