Could Science Build a Pet Translator?

Any fan of sci-fi shows such as Star Trek has probably longed for a universal translator, one which can translate any alien language on first contact, even if neither party knows anything about the other's language or culture. While a universal pet translator is probably not realistic, one scientist is studying animal language and behavior, with the idea of using artificial technology to develop a sort of "animal dictionary."

Dr. Con Slobodchikoff, a professor emeritus of biology at Northern Arizona University and the author of "Chasing Doctor Dolittle: Learning the Language of Animals," has spent much of his career studying the language of prairie dogs. His research shows that prairie dogs have a complex and sophisticated language, with nuances in their alarm calls to differentiate between predators: humans, coyotes, domestic dogs, and red-tailed hawks have their own call sound. The prairie dogs also have nuances in their vocalizations to describe the size and shape of an individual predator.

Dr. Slobodchinoff and a computer scientist colleague created an algorithm that turns the prairie dog vocalizations into English. In 2017 he founded a company called Zoolingua. That company's goals, according to the website, are to create a device that is as easy to use as a cell phone, that will tell the human user what the dog is trying to say based on the dog's vocalizations, facial expressions, and body language. Slobodchikoff told NBC News in January 2018, "if we can do this with prairie dogs, we can certainly do it with dogs and cats."

Slobodchinoff told NBC News that his work is currently based on analyzing thousands of videos of dogs. The videos show a dog barking, along with specific context clues such as what is going on at the time the dog is barking. He's developing an artificial intelligence (AI) algorithm to interpret these communication cues.

If the AI sees a certain bark along with a certain tail wag that happens consistently, for instance, the theory is that the AI would be able to "read" these communication signals and deliver the human an interpretation of what it means. The work is still in the early stages, but Slobodchinoff is clearly hopeful.

"Behavioral futurist" Will Higham, who wrote a report on "the next big things" for Amazon, predicted that Slobodchinoff's technology would be in use within the next 10 years.

In 2014, ABC News published a story titled "Talking Dog Device Ready to Hit Market Soon." That story reported on an assertion by The Nordic Society for Invention and Discovery (NSID) that they were developing a headset that distinguished canine thought patterns and then spoke them as short sentences via a microphone.

The product, called "No More Woof" combined EEG (electroencephalography) sensoring, micro computing, and brain-computer interface software. At the time of the report, the NSID was selling the headsets as a pre-purchase on Indiegogo, with three different versions that ranging in functionality and price from $65 for the micro, $300 for the standard version, or $1,200 for the Superior customizable mini-speaker.

That product seemed to go nowhere, as the Indiegogo fundraiser is now closed, there is no date on the NSID website, the product is referred to as a working prototype but is not available for purchase, and a website that visitors are referred to for further information is expired.

Back in 2002, the satirical award Ig Nobel Prize awarded a prize to Japan Acoustic Lab and some researchers for "promoting peace and harmony between the species" by inventing Bow-Lingual, a computer-based automatic dog-to-human language translation device. Reviews on the accuracy of the product are pretty negative, and the website given on the Ig Nobel award page no longer exists, although the Bow-Lingual blog is still active as of February 2020.

Still, perhaps now, Dr. Slobodchinoff is the pet translator's next best hope.

It's not preposterous to think that Slobodchinoff's dream of allowing dog owners to become their own versions of Dr. Doolittle might actually come true. In reality, we are already using AI and machine learning (ML) in many aspects of our daily life.

Spam filters, for one, analyze language to help weed out emails the filter is pretty sure you don't want to get. Delivery services like Uber and Lyft determine the pricing of your trip based on AI and ML. Social media providers use AI facial recognition and algorithms that suggest people you might want to be friends with based on who you know. Remote check deposits, self-driving cars, and automatic financial investing are just some of the ways AI is already being used.

It's not hard to imagine that an algorithm can take similarities between barks, tail wagging, body posture, and other physical cues, and turn them into a dictionary of sorts. If there is any time in which we have the technology to actually make this happen, it's now, with machines and robots getting smarter by the day.

Slobodchinoff's website doesn't show any updates in research, but he has previously stated that it's in the early stages. It could take a while to analyze and interpret thousands of videos of dogs acting within different contexts. But stay tuned! If the "futurist's" prediction is correct, by the turn of the next decade you could know exactly what your dog is thinking.