Abstract
Are memory capabilities of humans and monkeys similar or does one species have superior abilities? In particular, does language afford better memory facilities? We compared monkey and human memory capabilities in a delay-match-to-multiple-item memory task. In each trial, a series of samples was presented and participants detected and responded to a repetition of any previous item seen in the same trial. Two difficulties are included that are not present in standard delay-match-to-sample tasks: The repetition can be for any image in the trial, not only for the first, so participants must remember all seen images. Secondly, repetition of images that appeared in a previous trial are not considered valid, and should be ignored. Thus, participants need remember all the images of the trial and if the current repeated image was seen in the current trial. In general, we used novel images that had not been seen before, introducing a few catch images from previous trials, which should be ignored. Performance Hit rate is similar for monkeys and humans, about 90% except for the longest trials. The False Positive (FP) rate is very different, however, about 80% for monkeys, 30% for humans, for images from the preceding trial. When monkeys are intensively trained with a limited set of images, they necessarily develop a reset mechanism allowing them to reject the frequent presentation of images seen in earlier trials. Interestingly, they are able to transfer this reset capability to task performance with novel (and catch) images, reducing the FP rate to below 20%. Thus, there is a surprising similarity between human and monkey performance following intensive training with a limited set of images, forcing acquisition of a reset mechanism. This similarity is found even for human performance without such prior training. We conclude that human participants have an inherent reset mechanism before visiting our laboratory.
Meeting abstract presented at VSS 2015