Motion capture and visual attention: a reply to Ramachandran (1996)

In our previous paper, we proposed that when attention is used to track salient features of a moving object, other weaker motion signals can be captured and appear to move in synehrony with the tracked features (Culham & Cavanagh, 1994b). In one experiment, when observers used attention to track the bars of an ambiguous counterphase grating in either direction, superimposed dynamic luminance dots followed in the direction of tracking. This led us to suggest that motion capture may be attention-based.:} Ramachandran (1996) now provides examples which both support and contradict our claim of attentionbased motion capture. We are encouraged by Ramachandran ' s supporting examples of attentional shifts which cause motion capture and we provide further evidence that these shifts do not generate capture simply by modulating low-level motion signals. However, after first agreeing that attention plays a role in motion capture, Ramachandran then offers evidence for the contrary view that capture is preattentive and asks how the effect can be "both preattentive and attentionbased at the same time?" We do not find the evidence for preattentive motion capture (Plummer, 1992) compelling. Nonetheless, preattentive, low-level motion interactions may also be involved in motion capture, along with the attentional factors that we and Ramachandran have demonstrated. That is, attentive and preattentive influences on motion capture need not be mutually exclusive as suggested by Ramachandran ' s