Scientists discover bees understand the concept of zero

June 7, 2018, RMIT University
Trained to pick the lowest number out of a series of options, a honeybee chooses a blank image, revealing an understanding of the concept of zero. Credit: RMIT University

Scientists have discovered honeybees can understand the concept of zero, putting them in an elite club of clever animals that can grasp the abstract mathematical notion of nothing.

By demonstrating that even tiny brains can comprehend complex, , the surprise finding opens possibilities for new, simpler approaches to developing Artificial Intelligence.

In research published in the journal Science, Australian and French researchers tested whether honey bees can rank numerical quantities and understand that zero belongs at the lower end of a sequence of numbers.

Associate Professor Adrian Dyer, from RMIT University in Melbourne, Australia, said the number zero was the backbone of modern maths and technological advancements.

"Zero is a difficult concept to understand and a mathematical skill that doesn't come easily—it takes children a few years to learn," Dyer said.

"We've long believed only humans had the intelligence to get the concept, but recent research has shown monkeys and birds have the brains for it as well.

"What we haven't known—until now—is whether insects can also understand zero."

As well as being a critical pollinator, the honeybee is an exceptional model species for investigating insect cognition, with previous research showing they can learn intricate skills from other bees and even understand abstract concepts like sameness and difference.

Trained to pick the lowest number out of a series of options, a honeybee chooses a blank image, revealing an understanding of the concept of zero. Credit: RMIT University

But bee brains have fewer than 1 million neurons—compared with the 86,000 million neurons of a human —and little was known about how insect brains would cope with being tested on such an important numeric skill.

RMIT Ph.D. researcher Scarlett Howard set out to test the honeybee on its understanding, marking individual honeybees for easy identification and luring them to a specially-designed testing apparatus.

The bees were trained to choose an image with the lowest number of elements in order to receive a reward of sugar solution.

For example, the bees learned to choose three elements when presented with three vs. four; or two elements when presented with two vs. three.

When Howard periodically tested the bees with an image that contained no elements versus an image that had one or more, the bees understood that the set of zero was the lower number—despite never having been exposed to an "empty set".

Schematic representation of how over a period of time bees learn to choose between combinations of numbers such that the lower number is correct, and then when presented with a problem of zero elements versus the higher numbers bees understand that zero is at the lower end of a numerical sequence. Credit: Composite image by Scarlett Howard, Jair Garcia and Adrian Dyer

Dyer, a researcher in the Bio Inspired Digital Sensing-Lab (BIDS-Lab) in RMIT's Digital Ethnography Research Centre, said the findings opened the door to new understandings of how different brains could represent zero.

"This is a tricky neuroscience problem," he said.

"It is relatively easy for neurons to respond to stimuli such as light or the presence of an object but how do we, or even an insect, understand what nothing is?

"How does a brain represent nothing? Could bees and other animals that collect lots of food items, have evolved special neural mechanisms to enable the perception of zero?

"If bees can learn such a seemingly advanced maths skill that we don't even find in some ancient human cultures, perhaps this opens the door to considering the mechanism that allows animals and ourselves to understand the concept of nothing."

An individually marked honeybee inspects stimuli with either 3 or 4 elements before choosing to land on the correct “lower” number. After learning this type of rule with many combinations, bees understand that an unfamiliar presentation of an empty set is less than any other number of elements presented. Credit: Scarlett Howard

One of the problems in the development of is enabling robots to operate in very complex environments, Dyer said.

"Crossing a road is simple for adult humans, we understand if there are no approaching cars, no bikes or trams, then it is probably ok to cross," he said.

"But what is zero, how do we represent this for so many complex object classes to make decisions in complex environments?

"If bees can perceive zero with a brain of less than a million neurons, it suggests there are simple efficient ways to teach AI new tricks."

The research was conducted in both Australia and France and involved many control experiments to validate the findings.

Study co-author, Dr. Aurore Avarguès-Weber from the University of Toulouse in France, said: "The discovery that can show such elaborated understanding of numbers was really surprising given their tiny brain."

"Large brains are thus not necessary to play with numbers. This capacity is therefore probably shared by many other animals."

The paper "Numerical ordering of zero in honeybees" is published in Science today.

Explore further: Inside the brains of killer bees

More information: S.R. Howard el al., "Numerical ordering of zero in honey bees," Science (2018). science.sciencemag.org/cgi/doi … 1126/science.aar4975

Related Stories

Inside the brains of killer bees

June 6, 2018

Africanized honeybees, commonly known as "killer bees," are much more aggressive than their European counterparts. Now researchers have examined neuropeptide changes that take place in Africanized honeybees' brains during ...

New Insight Into How Bees See

January 23, 2009

New research from Monash University bee researcher Adrian Dyer could lead to improved artificial intelligence systems and computer programs for facial recognition.

Recommended for you

Researchers have discovered a new cell structure

October 23, 2018

A new structure in human cells has been discovered by researchers at Karolinska Institutet in Sweden in collaboration with colleagues in the U.K. The structure is a new type of protein complex that the cell uses to attach ...

Breakthrough test screens for all known bacterial infections

October 23, 2018

Scientists at the Center for Infection and Immunity (CII) in the Columbia University Mailman School of Public Health have developed the first diagnostic platform that can simultaneously screen for all known human pathogenic ...

New technique promises more accurate genomes

October 23, 2018

University of Adelaide researchers have developed a new technique that will aid in a more accurate reconstruction of human genomes by determining the exact sections of the genome that come from each parent.

28 comments

Adjust slider to filter visible comments by rank

Display comments: newest first

ryuuguu
5 / 5 (4) Jun 07, 2018
I wonder if they could be tested for addition and negative numbers using one colour of dots for +1 and and another for -1 . Put the reward on the target with the largest positive sum. Then on to multiplication. I have no idea how to code the axiom of choice in bee :)
dsylvan
3.4 / 5 (5) Jun 07, 2018
Or are the bees simply recognizing "less" or "fewer"? "I get reward if I go to the place with the least dots."
ryuuguu
4.6 / 5 (5) Jun 07, 2018
I think recognizing that 0 is less then 1 was the point of this part of the experiment.That "place with the least dots" includes 0 dots is not a given, the smallest number could be 1.
4johnny
1.7 / 5 (3) Jun 08, 2018
For the record: Zero is not nothing.
ZoeBell
Jun 08, 2018
This comment has been removed by a moderator.
bruce33333
4.2 / 5 (5) Jun 08, 2018
I would have to take a much closer look at the protocol of these experiments to see if I would agree with the researchers. The humans here might be reading at little more into what they are seeing, or projecting a bit more from their own minds than they realize. For example, the relative locations of the cards to each other, and those locations relative to other objects nearby, and the light sources in the environment, might matter. And what about the nature of the "objects" supposedly being "counted?" A blank card doesn't have anything on it. This can mean other things besides representing a "zero" numerically. And what about the area ratio of the dark and light sections on each card? This might be very important. And this is just off the top of my head right now. There are many variables here that must be considered and accounted for, and I'm skeptical of their conclusion.
Eikka
3.3 / 5 (4) Jun 08, 2018
Or are the bees simply recognizing "less" or "fewer"? "I get reward if I go to the place with the least dots."


Who says it counts the dots? Maybe it's counting the area of white paper versus black paper, so what it's actally doing is "more white is better".

And, even if it is counting the dots, what could be happening is that rather than understanding it as a number, it understands the problem as "avoid black", because the right option always has less black, whether it's in area or in count, so, a paper with no black is the right answer regardless of whether you understand it as "zero".

That is a far simpler solution to the problem, and much more likely to happen.
kshanholtzer
5 / 5 (3) Jun 08, 2018
And what about the area ratio of the dark and light sections on each card?
I was thinking the same thing, though more along the lines of "the brightest area" will be more likely to have food, since bees are probably far more likely to respond to brightness than recognizing "least" or "none". I'd like to see this reproduced with a dark background and white dots. I'm willing to bet the results will disprove this study. I haven't read the source journal, so maybe they did do that...
ZoeBell
Jun 08, 2018
This comment has been removed by a moderator.
Cusco
3.7 / 5 (6) Jun 08, 2018
Maybe it's counting the area of white paper versus black paper


If you click the third image it appears that they've gone to some effort to randomize the percentage of white space in the images with different size dots.
ZoeBell
Jun 08, 2018
This comment has been removed by a moderator.
Ojorf
3.4 / 5 (9) Jun 09, 2018
Gee people, ya don't think these scientists thought to check if the bees are really counting or just looking at light or dark?
Turns out they really do count, no matter the size of the dots compared to the background or the lightness/darkness/colours.

Bees can count to four!
DavidMartin
1.7 / 5 (6) Jun 09, 2018
The researchers lack the psychological understanding required for the research. As a result they're projecting complex human thought onto the bees. But they could get the understanding they need from the human subconscious mind, which is similar to the bees' mind. The concepts only become complex when we have to define them - express them in words. Then they have to go via the rational mind, and you get mathematical concepts, and a whole lot of other complicated stuff. These French people didn't know that 'l'abeille n'a pas besoin de DIRE ce qu'il va faire, il a besoin seulement de le faire', and that's different. For that, he doesn't need concepts like zero. He needs a neural net that gets a holistic picture of the problem.
IconicOne
3.7 / 5 (3) Jun 09, 2018
This data does not accurately represent the recognition of nothingness as a mathematical zero, this data only suggests the possible recognition of < less than...or even > more than.
marshach
2.3 / 5 (3) Jun 10, 2018
I want to add my voice to those who question the interpretation of the bees' behavior. How do the researchers know the bees aren't simply distinguishing the area of white space rather than counting?
ZoeBell
Jun 11, 2018
This comment has been removed by a moderator.
Roland Giersig
1 / 5 (1) Jun 12, 2018
As has been mentioned by others, there are other explanations to the bees behaviour than "understanding zero". Maybe they just learn "black dots are to be avoided". This would also lead them to choose the empty slate. Occams razor says that this is a better answer than "bees unterstand zero".
Roland Giersig
not rated yet Jun 12, 2018
To bee or not to bee - that is the question

All we are saying
is "Give bees a chance"!!
dsylvan
1 / 5 (1) Jun 12, 2018
I think recognizing that 0 is less then 1 was the point of this part of the experiment.That "place with the least dots" includes 0 dots is not a given, the smallest number could be 1.


The concepts of "less" and "more" are the foundation of counting, with "none" being the least possible state. I'm suggesting that the bees can recognize, i.e. count, "none" without recognizing the far more sophisticated concept of "zero".
jonesdave
2.6 / 5 (5) Jun 12, 2018
Turns out they really do count, no matter the size of the dots compared to the background or the lightness/darkness/colours.

Bees can count to four!


Wow! They should sign up for the Electric Universe cult. That would put them at the upper end of mathematical achievers.
Ojorf
1 / 5 (3) Jun 13, 2018
Turns out they really do count, no matter the size of the dots compared to the background or the lightness/darkness/colours.

Bees can count to four!


Wow! They should sign up for the Electric Universe cult. That would put them at the upper end of mathematical achievers.


This paper does not seem quite conclusive.....BUT, if bees DO get the concept of zero, it would definitely nudge them ahead of the EU pack.
granville583762
4 / 5 (4) Jun 13, 2018
The relevant point is never having been exposed to an empty set!

[Trained BEES choose an image with the lowest number of elements in order to receive a reward of sugar solution. When periodically tested the BEES with an image that contained no elements versus an image that had one or more]

Trained BEES never seeing an empty set, the researchers must have less brain cells than bees
How do untrained Bees react to empty sets versus full sets?
How do bees react to the colouring of the set and its colouring difference when full?
BEES react to colouring and symbols also infra-red light – BEES have a completely different concept to remembering and recalling as their BEE dance demonstrates, you will find the BEES response to the blank canvas is more to do with how the BEES visually visualise the pigmentation of the canvas in their compound eyes rather than any concept of numeracy that we perceive!
granville583762
5 / 5 (3) Jun 13, 2018
THIS IS WHAT YOUR BEE IS ACTUALLY DOING
After rewarding your bee it does the wiggle dance to inform all the other bees the new source of honey!
Bees, after visiting old sites, they visit new sites for fresh honey, so after rewarding the bee it is going to fly to its hive to do its wiggle dance to tell all the other bees there is this weird human feeding bees for visiting blank canvas's!
granville583762
3.7 / 5 (3) Jun 13, 2018
So in actuality the bee is not counting, nor has it any concept of zero - it is simply visiting new sources of honey it has not already visited
granville583762
5 / 5 (3) Jun 13, 2018
Oh! To Bee a Bee!
May be the researchers should become a bee for the day, to get a real life hands on experience as to what it is actually like to bee a bee.
KurtFitzner
5 / 5 (1) Jun 13, 2018
Insects don't have the same kind of image processing or, indeed, even the same kind of eyes we do. The black spots that are being "counted" are quite large, and make up a sizable percentage of each page. Humans look at those pages and our image processing, which happens in the background, separates out the individual spots automatically. Bees see very differently, and do not have that kind of sophisticated processing. The number of neurons we have dedicated to that part of our brain exceeds by orders of magnitude the number of total neurons a bee has.

It is equally likely that the bees are being trained that where there is more white, there is more likely to be food. Considering that bees see well into the ultraviolet and that flowers that are white to us and which reflect into the UV (like most white paper will) are quite attractive to bees.

I see a huge interpretation bias here. People see spots and assume bees do too. Not the case.
torbjorn_b_g_larsson
1 / 5 (1) Jun 25, 2018
Cool!

Also interesting that so many are repeatedly voicing the same problems, implying the find was not to everyone's taste.

Of course behavioral studies are problematic, but that bees have good visual acuity and can see trivial image elements such as spots are well established by establishing their counting ability.

The paper is here: https://www.resea...ney_bees

Its supplement tell us that area is controlled for (except for the zero card): "To control for surface area, each stimulus presented a pattern of elements culminating to a surface area of 10 ± 0.3 cm2 regardless of shape, pattern, or number of elements; and each element was above the minimum resolution threshold for honeybee vision as based on previous psychophysics findings (24)." Since the bees have increasing problems to recognize "fewer",it is easy to see that zero fits the trend, so is not are dependent either (fig. 2 of the paper).
torbjorn_b_g_larsson
1 / 5 (1) Jun 25, 2018
[ctd] One typical way to exclude bias is to let some uninformed persons collect and some crunch the data (double blinded test). It was not done here, but bee response was simply touching the landing platform so little if any bias could be introduced there.

More concerns:

@David Martin:
The researchers lack the psychological understanding
.

Irrelevant, it was a behavioral study.

@granville: Hive behavior is irrelevant, it was a study on individuals.

Please sign in to add a comment. Registration is free, and takes less than a minute. Read more

Click here to reset your password.
Sign in to get notified via email when new comments are made.