In 2017, a team of scientists discovered that the click-like vocalisations of sperm whales closely resemble Morse code. It planted the seeds for a large-scale initiative called the Cetacean Translation Initiative, or Project CETI, which would employ AI to interpret these whale sounds so that people could understand them. In addition to improving our understanding of animal behaviour, technology has ironically helped humans identify our own species’ limitations. This might either encourage stronger conservation efforts or breed hubris that could turn the newfound understanding of animal communication against conservation efforts.
Initiatives to translate have been made with regard to more than only whale communication. While some researchers used audio and video records to uncover the context of bat calls, others developed an algorithm that can analyse the emotional states of pigs based on their cries. Up until recently, attempts to decipher animal communication were mainly based on observation. The scope of research investigations has been greatly expanded by the development of technology. In her book The Sounds of Life: How Digital Technology Is Bringing Us Closer to the Worlds of Animals and Plants, Karen Bakker wrote that when these digital devices are used together, they “function like a planetary-scale hearing aid: enabling humans to observe and study nature’s sounds beyond the limits of our sensory capabilities.” These gadgets have, however, also produced enormous volumes of data that are laborious and time-consuming to manually sort through.
Thus, to research the calls of a variety of species, from chickens and rodents to cats and lemurs, the scientific community is increasingly adopting technological instruments like drones, recorders, robotics, and AI.
According to The New York Times, machine learning algorithms are able to spot small trends that may elude human experts. These programmes are capable of differentiating between various animal voices. Additionally, they are able to distinguish between the many animal cries made in various situations and divide these calls into smaller segments, which is thought to be a crucial step in interpreting the message. For instance, the software DeepSqueak employs deep learning algorithms to recognise, interpret, and classify rodents’ ultrasonic squeaks, which are otherwise imperceptible to the human ear.
In new ways, this might help conservation efforts. The goal of Project CETI is to decode the syntax and semantics of sperm whale communication, which is regarded as federally endangered in the United States. The Earth Species Project (ESP), a non-profit organisation focused on utilising AI to interpret non-human communication, is cataloguing Hawaiian crow cries and even making attempts to develop new technology that would enable people to communicate with animals.
The ESP’s study on the highly endangered Hawaiian crows, raised in captivity, has various conservation-related implications in light of studies about how climate change is hurting bird populations, decreasing the richness and variety of birdsong, and causing “‘song culture’ to break down.” It might aid in the study of extinct bird sounds and the introduction of those thought to be most vital to captive birds. “The goal we are pursuing is to understand animal communication and identify non-human language… Aza Raskin, co-founder and president of the ESP, told The Guardian that we are currently creating technology to assist biologists and conservation efforts.
Understanding social dynamics within species can also benefit from the abundance of information hidden in animal communication. Researchers discovered that each mole rat had its own distinctive acoustic signature and that each colony has its own vernacular, which is passed down through generations, after using machine learning to analyse about 36,000 chirps of naked mole rats. These dialects were eliminated when a colony queen was removed. An entirely new vernacular would appear with a new queen.
Deciphering animal communication may also aid in our understanding of “the development, neurology, and cognitive basis of human language,” according to studies. Songbirds “display a human-like aptitude to learn complicated vocal patterns,” according to several scientists, refuting the common view that animal communication has little to do with the development of human language.
However, it can be challenging to translate animal language. Raskin pointed out that animals do not simply communicate by sound, thus we will need to translate between many forms of communication, such as the “waggle dance” that bees do to alert other bees to a nectar source. Additionally, such studies bring up a more fundamental query that has divided the scientific community over the years: Do animals even have the capacity for language?
In her work, Bakker emphasised how indigenous societies have been familiar with animal communication for a long time. The Western scientific community, which has “historically denied the idea of animal communication outright,” has voiced strong opposition to academics studying animal communication, she said.
Assuming that language is unique to humans and that in order to demonstrate that animals possess language, we would also need to demonstrate that they could learn human language, so many of the attempts to teach primates human language or sign language in the 20th century were motivated by this belief. And looking back, that was a pretty human-centered viewpoint, Bakker said Vox, noting that modern research is taking a totally different tack and has produced some amazing findings, such as how elephants signal differently for threatening and non-threatening humans.
While it’s possible that understanding animals could improve our ties with the environment and direct protection efforts, Bakker also brought up a number of potential ethical issues. The capacity to communicate with other species “sounds amazing and fascinating, but it could be utilised either to foster a stronger sense of family, or a sense of dominance and the potential to manipulate to domesticate wild species that we’ve never as humans been able to control,” she added. The complexity of animal communication may not be fully unlocked by AI, according to researchers. Even so, it offers a hint of what may be feasible in the future.
Technology, particularly machine learning, is sustaining this promise even though there is still a long way to go before there is a Google Translate equivalent for animal languages that can comprehend the subtleties of intra-species communication. Understanding animal languages could lead to a wide range of opportunities, including influencing conservation efforts, defining how we will interact with other species in the future, and even providing information about how human language evolved.
