A couple of months ago, a team of scientists from the University of Eötvös Loránd, in Hungary, published a study with a surprising conclusion: your dog can understand you. Or in part, at least. At the very least, he is able to know if you are speaking to him in your language, the one that he listens to you frequently and that you use daily to address him, or in another foreigner. It is not the only investigation that draws a certain space of understanding between our pets and us. Another recent study concludes that dogs are capable of understanding an average of 89 words and phrases, and Japanese experts have shown that cats, at the very least, recognize their names.
The question is: Can we really communicate with our pets? Are there ways to understand them clearly? And to make ourselves understood more precisely? For years there have been applications that claim they can help. The approach varies from one to another, but in general terms the objective is always the same: to facilitate communication with our dogs and cats. Meow Talk, for example, claims to be able to translate meows into “readable” language.
Is it possible to “translate” pets?
“Each cat has its own unique vocabulary that it uses to communicate with its owners […]. This isn’t necessarily a language, since they don’t share the same meows to communicate with each other, but we can use deep machine learning to interpret an individual cat’s meows and translate it into human-readable language. MeowTalk gives your cat a voice!”, says the app, which exceeds five million downloads on Google Play. The person who writes this has a couple of kittens at home and has been able to test it. The application captures the noises made by the animal and converts them into short messages, such as “I’m tired”, “I’m resting” or “Love me!”
The operation is somewhat different in the Dog Translator app. Among its functions, it includes a list of images that represent states of mind, such as anger, happiness or surprise. When you press each one, the app makes a sound. Human to dog translator ensures that “help translate” the language of people to that of dogs and Cat&Dog Translator is also presented as a resource for communication and play. Using it supposedly passes the sound of our pets to messages. When the description that they include in the App Store or Google Play is analyzed in detail, most of the applications slide that their objective is fun and entertainment.
The list goes on. And it is quite extensive. Cat Translate: Speak to your Kitten, Human-Cat Translator, iCat: Cat Translator, Pet Translator, Dog Translator Simulator, iDog: Dog Translator, Dog Translator: Game for Dogs, Translator For Dogs Prank… There is even a collar available, Petpuls, which thanks to artificial intelligence and the analysis of approximately 10,000 recordings of fifty dog breeds, ensures that it can help us evaluate our pet’s emotions. It also makes use of the IoT to know, for example, its level of activity.
Although not all apps work the same, they generally agree on some points: most allude to the concept of “translator” in their name and, in one way or another, imply that facilitate communication with our animals. Even when, analyzing their description, they recognize that they cannot decode barks or meows. Some offer to “translate” human phrases and pet noise, others just one of those options; there are those that emphasize its fun character, to have a good time with dogs or cats, and also others that emphasize the potential of its automatic learning system and ability to, with training and time, “give a voice” to the animal.
The question is: Are they really effective? Can they “translate” our pets? The experts we’ve consulted are skeptical to say the least and encourage taking apps with a grain of salt. Especially if we intend to rely on them to treat or train our pets.
Aoife Ortega, a veterinary ethologist, questions their usefulness as real translation tools. “If you tell me that they give you a graph of facial expressions, body expressions, in different situations, they give you examples… you can get an idea, but they are useless that way,” explains the expert, who insists on a starting error: pretend transfer communication from cats or dogs to humans.
“Extrapolating human behavior or feelings to animal emotions is incompatible. They are totally different species. It is impossible for you to simply translate what it means to you with a cat’s meow,” reflects Ortega, who insists that another of the keys What apps often do not take into account is the context or the animals’ own bodily expression, vital to understand them correctly. “It’s too complex to do it with an app.”
“Barking is part of the ethogram, but if you only evaluate the barking and not the context or the animal on a physical level, it is impossible to recognize the behavior. We are extrapolating behaviors such as reasoning and animals do not reason. We must start there. Animals they associate, they don’t reason. It’s impossible for an animal, cats and dogs, to reason. If you’re using languages that require reasoning, you know it’s impossible. It is completely incompatible“.
Ortega warns about the risks of basing the treatment of our pets on what the applications tell us. “By extrapolating people misunderstand behaviors and the main problem that exists in the behavior of animals is due to misinterpretation,” he stresses. Whoever wants to go deeper, the ethologist encourages them to go to professionals and request proven scientific literature.
Skeptical is shown too Eduardo Polin, doctor in Psychology and expert in learning and animal behavior. “I do not know them in depth, but at first I would dare to say that there is not much basis to support these claims. There is no literal translation. When dogs bark, for example, they are not doing the same thing as when we speak in another language. There is no such equivalence. I’m certainly very skeptical,” she stresses.
“My suspicion is that they make a major mistake, which we sometimes make with humans as well, and that is that every time we make a noise or vocalization we automatically attribute a communicative intent to it. And that doesn’t have to be the case. Animals and humans humans many times we speak, we make noises, and we are not necessarily wanting to tell another individual something specific. That is, the vocalization does not necessarily have a translation”, he abounds. An example would be the cry from an injury. It is spontaneous, involuntary, almost a reflection of the pain we feel; but it certainly does not seek to convey a message to other people.
The “crucial point” for him, however, consists in the importance of the “function of behaviors”, which requires analyzing the environment, the consequences… a set of aspects that applications cannot assess. “What I would say is that the concept of translation in this sense is wrong. What the tools do It’s not translating, it’s something else. You have to be very careful when interpreting. That it can be used for some things… I don’t get into that; but that it is a translation is difficult for me to accept. That’s for sure,” she stresses.
What do the apps say? MeowTalk, for example, acknowledges that each animal has its own peculiarities, but insists on the potential of machine learning to interpret meows, “translate them into human-readable language” and frame them in nine basic categories based on their mood or activity. “When you give the app five to 10 examples of a specific meow from your cat, it can begin to recognize it when it hears it,” the company notes.
Petpuls —which includes among its functions a system that interprets our dog’s state of mind— also emphasizes its experimental base, created with over 10,000 recordings fifty breeds with different characteristics. During the process, he explains, its developers turned to veterinarians and pet experts, as well as research centers. His tool, which also offers other services, is based on artificial intelligence and deep learning and ensures that it has a recognition rate that exceeds 90%.
Although they play with the term “translator”, a word that they often include in their own name, many apps slip into their description that their objective is basically playful. This occurs, for example, with iCat: Cat Translator, which concludes in its Google Play tab: “Remember: this is not a real cat language translator, it is just a game, a simulation. Most likely, it is joke, but 95% of cats react and enjoy playing together.” Another example is Dog Language Translator Simulator, which specifies: “It was created just for fun.”
Apart from the apps and devices that claim to help us interpret our animals, the truth is that for some time in the academic field we have been working to better understand how do they communicate and to what extent can animals understand us. In addition to the research carried out by the University of Eötvös Loránd or those that analyze how many words dogs and cats recognize, attempts have been made to decode the sounds emitted by dolphins or birds to facilitate their study and we have delved, for example, into how they interrelate and express the apes.
None of them, however, claim to be able to literally “translate” animals into human language.
Images Chewy (Unsplash 1 and 2) and Andrew S (Unsplash)
George is Digismak’s reported cum editor with 13 years of experience in Journalism