November 21, 2016

Effects of selection for cooperation and attention in dogs


This study was interested in the way different breeds respond to human gestural cues. The researchers wanted to investigate whether or not there would be a significant difference in performance between breeds that were bred to work in cooperation with humans in tasks that require constant visual contact (i.e.  herding dogs, gun dogs/retrievers), and those who were bred to work independently (i.e. hounds, earth dogs, and sled dogs), in their ability to be successful in a “two-way object choice” test. They also were interested to see if there would be a difference in performance between brachycephalic (short-nosed) and dolichocephalic (long nosed) breeds. Researchers believe this difference would stem from the fact that brachycephalic breeds have been shown to also have more frontal facing eyes, and therefore have more centralized vision and are less likely to be distracted by stimuli in their peripheral vision.

Both of these questions were answered by conducting a two-way object choice test with 140 dogs, including a relatively large sample of Siberian huskies.  Although previous studies have shown that success in two-way object choice tests does not depend on age, sex, or various environmental factors such as special training or differences in amount of time spent with owner, all the dogs involved in this study were socialized at about the same level in human families, and all were walked regularly or had gone through basic obedience training. For Study 1, the subjects were separated into three groups: cooperative workers, independent workers, and “mongrels” (not crosses of two known breeds, usually the pups of other mongrels). Siberian huskies were categorized as independent workers. For Study 2, the same dogs were divided into two groups based on their morphology, the brachycephalic breeds and the dolichocephalic breeds. They only used those breeds that could be clearly defined as being on one extreme or the other, so the sample size was slightly smaller. Then the dogs were subjected to a test in which an experimenter set up two empty bowls, both scented with food, in front of her. Then, with her back turned, she placed a piece of food in one of the two bowls. Next, she turned back to the dog and in one very clear, but very swift motion, while keeping direct eye contact with the dog, she pointed directly at the bowl with the food in it and then re-crossed her arms. As soon as this motion was completed, the dog’s owner released the dog and researchers recorded which bowl the dog chose to go to, and removed the other bowl. Each dog did this 20 times, and half of the time the food was in the right bowl, and the other half the food was in the left one.

In both studies, the dogs’ success rates were better than chance. Study 1 supported the researcher’s initial hypothesis, showing that dogs from breeds characterized as “cooperative working breeds” performed better on the test than both of the other two groups. One example can be seen in the difference between the Australian Shepherd, a herding dog who was correct 19 of the 20 trials, and the 4 Siberian Huskies, who respectively were correct 14, 9, 9, and 16 times. Study 2 also supported their original hypothesis, showing that brachycephalic breeds were more successful than dolichocephalic breeds.

The researchers attribute the results of Study 1 to the fact that these cooperative working breeds have been artificially selected to be able to work well with humans through visual contact to be successful at whatever task they have been bred for. They explain that this “may not be attributable to differences in cognitive abilities per se, but rather reflects a genetic tendency to be responsive to social stimuli…it is more likely that direct selection for utilizing human visual signals endowed cooperative breeds with an ability to inhibit their own spontaneous behavior and benefit from human social cues” (Gasci, McGreevy, Kara, Miklosi). They also point out that the differences between the groups cannot be explained by differences in socialization because all dogs involved in the study were raised under very similar living conditions. Another part of discussion was about the surprising lack of success of the “mongrel” group. They suggest that because the mongrels likely had genetic histories that selected for independent survival skills to reproduce without the help of humans, they developed traits similar to the independent worker dogs.  

 As predicted, the brachycephalic breeds tested in Study 2 were more successful than the dolichocephalic group. They explain that this is due to the short nosed breeds’ ability to have much more focused attention on the experimenter. According to this paper, other sources say that humans may not have selected for this trait to its ability to help in communication, but instead because they preferred dogs that looked directly at them for longer amounts of time.

At the end of the article, the researchers also point out that there are some limitations on this experiment and its results. For example, they analyzed only the success rates of the dogs, not the behaviors they exhibited at other points of the experiment, such as how long they looked at the experimenter before choosing a bowl. They also explain that they only observed differences in head shape and their effects on performance in breeds on the extreme ends of the brachy-dolicho spectrum. Lastly, they use this data to caution future researchers against making generalizations about dog performance in “two-way object choice” tests when comparing dogs and wolves. 

Reviewed by Sarah Slayton 

  • Márta GácsiEmail author,
  • Paul McGreevy,
  • Edina Kara and
  • Ádám Miklósi
Behavioral and Brain Functions20095:31

DOI: 10.1186/1744-9081-5-31

Received: 02 February 2009

Accepted: 24 July 2009

Published: 24 July 2009