Attention and Long-Term Memory in System and Process Theories of Working Memory: A Reply to Soemer’s (2015) Comment on Schweppe and Rummer (2014)

Abstract

Recently, Soemer (2015) criticized our theoretical review paper published in this journal (Schweppe and Rummer 2014). In our article, we argued for a new perspective on (working) memory within theories of multimedia learning. Our main idea was that a theory that focuses on interactions between attention and long-term memory (as does Cowan’s 1999, embedded-processes model) provides a more unitary account of the multi-faceted principles associated with working memory in multimedia research than those models currently incorporated in multimedia theories. Even though many of the recommendations for how to design multimedia learning materials relate to reducing cognitive load in working memory, they refer to attention and long-term memory at least as much. Our article had two core messages: (1) Based on a description of the multicomponent model of working memory (e.g., Baddeley 2000; Baddeley & Hitch 1974) and how it is adapted in theories of multimedia learning, we concluded that the most important principles for multimedia learning are neither clearly attributable to the basic model (i.e., the multicomponent model) nor to the applied models (i.e., the cognitive theory of multimedia learning, e.g., Mayer 2009, and the cognitive load theory, e.g., Sweller 1999). (2) As an alternative, we suggested incorporating a different model that conceives of working memory as the interplay of attentional processes and long-term memory and thus allows for explaining principles related to attention, working memory, and long-term memory in a unitary manner (i.e., the embeddedprocesses model by Cowan 1999). Soemer’s (2015) comment mainly concerned our description of the multicomponent model of working memory. More specifically, he criticized that, unlike suggested in our article, the current version of the multicomponent model of working memory (Baddeley 2000, and later) does not postulate structurally independent working and long-term memory systems but rather Educ Psychol Rev (2016) 28:201–204 DOI 10.1007/s10648-015-9336-0