The elemental-configural distinction: a problem of two dimensions: commentary on Melchers, Shanks, and Lachnit (2007)

Melchers et al. (2007) show that learning is flexible with respect to a continuum between elemental and configural. Furthermore, they show that various factors, including the past experience of experimental subjects and the properties of the stimuli used, can determine whether elemental or configural strategies are used. In this article two different usages of the term configural will be highlighted. It is argued that flexibility in processing could occur in terms of either meaning. Unfortunately, many studies which show flexible processing have used designs that do not enable the specific locus of their effects to be identified and a suggestion for the use of one design, a feature negative design, is made. In principle use of this design could help to better characterise movements along the elemental-configural continuum