On Input Processing, Processing Instruction, and the Nature of Replication Tasks: A Response to Salaberry

Abstract

We ar e pl eased t o h ave this opportunity t o r espond to M. Rafael Salaberry's articleThe role o f input and output practice in s econd language acquisition' (CMLR 53:2). Because we believe he has misunder- stood t he n ature o f w hat has now come t o be calledprocessing instruction,' we feel obliged to comment. As m ost readers are p robably a ware, p rocessing inst ruction is a relatively ne w concept, and as VanPatten h as said i n hi s most recent work ( VanPatten, 1996), we ne ed t o k eep r esearching pr ocessing instruction before we jump on what can be called aprocessing band- wagon.' Moreover, the need for replication studies in a young field such as SLA is great, and so we were happy to see that our work on process- ing instruction had already started to be replicated. We underscore our enthusiasm for this development because we do not wish our comments on Salaberry's work t o be viewed i ncorrectly as a def ence o f a w eak position. Instead, our comments - largely critical - should be interpreted as part of professional discourse whose goal is to clarify and elucidate misunderstandings and misconceptions. Our comments are critical for two reasons. First, we believe there are a number of important misunder- standings of critical theoretical issues in bot h i nput processing and processing instruction that appear in Salaberry's work. Second, we find substantial m ethodological l imitations in his w ork that re nder his conclusions invalid. Most important, we observe that Salaberry's study is not a replication of VanPatten and Cadierno (1993), Sanz (1994) or 1