An event sequence to remember: Abstract temporal structure influences memorability
Poster Presentation: Monday, May 19, 2025, 8:30 am – 12:30 pm, Banyan Breezeway
Session: Visual Memory: Memorability
Schedule of Events | Search Abstracts | Symposia | Talk Sessions | Poster Sessions
Niels Verosky1, Brian Scholl1; 1Yale University
Much work in recent years has explored how some stimuli are intrinsically more memorable than others, often due to distinctive sensory or semantic features. But what determines memorability beyond such features, e.g. in sequences of musical tones? Here we explored the possibility that memorability also depends on abstract temporal structure, in both vision and audition. Participants were presented with short temporal sequences of three items varying along a scalar dimension--either three successive tones of different pitches (in auditory experiments) or three successive circles of different sizes (in visual experiments). Items were sampled from five fixed points along the relevant scalar dimension, creating a combinatorial space of 60 possible sequences in each modality, with a one-to-one mapping between analogous auditory and visual sequences. (Examples of possible sequences would thus be 1-2-3, 1-3-2, 1-3-5, and 5-2-4--with the items 1-5 mapping onto either the pitches of tones or the sizes of circles.) To comprehensively characterize memorability, we presented each participant with sequences that were randomly sampled from the full combinatorial space. During stimulus presentation, participants made an orthogonal judgment with no mention of memory. Subsequently, they completed a surprise recognition task. Though all sequences were constructed from the same small library of meaningless items, some were nevertheless reliably more memorable--and some particular sequences were exceptionally memorable across both modalities. For example, among all sequences in the combinatorial space, those that stood out as especially memorable were monotonically increasing sequences consisting either of three consecutive items (1-2-3, 2-3-4, 3-4-5) or of the lowest-magnitude plus the two consecutive highest-magnitude items (1-4-5). Intriguingly, this same pattern emerged independently for both vision and audition. This work thus adds a new dimension to the study of memorability: beyond distinctive sensory and semantic properties, what we incidentally remember is also shaped by abstract temporal structure.