In three experiments, we examined the internal processing mechanisms of relatively independent visual-form subsystems. Participants first viewed centrally presented word pairs and then completed word stems presented beneath context words in the left or right visual field. Letter-case-specific priming in stem completion was found only when the context word was the same word that had previously appeared above the primed completion word and the items were presented directly to the right cerebral hemisphere. This pattern of results was not found when participants deliberately recollected previously presented words when completing the stems. Results suggest that holistic processing, not parts-based processing as assumed in many contemporary theories of visual-form recognition, is performed in a subsystem that distinguishes specific instances in the same abstract category of form and that operates more effectively in the right hemisphere than in the left hemisphere.The ability to distinguish specific instances in the same abstract category of visual form is an important human faculty. One can differentiate a particular cup from other cups as well as an individual's signature from other written versions of the same set ofletters. An interesting aspect of this ability is that most theories of visual-form recognition, aside from theories that address only face recognition (e.g., Bruce & Young, 1986;Rhodes, 1985), do not focus on how it is accomplished. Most theories focus instead on the ability to distinguish different abstract types of visual forms (e.g., Biederman, 1987;Morton, 1979). Of course, the latter ability is also important and interesting. One can distinguish a cup from a pen, as well as the word form "cup" from the word form "pen," even when the input is an instance that has not been seen before. In the research reported in this article, we tested whether at least relatively independent processing subsystems support the two abilities. More importantly, we examined form-specific priming for novel holistic structures to test whether holistic processing, as opposed to parts-based processing of the sort assumed in many contemporary theories of visual form recognition (e.g., Biederman, 1987;Morton, 1979), characterizes the operations in a subsystem that distinguishes specific instances in an abstract category of form.
VISUAL-FORM SUBSYSTEMSWe hypothesize that a specific visual-form (SVF) subsystem underlies recognition of specific instances of forms and operates more effectively in the right cerebral hemisphere (RH) than in the left cerebral hemisphere (LH). In contrast, an abstract visual-form (AVF) subsystem supports recognition of abstract categories of forms and operates more effectively in the LH than in the RH. These subsystems likely focus on different properties of visual-form inputs to achieve different goals.
SVF SubsystemIn order to produce different outputs when different instances in the same abstract category appear as inputs, an SVF subsystem must process a form's visually distinctive information ef...