The iconic memory system stores a snapshot of incoming visual information, and is characterized by its brief du- ration but large storage capacity. Information from iconic memory is transferred to visual working memory (VWM), which is characterized by its very limited storage capacity. The limitation of information storage in VWM is often characterized as an all-or-none process, whereby information is either successfully stored or it is lost completely. This view raises a simple question: If the majority of perceived visual information is initially stored in iconic memory, but one second later most of it is completely absent from VWM, what happened to it? Characterizing the nature of how information is lost during the process of iconic memory decay is a critical step toward understanding this memory system, which plays a fundamental role in almost all aspects of vision, including our fluid visual percep- tion of the world across time and eye movements. Moreover, deficits in iconic and working memory have been implicated in may psychological disorders including Schizophrenia, Alzheimer?s Disease and children with ADHD, and a better understanding of how iconic memory works and how to measure it may provide a basis for studying several psychological disorders in ways that are not currently possible. Here we first seek to characterize how the precision and storage capacity of iconic memory change over time. In preliminary experiments we observe a clear dissociation: Iconic decay results from a rapidly shrinking capacity, evidence by the complete loss of more and more items over time. However, the precision of those items that are retained in memory does not decrease over time. This result implies that the way in which visual information decays in iconic memory is a highly complex process that involves higher-level visual representations, and that the discrete-capacity limit often attributed to working memory may in fact result from iconic memory processes. In this proposal we systematically character- ize both the behavioral and neural signatures of this capacity-limiting iconic memory decay process. Part of this work involves the development of new mathematical models, capable of characterizing how different pieces of information decay differently in iconic memory. This modeling framework will provide a new approach for assess- ing how different types of information are represented in and lost from iconic memory, and will allow for a more accurate characterization of the time course of decay than is currently possible. In addition, a novel neuroimaging approach is developed that allows the neural signatures of information loss in iconic and working memory to be tracked with high temporal resolution using EEG. Taken together, the developed behavioral paradigms, mathe- matical models, and neuroimaging techniques comprise a new framework that is capable of characterizing the mechanisms that underlie the iconic and working memory systems with a far greater resolution than is currently possible. By using this framework to shed light on the nature of information representation and loss in iconic memory, our results will provide new avenues for conceptualizing, testing and diagnosing the many psychological disorders that are associated with this fundamental part of our perception and cognition.
Although we are seldom aware of it, our brains retain an almost perfect visual memory of everything we see for about half of a second after we see it, like a photograph that quickly vanishes. This so called iconic memory system plays a fundamental role in our normal perception the world, and deficits in such early memory processes have been implicated in many psychological disorders, including Schizophrenia, Alzheimer?s Disease and children with attention deficit hyperactivity disorder (ADHD). In this work, we develop new mathematical modeling and neural recording approaches that, together, provide a framework that will allow us to study the iconic memory system with a resolution that is not possible with current technologies.
| Pratte, Michael S (2018) Iconic Memories Die a Sudden Death. Psychol Sci 29:877-887 |
