Abstract
Though much evidence supports object based attention, how an object is defined is still uncertain (Logan, 1996). Most work on object based attention uses bottom-up factors to define objects. In a classical demonstration, Egly, Driver, Rafal (1994) used two parallel rectangles arrayed horizontally or vertically, to contrast space-based and object-based attention. In a critical comparison of the invalid cue conditions, they found that RT was faster when the target was in the cued rectangle than when the target was in the other rectangle, though the distance from the cue and the target was the same in these two conditions. Using the same paradigm as that used in Egly et al. (1994), we found an object-based attention effect when the objects were defined top-down by Chinese words. We presented four Chinese characters spaced equally in the four corners of a square to control bottom-up organization. Each display contained two two-character words, arrayed horizontally or vertically. We cued one of the characters and presented a test probe in the cued location (cue valid) on 75% of the trials. On 25% of the trials, the probe appeared in an invalid location in the same word or in the other word. If skilled Chinese readers perceived these words as objects, RT on cue-invalid trials should be shorter when the target is in the cued word than when it is in the other word. The results confirmed this hypothesis. RT on cue-invalid trials was shorter when the target was in the cued word than when it was in the other word. Since we controlled for bottom-up organization, our results show that top-down factors alone can define objects and constrain the deployment of attention.
This research was supported by grant number R01-MH073879-01 from the National Institute of Health and by grant number FA9550-07-1-0192 from the Air Force Office of Scientific Research.