Two-component model of memory stability

Jump to: navigation, search

This text is part of: "I would never send my kids to school" by Piotr Wozniak (2017)

Two-component model of memory

Long-term memory is characterized by two essential variables: stability and retrievability (see: two-component model of long-term memory). New data from molecular neuroscience hints at the possibility that memory stability itself has two components: molecular and structural. The two-component model of memory is based on statistical data obtained in the process of learning based on spaced repetition. The new hypothesis on the two components of stability is largely theoretical and hard to confirm with presently available statistical data.

Two-component model of stability

Historically, we have always hypothesized that memory stability would be based on a buildup of proteins in the PSD. Over decades, AMPA receptors became the strongest candidates for that role and the evidence keeps adding up. In parallel, research has shown more clearly that new synaptic connections may form in the wake of learning. New connections seem too costly to play a role of propping up memory stability. Protein buildup is simply cheaper in metabolic terms and easier to mold with evolutionary forces. New connections, on the other hand, could play role in effectual connectivity. New concept maps established in memory should have the ability to establish more effective connections that could serve information processing. In other words, new synapses would contribute to connectivity, not to stability. However, a side effect of new connectivity might be increase in coherence which in turn is known to have a dramatic impact of survival of memories. Stability might be propped up in processes that occur in sleep, and which seem to depend highly on memory coherence.

This leads to a hypothesis that stability may have two components: molecular and structural. At this point, we have no way to measure those components outwardly other than in full convolution. Properties of memory in spaced repetition make it only possible to see stability as a single variable. However, it is worth noting that stability tends to increase dramatically with review early in the process, and approach asymptotically very low levels of increase decades later. This could be explained by the fact that a memory finds its place in the overall knowledge structure of the individual, achieves high stability, and is then minimally responsive to further retrieval. In other words, new memories might stimulate sprouting of new synaptic connections, while continuing the build up of receptors responsible for molecular stability.


  • in two component model of long-term memory, stability might itself be a two process phenomenon: molecular and structural
  • optimum timing of review will have a decreasing impact on memory stability over the lifetime of a memory
  • high stability memories may be best established with the help of: optimum review, high coherence, and sleep, which may have best effect due to the combined impact of both review and coherence