ABSTRACT
This chapter discusses the alternatives that have been proposed by distinguished researchers to our hypothesis of a time-based resource sharing in accounting for the pace effect. SOB-CS is a distributed neural-network model for complex span task. As Vergauwe demonstrate, and in line with the predictions of the TBRS model, increasing memory load in a Brown-Peterson paradigm results in slower processing as long as the maintenance process relies on attention. In Cowan's embedded-processes model, this capacity limit would correspond to the size of the focus of attention, which determines the maximum number of chunks that people can attend at one time. The TBRS model assumes a one-chunk focus of attention selecting a single working memory representation for the next cognitive operation among a limited number of representations held in an episodic buffer. Reintroducing operative aspects, Pascual-Leone reanalysed the paradigms mentioned by Cowan and found results consistent with Miller's number seven as an estimate of demand on working memory.
