There is significant work on how we recognize structurewithin objects and scenes that we have previously encountered. But wehave strikingly little understanding of how the visual system flexiblyrepresents spatial structure across arbitrary arrangements of objects(e.g. displays such as graphs or diagrams). I will review a set ofpossible visual mechanisms and argue for a counterintuitivepossibility - that relations across space are coded by changes in thelocation of the attentional "spotlight" over time.
Background papers:
http://dl.dropbox.com/u/47794/SpatialRelations_Cognition2012.pdf
http://dl.dropbox.com/u/47794/JON_XuFranconeri.pdf
