September 2019
Volume 19, Issue 10
Open Access
Vision Sciences Society Annual Meeting Abstract  |   September 2019
The contributions of visual details vs semantic information to visual long-term memory
Author Affiliations
  • Kelvin Lam
    Department of Psychology, University of California, San Diego
  • Mark W Schurgin
    Department of Psychology, University of California, San Diego
  • Timothy F Brady
    Department of Psychology, University of California, San Diego
Journal of Vision September 2019, Vol.19, 292. doi:
  • Views
  • Share
  • Tools
    • Alerts
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Kelvin Lam, Mark W Schurgin, Timothy F Brady; The contributions of visual details vs semantic information to visual long-term memory. Journal of Vision 2019;19(10):292. doi:

      Download citation file:

      © ARVO (1962-2015); The Authors (2016-present)

  • Supplements

Visual long-term memory can store thousands of objects with significant detail. However, the contribution of semantic information to this memory for visual details remains relatively understudied. To investigate this issue, we utilized stimuli known as “texforms” – spatially-constrained textures generated from objects, which contain the same low- and mid-level feature information but obscure object identity (Freeman & Simoncelli, 2011; Long & Konkle, 2017). Since participants cannot recognize the texforms, they cannot use semantic information to inform memory. In Experiment 1, participants memorized texforms presented sequentially for 3 seconds each. After a delay, they were shown old and new images of texforms and had to judge how confident they were the image was old or new. Participants were able to remember texforms well above chance (d_a=0.45), suggesting they had some ability to remember visual information without semantic information; and this pattern was replicated in a control with verbal suppression at encoding. Experiment 2 aimed to discern the role of semantic information in memory. Specifically, semantic knowledge may improve long-term memory either (1) by directing attention and encoding resources to the relevant and informative features of a particular object, or (2) by providing a relevant and distinctive retrieval-cue (Eysenck 1979). Thus, in Experiment 2, texforms were primed with an image of the original object for 100ms at encoding. This allowed participants to briefly recognize the object from the texform but not to hold onto this interpretation indefinitely. With priming, participants remembered texforms, (d_a=0.49) but no better than without priming (compared to Exp. 1: t(38)=0.39, p=0.70). These results demonstrate a new method for isolating the contributions of different kinds of information in visual long-term memory. Moreover, they suggest that the benefit of semantic knowledge to long-term memory may arise more by providing a useful retrieval-cue than by affecting how items are encoded.

Acknowledgement: NSF CAREER BCS-1653457 to TFB 

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.