Yes, and the way this one was tested is the neatest of the three we have written about. Two researchers went into a Paris park in different colored lab coats. One fed the pigeons. One chased them off.
Then they swapped coats.
The birds were not fooled.
What the experiment did
Live Science describes the setup in one sentence: "At a park in Paris, two researchers of similar build and skin color, but wearing different-colored lab coats, fed a group of pigeons."
Then the two behaved differently. One kept feeding. The other, in the words of the release, "was hostile, and chased them away."
- Where
- A park in central Paris, with the local feral pigeons
- Who
- Two researchers matched as closely as people can be matched. Both women, and of similar age, build and skin color
- What differed
- One fed the birds. One drove them off. Nothing else about them was meant to be distinguishable
- What the birds did
- They learned to avoid the hostile one, and kept avoiding her even after she stopped chasing them
Matching the two people on sex, age, build and skin color is doing the same job that the rubber masks did in the crow work. It removes the easy explanations one at a time until the face is what is left.
The coat swap is the whole test
Feeding one bird and chasing another only shows that pigeons can tell a good day from a bad one. The interesting move was changing the most obvious visual cue in the scene.
The release puts the result plainly: "Swapping lab coats during the experiments did not confuse the pigeons."
Dalila Bovet, one of the researchers, spelled out why that matters: the pigeons "spontaneously used the most relevant characteristics of the individuals (probably facial traits), instead of the lab coats that covered 90 percent of the body."
Ninety percent of the visible person changed and the birds did not update. Ten percent stayed the same and that was the part they were using.
These were not laboratory birds
Pigeons have been used in vision experiments for decades, and a trained laboratory pigeon sorting photographs is a different claim from a park bird sorting people.
Live Science notes the difference directly: "the new study was conducted on untrained feral pigeons."
Nobody rewarded these birds for getting it right. Nobody ran them through trials until the responses settled. They were pigeons in a park who worked out, on their own, which of two similar looking women was the problem.
The other pigeon face study, and why the two are not the same claim
There is a second body of work on pigeons and faces, done indoors with trained birds, and it is worth putting beside the park study because they answer different questions.
At the University of Iowa, pigeons were shown photographs of human faces that varied in who the person was and in what expression they wore, and were trained to sort them along one dimension while ignoring the other. What came out was a pattern already familiar from people. Ed Wasserman, who led the work, described it this way: "This asymmetry has been found many times in experiments with people and it has always been interpreted as the result of the unique organization of the human face processing system."
- The park study
- Untrained wild birds, real people, no rewards, and the question is whether they tell two individuals apart at all
- The laboratory study
- Trained birds, photographs, controlled trials, and the question is how face information is organized once they are attending to it
- Why you need both
- The first shows the ability exists outside a laboratory. The second says something about the shape of the underlying processing. Neither substitutes for the other
- The honest framing
- It came from the researchers. Fabian Soto, the graduate student on the Iowa work, wrote that "The point of the project is not that pigeons perceive faces just as we do or that people do not have specialized processes for face perception"
We are quoting that last line in full because it is the kind of sentence that usually gets cut when this research is retold. The finding is a similarity in one measured pattern. It is not a claim that a pigeon sees your face the way another person does.
Three species, three different ways of ruling out the easy answer
Read together, the three studies we have covered are less interesting for what they say about bird intelligence than for how each one closes off the obvious objection.
- Crows
- Keep the face and change the person wearing it. The mask moves between people of different height, build and gait, and the birds still react to the mask. Covered in the article on crows and human faces
- Mockingbirds
- Keep everything and change the person. Same nest, same approach, same season, new individual, and the response drops. Covered in the article on mockingbirds
- Pigeons
- Keep the people and change the clothing. The most visible thing about each researcher swaps over, and the birds do not follow it
- What the three share
- None of them asks you to take the researcher's word for it. Each one is built so that a skeptic's first objection has already been tested
Why it might be pigeons that are good at this
The researchers offered an explanation, and we are labeling it as a theory because that is how it was offered. Live Science reports that "they theorize that this recognition ability may have come about over the long period of association with humans, from early domestication to many years of living in cities."
The Cornell Lab describes what that association looks like now. Rock Pigeons are "a common sight in cities around the world," they "crowd streets and public squares," and they "may gather in large flocks in urban parks where people feed them."
A bird whose food supply depends on which humans are generous has a practical reason to keep track of which humans are which. Whether that is why they can do it is a separate question, and not one this study answers.
The one thing pigeons do differently at a bird bath
This has nothing to do with recognition, and it is the most useful fact in this article if you own a bird bath.
Almost every garden bird drinks by taking a billful of water and tipping its head back to let gravity carry it down. Pigeons do not. The Cornell Lab records that they "drink by placing their bill in water, using it like a straw."
- What that means at the basin
- A pigeon needs to get its bill properly into the water and keep it there, rather than dipping the tip and lifting
- Why you notice it
- The head stays down for several seconds at a time. It looks like the bird is stuck, and it is the easiest way to tell a pigeon drinking from anything else drinking
- The depth question
- A basin shallow enough to be safe for small songbirds is still deep enough for this. We are not suggesting you change the depth
- What it does not mean
- This is a drinking method, not a preference for deep water. Do not let anyone sell you a deeper bird bath on the strength of it
If pigeons have taken over the bird bath
This section is our judgment rather than anything from the research above, and we are marking it as such.
- Work out whether it is the water or the food. Cornell lists the diet as "seeds, fruits, rarely invertebrates" plus "food intentionally or unintentionally left by people." A flock parked in a garden is usually there for something other than a drink.
- Do not make the water deeper or shallower to exclude them. Depth is set by what is safe for the smallest birds using the basin. That constraint does not move.
- Expect to clean more often. Cornell describes flocks rather than pairs, and a flock fouls a basin faster than a few songbirds. The method does not change, only the interval. It is in the guide to cleaning a concrete bird bath.
- Accept that they may know you. If the research above carries over at all to a garden, the flock has probably already sorted the people in the household into categories.
What we could not check
This work was presented at a conference before it appeared as a paper, and the reporting we could open does not give the numbers.
- What we read
- The conference release as carried by ScienceDaily, the Live Science report of the same work, and the Cornell Lab species account
- What is missing
- How many pigeons, how many sessions, and how long the avoidance lasted. None of the accounts we could open states them
- So what we are not telling you
- Anything about how durable this is. The crow work has a figure for that and we quoted it there. This one does not, and we are not going to borrow the crow number and attach it to a pigeon
Frequently asked questions
Could the pigeons have been going by something other than the face
That is what the design was built to rule out. The two people were matched on sex, age, build and skin color, and the clothing was swapped. Bovet's own wording was "probably facial traits," which is appropriately careful and we will not upgrade it.
Do pigeons remember for as long as crows
Unknown, at least from anything we could read. The crow work reports close to seven years. Nothing in the pigeon reporting gives a duration.
Will feeding them make them like me
In this study the friendly researcher did feed the birds, and they treated her differently from the hostile one. That is closer to evidence for the friendly version than anything in the crow or mockingbird work. It is still one park, one pair of researchers, and pigeons.
Are pigeons a problem at a bird bath
They are larger and arrive in numbers, so they use more water and dirty it faster. They are not a hygiene category of their own and we are not going to invent a scare about them.
Which bird is smartest
Not a question these three studies can answer. They measured different things, in different places, by different methods.
References
- Reference: Pigeons never forget a face | ScienceDaily, carrying the Society for Experimental Biology conference release
- Reference: Pigeons Recognize Human Faces | Live Science
- Reference: Rock Pigeon Life History | The Cornell Lab of Ornithology, All About Birds
- Reference: Some features of human face perception are not uniquely human, pigeon study shows | ScienceDaily, on the University of Iowa laboratory work