Honeybees May Understand Numbers Better Than Critics Thought

We're still trying to understand how exactly to study bees.

Bees are better at math than you think. They can count, grasp the concept of zero, and even understand symbolic patterns like Morse Code. But not everyone is convinced.

One previous critique argued that honeybees may not understand numbers at all. Instead, the insects might be taking a visual shortcut, relying on texture, edges, density, or other patterns that tend to change as more objects are added to an image.

Now, a new study pushes back. The researchers argue that the critique missed a crucial point: bees don’t see like humans. When the same visual patterns were reanalyzed through the limits of honeybee vision, the supposed shortcut largely vanished.

It may sound like a small technical dispute. But it gets at a much bigger problem: studying animal minds is incredibly hard, especially when those animals experience the world so differently from us.

How Do We Study Bees?

We used to think intelligence required a mammalian brain, or at least something strange and complex like the nervous system of an octopus. But in recent years, bees have forced scientists to rethink what “smart” really means.

Bees don’t have human-like intelligence. They also differ from mammals in many important ways. But study after study suggests that their tiny brains can learn rules, compare quantities, navigate complex environments, recognize patterns, and sometimes solve tasks that look surprisingly abstract.

In some experiments, honeybees have been trained to choose “more” or “less,” discriminate between small quantities, and even respond meaningfully to an empty set. That is where the idea that bees understand zero comes from.

But these studies are delicate.

×

Get smarter every day...

Stay ahead with ZME Science and subscribe.

Daily Newsletter
The science you need to know, every weekday.

Weekly Newsletter
A week in science, all in one place. Sends every Sunday.
No spam, ever. Unsubscribe anytime. Review our Privacy Policy.

Thank you! One more thing...

Please check your inbox and confirm your subscription.

Bee studies usually work by training free-flying honeybees to choose between visual stimuli in a controlled setup, often a Y-maze or rotating-screen apparatus. The bee sees patterns with different numbers of shapes, and one choice is rewarded with sugar water while the other may be unrewarded or paired with a bitter solution. Bees like the sugar so they try to get it right.

Over repeated trials, researchers test whether the bee learns a rule such as “choose more” or “choose fewer,” then they change the shapes, sizes, spacing, or total area to see whether the bee is responding to the number itself or just visual shortcuts like brightness or density. Researchers try to eliminate everything that could skew the measurements or influence bees.

That is the heart of the critique. Some researchers have argued that bees might not be responding to number at all. More items can mean more visual “busy-ness,” more edges, or more contour. A bee could, in theory, learn that instead of learning numbers.

The new paper, published in Proceedings of the Royal Society B, takes that criticism seriously. But it adds an important twist: what matters is not how the images look to us. It is how they look to bees.

RelatedPosts

These Cockatoos Prepare Their Food by Dunking it Into Water
Could Claude Monet See in UV Like a Bee? Something Strage Happened After His Eye Surgery
A colony of bees has been discovered living in the lip of an active volcano
Most people hate wasps — but they deserve to be loved just as much as bees

Bee Sight

Spatial frequency sounds technical, but the idea is simple. Wide black-and-white stripes have low spatial frequency. Thin, closely packed stripes have high spatial frequency.

Human eyes can resolve fine patterns that blur together for many animals, while bees sit on the blurry side of that divide. The new study points to classic work showing that free-flying honeybees struggle to resolve patterns around 0.4 cycles per degree of visual angle or higher, with performance already dropping sharply near 0.25 cycles per degree. In plain language: bees are best at seeing coarse patterns, not fine ones.

The new study was mostly a reanalysis paper, not a new bee-behavior experiment.

The earlier critique had removed the lowest spatial frequencies from its analysis, but those are exactly the kinds of broad visual patterns bees are most able to see. So, the new authors converted the images into bee-relevant visual units, recalculated the old stimuli while keeping those low frequencies in, and found that the critique’s arguments disappear. In other words, the visual shortcut no longer explained the bees’ choices, making it more likely that the bees were responding to the number.

In other words, the critique had the right idea to check, but didn’t implement it properly. Bees can count, but probably not like humans count. This study says bees likely process numerosity, a basic sense of “more,” “less,” and even “none,” rather than having a refined understanding of numbers.

“There has been a debate about whether bees are really ‘counting’ or just reacting to visual patterns. Our results show that this criticism doesn’t hold when you consider the biology of the animal,” said Dr. Mirko Zanon, from the Center for Mind Brain Sciences at the University of Trento and first author on the study. “When we analyze the stimuli in a way that reflects how bees actually see the world, what remains is actual sensitivity to number.”

Science Is Really Complicated. So Are Animal Minds

All these experiments, analyses, and reanalyses show one thing: in science, proving something is rarely straightforward.

A bee’s choice at the end of a maze can look like evidence of number, or evidence of texture, or evidence of something else researchers have not yet thought to measure. Each interpretation should be analyzed critically and put to the test. A claim survives only by passing through doubt, correction and sharper tests. This is the absolute core of scientific progress, it’s the whole point of the entire process.

But ultimately, the paper is not only about bees. It’s a warning about animal cognition experiments in general.

A task that seems simple to a human observer may look very different to another species. A bird, fish, insect or mammal brings its own sensory world to the experiment. It may see different colors, resolve different details, hear different frequencies or move through the apparatus in ways that change the information available to it.

For that reason, the authors call for a “biology-aligned” approach. Stimuli should be designed and analyzed with the animal’s perceptual limits in mind. For bees, that means accounting for compound-eye resolution, viewing distance and the spatial frequencies they can actually detect. For other animals, it may mean different constraints altogether.

Monash University Senior Lecturer Dr. Scarlett Howard concludes:

“We must put the animal’s perspective first when assessing their cognition or we may under or overestimate their abilities,” said Dr. Howard. We see and experience the world quite differently from animals, so we must be careful of centering human perspectives and senses when studying animal intelligence.”