The contribution of meaning in forming holistic and segmented based visual representations

Abstract

Three paradigms were used to compare performance across meaningful and meaningless stimuli. Within each type of stimulus the relative performances with simple and complex objects were examined. The complexity was used as an analytical tool. In a segment-based representation, with relatively independent parts, there would be more parts in the representation of a complex stimulus that in that of a simple stimulus. In holistic representations the parts would be less independent; the number of parts would be less influential that the relationship among them.

The first study used a mental rotation task. A segmented representation would show an interactive effect of orientation and complexity, whereas a holistic representation would not. The findings suggested that the meaningful stimuli were rotated part by part, whereas the meaningless objects were rotated holistically. The second study used a part search task. It was predicted that the dependence among the parts in a holistic representation would result in a greater difference in performance across complexities than in a segmented representation with independent parts. The complexity of the stimulus showed a greater difference in the meaningful stimuli than in the meaningless stimuli. Assuming that holistic representations made more use of configural information than segment-based representations, the final study tested the contribution of configural information to the representation during a binary forced-choice probe task. The meaningful stimuli showed a greater advantage when configural information was present, especially the complex stimuli, relative to the meaningless stimuli.

The findings suggested that meaningful objects were represented more holistically than comparable meaningless objects; this difference is greater in the complex stimuli.

More information

Identifiers

Catalogue record

Export record

ASCII CitationAtomBibTeXData Cite XMLDublin CoreDublin CoreEP3 XMLEndNoteHTML CitationHTML CitationHTML ListJSONMETSMODSMPEG-21 DIDLOpenURL ContextObjectOpenURL ContextObject in SpanRDF+N-TriplesRDF+N3RDF+XMLRIOXX2 XMLReferReference ManagerSimple Metadata