# Challenging Classical Theories in Spatial Cognition: Contrasting Translator and Comparator Models of Human Retrosplenial Function

> **NIH NIH F32** · UNIVERSITY OF CALIFORNIA-IRVINE · 2023 · $73,942

## Abstract

Project Summary
 Everyday navigation behaviors that may seem mundane to healthy young adults are, at their core, quite
complex. Seemingly rote navigation such as driving to the office require coordination of multiple streams of
sensory information as well as memory for the global structure and layout of the environment (i.e., a “cognitive
map”). Representations of space are often classified into two distinct frames of reference: egocentric –
viewpoint-dependent relationships – and allocentric – observer-independent landmark relationships. The
inability to reconcile egocentric and allocentric representations leads to disorientation, even in familiar spaces,
and is associated with damage to spatial processing networks and regions of the brain including retrosplenial
cortex (RSC). This type of disorientation also serves as a marker for preclinical stages of dementia. However,
current spatial models of RSC function remain poorly connected with parallel lines of research from episodic
memory. This narrow theoretical focus is problematic given the wide range of cognitive processes ascribed to
RSC. The most prominent spatial model of RSC function, referred to as the BBB model, posits that RSC
primarily supports the flexible use of egocentric and allocentric reference frames by translating between
egocentric and allocentric spatial reference frames. Such spatial translator models of RSC are computationally
plausible and generally compatible with findings in the navigation and spatial cognition literature, but there has
been no direct test of this model in humans. Moreover, few functional neuroimaging studies are designed a
priori to directly study RSC function. An alternative model posits that RSC is part of a predictive coding
hierarchy. In this comparator model, a general function of RSC is to make predictions about the world based
on learned experiences and evaluate the accuracy of these predictions against actual sensory information to
resolve any discrepancies and update future predictions (i.e., Bayesian inference). Assumptions for each
model will be evaluated during memory for previously learned spatial experiences (slow, Aim 1) as well as for
the spatial demands of the previous trial (fast, Aim 2). To accomplish this, state-of-the-art immersive virtual
reality techniques will be combined with functional magnetic resonance imaging (fMRI). Receiver operating
characteristic approaches as well as advanced univariate and multivariate fMRI analyses will be used to
analyze data from novel behavioral paradigms comprising episodic and spatial memory elements. Functional
MRI will allow for accurate and non-invasive imaging of RSC during behavior, and novel task manipulations will
pit the predictions of RSC comparator and translator models against one another. These experiments will
challenge long-standing spatial theories that have not undergone a rigorous test in humans. The findings will
close a gap between poorly connected theories of RSC function in sp...

## Key facts

- **NIH application ID:** 10569490
- **Project number:** 1F32NS129626-01
- **Recipient organization:** UNIVERSITY OF CALIFORNIA-IRVINE
- **Principal Investigator:** Michael James Starrett Ambrose
- **Activity code:** F32 (R01, R21, SBIR, etc.)
- **Funding institute:** NIH
- **Fiscal year:** 2023
- **Award amount:** $73,942
- **Award type:** 1
- **Project period:** 2023-05-01 → 2025-04-30

## Primary source

NIH RePORTER: https://reporter.nih.gov/project-details/10569490

## Citation

> US National Institutes of Health, RePORTER application 10569490, Challenging Classical Theories in Spatial Cognition: Contrasting Translator and Comparator Models of Human Retrosplenial Function (1F32NS129626-01). Retrieved via AI Analytics 2026-05-22 from https://api.ai-analytics.org/grant/nih/10569490. Licensed CC0.

---

*[NIH grants dataset](/datasets/nih-grants) · CC0 1.0*