Forget all that nonsense: The role of meaning during the forgetting of recollective and familiarity-based memories

- The neocortex integrates related representations in overlapping neural codes.
- The hippocampus codes sparsely, keeping representations independent in neural coding.
- Semantics should influence the ability to retain information in the neocortex.
- We find words may be more easily retained than nonwords in neocortex (familiarity).
- We find words and nonwords are equally retainable by the hippocampus (recollection).

Memory can be divided into recollection and familiarity. Recollection is characterized as the ability to vividly re-experience past events, and is believed to be supported by the hippocampus, whereas familiarity is defined as an undifferentiated feeling of knowing or acquaintance, and is believed to be supported by extra-hippocampal regions, such as the perirhinal cortex. Recent evidence suggests that the neural architectures of the hippocampus and neocortex lead information in these regions being susceptible to different forgetting processes. We expand on these accounts and propose that the neocortex may be sensitive to the semantic content of a trace, with more meaningful traces being more easily retained. The hippocampus, in contrast, is not hypothesized to be influenced by semantics in the same way. To test this new account, we use a continuous-recognition paradigm to examine the forgetting rates words and nonwords that are either recollected or familiar. We find that words and nonwords that are recollected are equally likely to be forgotten over time. However, nonwords that are familiar are more likely to be forgotten over time than are words that are familiar. Our results support recent neuropsychologically-based forgetting theories of recollection and familiarity and provide new insight into how and why representations are forgotten over time.