A brief history of processing approaches to SLA: reply to Mellow

In his article Mellow claims that the 'predictive model' is based on a transformational analysis of word formation. We will show that this is incorrect and that instead, the model is based on the psychological concept of exchange of linguistic information. Therefore, Mellow's evaluation of various types of transformational analyses is quite irrelevant for the model.To demonstrate our point we will reproduce the full text of the original description of the 'predictive model'. We will show that the transformational connection should properly be traced to Clahsen's strategies which apply to syntax only. The strategies approach will then be described in contrast to the 'predictive model' in order to characterize the two approaches and their historical context. This will provide the background to a brief critique of the strategies approach. Theory formation in later processing approaches to SLA will then be sketched out briefly before addressing several points of detail in Mellow's article.