Crows "count" the number of self-generated vocalizations | Hacker News Reader

Crows "count" the number of self-generated vocalizations

science.org

72 pointsgnabgib53 comments

I used to have long conversations with the crows in the back yard at my old house. If you feed them they will trust you a little more than normal, such that when you caw at them they will caw back. They are happy to repeat the number of caws. So if you caw 5 times, they’ll caw back 5 times. Also sometimes I let them go first and I copy them. We did this for sometimes ten minutes straight.

I did this once with a noisy jay.

the jay would screech, and I would copy it. I did this for enough time for the jay to realize this, and then it went from screech to a pleasant coo type sound.

I find it funny that observations of the cognitive capabilities of animals like in the featured article always induce people to come up with convoluted alternative ways to word the "anthropomorphic" descriptions of the behavior. "Oh that crow isn't counting, it's just basic pattern recognition associating specific quantities with specific sounds", "oh that octopus isn't playing, it's just throwing the bottle around as a stress response to boredom", etc. I get wanting to be careful not to misinterpret the behavior of specices which have evolved different social structures (if any) and under different selective pressures, but sometimes the simplest explanation really is that the animal is doing a "human behavior".

What if humans are “playing” because they’re bored?

I think that's Grant's point - there's a (potentially) false dichotomy in those descriptions.

I don’t know if he was he was making that point, but rather he was expressing the desire to avoid anthropmorphosizing by describing the same thing in different words.

(Kind of like what I did now, hehe)

Being straightforward about it would force us to question a lot of assumptions we have about animals and how we (mis)treat them. We can’t have that!

Because straight forward language isn't scientific language and lacks precision.

Sure, but the idea that animal cognition is radically different or lesser than ours is itself an assumptive default that science is comfortable making. We should question that as the default assumption.

As a pet owner, I find it absolutely ridiculous and insulting that animals intelligence is assumed to be so low. My firsthand experience is that there is a lot more intelligence there than people generally give them credit for, even if science hasn't delineated every single aspect of it yet. As scientists keep confirming these "discoveries", I think the default assumption (and language) should shift to assuming more similarity than differences unless proven otherwise.

>"but sometimes the simplest explanation really is that the animal is doing a "human behavior"."

If it looks like duck and quacks like a duck ...

I think you're right, but historically there have been problems. Frans de Waal has written about how in his field there was a lot of early egregious anthropomorphizing. As people realized how flawed that was, they swung around the other way, to refusing to consider analogous human behavior. But things like de Waal's work demonstrating moral sense in animals is hard to explain in any other way. [1]

Even so, it can be tricky. In Bernd Heinrich's excellent "Mind of the Raven" [2], he points to a story where a woman, lost in the wilderness, noticed a raven crying near her and spotted something like a mountain lion creeping up on her. In the news story she gave the raven credit for warning her and saving her life. But Heinrich points out that ravens have a demonstrated history of leading predators to prey, because that way they get the leftovers with much lower risk. So although the woman was correct that the raven understood her danger and was demonstrating advanced social cooperation, her anthropomorphic bias led her to perhaps be dangerously wrong about who the raven was trying to help.

So although I think shorthand like, "The octopus is playing" is fine for casual use, I think serious people have to always be challenging those tempting assumptions.

[1] e.g., this book https://www.amazon.com/Good-Natured-Origins-Humans-Animals/d... or this TED talk: https://www.youtube.com/watch?v=meiU6TxysCg

[2] A stellar book, and I recommend it strongly to anybody fascinated by corvids or interested in decoding animal behavior more generally: https://www.harpercollins.com/products/mind-of-the-raven-ber...

>story where a woman, lost in the wilderness, noticed a raven crying near her and spotted a mountain lion creeping up on her. She gave the raven credit for warning her and saving her life. But Heinrich points out that ravens have a demonstrated history of leading predators to prey.

That's a hilarious (and scary) story. I made the same cognitive mistake when reading, i.e. assumed that either raven was helping her, or was crying for unrelated reason (because I suspected a twist). It completely didn't occur to me that the animals could have cooperated among themselves.

> So although the woman was correct that the raven understood her danger and was demonstrating advanced social cooperation, her anthropomorphic bias led her to perhaps be dangerously wrong about who the raven was trying to help.

This is such an excellent insight, thanks for sharing!

There is also similar story about Komodo Dragon. Guy buys a big lizard, tries to communicate with it, teach tricks, all to no avail. Then one day the lizard bites him out of the blue. Guy is angry at the lizard, swears to give it back to the shop the next day. Then he notices the lizard is following him with it's head down. Guy warms up to the lizard "Finally you show remorse! Understood that biting me was a bad idea, huh?" After several days of that following he decides to call the vet about it. And the vet explains to him that the lizard is simply waiting for him to die from the bite.

>>sometimes the simplest explanation really is that the animal is doing a "human behavior".

Or, perhaps more accurately, humans are one of many intelligent animals and we do behaviours typical of many intelligent animals, and are now noticing that they also do these behaviours (but notice it from a self-centered point of view).

Eh, with the crows example you could think of it this way for a counter example. Listen to the drum intro to wipeout. Now play a measure of it with your hands banging on your desk. You recreated the sound but how many times did you hit the desk? I find it very plausible that they aren't actually counting. This isn't a counter example trying to prove they're not, but they can certainly repeat or "know" things without doing arithmetic.

And on the other hand, if you start recognizing 'play' in other species, it can help you learn what 'play' actually is and why we do it. What was frivolous is now how we learn and discover and invent and develop.

I wonder why the quotes around "count" are required.

> The crows flexibly produced between one and four vocalizations for corresponding cues associated with numerical values. Furthermore, they used different calls for different numbers.

Is this because the crows might just remember four separate stimuli, and did not demonstrate an understanding of integers in the Peano arithmetics style? I still think that producing a specific number of vocalizations, well, counts as counting.

"between one and four vocalizations for corresponding cues associated with numerical values"

Counting to 4 seems awfully limited

Now imagine how limited computers are with their ability to count to only 1.

Most computers can count to 2^32-1, and many can count even higher! They can also string words together grammatically, which means they are one step away from an artificial general superintelligence explosion.

But when non-human animals use AAC systems, or vocalise words directly, to express wants or needs, they aren't really talking, because they don't use tree-like grammars.

What? Even really simple MCUs can output more symbols than 0 and 1 and of course they can count higher.

I got paywalled but as a layperson I'd define actual counting as e.g. croaking the same number of times as how many treats are in front of you.

So I'd want to see 3 treats -> 3 croaks, not square -> 3 croaks and circle -> 4 croaks, to remove the quotes.

At the same time, teaching an animal to repeat a behavior an exact number of times is actually very tricky. For example, a dog will likely over-eagerly repeat the action until rewarded.

I think so (not the expert on biology, of course). Depending on the study and setup, crows can count even a bit more, maybe to 6-8. The thing is, humans also can only "intuitively" count only to 7-11 (depending on the person). Counting higher requires some abstract thinking, not just intuition.

Crows (and other birds, like parrots) are pretty smart.

There’s a school of thought/fantasy that the dinosaurs actually had an advanced theropod civilization.

Cool idea. Not sure I buy it, but they might have been quite intelligent.

I used to like seeing the Japanese crows, in Tokyo. I’ve heard they can be real pests.

They are big. Like, raven-sized.

Ravens are crows, crows are ravens. They exist in the same taxonomic area as frog/toad, bush/tree, etc.

The difference between crows and ravens are who found it first

They may be ravens, for all intents and purposes.

They have heavy beaks, and about a 4-foot wingspan.

They also have a different “accent” from American crows.

Wikipedia says they are crows: https://en.m.wikipedia.org/wiki/Large-billed_crow

Ignoring size, you can often distinguish between Corvus corax (common raven) and Covus brachyrhynchos (American crow) by the tail feathers; ravens tails in flight have a wider angled cone shape, crows tails have a narrower fan shape.

Crows and ravens, taxonomicaly speaking, are of the same genus, whereas frogs and toads are in different families--not exactly the same area perhaps.

> Ravens are crows, crows are ravens.

I would argue they are distinct.

https://en.m.wikipedia.org/wiki/Toad

https://en.m.wikipedia.org/wiki/Raven

There’s no real feature that separates frogs from toads or ravens from crows.

Taxonomy is more art than science

From the Wikipedia article: "The related term 'raven' is not linked scientifically to any certain trait, but is rather a general grouping for larger species of Corvus."

Only tangentially related yet quick obligatory shoutout to Cornell Lab's stellar Merlin app if you are interested in deep diving into birds / their communication in general.

https://merlin.allaboutbirds.org/

Of course crows are base-4. A crow doesn't have hands. So if they stand on one foot they have four digits on their other foot for counting. If we had only four digits at the ends of each arm we would be base-8 rather than 10.

Corvidae are very intelligent, and can readily recognize individual people.

Over the winter we tossed out a few peanuts while wearing the same baseball cap, and whistling a specific song. After awhile, the crows changed their dialogues when we walked in the area, and indeed they certainly can communicate with each-other about their environment.

We found this rather surprising, and gained a deeper appreciation for their antics. As a consequence, the birds often indicate when someone/something new is in the area.

Fascinating little creatures, =)

Crows are intelligent, There are some videos out there showing crows playing simple games and stuff...

I wonder if crows being intelligent has it linked with them being ominous in many litrature

Crows are ominous in literature and folklore because they are carrion eaters. So, if you see crows, especially in large numbers, that will mean it's likely there some large amount of dead animals or even people around: a very ominous sign almost be definition.

I've seen pigeons eating corpses carcasses of other birds. Those winged rats (I don't like pigeons) will eat anything.

As for crows, I live in a country where sometime it snows, and I make sure that when it is very cold (snow, ice, or generally below 5 C)I feed the (3) crows of my building. Every time I am outside, walking, they are always looking after me. Flying near me, standing and looking at me, nodding at me. And I always speak to them when they come close, with calm voice (phrases like "I'm friend, you are good, thank you", etc.).

I want to find a way to 'monetize' on this relationship and teach them how to bring me shiny things (I keep thinking but never actually googling on how other people have done this).

I've thought about teaching them this, but ultimately I decided against. While they may find unclaimed loose change for you, they could just as easily figure out a scheme to distract and rob some poor beggar or busker, and you wouldn't ever know.

There's probably some kind of crow-to-crow "these are friends, these are enemies" abstract communication, but I don't think you'd have much luck communicating the concept "to me, humans are friends, so don't rob them on my behalf" across the species barrier.

The prize is the crow's reciprocal gratitude, not the actual object. I just find it cool that they like me.

Buy a vending machine that dispenses crow snacks when money is put in.

Place the vending machine outside

Teach them that putting money in the vending machine dispenses crow snacks.

Practice placing dollar bills and other bits of paper around your property til they get the hang of it and figure out what works for the vending machine.

... profit?

Lots of animals are carrion eaters, but aren't given the same ominous treatment. Some examples would be beetles, eels, alligators, etc.

You cant see those from a long distance away. Also factor in where advanced cultures evolved that were our ancestors.

Most of those animals aren't large, loud, dark black, and can fly around your head and face.

So probably that as well?

Indeed!

The point I was emphasizing was that it is not just:

>because they are carrion eaters.

as the parent I replied to asserted.

If you read the Game of Thrones books, there's recurring mentions of "a feast for crows" after major battles. It really sets the tone, whenever crows gather there's devastation.

I don't have it to hand but they're able to trade and are wise enough to avoid people who cheat them. Either them or family member kinda trains/works with wolves as well.

It probably is why crows and ravens get that treatment. Not sure if they're not that fussy about food though, media/literature always makes them seem like vulture like scavenge

I'v seen a video where a crow use a series of tools in a specific order to obtain a reward

Ze Frank (True Facts YouTube Channel) recently released a video about crows crafting tools to solve problems. It's quite amazing and, of course, very funny: https://www.youtube.com/watch?v=B-HF-wBwQsc

Welp, here I go again, off to watch random episodes of The Show. I miss those times.

It's interesting (and cool) how Science has one of their editors write a summary for papers

there was this crow in front of the house every morning, training its cawing, I tried to imprint some HarrHo to get it to say hallo, but it just looked like it became self-conscious and flew away, cawing in the distance

There are some Common Myna [1] that live on my land, related to crows in level of intelligence. I have observed them seemingly playing games with betelnut. They pick them, bring them to the top of the roof, release, watch it roll down into the gutter and drain spout, then retrieve it and do it again.

[1] https://en.m.wikipedia.org/wiki/Common_myna

Five is right out.

Reply on news.ycombinator.com