Do birds understand us?

0:00This BBC podcast is supported by ads outside the UK.

0:30at Whole Foods Market. Doja Cat. My V World Tour.

0:38Live in concert. With special guest, Lotto.

0:48Doja Cat. Live across North America in 2026. Get tickets now at DojaCat.com.

1:04Ready? What are we doing? So what we're doing is we're recording what you say. Recording means listening. Listen. Listen. Welcome to CrowdScience from the BBC World Service, the show that answers your science questions. Listen. My people are going to listen. Listen. To Ellie. Ellie. Yeah. I'm Caroline Steele and you're listening to a conversation between Jen and Ellie. This is a microphone. Microphone, microphone, microphone. It's okay. Jen is a scientist and Ellie is a bird. A cockatoo. A very clever cockatoo speaking via a computer. What do you think? Move and dance. Oh, do you want to do some playing?

1:53Peekaboo. You might be wondering why you're listening to a talking bird. Peekaboo. Peekaboo. And wait, can birds really talk? Peek-a-doo! Peek-a-doo. You're so good! It's all thanks to a question from crowd science listener David in Guinea.

2:16My name's David Gilavotier. I'm an aviation meteorologist and I'm from West Africa, Guinea, Conagry. When I was growing up, my friends and I would go for bird hunting. David and his friends would hunt birds that were eating crops on the family farm. And he found a cunning way to lure them in. Simply making the sound they used to communicate, you can make them react, OK? You can make that species of birds get closer. For instance, it was easier for us to catch a black-headed weaver. So you're particularly good at the black-headed weaver call? Exactly. Now, this is what a black-headed weaver call sounds like.

3:01You know what I'm going to ask you, David? Uh-huh. Can you do it for us now? Could you mimic the black-headed weaver call? Yeah, we would make it sound like this.

3:14Whoa. I just think that if we use their exact way to communicate, they can understand us, But the problem is we may not understand exactly the message we are trying to send to them. Yes, you don't know what you're saying. You might be saying something like, don't forget to wash your socks on Wednesday or something like that. You have no idea what you're saying. Exactly. So what's your question for CrowdScience? I'd like to know whether a birther can really understand us.

3:50David wants to know if we can speak to birds. Do they really understand us? Was he actually communicating with the black-headed weavers on his family farm? But before we work out whether a bird can understand us, I think we need to find out whether they can understand anything. That's Mogritsa, that's Melva, that's Ukraine. Nestled deep in the Austrian Alps is a flock of nearly 100 geese. and Andrew Katsis knows them all by name. Qualia, quartz, hansi, amber and that's quiz up the back. Producer Harry went to meet him, didn't you Harry? I did and not to make you jealous Caroline, but I also got to feed the geese.

4:34Oh, very jealous. So where exactly were you? I'd gone to visit this facility called the Conrad Lorenz Research Centre for Cognition and Behaviour. So it's stationed at the bottom of this valley with a wide turquoise river running parallel to it. It's actually home to a permanent gaggle of grey lag geese, as you've already heard. And Andrew introduced me to them. My favourite is named Vivian. Vivian? Yep, he's a yearling, so he was hand-raised last year. A few days ago, I was watching the flock and Vivian was aggressive towards Floki. but then he turned around and did a very dramatic triumph ceremony in my direction and the triumph ceremony is something that usually only couples do between each other.

5:21Okay and is that because you've been saying nice things to Vivian? I think it's because I'm here every day and he just got used to me being around but there will never be a happier day when he rejects me for another goose. Let's hope Vivian finds a more feathered mate soon. So Harry, Andrew has been studying these geese right? He has for quite some time now. I'll let him explain what he's been up to. I'm a postdoc studying goose behaviour here at the KLF and specifically goose personality. Goose personality? Yes. What does that mean? In animal behaviour research, traditionally researchers tended to assume that every individual within a species behaves in basically the same way.

6:06but over the last 20-25 years we've started to realise that no, you have different individuals, some are very bold, some are very shy and they may serve different functions within their animal society. It seems wild to me that we've only relatively recently realised that birds have different personalities. I guess what we have to look at is the history of how scientists have treated animals and looked into animals in the past and in the early 20th century animals were really only tested in the lab so it was sort of believed that they were like machines or computers they could be provided with an input or an environment and they would just react in the moment to whatever happens and it wasn't until really people like Conrad Lorenz and he's the scientist that this centre in Austria is named after came along that that all started to change.

7:03So what did he do differently? Right. So Conrad was a naturalist, a lover of animals, and he basically just immersed himself in their world. He spent most of his life sat in and amongst the geese, which sounds rather lovely to me. And he completely sort of transformed the way that we view bird, I guess, cognition. Right. Yeah. So he suddenly found that or made observations that would seem like these geese had instincts, they had problem-solving abilities, they could communicate with one another, they had personalities basically. And that completely dismisses the idea that animals are like machines.

7:42So Harry, what does studying these geese look like today? Are they putting them through tests or are the scientists observing the geese like Conrad did? Yeah, it's a bit of both. There are some experiments but there still appears to be a lot that can be learned just from observing them.

8:02That is a forward neck by Cocosnus towards Chonky. You can see that Cocosnus is one of the more dominant geese in our flock and he actually has two fledglings at the moment and so he's very defensive of them. There we go again. There's Cocosnus again. That is a chase by Cocosnus towards Bacardi. red AU, red, right, red on the left, nothing on the right What are you up to right now Andrew? You've got your iPad out At the moment I'm recording agonistic interactions between the geese Sorry, I'll just record that one That was a peck on demand blue AU, white, yellow, yellow by lumpy, black AU, white, orange, red, white, red So when you say those colours you're pointing out the bird tag Yes, so that's how we identify the birds most of the time by their colour bands and each bird in the flock has a unique combination of colours.

8:58So just by looking at its legs, we can say, oh, that's Luca or that's Severin or that's Bruce Springsteen. There's one called Bruce Springsteen. There is a Bruce Springsteen. Oh, Bon Jovi's gone. It's just the boss left. So at feeding time is when all the petty inter-goose squabbles come to a bit of a head. and so by recording all the interactions that we observe we know who the aggressor is we know who the receiver is and we can start to build up a dominance hierarchy so what you might call a picking order that sounds to me like a very sort of slow geese focused dating show it is just like a goose reality tv show in a way and at the end of the day all the researchers come around the table for dinner and they'll discuss that gossip you know they'll discuss who's been flirting with who you know who's made a great new friendship who's had a big bust up and all of that stuff all of those relationships can be incredibly insightful well we know that the geese can recognize each other they recognize each other's voices and they can recognize each other's faces so for example we did one experiment where we printed life-size photographs of the geese and we posted them around the feeding area in front of a bowl of food and we found that a goose was more likely to spend time feeding at a certain poster if the goose represented there was their partner versus a non-partner.

10:31So even from a completely static photograph they seem to be able to recognize at least is this my partner or is this somebody else. And how far can you extrapolate that? To friends? To friendships? They have these wonderful beefs especially around the breeding season where they'll be foraging perfectly happily. There'll be 20 geese all around them, but they will see another goose about 30 metres away and they will immediately make a beeline for it and chase it away from the entire feeding area. And you can see these disputes playing out over days and weeks. And so I think there's no doubt that they can recognise their friends from their Okay, so Harry, this is basically evidence that geese can recognise and respond to each other.

11:16But our listener David wants to know if birds can recognise and respond to us. Yes, you're right. So before I headed home, I took the chance to speak to Thomas Bugnier. Now he's a professor in cognitive biology, which means he basically tries to get inside the heads of other animals and figure out what they actually can understand. If you would like to use the word understanding, then I would say you can design studies where you can differentiate between intentional and unintentional actions if you want to. So could you give me an example of how that might play out in intentional behavior? I'll give you an example from the primate world.

11:58In this case, you show the animal a piece of food which it either likes very much or it doesn't like so much. and then a human experimenter is acting towards either the preferred or the non-preferred food and you can see how the animal is responding to the human as are going to grab the preferred food and it's gonna deliver it to the animal or he's gonna grab the unpreferred food. So Thomas is describing this experiment where the primate, so a primate could be a monkey, the monkey is sat down and in front of it out of reach are two different choices of food. So one of those is like an all-time favourite and one of them is rubbish.

12:39Doesn't want that. So banana, broccoli. A banana is the perfect illustration. Thank you very much, Caroline. So there's a banana and there's broccoli. Now, what we can do is we can watch that monkey's behaviour as the keeper comes in and, say, starts to go towards the broccoli, the food that the monkey doesn't really fancy. OK, so the monkey might sort of see the keeper go for the not as nice food. and I don't know, do some gesture or make a noise to say, Oi, go for the banana, please. Precisely. And it could be anything. So as long as it's making that gesture, what this monkey is demonstrating is this intentional behaviour.

13:20Right. It's trying to influence the situation. So in this circumstance, the monkey would appear to be thinking rationally. It's predicting what's going to happen and it's trying to make sure that it gets what it wants. OK, impressive and interesting, but that's for primates, and we want to know about birds. Fortunately, Thomas does most of his research with ravens. So we have observations here in the field. We can see one raven, for instance, gets a beat up from another one, and then he is giving defensive calls, basically to appease the attacker. I didn't realise ravens were aggressive at all.

13:59Yeah, there are quite strict hierarchies within raven populations. And so individuals that are dominant will tend to assert that dominance onto subordinates. And by subordinates, you mean ravens they look down on? Yeah, just a bit more shy, a bit more timid. But then we had the impression that sometimes they call like mad and they seem to overdo it. And sometimes they seem to be super shut up. And so we were interested in what's going on here. and then it turned out that it's the audience which is important. So if their body was in the audience, they were overdoing it, so come and help. And if the body of the aggressor was in the audience, they were shut up, suggesting that they were taking their own relationship to the individuals in the nearby audience into account, but also the relationship of the aggressor, which would be third-party understanding.

14:51Okay, so if a raven's getting bullied and it has a friend nearby, it will call out loudly for help. But if instead the bully has a friend nearby, the raven that's being attacked will keep quiet. Yeah, that's exactly right. So the victim here is demonstrating that not only can it recognise others, but also it has this situational awareness. It kind of knows what the repercussions of its actions might be. There was also, Caroline, another story that Thomas told me that I want to share with you. This occurred to me in my very first years when I was studying ravens. I had a raven and I was visiting him every day to give him some company during lunch basically.

15:35So he was sitting on my lap when I was eating my lunch and I was sharing it a little bit. And then I was gone to a conference and I was gone for about a week or two. And when I came back and I was putting my hand into my pocket to take out a little bit of the rewards and when he saw it, he came very fast towards the cheese. And I pulled back my hand. And then he looked at me and then he said, ouch. Hold on a second. So a raven said, ouch. Yeah, I probably should have mentioned this earlier. Ravens are sort of like parrots. So they can produce a wide range of vocalizations. And one of those is the ability to mimic humans.

16:15And I said, no, ouch, almost ouch. And so only afterwards, I figured what could have been going on in his brain. So he wanted to take the reward. I pulled back. So most likely he could recall previous instances where I didn't pull back and where he hit my finger and I said, ouch. So that was my interpretation. And so he must have somehow tried to reason what's going on here.

16:45Perhaps birds do understand what's going on in their environment. Thomas and geese personality expert Andrew certainly thinks so. But our listener David wants to know if they can understand us humans. Was he actually communicating with the black-headed weaver on his family farm? Well, the simplest way to find out would be to just ask a bird. And as we heard at the start of the show, that idea might not be quite as far-fetched as it sounds. Ellie the chatty cockatoo is coming up next. This is summer at its peak. Whole Foods Market Summer Fruit Fest is your invitation to eat the season. Fresh, organic, and bursting with flavor.

17:27Start your day with peaches and organic blueberries and yogurt. Build a grazing board with fresh fruit, prosciutto, and artisanal cheese. Then fire up the grill with no antibiotics ever proteins and fresh produce. Savor the season. Shop Summer Fruit Fest at Whole Foods Market. Doja Cat. My V-World Tour. Live in concert with special guest Lotto.

18:00Doja Cat. Live across North America in 2026. Get tickets now at doja cat dot com.

18:18You're listening to CrowdScience from the BBC World Service, the show that listens to your science questions and tries to understand them. Thanks to a question from listener David in Guinea, we're trying to find out if birds understand us. And I think it's about time I tried to have a chat with one. Hi, Ellie. Hi. Are you ready? So you're holding up a tablet in front of her? Yeah, so this is a tablet. Hello. Caroline. How are you? What? Wow. I'm well, Ellie. Thank you for asking. How are you? What do you think? Feel. Uh-huh. Be excited. Yay! I'm glad you're excited. I'm excited to meet you. What do you do?

19:06I make science radio shows. And this show is about do birds understand us? And I think you might be able to help us answer that question. Like. She says like. I'm talking to Ellie the cockatoo. In front of her, she has a touchscreen computer, a tablet, which she uses to communicate. When she wants to talk, she uses her beak to navigate it. Next to Ellie is her caregiver, Jen Taylor O 'Connor, an animal communication researcher at Texas A &M University in the US. So I got her as a baby from a really excellent breeder who allowed her to grow up with her parents and siblings. I thought this would be a really great start to life for her.

19:53And then it turned out when she arrived home as a baby, she was angry. She would scream all the time, and she had this anguished scream that really pierced your heart. She tried to chew through screens. She'd bite me. She liked other people. She'd go to other people, but she wouldn't come to me that much. So how did you try and help Ellie to stop being furious? The only thing that really reached her were the days that she had training on new topics. She would just kind of hunker down and want more and more and more training. So in a desperate moment, it was an inspired moment. It was a desperate moment when she was throwing my glasses off the counter and throwing glass off the counter and screaming and chasing her sisters.

20:39I grabbed some foam letters I'd gotten from Walmart and started teaching her the phonetic alphabet in association with other objects like A, A, A for apple and B for banana in what they call a two-task test, you know, like which is which. Could you describe that test? Sure. So I did the association with the letter A, A, A, A. So I taught her A, A, A with a foam letter A. and then I showed her an apple and I said a apple a a a letter and then I hold up both I held up the apple and I held up the letter and I said which one's a a a and she touched the letter and which one's the apple and she touched the apple before long Ellie had learnt the whole alphabet but that still wasn't enough she wanted to learn more she got bored you know with just letters so we got a very ordinary tablet from Target and then we loaded it with games and a speech board and a writing pad.

21:42She not only started expressing, now she had a whole board she could navigate. And she, you know, she lit up and she took off. Ellie's tablet now has 1 ,500 different pictures, all associated with words. And the images are arranged into different menus, like emotions, activities or weather. And she can navigate between these menus herself, which she does to tell Jen how she's feeling, what she wants or even to share a story. Do you want her to just navigate and chat with us? Yeah, that'd be great if she's up for it. What do you think? Dreams and thoughts. Tell a story. We've been exploring with her the notion of story because she seems to enjoy books.

22:33The idea of the elements of what stories are. Do you want to? Characters. Are you going to tell us a story? What do you think? A unicorn. Yeah? A unicorn. You want to tell us a unicorn story? This is very sweet. So only recently has she started initiating this menu herself to kind of like, I don't know. We don't know, right? We don't know what they know and what they think. We just follow. I'd love to hear about the unicorn. Do you want to tell a little bit more about the unicorn? How was the unicorn? What do you think? And then? Uh-huh. Temp and weather. How was the unicorn? Warm. The unicorn was warm?

23:09Back. Yeah. And then what? Temp and weather. Uh-huh. Like. Oh, did the unicorn like the warm weather? That's good. Thank you, Ellie. High five. Boom. I have to say, I wasn't blown away by what Ellie could do over our video call. At points, it felt like she was just imitating Jen. And when she told a story, she might have just been pressing random buttons. But then Jen started to describe the science. Last year, we did some testing about what do you do if she requests something or selects something and then you give her the wrong thing. The very first time she bit me for giving her the wrong thing.

23:48Another time, she selected a sunflower seed on her speech board for a treat. And I brought her broccoli and she looked at me like, are you OK? This is not usual. And then what's remarkable is that she exited her speech board on her tablet. She brought up the home screen, went to her photo gallery and opened the picture of the sunflower seed. Like, this is what I was asking for. And in case you needed it larger, here it is on the entire screen. This is a sunflower seed. This is what I wanted. This is another example of intentional behaviour, like Thomas described in Primates. It caught the attention of a lot of researchers, from the Open University, MIT and Northwestern, who have now worked alongside Jen.

24:33They've tested for randomness. Maybe Ellie is just pressing buttons, but they found that she actively engages with what she's asked for. Jen has even investigated whether Ellie prefers touching certain parts or colours on the computer screen, but that's not the case. I thought that maybe Jen subconsciously influences what button Ellie presses, but in many of the experiments, Jen sits where Ellie can't see her. So the research does seem to suggest that Ellie is intentionally using her speech board to communicate. To what level do you think she understands what you say? There's an idea that maybe animals are kind of like a vending machine, that they press a selection and they know, like, if I press this, I'm going to get a sunflower seed.

25:23But anecdotally, she seems to have a lot of flexibility in how she expresses her selections. Her sister, Isabel, was very sick. And actually, I didn't know it yet. It was two days before I found out. But she used the combination of fix Isabel, no like, no like, fix Isabel, mom, mom, mom, Isabel, fix, fix, no like. And then two days later, we found out that Isabel was very sick. So we have these, these kind of persistent combinations, I'll say, that seem pretty amazing.

26:08Jen's research and anecdotes are very impressive. We've examined cockatoos, ravens and geese. So it's finally time to try and answer listener David's question. Do birds understand us? What do you reckon, Harry? I'm just really impressed, basically. I'm surprised by what we've learned along the way. Yeah, you're impressed by how intelligent birds can be. Sure. I used to think of them, as Jen said, rather naively as vending machines. I never thought there was any kind of complex thinking going on there, but the science would appear to suggest otherwise. Yeah, so we know that they can recognise and respond to each other.