What's That, Lassie?

Student research taps human-dog communication

by MaryAlice Bitts-Jackson

The pet-human relationship is an important part of many Americans' lives—and more so than in decades past. Fully 85 percent of dog owners say they consider their pet to be a member of their family, according to a 2006 Pew Research survey, and most feel closer to their pets than to their parents. And it’s not just lip service: According to the American Pet Products Association, American pet owners spent a total of $58.51 billion on their pets in 2014, much of it on pet pharmacology.

Dogs, in particular, are adept at forging these close bonds by correctly “reading” humans (one recent study, published in Current Biology, indicates that the human voice elicits similar responses in canine and human brains). The question is: Do we, in turn, understand dogs, based on their vocalizations?

Not so much: Cross-species translations might be trickier than you think.

That’s according to a survey conducted by a trio of biology majors as part of a Senior Biology Seminar last fall. Instructor of Biology Gene Wingert assigned the project to Steve Collins '15, Ashley Kinney '15 and Grace Mulcahy '16, based on their research interests, and gave the students a semester to explore canine-human communication by determining dog-owners’ and non-dog-owners’ ability to understand the vocalizations of several different species of dogs.

After reviewing available related research, the students recorded 10 vocalizations by a total of five dogs—including Kinney’s German shepherd, Bailey—as they responded to different stimuli. (One professor volunteered the family pooch; its spirited reaction to a passing skateboard made a great clip.)

The team then distributed 100 surveys in classrooms across campus; 61 percent of respondents owned, or had owned, a dog. Each participant was asked to listen to the recordings and identify the emotions the dogs were trying to convey (fear, joy, anger, play or despair). Seventy-two respondents correctly identified anger/aggression, 64 recognized despair, 20 identified fear, 23 picked up on joy and 30, play. (“Many confused joy and play,” said Kinney, noting that this is understandable, since dogs are joyful when playing, and they play when joyful, and that more respondents may have guessed correctly had these options been combined into one.)

After submitting their surveys, the participants viewed video recordings of the same vocalizations to see if the dogs’ visual cues informed their answers.

“Most participants expressed their willingness to change their answer after the visual representation,” said Collins. “People often explain that they can understand their dog, but if you took them out of the environment in which the dog was barking or the body language of the canine, most likely the owner will have no idea.”

The student researchers also found that when listening to the recordings, the dog owners were correct 42 percent of the time, while non-dog owners were correct 37 percent of the time, indicating only a slight advantage for dog owners. “Therefore, our data showed that humans cannot correctly identify a dog's emotion based on the bark alone, regardless of ownership status,” said Kinney, who plans to be a veterinarian. The students presented their findings during Dickinson’s recent Biology Student Research Symposium, one of many student symposia held on campus throughout the academic year.

“The team did a great job,” said Wingert, who got the idea for the semester-long research project after watching a NOVA special on dogs and was surprised by the results of the students' investigation. Based on the findings, he’ll fine-tune the assignment in the coming months, so that a new crop of dog lovers will have a chance to further the line of inquiry next year.

Learn more

• “Barks and Beetle Guts”
• Research at Dickinson
• Students as Scholars
• “Top Dog”
• Department of Biology
• Latest News

Published March 18, 2015