Can Dogs Cooperate With Each Other and With A Human?
Being able to communicate with people is just one of the ways dogs show how clever they are.
By Zazie Todd, PhD
By Zazie Todd, PhD
In the process of domestication, it seems that dogs have become
especially attuned to human communication. Does this mean they can cooperate
with a human to solve a problem? And what if they need to cooperate with another
dog instead? A study in press by Ostojić and Clayton investigates.
![]() |
Photo: Jim Parkin / Shutterstock |
The study is based on a “string task” in which two dogs (or
a dog and a human) have to pull each end of a string in order to gain access to
food that is otherwise out of reach on a platform.
The dogs were trained initially on a task they could solve
on their own, because both ends of the string were close enough together. Some
dogs opened their mouths very wide to get both ends in at once, while others
used their paw to bring the ends closer together to make them fit in their jaw.
When they pulled, a treat fell off the platform for them to eat.
Twenty-nine dogs were trained on this initial task. They
included a number of search-and-rescue dogs, some with agility training, and pet
dogs who had been trained in basic commands. They were a mix of breeds,
including Golden retriever, Labrador retriever, and Parson Russell terrier, and
of a range of ages.
The training for the study was very intense. Of the initial
twenty-nine dogs, fifteen dogs could not complete the experiment because their
owners did not have time, and three dogs did not take part because they were
not interested in the initial task. This left eleven dogs that took part in the
full experiment.
The dogs were from households with more than one dog, so
that when they had to work in pairs it was with a dog they already knew.
The cooperative task was very similar to the initial string
task, except that the two ends of the string were too far apart for one dog to
solve it on their own. Each dog took part paired with another dog, and then again
but paired with a human. When they set off from the starting point, each
partner had to go to their end of the string, and when they pulled food fell
off the platform. In the case of two dogs, two pieces of food fell off so that
each got a reward.
In another version of the experiment, the scientists erected
barriers so that one partner (dog or human) was delayed in reaching their end
of the string. This meant the dog that was released normally had to wait for
their partner to get to their end before they pulled the string; if they pulled
it too much then the string would come all the way through and their partner would
not be able to reach it.
Each trial lasted until either the problem was solved, the
string was pulled so far through that the problem became unsolvable, or 2 minutes
had elapsed. Each pair had up to sixty trials, or until they had solved it twenty
times in total, to show they really had understood the problem and their
success wasn’t a fluke.
Of the eleven dogs that had completed the initial training,
all of them could solve the puzzle by cooperating with their partner. This was
the case whether their partner was another dog or a human.
In the ‘delay’ condition, the non-delayed dog learned to
wait for their canine partner. Sometimes they tugged a little bit on the string
and then waited, before tugging hard once their partner was also tugging. There
was hardly any difference between the ‘delay’ condition and the regular
condition when the pair was made up of two dogs.
When the ‘delayed’ partner was a human, although dogs could
still solve the problem, they did not do as well. This is probably because the
length of the ‘delay’ was much greater; on average the human took an extra 13
seconds to reach their end of the rope, compared to just an extra 2s for the ‘delayed’
dogs. The task was therefore much harder with a human partner, since it
required a greater degree of self-control.
The authors point out that the social cue – of their partner
arriving at the rope – may not have been the only cue they used. They say the
feeling of pressure on the rope may also have been a factor, as could the sight
of the food moving closer to them when both partners had hold of the rope.
However, a social cue was definitely part of the solution, because even on the
very first ‘delay’ trial dogs waited longer to pull the string than on the
previous regular trials. Future studies could use a design in which the social
cue was the only one.
This study showed that once they had learned the puzzle
during training on their own, dogs were able to learn how to solve the task
when it required them to cooperate with a partner. This was the case whether
the partner was canine or human. They could even still solve the problem when
their partner was delayed in reaching the rope, and it required them to wait
before taking action.
The reasons why dogs can do this are less easy to understand.
The authors say, “In the case of cooperative problem-solving, it is not yet
clear whether the dogs’ ability to solve such tasks arises from group hunting
shown in other social carnivores and, in particular, in wolves, or from
abilities evolved during domestication.”
Animal cognition research keeps finding ways to show just
how clever dogs are. As ordinary dog owners, perhaps the thing to take from
this is that our dogs might like more activities that use their brain power,
whether it’s teaching tricks, agility, giving them problems to solve, or simply
letting them enjoy their regular walks.
What do you do to keep your dog’s brain active?
Reference
Ostojić, L., & Clayton, N. S.
(2014). Behavioural coordination of dogs in a cooperative
problem-solving task with a conspecific and a human partner. Animal cognition, 17, 445-459.