PROJECT SUMMARY
The details we bring to mind about past positive and negative events can have important implications for
our mental wellbeing, affecting how we see ourselves and maintain relationships with others. Do we vividly
remember an argument with a friend, or the heartfelt apology that followed? Do we remember the feeling of
accomplishment after giving a talk, or our struggle to answer an audience member’s question? The details we
bring to mind also can affect our decision-making, with consequences for our social networks and career
success: Will we call that friend again? Will we agree to give a talk at an upcoming meeting?
There are fascinating examples that suggest that what emotional details we remember from past events
may vary, in part, with a person’s age. Older age is associated with a greater use of positive (vs. negative)
words in descriptions of past events (Pennebaker & Stone, 2003; Ford et al., 2016), and with a greater
reported focus on positive aspects of otherwise challenging situations (Ford et al., 2018a, b). Yet we
understand little about why such differences arise, and the answer is likely to have implications for
understanding how cognition-emotion links differ across the adult lifespan and how adults of various ages
achieve mental wellbeing. The first aims of this research are to understand these differences upon initial
retrieval, focusing on memory phenomenology (Aim 1) and neural representations (Aim 2). Using multivoxel
pattern analysis (MVPA), we examine two types of recapitulation: trial-specific (tied to each specific encoded
event) and valence-generalized (shared across stimuli of a particular valence category). We hypothesize that,
with older age, memory representations will have more valence-generalized vs. trial-specific recapitulation, and
this age-related difference will be most pronounced for memories of positive content. We anticipate this will be
reflected in positive memories feeling subjectively vivid to older adults, while lacking in specific detail.
A key feature of episodic memory is that it is not a static representation. The details we remember now are
not necessarily the details we will remember later, though the way we represent a memory now can tip the
scales, making some details more likely to come to mind later and others more likely to be omitted. For
emotional memories, our recent research and preliminary data suggest that retrieval at one time-point may
trigger the start of what could colloquially be thought of as a “virtuous-memory-cycle”, with positive details
emphasized and negative details de-emphasized, or the opposite, a “vicious-memory-cycle.” Critically, we
hypothesize that older age increases the likelihood of a “virtuous-memory-cycle.” We propose that age-related
differences in memory representations make it particularly likely that older adults’ initial retrieval of positive
content impairs their ability to later remember related, negative content (Aim 3) and leads to broader age-
related differences in the details accessed over successive retrievals such that, over repeated retrievals, older
age is associated with memories that become more positive (Aim 4).