Oded Bein: The direction of associations: prior knowledge promotes hippocampal separation, but cortical assimilation
How do we learn new information? Typically, we do not learn something anew, but rather we add a piece of information to our existing knowledge. For example, we learn that our best friend started dating a new person, or that our favorite singer has a new song out, and so we rush to hear it hoping we will like it. How does our brain add this new chunk of information to our previous knowledge? To ask this question, we repeatedly presented to our participants pairs of faces: either a famous face for which we have a lot of knowledge, for example, Madonna, with a novel and unfamiliar face, or two novel faces, while we scanned their brains. Through this repeated exposure, participants learned to associate Madonna and the novel face, or the two novel faces. This way, we imitated a real-world situation in the lab, like learning that Madonna has a new friend, compared to learning that two unfamiliar people became friends. We found that the hippocampus, a brain region important for making new memories, made distinct representations of Madonna and her friend. In other words, after learning, brain activity in the hippocampus looked very different for Madonna versus her friend. That is in contrast to brain activity for the two novel faces, which became similar after learning. Meanwhile in the cortex, another brain region that is important for storing our prior knowledge, we found that brain activity for Madonna and her new friend were actually more similar after learning, suggesting that new information was incorporated in our prior knowledge structures in the cortex. This tells us that adding new information to prior knowledge is a very different learning processes from learning something completely novel, that leads to different patterns of brain activity in different parts of the brain. It also tells us that different brain regions are doing different things to allow us to learn new information: the hippocampus has distinct activation patterns, while the cortex has similar activation patterns.
Article can be found here: https://www.biorxiv.org/content/10.1101/851204v1
Oded Bein (ob598@nyu.edu)