Why Is Our Working Memory and Processing Capacity so Limited?

We show that spurious synchronization, as suggested by Luck and Vogel 6], is not suucient to explain the limited storage capacity of working memory. Instead we argue that the sharp breakdown of error free recall suggests a dynamic mechanism behind working memory. We summarize some suggestions for the dynamic basis of working memory and argue that the reason behind our limited processing capacity lies in the enhanced ability to process highly complex tasks by the collaboration of mental subsystems. 1 Limited storage capacity of working memory Cognitive science has formulated the concept of working memory which seems to be central to a vast amount of higher mental processes such as language comprehension, reasoning, problem solving and consciousness. However, the understanding of underlying functional mechanisms and neural correlates have been rather scattered. One of the long recognized features of working memory is that of the limited capacity of keeping information from several independent objects available. The capacity of working memory has been studied extensively for verbal information 1]. Miller 2] stressed the magical number 7 2 as capacity limit, which coincides with the capacity limit for list items shown already in 1928 by Oberley 3] (reproduced in 4]). However, the precise number of objects which can be recalled without error can vary depending on the detailed task, and mental tricks such as the phonetic memorization of a string of letters can shift the observed memory limit towards higher numbers. Luck and Vogel 5] have recently demonstrated in a purely visual task that the memory capacity of this working memory is about 4 objects. It seems puzzling that we have such a severe limit in our short term memory and should ask why evolution was not able to increase our working memory since a larger working memory should make us tter to survive. This is our main motivation for this discussion. A second major contribution of Luck and Vogel is the demonstration that the visual working memory is apparently for integrated objects (independent of the number of features) 5]. In a recent comment 6] they suggested that synchronization of spike trains can explain both, the limited storage capacity as well as it's independence of the number of features. Their argument follows the popular suggestion that diierent features of an object processed in diierent parts of the brain are related by synchronized spike trains (which solves the binding problem). The limited storage capacity …