Properties of the organization of memory for people: evidence from dream reports.
Richard Schweickert
Psychonomic Bulletin & Review April 1, 2007 DOI: 10.3758/bf03194063 (opens in new tab)
Study at a glance
AI-extracted from the abstract| Characteristics | Observational network analysis of dream reports Peer reviewed |
|---|---|
| Sample size | 3 |
| Population | Dream reports from three individuals |
| Keywords | Humans Interpersonal relations Dreams Memory Semantics Vocabulary |
| Key findings | Argues that if short path lengths, high clustering, and power law degree distribution are general properties of memory organization, they should appear in relational networks of dream characters. Networks built from three individuals' dream reports showed all three properties, with the power law holding only when low degrees were omitted; one network had a tree-like outline distinct from the other two. |
Abstract
Steyvers and Tenenbaum (2005) showed that semantic networks for words have three organizational properties: short average path lengths, high clustering, and power law degree distribution. If these are general properties of memory organization, they would apply to memory for other complex material, including people and relations between them. In addition, if during dreaming, characters are generated via knowledge in the dreamer's memory, the three properties would be found in a relational network of characters in dreams. In dream reports from three individuals, two characters in the same dream were considered affiliated. Resulting social networks have the three properties, with the power law holding when low degrees are omitted. One network with a tree-like outline is different from the other two. Results suggest associative memory has the three properties, and demonstrate that dream reports are a potentially valuable source for information about social networks.