Living World / Animal Intelligence

The DISCOVER Dancing Pet Challenge

Neurobiologist Aniruddh Patel says he often gets e-mail from people who claim that their cats, horses, or other nonfeathered pets have rhythm. Patel suspects that these cases are akin to the dancing dogs featured in YouTube videos, in which the animals don’t innately respond to the beat but instead react to cues from their human partners. However, if your pet really does have rhythm, he wants to know about it. “If someone has a dog that can dance to the beat, it will totally refute my hypothesis,” he says, “and that’s progress in science.”

If you think your pet proves Patel wrong, collect some video evidence, upload it to YouTube, and e-mail the link to webmaster@discovermagazine.com. We will post the best videos on May 1 (along with footage of Snowball shaking his groove thang).

So you think you can dance? You probably can, thanks to a brain that is remarkably adept at perceiving rhythms and synchronizing our body movements to what we hear. The ability to get into the groove—to step to the beat—is a hallmark of our species, raising the question of why we might have evolved this ability. Neurobiologist Aniruddh Patel at the Neurosciences Institute in San Diego looks for answers in brain scans and 
laboratory tests and also in the fancy footwork of what seems to be another dance-loving species, the sulfur-crested cockatoo. By monitoring the brain regions that activate when people hear a beat, Patel and colleague John Iversen find evidence that our hearing system is entwined with the motor control systems that guide our muscles. Patel proposes that these connections are a happy accident of evolution, a by-product of the brain development that allowed humans to learn to speak.

We take our ability to groove for granted, but it turns out that scientific studies show it’s quite rare. How many other animals can rock out?

It’s a behavior that seems so simple at first glance: How complicated can it be to bob up and down to music? You can do it while you’re drunk. And yet almost no other species can. In 2009 a team of Mexican researchers conducted a study with monkeys, trying to train the animals to tap to the beat of a metronome. Even after a year or two of training, the monkeys couldn’t do it. The results surprised people, because these monkeys are routinely taught really complicated things, and following a rhythm seems as if it should be easy.

That finding appears to support the theory that music gave humans an evolutionary edge over our primate relatives. So why are you unconvinced?

One theory is that music promoted group bonding when people danced together, and that in turn promoted the survival of groups. The counterargument is that maybe our sense of musical rhythm is just a by-product of another cognitive function that gave us an advantage. But what other function could that be? In 2006 I wrote a paper suggesting that rhythm could be a by-product (pdf) of another function that requires us to tightly integrate sound and movement: vocal learning, or the ability to mimic sounds made by other individuals. That led to the hypothesis that you need a vocal-learning brain to be able to move to a beat. The implication is that dogs and cats can never do it, horses and chimps can never do it, but maybe other vocal-learning species can do it. I proposed that idea, but it was purely hypothetical until a few years after, when along came Snowball.

Snowball is the dancing cockatoo who gained Internet fame via a video showing him bobbing in time to a Backstreet Boys song. What did that video tell you?


So I thought, let’s make this into a testable hypothesis. Let’s see if a cockatoo, which has a vocal learning brain, responds to a beat in the same way humans do. So we got in touch with Snowball’s owner and conducted carefully controlled experiments. We took the song he danced to on YouTube and slowed it down and sped it up without changing the pitch, and he synchronized at many different tempi. Another research group searched through YouTube videos and found 14 other species—parrots and one Asian elephant—that showed evidence of rhythm perception, suggesting that Snowball is not a freak. We’re also working on a collaboration with a research group in Germany on dolphins, another species that engages in vocal learning.


Are you still doing research on Snowball?

Yes. In those first experiments we made sure that Snowball wasn’t just responding to cues from his owner, that he was reacting to the beat on his own. But his owner told us that he really preferred to dance with a partner. So we did an experiment that we haven’t published yet in which we had Snowball dance by himself; dance with someone in the room encouraging him, saying “good boy” and so forth; and then dance with someone else. We found a huge and enormously clear effect. He dances least when he’s by himself, more when hears vocal encouragement, and most when he has a partner. This suggests that dance is a social thing. It’s possible that vocal learning capacity is necessary for rhythm perception but that it’s also important to have a brain that seeks out tight social bonds.

What is the actual mechanism that allows a person to move to a beat? How do we know when the next beat is coming? 

When you move to a beat, you are doing something very sophisticated: You’re moving to a model in your head. What do people do if you ask them to tap to a metronome? They tap right on the metronome beat or a little bit ahead of it, not a few milliseconds after it. It’s not a reaction to stimulus, it’s anticipation. And this is exactly what the monkeys in that study I told you about were not able to do. They reacted. They always tapped 200 or 300 milliseconds after the metronome click. So what are you doing when you tap exactly on the metronome beat or just ahead of it? You’ve built a model of the time interval in your head, and that’s what you’re synchronizing with.

My colleague John Iversen is doing some really elegant work now with brain imaging, looking at syncopated rhythms, because with those the sound is shifted off the beat, but you still feel the beat. With brain scans we can look at what’s going on in the brain as you internally construct that feeling of a beat without any brain activity related to processing sound. So it’s a very clean, clear way to decouple the beat sensation from the processing of sound.