Dynamic aspects of visual working memory storage; toward a complete account of representational changes in visual working memory over time.
MetadataShow full item record
Working memory (WM) is a dynamic and flexible storage system. It needs to constantly adapt to maintain and optimize WM storage in the face of interference and changing internal behavioral goals. Yet current research often focusses on analyzing WM as a stable entity. Here, WM representations are analyzed over an entire maintenance period. Furthermore, participants are often asked to perform a task about a few simple stimuli while looking at a grey screen. While this setup can provide answers about where information is stored and what might happen if one internal or external factor is added, it leaves out the when. This is problematic when dealing with a dynamic entity as WM, as representations can change or adapt over time. I could lead to null-results, as shifts in representation add noise to a block-analysis. Moreover, it does not provide a clear account of what WM does over time and how different mechanisms are employed to sustain human behavior. In this review, current findings about dynamic WM mechanisms are discussed. We discuss intrinsic characteristics of WM storage and the effects of external and internal factors on WM representations. This includes dynamic shifting, neural drift, visual interference and attentional and behavioral goals. Additionally, some theories about the role of the early visual cortex in WM and possible activity-silent codes are discussed. We find that WM is highly dynamic, as it is constantly adapting and modulating neural representations of visual stimuli in the EVC in response to interference. We also find that WM can change representational formats depending on attentional demand to optimize maintenance of multiple items without having them interfere. To understand when these changes precisely happen, fMRI studies could benefit from temporal generalization methods. Here, a timeseries is analyzed and compared over multiple time intervals. Future studies could employ this analysis method, as well as combining intermixed and blocked experiment designs to isolate strategic and responsive WM adaptations. Finally, WM research could benefit from more complex research paradigms, where multiple tasks and/or interference methods are used with more complex stimuli. This would increase the ecological validity of WM studies, and help understand how WM adapts in our stimulus-rich environment. Through this new dynamic focus on WM storage, we could understand more about this complex system that supports our daily functioning.