Project Summary
Episodic memory binds together the people, objects, and locations that make up the specific events of our
lives, forming multi-element traces of experiences that can guide ongoing behavior, help imagine the future,
and enhance well-being. Episodic memory is poor in young children, a phenomenon called childhood amnesia.
It gradually improves over the preschool years, with further refinement in elementary school. Understanding
this developmental change requires linking research at three levels of analysis: (1) dissecting component
processes of episodic memory in controlled laboratory tasks; (2) assessing underlying neural changes; and (3)
evaluating children's naturalistic behaviors, both in the memory domain (self-related episodic memory and
memory conversations with adults) and more generally. We propose a longitudinal study of children from 4 to 6
years and from 6 to 8 years, i.e., across the offset of childhood amnesia, with rich evaluation of all three
components at three points in time. This multi-componential study across levels of analysis will be the first of
its kind.
Relational binding of multiple elements within an event may be conceptualized as the formation of coherent
multi-element episodes that allows for pattern completion, i.e., the elicitation of all components of the
experience from a subset. Another process involved in episodic memory is pattern separation. The
Complementary Learning Systems (CLS) model, as well as earlier mathematical models of the hippocampus
and later refinements of CLS, proposes that the hippocampus (HC) is specialized for the rapid and automatic
acquisition of information that is then orthogonalized to reduce potential interference among similar memories.
We will test children's abilities to retrieve all elements of an episode via pattern completion and children's
abilities to discriminate between overlapping episodes containing similar people, objects or spatial contexts via
pattern separation. In addition, a rich research tradition explores self-related episodic memories and their
retention over time, as well as their linkage to environmental support. We will link findings regarding children's
naturalistic behavior to laboratory measures of episodic memory and to neural development. These relations
are likely bi-directional in causality, and hence longitudinal evaluation is essential.
The behavioral work will be complemented by work at the neural level. Both pattern completion and pattern
separation depend on the HC and are associated with maturation of its subfields, and the maturation of
connectivity between the HC and cortical control structures. Using high-definition MRI and cutting-edge
diffusion imaging parameters, we will test the hypothesis that specific hippocampal subfields perform pattern
completion (CA3) and separation (dentate gyrus). In addition, we will test whether white matter pathways, both
within the HC (e.g., perforant path) as well as long-range tracts linking the HC to neocortical areas, relate to
pattern completion/separation behavioral performance.