Less than 20 years ago, the lion’s share of work in the cognitive neuroscience of memory, including our own, examined neural activity during an experience (encoding time) or during memory retrieval. Using the subsequent memory effect, whereby trials were sorted by indices of memory behavior, this approached revealed was pivotal in revealing a network of brain regions where encoding activity predicted successful encoding or retrieval. Furthermore, it became apparent that the hippocampus supported domain-general memory formation (irrespective of content) whereas other regions’ activity profiles were supporting memory encoding for particular kinds of content (domain-specific).
Our research focus took a pivotal turn when we reasoned that the ultimate fate of a memory was only partially dependent on encoding activity and much of what we were measuring in the brain during encoding was related to perception, attention and control processes. Thus, in a conceptual and methodological leap, we ran the first study to examine post-encoding resting brain activity to isolate memory processes. Specifically, in this initial investigation, we found that hippocampal-cortical functional connectivity increases from pre- to post-experience and that specific multivariate hippocampal patterns of activity of each trial reactivate during post-encoding time periods. Critically both of these measures were shown to be biomarkers of memory consolidation as they both were found to be related to individual subjects memory for the presented materials. We believe post-encoding persistence is a ubiquitous mechanism to support long-term memory irrespective of the type of memory as we have now demonstrated these effects in the amygdala and various cortical regions important for the specific task at hand.
While we have mostly examined the effects of neural persistence, as described above, as a mechanism for memory consolidation, this work begged the question of whether the persistence of activity might influence ongoing perception and memory. Using a clever design, we showed that emotional brain states persist into more neutral encoding blocks and ‘color’ the neutral stimuli with the emotional brain state leading to improved behavioral recall and neural substrates associated with emotional memory encoding. A related question is whether new learning can retroactively augment memory for related past experiences. We were the first to show that retroactive memory enhancement is selective and to identify both offline and online reactivation as the critical mechanisms.
