Abstract
We intuitively believe that when we attend to an attribute (e.g., color or identity) of a highly visible stimulus, we will be able to report it immediately afterwards. Conventional theories of cognition also suggest attention as a gateway to working memory (e.g., Awh, Vogel, & Oh, 2006), or regard working memory as an interface in which attention can operate in the absence of visual input (Chun, 2011). Despite extensive evidence showing the crucial contribution of attention in the storage and maintenance of information, recent experiments have revealed striking demonstrations that a more important determinant of robust memory storage is the expectation of the observer that the attended information needs to be remembered (Chen & Wyble, in press at Psychological Science; Eitam et al., 2013). In our experiments, participants were asked to locate a target among distractors by using one particular attribute (e.g., letter identity) for several repetitions and were then surprisingly asked to report that attribute. The results showed that participants could not report the attribute of the target, even though that attribute had been attended and used by participants to perform the primary task (e.g., localized a target letter among digits) just prior to the surprise question. Followup control trials ensure that this effect is not a failure of perception. We term this effect attribute amnesia. In an extension, participants counted passes of a colored ball in a video recorded from a real-world context. In a surprise trial, 35% of participants could not report the color of the ball they had just been tracking for 40 seconds. These results change our understanding of the link between attention and memory. Instead of memory being a record of attended information, we must view memory storage as a resource that is used sparingly in strict accordance with the goals of the observer.
Meeting abstract presented at VSS 2015