Going to the Dogs
The Takara Co. of Japan has a line of products that facilitate communication between animals and humans. Bow-Lingual, the first of those products to be released, is for dogs. Its a small transmitter that is worn on a dogs collar that sends signals to a hand-held device. Bow-Linguals Bark Translation module analyzes the dog's barks, determines which emotion was expressed (frustrated, happy, needy, on-guard, sad, showing off) and assigns a phrase to the bark that matches the emotion. The Body Language module helps interpret the dogs non-bark behavior. The product also keeps a history of past barks and can be used to record barks while the dogs owner is away. Dr. Matsumi Suzuki, president of Japan Acoustic Lab, and Dr. Norio Kogure, executive director of the Kogure Veterinary Hospital, developed the science incorporated into Bow-Lingual for Takara. Dr. Suzuki was kind enough to grant an interview to Speech Technology Magazine
(STM). STM: What made you first think about doing something like analyzing dog barks?
I was asked to examine if it were possible to understand dogs feelings from their barks for an animal-related television program. After that, several similar requests came in and I began to see basic emotional patterns in the barks. STM: Would you describe the approach that you used for this research as using a linguistics/philosophy approach or an experimental science approach?
Dr. Suzuki: Japan Acoustic Laboratories (JAL) has approached dog and cat translation using acoustic science techniques to analyze and establish the basic traits of these animals voiceprints. After that, animal behavioral science was used to determine and verify the final emotions. STM: Dogs and cats cant smile or reflect on what theyve just said. What independent variables/indicators did you use to determine that a specific bark expressed a given emotion? Was this only based on Dr. Kogures interpretations?
Dr. Suzuki: The six basic emotions are the culmination of expertise built up through my voiceprint analysis on dogs, done initially for various media, then specifically and in-depth for work on Bow-Lingual. Verification of the meaning of barks was done to a very high degree of certainty by using both the opinions of dog owners and animal behavior specialists, including Dr. Kogure. STM: What is your reaction to some of the frivolous translations that have been used in the promotional material for the Bow-Lingual?
Dr. Suzuki: Bow-Lingual, was developed by a toy company and was purposely meant to be entertaining. The phrases used are certainly within the interpretation of the emotional category, and they make the product fun for the user, I think thats fine. STM: In 2002, Harvard University awarded you, Dr. Norio Kogure and Keita Sato, (president of Takara Co.) an Ig Noble Peace Prize for promoting peace and harmony between the species by inventing Bow-Lingual. Ig Nobel Prizes are presented for achievements that cannot or should not be reproduced. What is your reaction to receiving this award?
Dr. Suzuki: I was a little surprised to have won the Peace prize, but then realizing the humorous nature of these awards, I now consider it not strange at all. STM: What is your reaction to humorous news reports?
Dr. Suzuki: One report took me by surprise: I read an article that said police in South Korea used Bow-Lingual in an investigation. Its probably fortunate that I dont know many details. STM: Can you extend your work on the Bow- Lingual and Meow-lingual to identification of the voice of a specific dog or cat?
Dr. Suzuki: I give priority to working on identifying voices with a little more social significance. I havent thought about applying my research to the identification of dogs or cats. But given the current technology, I think what you are proposing would be rather difficult. Im still working on overcoming problems of distinguishing emotions, which are conveyed with both voice and behavior.
Dr. Judith Markowitz is the associate editor of Speech Technology Magazine and is a leading independent analyst in the speech technology and voice biometric fields. She can be reached at (773) 769-9243 or firstname.lastname@example.org.