Short-term memory not 'all or nothing'
As a basketball player scans he court, looking for an opening, his brain captures the position of other players for a few seconds, allowing him to choose the best path to the basket. Without those few seconds captured in memory, he would forget where each player is as his attention shifts.
This visual short-term memory is like a "buffer" in a computer, allowing you to retain important pieces of information that will inform your future actions –whether it is on the basketball court, the freeway or just walking down the street. Visual short-term memory is usually thought to be limited by a "magical number" of objects that can be stored – usually about four, said Dr. Wei Ji Ma, assistant professor of neuroscience at Baylor College of Medicine. "According to this theory, any additional objects will be discarded completely."
In a new report published in the current issue of the Proceedings of the National Academy of Sciences, Ma and his colleagues describe a new theory that describes short-term memory as a graded or continuous resource that is distributed over all objects that need to be remembered. On top of that, it fluctuates.
Ma compares it to watering plants in a garden.
"The more plants you have, the less water each will receive. By analogy, the more objects you have to remember, the worse each object will be stored. It’s a matter of quality, not quantity," he said.
There is another effect going on as well.
"If I give you only little time to water the plants, not all of them are going to get the same amount. You cannot evenly spread your resource," he said.
The new theory is conceptually a big departure from the accepted view.
"You don’t remember everything perfectly or not at all," he said. "It’s a spectrum. On the extreme end, you are guessing. Or you can remember a little or you can remember near-perfectly."
He and his colleagues tested the new theory using computer tests. In one study, they showed a group of items with randomly chosen colors for just an instant and then asked them what color a particular item in the display was. As the numbers of items increased, the subjects’ ability to answer the question decreased.
In a similar experiment, they asked subjects to identify the orientation of a particular object. In two other experiments, they showed subjects a set of objects on the screen, then they showed another similar screen in which one item had been changed. They asked the subject to locate the change.
They evaluated each current theory about how short-term memory works against the results of these experiments. The theory of visual short-term memory as a continuous and fluctuating resource was consistent with the results. The other theories fell short of explaining the results.
"We have to move away from the notion of a magical number," he said. "We need to think of the quality of encoding in visual short-term memory – not the quantity," he said.
Understanding how short-term memory works might have implications for understanding impairments in diseases such as Alzheimer’s and schizophrenia. It might also help understand how such memory – which enables people to accomplish countless tasks – is affected by distraction or competing sensory inputs that can also make demands on people’s short-term memory.
"This is not necessarily the final word in the debate," said Ma. "It does open the debate and there may be proponents of other theories that can come up with new models."
Ma credits postdoctoral scholar Dr. Ronald van den Berg and graduate student Hongsup Shin with the bulk of the work. Others who took part include Wen-Chuang Chou and Ryan George, who was an undergraduate at Rice University during his contribution to the study.
Funding for this study came from the National Eye Institute (R01EY020958) and the Netherlands Organisation for Scientific Research.