Project Awarded: $7,700
We are exposed to a vast amount of information over the course of a day, and yet our memory system is capable of encoding and later retrieving relevant details about our environment. This would be impossible unless stored information were structured - that is, organized across various dimensions such as space, time and semantic content. Likewise, the most prominent example of memory success, the Method-of-Loci (MoL), can be understood as a way to structure incoming information according to a spatial format. In the MoL, subjects typically memorize a list of unstructured items such as random words or phone numbers by imagining these items in a structured environment such as a familiar childhood walk through the neighborhood. The MoL thus appears to be a way of using a useful feature of spatial memory to store non-spatial content. However, how this process actually works, and why it works so well is poorly understood. Our aim in this study is to investigate how different kinds of structures work together in human memory at the neural level, as a step towards understanding how our memory systems succeed. More specifically, we aim to do by designing a set of memory experiments to be conducted while taking iEEG recordings from neurosurgical patients implanted with electrodes for the treatment of epilepsy. These electrodes reflect the activity from many local neurons, allowing for more precise measurement of localized neural activity than is standardly allowed in investigations of human memory, where invasive recording is not otherwise possible.
The explicit process in the MoL involves two stages: a strategy for encoding items through visualization, and a strategy for retrieving items through a parallel visualization. For example, I would encode a list by imagining a walk through my childhood home, and then later recall the items by imagining walking through the home again and picking up the items. The parallel between encoding and retrieval raises a broader question about memory: how much does the relationship between remembered items at encoding predict subsequent relationships at retrieval? Our task encourages participants to encode lists of items in a spatio-temporal sequence, but allows them to choose freely how to retrieve the items. By varying the semantic cohesion of the lists, we are effectively providing subjects with three possible structuring dimensions: space (a map-like representation of a museum), time (an ordered path through the museum), and semantic content (relations of meaning between items on the list). Based on previous work, we expect for at least spatial structure to be clearly distinguishable at encoding when recording with iEEG from hippocampal areas, and likely temporal structure as well.
The question of how structuring information via mnemonics facilitates remembering is of interest to both philosophers and neuroscientists. Neuroscientists are interested in how the structural features of spatiotemporal encoding and retrieval, as well documented in models of rodent spatial navigation, are conserved and used in human remembering, which makes use of the same neural structures (i.e., the hippocampal formation). The question is also of interest to philosophers as a phenomenon to be explained when theorizing about the nature of memory and exploring relations between levels of explanation in philosophy of science.