It’s been well established that people have “massive” memory for pictures and scenes—the ability to remember thousands of images after only a few seconds of exposure.
Now, researchers at Brigham and Women’s Hospital have shown more specifically that humans have a spatial massive memory—they can remember where they saw objects presented briefly—and a temporal massive memory—they can remember when they saw the objects.
Jeremy M. Wolfe, PhD, director of the Visual Attention Laboratory in the Department of Surgery at Brigham and Women’s Hospital, and colleagues describe their research and findings in Current Biology.
Massive Spatial Memory Storage
In the first experiment, 167 study participants viewed a grid of 7 x 7 squares in which five, 15, 25, or 49 squares featured pictures of diverse objects. Each item was highlighted in red for two seconds.
After the items were presented they disappeared and one new item appeared. Observers were tested on the following:
- Immediate recognition—They clicked a box that said “new item” if they didn’t remember seeing the item
- Location memory—If they thought they had previously seen the item, they clicked the square in the grid where they remembered seeing it
New screens of five, 15, 25, or 49 items were presented, and testing was repeated until the observer had seen 300 items. Many observers were able to recall the location of more than 100 items with an accuracy of ±1 cell.
In a second experiment, the participants viewed 200 items, all faces or doors (40 screens of five items each). Their performance was very poor, suggesting the accuracy of spatial memory depends on the stimuli. For example, spatial memory may be poor when items are similar to one another.
Massive Temporal Memory Storage
Twenty-three subjects participated in a third experiment where there were no separate viewing and testing periods. In the same grid, one item appeared at a random location for three seconds. When it appeared, observers clicked “new item,” if applicable, and the remembered location. New to this experiment, if they thought the item was “old” (previously seen), they were also prompted to estimate when the item first appeared.
Observers did that by clicking on a time bar. The beginning of the experiment was marked as zero, and a green progress bar grew longer as the experiment proceeded, indicating to observers their current temporal location. Observers were instructed to click the location on the bar corresponding to when the current ”old” item first appeared.
The researchers estimated 40% of items could be accurately placed in time by guessing, but many observers associated 60% to 80% of items to within ±10% of their correct time.
Spatial and temporal performance were not correlated, although that could have been an issue of statistical power.
Perfection Is Not Required
Food-hoarding birds and squirrels are still preeminent when it comes to specialized memory. However, this study clarifies that human spatial and temporal memory for objects can be massive.
If localization to the exact cell in the 7 x 7 grid had been required for a “correct” answer, estimates of spatial memory would have been much smaller. However, knowing that an object is “over there” is a genuine memory and, often, all that’s necessary to be useful for a real-world task.