Bees and the thought of naught

Authored by www2.cnrs.fr and submitted by drewiepoodle

Though some vertebrates had already been shown to master complex numerical concepts like addition and zero, no evidence of this existed for insects. Since bees can count to five at least, the researchers taught them the inequality relations “greater than” and “less than.”

First they trained them to drink sweetened water from an experimental setup where platforms were paired with images. Their task was simply to choose the image depicting the smallest number of elements. If they selected the correct one, they were rewarded with sweetened water. Otherwise, they got bitter quinine solution. Once the bees grasped the exercise, the researchers showed them two images at a time: one was blank (representing zero) and another had one or more dots (representing a whole number). The insects selected the blank image as representing the least number of elements. This shows they had extrapolated their understanding of “less than”—as applied to whole numbers (1, 2, 3, 4, or 5)—to zero, which they assigned the lowest rank of all.

Bees have only a million neurons (100,000 times fewer than humans), yet both species can grasp zero. Zero was a major mathematical breakthrough for humans and is a particularly abstract concept. We can use this symbol of “nothing” to designate the absence of objects, while manipulating it as if it represented a quantity. The brain, which evolved to process sensory stimuli, can also perceive the absence of stimuli as a construct itself. By suggesting that the concept of zero is useful even to pollinating insects, the scientists' discovery invites inquiry into its symbolic importance more generally.

Bibliography: Howard SR, Avarguès-Weber A, Garcia J, Greentree A, Dyer AG. Bees extrapolate ordered relations to place numerosity zero on a numerical continuum. Science. 2018. doi:10.1126/science.aar4975.

MadroxKran on June 9th, 2018 at 01:55 UTC »

What's the dance for zero?

gyroscape on June 8th, 2018 at 21:30 UTC »

I'm deeply skeptical of this claim. Based on the images that they used, it seems like there is a huge potential for error. It looks like images with a larger number of spots on them had much more black shading by area than other images.

So, the "zero" version was perciptly brighter than the "one" version, which was brighter than the "two" version, and so on.

How did they prove that the bees were not just being trained based on brightness, and were actually counting?

DirtysMan on June 8th, 2018 at 20:17 UTC »

tl;dr: First they trained them to drink sweetened water from an experimental setup where platforms were paired with images. Their task was simply to choose the image depicting the smallest number of elements. If they selected the correct one, they were rewarded with sweetened water. Otherwise, they got bitter quinine solution. Once the bees grasped the exercise, the researchers showed them two images at a time: one was blank (representing zero) and another had one or more dots (representing a whole number). The insects selected the blank image as representing the least number of elements. This shows they had extrapolated their understanding of “less than”—as applied to whole numbers (1, 2, 3, 4, or 5)—to zero, which they assigned the lowest rank of all.