Is subliminal priming visual? Is it translationally invariant?

Moshe Bar, Irving Biederman

Research output: Contribution to journalArticlepeer-review


Purpose. A number of experiments have demonstrated that words that are not recognized on a first, very brief, exposure are, nonetheless, recognized on subsequent exposures. Does this phenomenon extend to objects? If so, is this priming visual (i.e., specific to the shape of the image) or is it conceptual? If the object is presented at a different position in the visual field from where initially presented, is the magnitude of the priming reduced? Method. Following several training trials, subjects viewed 25 line drawings of common objects that could be readily given a basic-level name. The objects averaged 5.7° in extent. The images were exposed for 14 msec each, at one of 9 locations with an average separation of 10.7°. Each object was immediately followed by a mask designed for that particular object. Subjects often claimed that they did not see anything or were just guessing. After each of the 25 objects were shown once, there were 25 additional presentations of objects that bore one of the following relations to the objects in the first block: 1) identical image, same location, 2) identical image, different location, 3) similar-shaped exemplar, same location, 4) similar-shaped exemplar, different location, and 5) different names (and basic level concepts). Approximately 10 minutes and a mean of 25 presentations elapsed between the first and second presentation of an object. Results. Repetition of the identical image resulted in a 20% increase in accuracy. There was no effect of changing the position of the image. The similar exemplars showed a gain of approximately 5%, a value that was approximately equal to that for the different name objects. Conclusions. The large priming advantage of identical over similar-shaped images indicates that the priming was visual, and not just verbal or conceptual. That the priming was translationally invariant over visual angles that could reach 15.8° suggests that the representation mediating the priming might be in areas that have relatively large receptive fields, such as V4 or IT.

Original languageEnglish
Pages (from-to)S178
JournalInvestigative Ophthalmology and Visual Science
Issue number3
StatePublished - 15 Feb 1996
Externally publishedYes


Dive into the research topics of 'Is subliminal priming visual? Is it translationally invariant?'. Together they form a unique fingerprint.

Cite this