Introduction
This is Kipper. As a 4 year old miniature schnauzer, knows a fair amount of commands: sit, down, crawl, roll over, bang bang, etc. Yet these are just words. Think about the immensity of all of the components of human language. We have a vocabulary of thousands of words, and an ability to develop universal grammar and automatically map out rules in our heads as we utter new combinations of words every time we speak.
Human to animal communication is much more limited than human to human communication, due to a few reasons. First, dogs cannot speak back to us. The sounds that a dog can make are extremely limited in comparison to people. Because a human stands vertically, evolution has allowed for a 90 degree bend in the airway. This permits the larynx to lengthen and for additional sounds to be made. Also the combination of a rounder tongue and greater complexity in the facial muscles allows for more vocal control and a greater variety in movement.
Man's best friend is much smarter than what can sometimes appear. Dogs listen, they interpret, and they perform. Dogs are typically trained with operant conditioning, a method in which a dog is rewarded for performing a command correctly. Often in training, they begin to understand that if something is done right, they will get a treat (positive reinforcement) and if they perform incorrectly, the treat is taken away (negative reinforcement). This method, combined with vocal commands and body language, is how a human communicates to a dog.
But how capable are dogs in understanding language as humans do? Dogs do have the sound discrimination ability to develop receptive language. They can actually pick up quite subtle distinctions in the pronunciations of human words. It is also hypothesized that dogs understand a vocabulary of about 200 spoken words and sounds, the equivalent to a small human toddler.
With these facts in mind, I decided to experiment on my own dog. I created eight different recordings of myself saying the command “sit,” in a variety of lengths, tones, and intensities. Kipper knows the command “sit” very well, but it is typically spoken in a brief and intense tone (sound clip 1). I wanted to test her to see whether she would be able to distinguish among different sounds. For example, if Kipper was able to respond to a more drawn out “ssssiiiiittttt,” that would mean she had the capability to interpret phonetics. If I created a shorter pitch but with the same intensity, Kipper's correct response would mean that she responded to the force of the command. I recorded her reactions on camera.
Here are the results.
Transcription
1: Clear and firm (control) |
2: Short and firm |
3: Off tone |
4: Long and pronunciated |
5: Higher and longer |
6: Deeper and medium |
7: High and short |
8: 2-toned |
Analysis
Just to note, the eight varieties of “sit” that I featured for this project were quite random. There are probably hundreds of varieties of pronouncing the word that could be used in such an experiment. Also, one thing I made sure to do was use Kipper's name. It was important for Kipper to know that these commands were directed at her, even if she was confused in interpreting them.
Kipper only responded to my first two trials, in which my pronunciations of “sit” were firm, short, and clear. This was how she was trained. My experiment shows that canine intelligence, at least that for a dog on Kipper's level, cannot distinguish just any tone or pronunciation of the same word. Thus, I can conclude that it is not what you say to a dog, but how you say it that results in successful communication.
I wanted to focus on a particular experiment featured in Stanley Cohen's book “How Dogs Think.” Patricia McConnell, from the Department of Zoology at the University of Wisconsin Madison, interviewed and recorded 104 dog handlers speaking a variety of different languages (English, Spanish, Swedish, German, Polish, Basque, Peruvian Quechua, Finnish, Chinese, Korean, Arabic, Farsi, Pakistani, Navajo, Shoshone, and Arapaho).
“McConnell was most interested in the kinds of signals that were used to change the dog's activity level, either exciting him to increase activity or causing him to slow down and inhibit activity. She found that the trainers, regardless of their language and culture, used signals that reflected the dog's genetically programmed rules of duration and repetition in a fixed combination: short sounds that were repeated several times increased a dog's activity level, while long, drawn-out, single sounds tended to slow activity or get the dog to remain still…Out of 2,010 signals that McConnell analyzed, these kinds of repeated short signals were never used to stop activity or to get the dog to stay in place. To slow or stop a dog's activity, longer, single signals were used, such as ‘Down,' ‘Stay,' or ‘Whoa' in English.”
McConnell's results on pitch:
“She has found that human signals to dogs vary in the number of different pitches in the signal (technically referred to as the “bandwidth” of the sound). Claps, tongue clicks, and kissing sounds have a broad bandwidth with both low- and high-pitched sounds. If a sharp sound like ‘k' or ‘t,' or a hissing sound like ‘ts,' ‘sh,' or ‘ch,' is part of a word, especially at the end, it also adds a strong dose of higher-frequency components. So phrases like ‘Take it,' ‘Let's go,' or ‘Go back' fit the pattern of multiple-broadband sound bursts. Contrast this to a word like ‘Down,' which has a narrower bandwidth, with a much smaller contribution from the higher-pitched sounds. So it appears that dog trainers naturally resort to more complex sounds with higher pitches to get dogs moving and lower pitched sounds to slow or stop them.”
Although dogs are able to be trained across various cultures with different languages and dialects, humans have different interpretations of the sounds a dog makes (in Spanish jau-jau, in French woa-woa, in Hebrew chav-chav, in Chinese wung-wung, etc).
In recent years, psychologists have coined the term “Doggerel,” the dog equivalent to Motherese. Doggerel features the following characteristics when humans communicate with their dogs:
There is no evidence that speaking in Doggerel to a dog helps it understand what we are saying. There is evidence that talking to dogs in a normal voice, meaningfully, improves a dog's receptive language capabilities.
If we could walk with the animals, talk with the animals what a wonderful place this world would be.
-Dr. Dolittle
Sources:
Coren, Stanley. How to Speak Dog . New York : The Free Press, 2000.
Coren, Stanley. How Dogs Think . New York : The Free Press, 2004.
About the Author
My name is Mindy and I've lived in Chapel Hill since I was three years old. None of my parents or grandparents speak a second language. Throughout middle school and high school I studied Latin up to the AP level, only to realize that I wanted to do nothing in life that would involve the classics. I longed for the usefulness of a spoken language in today's communication age. This inspired a three week trip to the beautiful city of Oaxaca, Mexico this past summer, where I balanced my study of Spanish with immersing myself in Oaxaca's sights and culture.
I feel as though my information technology abilities are average for a college student, basically meaning I can utilize most programs on my laptop. My past expressive art experience, which included piano, saxophone, ballet, and tap dancing, quickly died out before the age of twelve. Currently I enjoy learning about art and am considering a visual communication major at UNC and a career in graphic design. My music preferences have definitely been influenced by living in Chapel Hill, where I am lucky enough to see many amazing alternative and indie rock bands in concert.
Since my study of Spanish began, I was amazed not only to notice the influence of Latin, but also to realize that all of that “wasted” time could be put to use by studying the development of the romance languages. Language is such an amazing and complex part of human existence; in this class I want to explore in depth its history, evolution, and role in our world. I am planning to continue my Spanish education at UNC, so that studying abroad could be an option for me. I think this class will be very insightful, especially if I travel to an area where language is extremely diverse such as Europe.
After reading “The Expert Mind,” I thought about something that I would like to be an expert at not involving my academic goals. One of my addictions happens to be the number puzzle sudoku. Although sudoku is a one person game, a true expert would be able to fill in the empty boxes in seconds. At the moment I have a book full of 365 puzzles, all of different levels, and with the right strategies I have been able to solve all that I have gotten to so far. I am not a math genius by any means, but I do have the motivation for sudoku when it occupies my free time.