Dogs and babies prone to same classic mistake

Not Exactly Rocket Science
By Ed Yong
Sep 4, 2009 7:00 PMNov 5, 2019 12:13 AM


Sign up for our email newsletter for the latest science news

Domestic dogs are very different from their wolf ancestors in their bodies and their behaviour. They're more docile for a start. But man's best friend has also evolved a curious sensitivity to our communication signals - a mental ability that sets them apart from wolves and that parallels the behaviour of human infants. Dogs and infants are even prone to making the same mistakes of perception.

Like infants less than a year old, dogs fail at a seemingly easy exercise called the "object permanence task". It goes like this: if you hide an object somewhere(say a ball under a cup) and let the baby retrieve it a few times, they will continue to search for it there even if you hide it somewhere else (say behind the sofa) and even if you do so in front of their eyes. Piaget, the legendary psychologist who discovered this behaviour, thought that it reflected a wildly different way of seeing the world.

More recently, Jozsef Topal suggested that it's the influence of the adult experimenter that's the key. By repeatedly pointing at the ball in the first hiding place, the adult enshrines a generalised rule in the infant's mind. And infants, being programmed to learn from communicative signals, come to believe the adult's instructions over the evidence of their own eyes (some people apparently never grow out of this, but I digress). Topal demonstrated this by showing that infants were much better at the task if the experimenters avoided social cues like calling the child's name or eye contact.

And the same is true for domestic dogs. Topal tested a dozen adult dogs with a version of the hidden-object challenge, concealing toy behind one of two possible screens. If he called to the dogs by name, made eye contact and waved, the animals made the same errors that infants make on 75% of the trials. Without any of these signals, their scores improved and they only failed to realise the ball's new location on 39% of the trials. Their error rate dropped even lower in completely non-social situations, where the ball was moved by pulling on a transparent string.

These results suggest that dogs and infant share a social mindset where certain cues prepare them to learn from humans. It's not the case that the gestures and facial signs were just distracting for that would lead the animals or infants to search both hiding places equally - instead, they both preferred the one that the object was initially hidden behind.

Dogs, it seems, have a particular breed of social smarts even as inexperienced puppies and some scientists have suggested that these skills are adaptations that have developed over the last 10,000 years to allow dogs to better interact with their two-legged partners.

Indeed, the true scope of the dog's mental abilities has probably been known to dog owners for some time, but has only really been confirmed through experiments in the last decade or so. If scientists point to one paper cup among many that hides a piece of food, dogs understand what they mean - they clearly understand the pointing gesture. If you're not impressed, bear in mind that wolves can't do it, great apes are flummoxed by the task and even babies fail before their first birthday.

Topal also found that wolves, even those reared by humans, don't make the same errors that dogs do on the object permanence task. When he tested 10 wolves that had extensive experience with humans, they passed with flying colours regardless of whether their human partners were gesturing and calling, standing impassively or entirely absent.

This difference between wolves and dogs is a striking one, but it doesn't mean that one subspecies is cleverer than the other (or better at logic, as LiveScience purported). Instead, Topal thinks that the two species have different biases in the way they perceive the world, and those of dogs are superficially similar to those of human infants.

Note that I said "superficially" - it would be folly to suggest that dogs are reacting to the human signals in exactly the same way as our own infants do. Topal demonstrated this by repeating the tests, but swapping the experimenters around before the object was moved to a new location. This time, the infants still made the same errors that they did before but the dogs did not.

This suggests that the two species are doing different things. Topal thinks that infants use the gestures as a sign that a learning opportunity is coming up. They're about to get some info that they can generalise to new situations. As such, it doesn't matter if the experimenter swaps around - they still think that the rule they've learnt about the ball behind a certain place will hold. Dogs, on the other hand, relate the signals they see to specific situations and to specific people - if the person making the funny actions changes, the rules of the game change too.

Reference: Topal, J., Gergely, G., Erdohegyi, A., Csibra, G., & Miklosi, A. (2009). Differential Sensitivity to Human Communication in Dogs, Wolves, and Human Infants Science, 325 (5945), 1269-1272 DOI: 10.1126/science.1176960

Images: Dog by John Haslam

More on dogs:

1 free article left
Want More? Get unlimited access for as low as $1.99/month

Already a subscriber?

Register or Log In

1 free articleSubscribe
Discover Magazine Logo
Want more?

Keep reading for as low as $1.99!


Already a subscriber?

Register or Log In

More From Discover
Recommendations From Our Store
Shop Now
Stay Curious
Our List

Sign up for our weekly science updates.

To The Magazine

Save up to 40% off the cover price when you subscribe to Discover magazine.

Copyright © 2024 Kalmbach Media Co.