Typing “Yoda” into Google will autopopulate a list of predictive searches, one of which inevitably includes Luke Skywalker. It’s no surprise that, to the internet, they’re an obvious pair. But scientists have just discovered that our brains perform their own “searches” in much the same way.
It’s already known that neurons in the brain fire off when we see certain things that we recognize, but we know less about how memories triggered by facts and events become cognitively associated as they’re encoded in the brain’s medial temporal lobe. New research shows that the brain works a lot like a search engine, storing up memories and rapidly making meaningful associations between concepts.
In the study, published Tuesday in Nature Communications, researchers describe how they studied the responses of medial temporal lobe (MTL) neurons in neurosurgical patients when they were asked to think of concepts, like people and places. They wanted to find out whether or not memories that triggered multiple associative memories represented random coincidences or meaningful connections. Researchers showed 49 people 100 pictures of celebrities and asked them to rate how closely related a variety of images were. To compare, the researchers plugged in the same celebrities into Google and Bing to see how the internet paired off celebs.
Turns out that their search engine hypothesis was right. If, for example, one of the subjects was shown different pictures of Luke Skywalker, a similar set of neurons fired up for each picture. But a photo of Skywalker also triggered neurons associated with the memories of similar characters like Yoda, Darth Vader, and Han Solo. When the researchers searched that celebrity name online, the search engines made the same associations between characters as the subjects did with their brains.
Even in your brain, you can't tear these two apart.
“We found that MTL neurons tend to fire to associated concepts, an effect that cannot be attributed to visually similarity between stimuli, familiarity effects, recall of associated items or broad semantic categorizations,” the researchers write. “Moreover, we show a non-topographically organized distribution of responses, which is ideal for the fast formation of new associations.”
The researchers told New Scientist that this “exciting study” brings further clarity to how “individual neurons store memories in a core memory center in the human brain.”
Additionally, this research raises the point that, while innovators continue to build computers modeled after the brain, the brain also acts more like a computer than researchers had previously realized.
