Dogs may understand human point of view, researcher finds

Feb 11, 2013
Dr Juliane Kaminski and her dog, Ambula.

(Phys.org)—Domestic dogs are much more likely to steal food when they think nobody can see them, suggesting for the first time they are capable of understanding a human's point of view.

Many think their pets are clever or that they understand humans but, until now, this has not been tested by .

Dr Juliane Kaminski, of the University of Portsmouth's Department of Psychology, has shown that when a human forbids a dog from taking food, are four times more likely to disobey in a dark room than a lit room, suggesting they take into account what the human can or cannot see.

Dr Kaminski said: "That's incredible because it implies dogs understand the human can't see them, meaning they might understand the human perspective."

This is the first study to examine if dogs differentiate between different levels of light when they are developing strategies on whether to steal food. It is published in the journal .

Dr Kaminski said: "Humans constantly attribute certain qualities and emotions to other living things. We know that our own dog is clever or sensitive, but that's us thinking, not them.

"These results suggest humans might be right, where dogs are concerned, but we still can't be completely sure if the results mean dogs have a truly flexible understanding of the mind and others' minds. It has always been assumed only humans had this ability."

The research is an incremental step in our understanding of dogs' ability to think and understand which could, in turn, be of use to those who work with dogs, including the police, the blind and those who use gun dogs, as well as those who keep them as pets.

Dr Kaminski ran a series of experiments in varied light conditions. In each test, a dog was forbidden by a human from taking the food. When the room was dark, the dogs took more food and took it more quickly than when the room was lit.

The tests were complex and involved many variables to rule out that dogs were basing their decisions on simple associative rules, for example, that dark means food.

There is no evidence on how well dogs can see in the dark, but the results of this research show dogs can differentiate between light and dark.

Dr Kaminski said: "The results of these tests suggest that dogs are deciding it's safer to steal the food when the room is dark because they understand something of the human's perspective."

Dogs' understanding may be limited to the here and now, rather than on any higher understanding, Dr Kaminski said, and more research is needed to identify what mechanisms are controlling dogs' behaviour.

In total, 42 female and 42 male aged one year or older took part in the tests. They were chosen only if they were comfortable without their owners in the room, even in complete darkness, and if they were interested in food. "Some dogs are more interested in by than others," Dr Kaminski said.

Previous studies have shown chimpanzees have a sophisticated understanding and seem to know when someone else can or can't see them and can also remember what others have seen in the past. It is not known how sophisticated dogs' is in comparison. Many earlier research papers have found that, for dogs, a 's eyes are an important signal when deciding how to behave, and that they respond more willingly to attentive humans, than inattentive ones.

Explore further: New camera sheds light on mate choice of swordtail fish

Related Stories

Dogs turn down extra food if a human provides the right cues

Apr 25, 2012

Dogs can be manipulated to choose against their preference by human cues, opting to turn down extra food in order to follow the human's choice, according to results published Apr. 25 in the open access journal PLoS ONE. The wo ...

Computer savvy canines

Nov 28, 2007

Like us, our canine friends are able to form abstract concepts. Friederike Range and colleagues from the University of Vienna in Austria have shown for the first time that dogs can classify complex color photographs and ...

Recommended for you

Slimy fish and the origins of brain development

18 minutes ago

Lamprey—slimy, eel-like parasitic fish with tooth-riddled, jawless sucking mouths—are rather disgusting to look at, but thanks to their important position on the vertebrate family tree, they can offer ...

Global importance of pollinators underestimated

21 minutes ago

(Phys.org) —Declines in populations of pollinators, such as bees and wasps, may be a key threat to nutrition in some of the most poorly fed parts of the globe, according to new research.

Meteorite that doomed dinosaurs remade forests

19 hours ago

The meteorite impact that spelled doom for the dinosaurs 66 million years ago decimated the evergreens among the flowering plants to a much greater extent than their deciduous peers, according to a study ...

User comments : 12

Adjust slider to filter visible comments by rank

Display comments: newest first

zorro6204
1 / 5 (2) Feb 11, 2013
I'm not convinced, the same reaction might exist if a machine that dispensed food issued the order. Could the dog be said to understand that the machine thinks? Or just that it's following a behavior more likely to produce food from the two-legged thing that provides it.
MandoZink
4 / 5 (2) Feb 11, 2013
I think this a bit too anthropomorphic an interpretation. If you observe, in nature many scavengers have rules of hierarchy they follow. You don't dare approach the food until the "top dog", or apex scavenger is through. This behavior is seen across a large variety of scavenging species. Your dog is aware of when the scavenging opportunity is now his.

Then again, animals may just have a quite different interpretation of the rules you think they were supposed to have learned. For instance, my cats absolutely know not to get on the kitchen counter when I am in the room. When I am not in the kitchen the rules apparently are different to them.

I'm not too sure if they have a continuing concept of your personal feelings, property or space when you are not present.
Myno
1 / 5 (2) Feb 11, 2013
The dogs likely come into the study pre-trained by prior life experiences. The light is on and the dog breaks a rule, and the dog gets punished. The light is off and the dog breaks a rule and gets away with it. The dog concludes, if the light is off, no punishment. This does not involve any ability of the dog to take the human's perspective. All it involves is an application of "stealing food" to "breaking rules". Any dog owner will tell you that dogs don't so easily generalize behavior patterns, but the only thing this study proves is, as the author says, "the results of this research show dogs can differentiate between light and dark."
Modernmystic
3 / 5 (4) Feb 11, 2013
I know when my dog has done something wrong 2-3 seconds after I come in the house....so do all dog owners....
ValeriaT
1 / 5 (2) Feb 11, 2013
dogs are much more likely to steal food when they think nobody can see them, suggesting for the first time they are capable of understanding a human's point of view.
They afraid of punishment, which doesn't mean, they really understand, why not do it. Trained dogs (don't) do many things, which the people taught them - despite they have absolute no idea, why they're doing it.
MandoZink
5 / 5 (2) Feb 11, 2013
I know when my dog has done something wrong 2-3 seconds after I come in the house....so do all dog owners....

I have no dogs myself but I have certainly seen that "Uh-oh, now I've done it" look in people's dogs. On the other hand, my next door neighbor's dogs are just as happy as larks no matter what they've wrecked, in spite of the attempts at training and the subsequent rage after another failure to communicate.

Here at my house my cats just don't give a shit. The simply have no shame.
Modernmystic
3.4 / 5 (5) Feb 11, 2013
Here at my house my cats just don't give a shit. The simply have no shame.


This is my experience with cats too. I don't know whether to admire them or scorn them for it....
NeptuneAD
1 / 5 (2) Feb 11, 2013
Dogs and Cats responses are just conditioning, trying to determine what is going on in their heads is absurd at best.

I can usually tell if my Dogs or my Cats have done something wrong, but I am also aware it is only because they associate their actions with reactions, it doesn't mean they are applying psychology or studying our behavior.
epicureous
3 / 5 (2) Feb 11, 2013
Dogs obviously can learn prospective. Rules of operation are not intrinsic to any species. (Although people will argue tooth and nail that humans somehow are born with magical abilities. oh the bible what a wonderful...) Nor can humans fully demonstrate (understand) the perspective of his fellow man (my girlfriend can't see in the dark, I know/learned this and use it to my advantage sometimes). Rules are learned. All animals make observational rules based on perceived perspective. You would have to be quite insecure with your place in the world to deny that. Being sneaky, is just that.

Dogs fully understand that humans can/do think. Dogs play to that. Yep dogs are smart. They can even tell time. They know when they've done something wrong (even if they've never been punished for it before). Dogs are even vindictive sometimes (I assume for the same reasons humans are- attention); my dog hides my cigarettes.

Let's talk about bovines and chickens...

Flame...
tpb
5 / 5 (2) Feb 11, 2013
When I was a kid, my dog would stay in the yard as long as he saw you were watching him. If you took your eyes off of him, he would leave the yard and not come back even if you called him just as he crossed the boundry.

While he was still in the yard he would always come if called.

He knew full well he would be punished for leaving the yard. Apparently he realized that once he left the yard he would be punished regardless of how long he was gone, so he made an extended trip.
sculpwnwiss
3 / 5 (2) Feb 12, 2013
they should have had a control group of humans (children) and seen how the dogs fared, humans act like dumb animals at times.
alq131
5 / 5 (1) Feb 12, 2013
This can't be new research. I saw a show on NOVA about dogs and they showed a more subtle experiment where the owners eyes were open or closed, which was the "trigger" for the dog. It didn't matter if the owner and dog were having direct eye contact, only if the dog could see that the owners eyes were open or closed.

They also detected subtleties in the side to which their tails wag. One side was a welcome/hello the other was a concern/warning.