Language can boost otherwise unseen objects into visual awareness

Gary Lupyan; Emily J. Ward

doi:10.1073/pnas.1303312110

Edited by James. L. McClelland, Stanford University, Stanford, CA, and approved June 25, 2013 (received for review February 20, 2013)

Abstract

Linguistic labels (e.g., “chair”) seem to activate visual properties of the objects to which they refer. Here we investigated whether language-based activation of visual representations can affect the ability to simply detect the presence of an object. We used continuous flash suppression to suppress visual awareness of familiar objects while they were continuously presented to one eye. Participants made simple detection decisions, indicating whether they saw any image. Hearing a verbal label before the simple detection task changed performance relative to an uninformative cue baseline. Valid labels improved performance relative to no-label baseline trials. Invalid labels decreased performance. Labels affected both sensitivity (d′) and response times. In addition, we found that the effectiveness of labels varied predictably as a function of the match between the shape of the stimulus and the shape denoted by the label. Together, the findings suggest that facilitated detection of invisible objects due to language occurs at a perceptual rather than semantic locus. We hypothesize that when information associated with verbal labels matches stimulus-driven activity, language can provide a boost to perception, propelling an otherwise invisible image into awareness.

Footnotes

↵¹To whom correspondence should be addressed. E-mail: lupyan{at}wisc.edu.

Author contributions: G.L. and E.J.W. designed research, performed research, analyzed data, and wrote the paper.
The authors declare no conflict of interest.
This article is a PNAS Direct Submission.
*This hit rate is substantially higher than the 50% QUEST threshold to which the staircasing procedure was set because participants’ performance tends to improve during the course of the experiment. The mean hit-rate for no-label trials during the start of the experimental session (first 20 trials) was not significantly different from 50%, t < 1.
^†Although Pylyshyn (1) explicitly excludes attentional effects as instance of cognitive penetration of perception, contemporary theories of attention—both spatial and feature-based—point to attention itself as modulation of perceptual processes (60).
This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1303312110/-/DCSupplemental.