In one of the first signs that my then new husband trusted me, he agreed, albeit reluctantly, to let me take William the cat to Albuquerque. I was beginning my postdoctoral work at the University of New Mexico, and he was staying behind at his faculty job in Ohio before joining me in the summer. I thought William would be a good companion as I started out on my own, though the cat had been my husband’s before we got together. Since I would take the cat with me when I flew down, I figured it would be a good idea to get him some kind of sedative to keep him calm during the journey. I duly inquired at the vet’s office about what they could prescribe. The veterinary technician looked thoughtful. “Is the cat very verbal?” she asked.
Verbal? William produced the usual assortment of meows and purrs, but had never actually, well, talked. I suddenly had a sense of inferiority, wondering if other people had had talking cats all this time and mine was just inadequate. Then I realized that of course she meant vocal, not verbal, and I heaved an internal sigh of relief. No, I assured her, he wasn’t particularly prone to making noise. The tech proceeded to give me some pills for William, and he remained suitably subdued during almost the entire trip, waking up to express his displeasure just as the plane touched down in New Mexico and yowling nonstop down the jetway, through baggage claim, and into the taxi home. He never, however, uttered a word.
Animals being verbal, and not merely vocal, is not just a matter of concern for those of us with a sudden fit of pedantry in the veterinarian’s office. Language is often viewed as the final wall between humans and other animals, and a characteristic whose evolution seems obscure. Even after tool use fell to the bees and crows, and the mirror test became something fish could pass, the Rubicon that remained was language. Raymond Bergner, the psychologist who fretted about our inability to define behavior mused:1
What is different with respect to animal behavior is not the parameters that apply but the capacities of infrahuman species. For example, aside from the modest sign-languistic ability of certain species such as chimpanzees, the great majority of animals seem to possess neither language nor a capacity to acquire it. Accordingly, they are not able to act on the enormous range of discriminations that are captured in human language and that are thereby available to human beings in their behavior.
After undergoing a religious conversion, author Andrew Norman Wilson became skeptical about the evolution of behavior, particularly in humans. From a very different starting point than Bergner, he comes to a similar conclusion:2 “Do materialists really think that language just “evolved,” like finches’ beaks, or have they simply never thought about the matter rationally? Where’s the evidence? How could it come about that human beings all agreed that particular grunts carried particular connotations? . . . No, the existence of language is one of the many phenomena—of which love and music are the two strongest—which suggest that human beings are very much more than collections of meat.”
In a more scholarly vein, although they generally agree that indeed, language “just ‘evolved,’ ” linguists, psychologists, and biologists have spent decades debating, often with an astonishing degree of acrimony, even for academics, whether it arose suddenly in our evolutionary history or more gradually. If the latter, what are its best counterparts in other animals, and which qualifies as similar to us? According to the view that language arose suddenly, humans—in particular, the behavior of humans, not their appearance or bipedalism—are distinct. That means if we want to settle the arguments about the inherent brutality of men that I brought up in the introduction, we are stuck with those brutes, because some behaviors, at least, evolved in a way unlike other characteristics. In a 2003 article3 in the New York Times, Nicholas Wade claims, “The only major talent unique to humans is language.” Whether language is a talent, akin to playing the violin or shooting hoops, seems questionable, but language seems to have remained as the last bastion of human behavior categorized as being special, that is, sudden. As should be apparent by now, I think this argument is false. Behavior, even a behavior as complex as language, evolves the same way that physical appearance does. This doesn’t mean that human language lacks unique components, but that those components are not somehow outside the rules governing the evolution of other traits.
This chapter takes a critical look at claims about who talks and who doesn’t. What constitutes language? Is it the same thing as communication, putting language on a par with screeching monkeys or mimicking parrots? Does it require what some researchers call mental time travel, the ability to imagine oneself in another time or place? And is language required for thought itself?
We Don’t Want to Talk about It
Although people have been speculating for millennia about where language came from and how it differs among cultures around the world, modern scholarly consideration of the origin and evolution of language got off to a rocky start when the Linguistic Society of Paris banned discussions of the beginning of language in 1866 because “it was an insoluble metaphysical problem.”4 Leaving aside that being insoluble, much less metaphysical, doesn’t seem to stop people, particularly academics, from discussing other topics, the society had a point. As is often said, behavior doesn’t fossilize, and we have no exact analogues of language among animals, though there are some interesting approximations, as I will discuss later. At the same time, the nearly seven thousand human languages of the world seem almost eerily similar: no human society has a language that can be described as less fully developed in its structure or its use of symbols than any other language. We therefore don’t have an obvious intellectual foothold to determine how language itself began, because there is no linguistic equivalent of trying to understand the eye by examining simpler light-sensitive organs in an invertebrate ancestor. How did we get to modern speech from primordial ancestors that lacked it?
Some evolutionary biologists see human language as representing one of the “major evolutionary transitions”—steps that made a huge difference in the way that life on Earth proceeded. Some of the other transitions include becoming multicellular (composed of many interacting cells instead of a unit with a single nucleus), and reproducing sexually, with genes from two individuals combining rather than being replicated as an identical copy. Like those other steps, language opened the way for many other abilities. Being able to communicate with symbols, to refer to things in the past and future as well as the present, means that people could form complicated social groups, with alliances that had reference to previous actions instead of only immediate responses.
In The Goodness Paradox, anthropologist Richard Wrangham suggests5 that the crucial function of language is that it allows gossip—the discussion of others’ behavior. That in turn means that someone’s reputation is made and can be used as a basis for social decisions—who gets rewarded? Who gets punished? Wrangham is particularly interested in the latter question, because a society that can impose moral sanctions based on prior behavior can ensure that delayed punishment is meted out. Chimpanzees, in contrast to people, do not seem to care what others think; an experiment in which one chimp sometimes could see another individual steal food from a third didn’t make any difference to the thief’s behavior. The same thing applied if the chimps could see another individual being helpful. In humans, of course, language allows people to explain their motives, deal with mistaken impressions, post inflammatory material on Twitter, and coordinate punishment, up to and including execution of those who violate social contracts. According to Wrangham, no human ability is “as special as the one that enabled conspirators to trust one another sufficiently to collaborate to kill a bully.”6
Whether or not one accepts coordinated and preplanned punishment as one of its primary outcomes, let alone benefits, language has made a huge difference in human evolution. Some scholars see the acquisition of language as part of a number of evolutionary innovations that allowed humans to become humans. These attributes include bipedalism, which in turn allowed early humans to carry objects in their hands rather than their mouths. Carrying things in one’s hands then paved the way for language, according to a paper7 by Charles Hockett and Robert Ascher published in 1964, because this freed the mouth for other activities. “What were the hominids to do with their mouths, rendered thus relatively idle except when they were eating? The answer is: they chattered.”
No one else seems to have picked up on this hypothesis of speech-as-side-benefit-of-hands-free-living. Hockett and Ascher also segue8 into the unsettling idea that speaking with the mouth isn’t required, and that if evolution had proceeded differently “speech sounds today might be anal spirants,” which are exactly what you think they are. That no other animals communicate via this channel did not seem to discourage them, but the notion does not seem to have gained much traction, which is probably all to the good.
As Old as Thought Itself?
The ban by the Linguistic Society of Paris has now been lifted, of course, and fraught discussions of the origin of language abound. Some scholars still maintain that language emerged all at once, and is a universe apart from communication in other species, while others see a more gradual development. As I also noted, I fall into the latter camp. Scholars’ thinking about this issue has been helped by developments on two fronts.
First is what we mean by fast evolution. Although Darwin emphasized the slow, incremental pace of natural selection in producing evolutionary change, a lot has happened since his time, and scientists now recognize that evolution—changes in the proportions of different genes in populations—can happen extremely quickly. The crickets that I study in Hawaii evolved that silent male form in fewer than twenty generations, and that’s not even a record. Admittedly, losing the ability to sing in an insect is far different from gaining the ability to not only sing but also recite poetry and rant about politics, but many, many examples of rapid evolution have been recorded. Microbes, of course, are the champions of speedy genetic change, and insects are not far behind, but fish, birds, and even humans can respond remarkably quickly to pressure from the environment. Guppies, for example, evolve to become sexually mature at an earlier age in places where their predators abound, which helps them reproduce before they get eaten; this is a genetic change in the way they develop, not a response to immediate conditions.
Part of the difficulty in thinking about language evolving is that it can be hard to understand how any complexity, language included, can arise from small changes, each building on the one before. Andrew Norman Wilson’s bewilderment9 about all human beings having “agreed that particular grunts carried particular connotations” may stem from a misunderstanding of the way evolution proceeds. That misunderstanding is common, and it is related to the widespread belief I have mentioned before, that behavior is different from physical attributes. Humans didn’t get together and agree on grunts having meaning any more than vervet monkeys discussed which call means “predator from above” and which means “predator from below,” though such differentiation is well established. For that matter, mammals didn’t get together and discuss how fur would be structured, or vertebrates the way that the immune system would recognize foreign substances and manufacture antibodies to remove them from the body. But just because something is complicated doesn’t make it impossible. Language evolved like other complex mental, and physical, characteristics.
The second development in how we look at language is in the ability to detect its precursors in ways that the Paris linguists probably couldn’t even have imagined. The fossil record was extremely limited in 1866, and back then we knew much less than we do today about the ancestors of humans. Many scholars think that language emerged perhaps one hundred thousand years ago, but a few anthropologists now suggest that language is detectable at least one million years ago, in our ancestors Homo erectus, and that it is also possible that Neanderthals had some form of symbolic communication, though this is speculative. The evidence, they say, comes from the use of tools such as hand axes that several people often used for a long time or hoarded with other items, a behavior that indicates a more sophisticated form of communicating than can be achieved with gestures or a few vocalizations. Long-distance ocean voyages, undertaken by people in islands in various parts of the world, also suggest that people had to discuss events in the future and coordinate their activity, again difficult to do without symbolic communication that can refer to events occurring in something other than the here and now. Symbols—items with meaning beyond their physical existence—do not just occur in rituals such as burials; they can be part of our day-to-day life. Thus, if people had a tool with an agreed-upon use, that might count as a symbol. What exactly constituted such early forms of language is, of course, unknown.
Another source of evidence for a long-ago and gradual origin of language is physical, but of a very different nature. No other primate besides the human speaks, and for many years scientists assumed that nonhuman primates were physically incapable of even approximating human language because of limitations in their mouth, throat, and brain. This apparent gap meant that at least our kind of language really did seem to have appeared all of a piece, without much in the way of ancestral foreshadowing in nonhuman primates.
Now, however, that assumption is being questioned. A group of European and US scientists performed some highly sophisticated brain imaging on a macaque,10 a kind of monkey that has not shared a common ancestor with humans for twenty-five million years. They found a structure that forms a pathway between the auditory part of the brain and the frontal lobe, similar to the component that is seen in humans and that is part of the machinery needed for speech. This link suggests that the seeds of human language are buried deep in the primate lineage. Even though the macaques cannot speak, the ancient origin of the pathway means that human language has come from the same kind of parts-lying-around-in-the-garage that other characteristics do.
Another anatomical theory about human speech and its uniqueness that has fallen in recent years is that of the descended larynx. The larynx, sometimes called the voice box, is a structure that sits in the neck and is important for both breathing and speech. Other mammals have a larynx, including nonhuman primates, but some scientists had thought that because of its position high in the throat, it was simply impossible for those animals to produce speechlike sounds, particularly vowels. In humans, the larynx descended—moved to a lower position in the throat—around two hundred thousand years ago, which used to be thought of as the absolute floor for the evolution of language. And indeed, efforts to try to teach chimpanzees, for example, to speak had resulted in failure, though apes, like the famous gorilla Koko, could learn to communicate with humans using symbols.
But more recent examination both of the anatomy of monkeys and apes and of the necessity for certain structures in producing vowels tells a different story. That descended larynx isn’t unique to humans after all. It is seen in other animals, including some primates. And those primates do produce calls in a way that is similar to the way people make vowel sounds, in a manner called “speech-ready” by some researchers. Interestingly, some of the best examples of speechlike sounds come from marmosets, tiny monkeys that are much less closely related to us than are the apes. Like humans, as marmoset infants get older, they change their sounds in a way that reflects their brain development. So speech may have its roots much further back than two hundred thousand years, just as the brain imaging studies suggest. The authors of the anatomical study conclude that “the idea of recent, sudden, and simultaneous emerging of speech and language is no longer plausible.”11
Finally, people sometimes derisively refer to talking as “flapping your lips,” which is a vivid if crude description. Scientists being what they are have measured the rate at which those lips are flapping, or more accurately, the number of times per second someone opens and closes their mouth. This turns out to be remarkably standard across all spoken languages, at between two and seven open-close cycles per second. In chimpanzees, which smack their lips to communicate, that rate is four cycles per second, and other apes are similar to them. This means that speech could have arisen from what a 2020 paper12 calls “ancient primate rhythmic signals,” which is a rather more dignified phrasing than lip flapping.
Monkey See, Monkey Point, Monkey Say
So if language did not emerge full-blown, all at once, then what did the first speech sound like? One possibility is that it was similar to pidgin, a term for a simple cobbled-together language that modern people devise when they do not share a language. For example, people on Pacific islands who come from a number of different places, including English-speaking countries, may speak pidgin English when they need to conduct business together. Such simplified languages may have arisen before the more grammatically complex tongues that are spoken today.
Another option is that gestures preceded speech, since spoken language is just one form of communication. Michael Corballis, who was a professor of psychology at the University of Auckland, suggested exactly that,13 and also focused on “mental time travel,” the ability to imagine what might happen in the future and what has happened in the past, as a crucial part of the evolution of language. Being able to refer to what he called “the nonpresent” is particularly important. This ability, he argued, while clearly well developed in humans, may also be seen in other animals, like jays that remember where they stashed nuts and return to the locations with precision much later. Corballis was skeptical of the idea that language appeared suddenly, and noted that trying to separate human language completely from anything animals do “can also lead to overly simple explanations for animal behavior, leading to smug superiority and the invocation of what seem to be miracles.”
Other animals, of course, use gestures, and both of our closest relatives, the chimpanzees and bonobos, have a rich repertoire of motions that seem to be shared at least with each other. A 2018 study14 characterized both species’ gestures, and tried to assign meanings to each distinct movement or set of movements using something called the “ ‘Apparently Satisfactory Outcome’ (ASO), the reaction of the recipient that satisfies the signaler as shown by cessation of gesturing.” In other words, if you wave at someone across the street and they wave back and you then stop waving, it’s an ASO. One is tempted to think of scenarios in which gestures are not followed by signaler satisfaction, per se, with communications among drivers in traffic coming to mind, but the idea is still clear. It turns out that chimpanzees in the wild have at least nineteen ASOs, and their repertoire has about a 90 percent overlap with that of bonobos. The gestures not only look similar but also seem to mean much the same thing, which in principle suggests that the two species could understand each other.
What about people? We often focus on other animals’ abilities to interpret our language, but how about our ability to interpret theirs? The Great Ape Dictionary15 project aims to answer that question by having people view videos of chimps and bonobos performing a gesture, maybe reaching an arm out or doing what the project leaders call a Big Loud Scratch. Led by primatologist Richard Byrne of the University of St Andrews in Scotland, the project enlisted the aid of thousands of people watching twenty videos of chimps and bonobos and choosing which interpretation they think fits the apes’ gestures the best. The data are still being analyzed, but the possibility of a kind of universal primate gesticulation, one that points to (pun intended) a very ancient common evolutionary history of communication, is appealing.
The analogy between ape gestures and human speech only goes so far, however. Human infants gesticulate before they can speak, and baby sign language has become popular among some parents eager to interpret the arm waving of their children long before it is realistic to ask them to “use their words.” In humans, those gestures affect the way a baby’s cognition develops as he or she grows up. But a careful comparison of ape and human infant gestures16 shows that some of the similarities at least are superficial; the nonhuman primates do not point at external objects the same way that babies do, and their gestures do not show the same kind of cultural changes over time. This does not mean that ape gestures are inflexible—they can be adapted to circumstance the way that human gestures can, and they can show intentionality, which means that the gestures are about something, rather than simply expressing emotion. They do not, though, contain all the elements of language.
It’s (Not) Only Human
Even with all of those connections between human speech and animal signals, no one would argue that animal communication is exactly the same as language. The Linguistic Society of America puts it even more starkly on its website:17 “No other natural communication system is like human language.” That depends on what you mean by “like,” since many species communicate in ways that are at least somewhat similar to the way that humans do, even if those communications are not identical to our own speech. Perhaps Sara Shettleworth put it best: “The old question, ‘Can animals learn language?’ has been replaced by appreciation that although human language is just that, human, other species share important components of it.”18
Part of the reason that people have seen such a gulf between human language and animal communication is that we have tended, for both good reasons and bad, to focus on primates. The good reasons include our recent common ancestry, which makes apes and monkeys an obvious choice to look for common structures or neural patterns that could have led to language. The bad reasons are similar, because as is the case with cognition more broadly, as I discussed previously, convergent evolution can mean that distantly related animals independently evolve similar characteristics, and the way those characteristics got there can be illuminating.
Truth be told, primates are a bit disappointing when it comes to language. They aren’t musical like birds or even whales, they don’t seem to have individual names for each other like dolphins, and while they have complex gestures that may even be interpretable by other species, those gestures mean rather ordinary things like “let’s have sex” or “climb on me.” I suppose one could argue that the vast majority of human communication is equally prosaic, but one might have hoped for more from our closest relatives.
Instead, we can learn a great deal about language and how it might have evolved in humans by considering a wide range of other creatures. Animals signal to each other with smells, sounds, and sights, and often do so with an impressive degree of complexity. That said, the link between having a lot of signals and animals’ cognition is not necessarily straightforward; we tend to assume that an animal with more complex communication must be smarter, but that is our anthropocentrism talking. Like humans, animals seem to know more than they can express. For example, a honeybee can remember many different colors, and can even see in the ultraviolet range, but a worker’s waggle dance, a complex form of communication with at least some aspects of language, only includes information about the location and distance of flowers, and not their color. Humans, of course, also have many concepts that go unexpressed (for which we can all be grateful, since presumably the alternative is an endless narrative of what is going on in all of our heads), and animals may also know more than they can say, as it were.
The idea that one can have concepts without the words to express them leads us to an often-considered problem in the evolution of language, namely whether language is necessary for thought to occur at all. Darwin believed that language was necessary for any kind of complex thinking and hence complex behavior, and assumed that animals lacked either one. But he was unaware of many of the impressive accomplishments of animals such as the New Caledonian Crows or even chimpanzees, since few detailed field or laboratory observations of behavior had been made at the time he was writing. Because we humans find that language helps us think, and because it is inextricably tied to our imaginations, it is hard to come up with a definition of thought that doesn’t require it. Peter Godfrey-Smith questions the idea that thought requires language, pointing out that the aforementioned birds do what can only be seen as complex internal processing of events. Language, he says, “is not essential to the organization of ideas, and language is not the medium of complex thought.”19 As Godfrey-Smith emphasizes, just because we do something a particular way doesn’t mean it is the only way. If you always assume things happen the same way in other animals as they happen in people, you end up with a circular and not very interesting definition of whatever you are trying to explain.
None of this is to say that bees can talk, or that dolphins, even though they can master concepts like “press the lever on the right to get food,” are thinking things like “I wonder when that stupid keeper will give me a test that lets me show off my real skills.” But it does mean that, as Corballis suggested,20 “The burden of explaining language evolution is lessened . . . if language is regarded as communication, not thought.”
Vocal Learning: If You Hear Something, Say Something
One of the hallmarks of language, and one that we share with at least some animals, is vocal learning. Infants are born crying, not declaiming Shakespeare, and take years to get from the former to the latter through a process that is a fascinating combination of genetic input and spongelike absorption from the environment. The jargon term for what the babies are doing is vocal production learning, which means the ability to make a sound based on what you hear, or change those sounds based on social feedback. It is found in three groups of birds: songbirds, parrots, and hummingbirds—and five groups of mammals: humans (but not other nonhuman primates), dolphins, bats, elephants, and seals. This somewhat erratic distribution, and the length of evolutionary time since the groups shared a common ancestor, means that vocal production learning arose more than once, buttressing the argument that complex communication is the result of convergent evolution. (A broader kind of vocal learning, auditory comprehension learning, means that the listener can learn what a particular signal conveys, and is the kind of learning found in dogs and other animals that respond to verbal commands.)
In birds, we are probably most familiar with vocal learning in the form of mimicry, with parrots and a few other species like mockingbirds being able to imitate a wide range of sounds by their own and other species, including human speech. The many birds that learn their own species’ songs are perfect examples of the exquisite interplay of innate predilections and learning. The degree to which each of the two is combined differs among birds, even closely related ones; White-crowned Sparrows, for example, need to hear a White-crowned Sparrow song during a critical period while they are in the nest in order to produce the correct song later, while Chipping Sparrows just have to be able to hear themselves sing and do not need a tutor.
One of the most remarkable aspects to songbird learning is that the young bird hears its song months before it produces it on its own, and even then goes through a period of refinement after it starts to sing. I can always tell when the juvenile White-crowned Sparrows have arrived in the spring because they seem to have a slightly screechy quality to the song, like a ten-year-old enthusiastically belting out a school anthem off-key. This learning-plus-genes process also means that in some species, including the white-crowns, so-called regional dialects can develop. A professor of mine, Barbara DeWolfe, was one of the pioneers in birdsong study, and she could walk through the campus at the University of California, Santa Barbara, where I was an undergraduate, and tell us where a given bird had come from just by listening to him sing. “San Francisco,” she would say, or “Oregon,” an ability I still think seems nothing short of magical.
Why vocal learning occurs in some animals and not others is not fully understood. The distinction isn’t binary—what you might call partial vocal learning is seen in a range of species, including, somewhat surprisingly, mice, which can tailor their vocalizations in the lab with training but do not seem to do so in the wild. We also do not know the function of full versus partial vocal learning: What do the Chipping Sparrows have that the white-crowns lack, or vice versa? Biologists have a number of theories, but much remains to be done. Even in species with robust vocal learning, some aspects of response to vocalizations are probably not learned—having to find out the hard way that a particular call means “a hawk is about to descend” wouldn’t be very practical, to say the least.
A somewhat unlikely candidate for future research on vocal learning was proposed in a 2019 paper21 by Sonja Vernes and Gerry Wilkinson that championed bats as a good model, at least for mammals. Most people think of bats, if they think of them at all, as fairly silent and solitary animals, but many species are quite social and will roost in large groups during the day. I once saw a tree in Australia that was festooned with fruit bats, sometimes called flying foxes. The bats were hanging upside down, more or less shoulder to shoulder, or what passes for that in a winged creature, and the sound was unrelenting: squeaks, chirps, and grunts as the bats shuffled for position. I can’t imagine that it was easy to sleep in the cacophony, and I fancied that if they could have spoken, it would have been a nonstop litany of “Move! Your wing is in my face!” “No, you move—I just need to stretch and your head is in the way!”
Vernes and Wilkinson22 do not suggest that bats learn to complain, but they do note that the mammals use individualized calls that let mothers find their babies when the adults return to the roost, and they also change their echolocation calls depending on feedback from other bats in their colony. Pups will match their calls to their mothers’ calls. In the disc-winged bat, pairs of individuals will produce alternating calls that Vernes and Wilkinson say resemble the swimming pool game of Marco Polo so that they can find each other in the roost. Finally, male bats in some species sing to attract females, which makes their vocalizations seem like the bird songs that are learned by the sparrows and others. A key difference is that unlike songbirds, bat fathers do not stay around the nest or roost while the pups are young, so the babies do not have a similar opportunity to hear their fathers sing, which suggests a stronger genetic influence on bat song. Still, bats may provide an intriguing test case for understanding vocal learning in mammals.
The Babbler’s Inadmissible Transgression
If, as the nineteenth-century scholar Friedrich Max Müller put it, “Language is our Rubicon, and no brute will dare to cross it,”23 at least a part of that impassable boundary is syntax. Syntax refers to the way that the different parts of language are combined into meaningful units, and has long stood as an impenetrable barrier between human and animal communication. Although bird and whale songs have ordered components in them, and even my crickets sing with distinctively repeated elements, these are not generally viewed as containing syntax because the units themselves lack independent significance and cannot be recombined to have different meanings.
As with so many other barriers, however, the one between human and animal syntax may also be permeable. Some birds, such as Japanese Tits and Southern Pied Babblers, produce calls that can in fact be recombined. The tits have “alert” calls that warn their neighbors about potential predators as well as “recruitment” calls that tell others to come to a good food source. Unlike many other birds, they can combine the two calls into one that means, in effect, “come over here and help me attack this predator,” which qualifies as syntax at least according to some scientists. The tits can even respond to the calls of a related species, the Willow Tit, in a way that suggests that the birds can assess the meaning of individual components rather than just hearing a generalized “Hey, a little help over here!” message. Other scientists are pickier, and according to a 2019 article,24 “perceive this [animals having syntax] as an inadmissible transgression.”
A novel approach to determining whether birds or other animals might have language-like components in their signals was undertaken by a group of researchers using a method that is generally applied to understanding species diversity. When we speak of the tropics as having a greater biodiversity than, say, the Arctic, we mean two things. First is the number of species: the Arctic has about 200 kinds of breeding birds, for instance, while Brazil has over 1,800. But we also want to take into account the relative abundance of those species, since if each of the 200 in the Artic had a population that was ten times the size of those in Brazil (they do not, but let’s say so for the sake of argument), that would make the Arctic comparably diverse at least in one sense of the word. Estimating those numbers is difficult, and doing so requires the use of sophisticated mathematical models.
The researchers, based in the United Kingdom, the United States, Israel, and Germany, applied those models to language,25 but instead of species number and abundance, they used the units of communication—components of either human speech or animal signals. It turns out that the information content of all human languages follows a mathematical rule, called Zipf’s law, which is in keeping with all languages also being able to express the same kinds of ideas. But it also turns out that the units of communication in songbirds, rock hyraxes (small mammals that are related, improbably, to elephants), and dolphins follow Zipf’s law as well, showing that the patterning of human language is roughly similar to that of the other animals.
A concept that is related to syntax, and is also a characteristic of human language, is called combinatorial signal processing, which means that smaller elements are grouped into larger ones, like syllables into words and words into sentences, in an ordered manner. Again, this is often thought to occur only in humans, but a very similar ability was recently discovered in a treehopper. Treehoppers are insects that look a bit like they have dropped in from outer space; oddly shaped, many of them also sport elaborate headgear, including one species named after Lady Gaga (Kaikaia gaga). The species under question doesn’t have quite such extravagant ornamentation, and instead resembles a rose thorn with legs. It even lacks a common name, going by the unprepossessing Latin title Enchenopa binotata. Like others of their kind, male E. binotata signal to prospective mates by rhythmically vibrating the leaf or stem on which they are perched. The signals are patterned, like tiny drumbeats, and a group of researchers from the University of Wisconsin-Milwaukee,26 led by Bretta Speck in the laboratory of Rafael Rodriguez, who has been studying the tiny creatures for many years, asked whether the female treehoppers would respond to the signals if the beat, as it were, was altered. In other words, if they usually responded to the equivalent of “Hey, come over here! I am a terrific treehopper!” would it work equally well to broadcast, “Here over terrific! Treehopper hey come am a!”?
The answer was no. Female treehoppers were quite discriminating in their responses, and seemed to have internal rules about how the parts of the signal should be combined. The authors do not claim that treehoppers are talking, or that grammar or syntax in human language arose from the same ancestor in both the insects and people. Instead, combinatorial signal processing, they suggest, “may represent a common solution to the problems presented by complex communication in a complex world.”27
It is worth noting that just because syntax has only been documented, even controversially, in a few species does not mean that all other species lack it. We simply haven’t looked at all of the perhaps ten thousand bird species, for example, or even just the more than six thousand kinds of songbirds, with anywhere near the level of scrutiny and careful experimentation required to establish how they communicate. This points to one of the many problems with declaring that only humans possess this, that, or the other characteristic: How can we possibly know for certain? Remember, the scala naturae is false, so we cannot assume that just because our closest relatives lack syntax, or any other signpost of language, all other species do as well.
At the same time, I am not arguing that all creatures are alike, or that they have the same abilities. Birdsong, complex though it is, does not seem to be related to other cognitive abilities in birds. This means that nightingales, for instance, do not tend to be any better at solving problems using tools than birds with less melodious songs. This contrasts with the situation in humans: people who are verbally skilled also tend to be good at other intellectual tasks. Bird songs also do not refer to objects that are away from the singer, meaning that birds do not have the capacity for mental time travel. Bird songs do show one intriguing similarity with human language, which is that at least some species, including Canebrake Wrens, can take turns while speaking, or, in their case, singing, like humans having a conversation. Usually it is the members of a mated pair that engage in a kind of call-and-response duetting, often with remarkable precision that rivals or even exceeds the way in which humans can intertwine their speech during a conversation. Whether one sex or the other interrupts more remains to be seen.
Like They Know the Score
In the previous section I referred to birdsong as melodious, and indeed the idea that it is like human music is quite popular. For some reason, people do not seem to feel that birds having music is nearly the encroachment on human exceptionalism as the idea that other species have language. Yet both music and language are seen as distinctively human, and theories about how musicality evolved abound, ranging from its being a by-product of evolution for dealing with sound in our environment to it being a sexually selected characteristic, like the flashy tail on a peacock. Just as all human cultures have a language, all of them have some form of music.
So how might birdsong be similar to music? One of the obvious commonalities is vocal learning. Birds can learn complicated rules about sounds—when each note should be produced, how long to wait between segments of a song, and so forth. The ability to memorize patterns in sounds differs among species, and not in ways that are easy to rationalize. For example, Budgerigars, which are a kind of small parrot, can perform some sound discriminations that a Zebra Finch cannot, even though Zebra Finches are part of the group of birds that sing their own complex songs.
Why the variation? Perhaps parrots are just different from other birds, as evidenced not only by their adeptness at mimicking human speech but also by their dance moves. Like Snowball the cockatoo, which I mentioned in chapter 7, Snowball’s dance feats spurred a flurry of interest in the idea that being able to move to the rhythms of music also indicated a preadaptation to language. This idea was particularly interesting since, at least in the early research, the species that danced also were good at vocal production learning. That bubble burst, however, when Ronan the sea lion was found to do much the same thing.28 Humans can teach sea lions to do tricks, but they certainly don’t have anything resembling language. So the real question is why some, but not all, of the species that show vocal learning also perceive musical rhythms. This will require more research, but in the meantime, being musical, or at least being able to respond to it, doesn’t mean that an animal is on the way to human language.
Doubt has also been cast on whether birdsong is music after all. Marcelo Araya-Salas from New Mexico State University examined a species29 already thought to be a virtuoso, the Nightingale Wren, a tropical songbird from Mexico and Central America. He analyzed the songs of eighty-one individuals for their adherence to the same rules that appear in human musical scales, and found a very poor match. By human standards, at least, birds don’t make music, at least if you want to be a strict constructionist about it. Perhaps, some scholars suggest, we turn the birdsongs into tunes in much the same way we fool ourselves into finding meaning in images in clouds or in hearing foreign languages and thinking we can understand them.
Talk to Dogs and Listen to the Casual Reply
Finally, whether or not animals have language, we have long been fascinated by the idea of talking with them ourselves, a la Dr. Dolittle, the fictional character who could communicate with every animal on Earth. I loved Hugh Lofting’s books as a child, though even at a young age I found it perplexing that animals could speak exactly as humans, but that no one else besides the good doctor could communicate with them. And what is more, that he never once thought to simply teach other people to do the same thing (I certainly fantasized about being his pupil and doubt I was the only one). Think of the trouble that would be saved at the veterinarian’s office—it would have made my question to the vet about my cat William obsolete. I also was slightly irritated by Gub-Gub the duck, one of the only female characters in the books, being portrayed as fussy and narrow-minded, though I was willing to accept the characterization given that I had never met any ducks myself.
My youthful literal-mindedness aside, whether and how much we can communicate with animals, and how much they understand what we say, has long been of interest for people with pets as well as scientists trying to understand how the brain processes information. I have already mentioned the study30 in which dogs were placed in an fMRI machine to see how they responded to neutral versus positive words. The dogs did indeed show different brain responses, but that does not mean that they actually interpreted the meaning of the words. Instead, the research suggests common brain functions in response to different kinds of sound, something that was probably at the root of the evolution of human language but says nothing about cross-species understanding.
Another study, conducted by members of the same Hungarian team that examined the dogs’ responses to words, used a technology known as electroencephalography. The technology measures electrical activity in the brain via sensors that are attached to the dogs’ heads. They then detected the dogs’ responses to familiar instructions, like “sit,” and to nonsense words. The dogs’ brains responded differently to known words than to different-sounding nonsense words, but didn’t seem to distinguish between known words and nonsense words that sounded similar.31 Whether one of the headlines32 on a story about the work—“Sorry, Folks—Your Dog Doesn’t Really Know What You’re Talking About”—was unnecessarily harsh or not, the results reminded me of an old Gary Larson cartoon in which the first panel is captioned, “What we say to dogs,” with the owner scolding Ginger in detail for getting into the garbage. The second panel, titled, “What they hear” has a balloon above the dog’s head saying, “Blah, blah, blah GINGER blah blah blah.”
Larson had a companion cartoon in which a woman took Fluffy the cat to task for shredding the furniture, but this time the balloon over the cat’s head was completely blank. But a 2020 study might provide, not a way to stop cats’ destructive tendencies, but a glimmer of hope for communicating with them. A group of psychologists in the UK decided to see if the anecdotal stories about cats responding to humans blinking at them would hold up in an experiment.33 Using twenty-one cats from fourteen households, the researchers allowed each cat to see either a stranger or its owner, and then the person was told to “slow blink” at the cat. In both cases, the cats responded by blinking back, and often approached the person who had blinked at them. The researchers concluded that the cats liked the interaction, and suggested that one could use this form of friendly communication with one’s own cats or those encountered on the street, saying, “You can start a sort of conversation.”
This discovery notwithstanding, William remained nonverbal his entire life. Maybe I should have tried blinking at him more often.