Purchase this article with an account.
Olivia Cheung, Jennifer Richler, Stewart Phillips, Isabel Gauthier; Does temporal integration of face parts reflect holistic processing?. Journal of Vision 2009;9(8):544. doi: https://doi.org/10.1167/9.8.544.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
Holistic processing of faces can be revealed in the composite task by failures of selective attention to a face half under instructions to ignore the other half. In the normal version of this task with both parts presented simultaneously, interference from the irrelevant half is due to holistic processing and not response interference (Richler et al., in press). However, failures of selective attention are also found when parts are separated briefly in time (Singer & Sheinberg, 2006; Anaki et al., 2007). Here we ask whether such temporal integration may reflect response interference, especially when irrelevant information is presented first. Participants learned to name faces, two “Fred” and two “Bob”. At test, composites were formed by top and bottom halves of different learned faces. The halves of each composite were presented either 50ms or 200ms apart. Participants named only the target halves. When the irrelevant half was presented 200ms or 50ms before the target half, naming was slower when the irrelevant half was from a different face with a different name vs. a different face with the same name, suggesting response interference. When the irrelevant half was presented 50ms after the target half, response interference was again observed, and naming was also slower when the irrelevant half was from a different face with the same name vs. the same face, indicative of holistic processing. No response interference or holistic processing was observed when the irrelevant half was presented 200ms after the target half. These results suggest that 1) temporally separated face halves are processed holistically only when the target half precedes the irrelevant half briefly (50ms), and 2) although the irrelevant half presented before the target half for up to 200ms is nonetheless processed, the face composite is not processed holistically and interference from the irrelevant half instead arises from response conflict.
This PDF is available to Subscribers Only