Humans are not the only species that can solve math problems. After learning that honeybees understand the basic concept of zero, researchers began a new study to figure out if bees could comprehend how to solve basic arithmetic. A study, conducted by a team from RMIT University in Melbourne, Australia, aimed to discern whether bees can do addition and subtraction. 

The Study

The researchers taught the honeybees to relate certain colors to mathematical symbols, such as addition and subtraction. While solving the math, humans use both their long and short term memories. The long-term memory relates the rules of adding and subtracting while the short-term memory is manipulating the given set of numbers. This is similar to the bees. Bees learn to recognize the difference in concept between adding and subtracting. They then used their short-term memories to solve the mathematical problem. To train the bees, they were brought to a Y-shaped maze. In the maze, the bees were given problems to solve. If they solved it accurately, they were rewarded with sugar water. Similarly, if they were wrong they were given quinine, which has a bitter taste. Once the bees learned about the positive reward, they repeatedly went back to solve the problem to get their nutrition which ultimately furthered their learning. As the bees flew into the maze they were given a problem and had to make a choice. If they saw yellow they were to add (blue meant subtraction) and were given two options to fly through. If they flew through the collect slot they were rewarded. If they got the problem wrong, the correction solution was written to teach the bees. Different problems were posted to prevent the bees from only flying on one side and simultaneously taught bees math. 

Importance

This study leads researchers to believe that there may be more non-human species who are capable of understanding arithmetic given the small size of the bee’s brain. Additionally, Adrian Dyer, RMIT’s Associate Professor, drew the conclusion that this study could relate to advances in AI. He summarized his studying saying, “If maths doesn’t require a massive brain, there might also be new ways for us to incorporate interactions of both long-term rules and working memory into designs to improve rapid AI learning of new problems.”

Resources:

https://www.sciencedaily.com/releases/2019/02/190206200358.htm

-Alyssa Schwertfeger ’20