Linear forgetting

Publication date: October 2019Source: Journal of Memory and Language, Volume 108Author(s): Jerry S. Fisher, Gabriel A. RadvanskyAbstractMemory retention and forgetting is typically captured by an Ebbinghaus curve in which there is a sharp initial decrease that follows a negatively accelerated function. This pattern, typically well fit by a power function and poorly fit by a linear function, has been observed across a variety of materials, tasks, and retention lengths. However, here we demonstrate, across three experiments, a set of retention patterns that are better fit by a linear function, which is not accounted for by any existing theory of memory retention and forgetting. This linear pattern was also observed, but not noted, in existing studies from the literature. Our assessment suggests that higher degrees of learning and meaningfully complex materials may be jointly needed to observe linear forgetting. A simulation is provided as a proof of concept that linear forgetting may emerge when there are (a) the multiple components of memory traces are lost at different rates, with each following a negatively accelerating function, and (b) memory responses may be made using degraded memory traces, such as through partial matching and/or reconstructive processes. Linear forgetting is important to the study of memory retention and forgetting, and it may apply to the bulk of everyday event memories that concern people over long lasting periods of time.
Source: Journal of Memory and Language - Category: Speech-Language Pathology Source Type: research