This book is about the origin of language, why it is special, and how it got that way. One reason language is special is that it allows human beings to name things and to use those names conversationally. No animal has this ability; only humans do. We are now beginning to understand when and how our ancestors began to talk and what it takes to get an infant to speak their first words.
Some people say language was simply created. End of story! That’s nonsense. Language evolved, just like all other biological and psychological processes. Until recently, however, nobody had any serious idea how it took shape from animal communication. I didn’t either, at least not until I tried to explain the failure of a project in which I attempted to teach a chimpanzee to use language.
The failure of that and similar projects comprise one of three themes in my book. The other two concern (1) an ancestor who likely produced the first words and (2) the non-verbal experiences an infant shares with their parents that are crucial for producing their first words.
Consider the failures of ape language projects. Linguist Noam Chomsky began his distinguished career with a scathing critique of behaviorism in which he claimed language was uniquely human. Many behaviorists reacted to that claim by starting projects in which they attempted to teach apes language. To get around an ape’s articulatory limitations, some projects, including my own, used American Sign Language, a gestural language used by hundreds of thousands of deaf people. The focus of my project was an infant chimpanzee we humorously named Nim Chimpsky. Other projects sought to teach language to chimpanzees and bonobos by training them to produce sequences of arbitrary visual symbols.
All those projects failed. In the case of sign language, my analyses of videotapes in which Nim signed with his teachers revealed that they had inadvertently prompted him to make signs they anticipated he would make. Sequences of signs that seemed spontaneous were, in fact, cued by Nim’s teachers.
Sequences that chimpanzees learned to produce in other projects could be explained by rote memorization, like the sequences people use to enter a password to obtain cash from an ATM. What those sequences have in common is they are motivated by reward. Requests for rewards, however, constitute a minuscule portion of human vocabulary. If such requests were all an infant learned, they would never learn language.
“Ape language” experiments failed because none of the subjects could learn to use symbols as names — the basic function of words. They showed why it is futile to teach an ape to produce sentences if it can’t even learn words. Although the results of Project Nim were negative, they showed why a theory of language’s evolution must begin with words, not sentences.
What ancestor might have produced the first words? Now that we know chimpanzees are unable to learn words, we must ask, which, if any, of our ancestors were the first to use them, why they might have done so, and what they might have said. Recent discoveries by paleoanthropologists suggest it is likely that the first words were invented by Homo erectus.
The critical question to ask about any purported inventor of words is how did they contribute to a species’ survival? If they didn’t, they couldn’t be naturally selected. Most candidates fail that test, e.g., using words to enhance pair-bonding and social bonds. Each of these behaviors develop without words. To survive, Homo erectus needed copious calories to feed its enlarged brain, the volume of which was almost three times the size of a chimpanzee’s.
The most efficient source of calories is meat, but Homo erectus lacked the required weapons to kill large animals. They could use stone tools to butcher animals that had been killed by other predators or that had died a natural death, but they couldn’t kill them outright. They had to use another approach. After one of their group, a “scout”, located a dead animal, he had to recruit colleagues to help butcher it where it lay and scare off other animals that might pick at its remains while they attempted to.
Because the dead animal was far away and out of sight, the scout had to invent arbitrary words to describe it and its location. The innate vocabulary of signals animals used to communicate would not suffice. We don’t know if the scout used gestures or a spoken utterance, or both, to get his colleagues to think about the animal they had to scavenge, but whatever form the gestures or utterance took, some linguists suggest it was the origin of the first words. Our vocabulary grew from there.
Now consider pre-verbal precursors of an infant’s first words. Something remarkable happens to every infant during their first year that distinguishes their history from that of every other primate. They experience two non-verbal relations with their parent that pave the way to language. During the first few months, human infants are cradled by their parents. That provides a basis for their sharing gazes and emotions in a stage of development called intersubjectivity.
Beginning at approximately six months, the infant begins to crawl and to point to external objects, often picking them up to show to their parents. Pointing to an object and sharing it with a parent takes place during a second stage of development called joint attention. When an infant and parent know they are attending to the same object, the infant can readily learn its name by imitating a parent’s comment.
Intersubjectivity and joint-attention, two uniquely human phenomena, are crucial for the development of language. Their absence in chimpanzees is the best explanation for their inability to learn language. Their partial absence in autistic children and in children raised in orphanages also explains why language development in those children is retarded.
As a graduate student at Harvard University, I worked with B. F. Skinner, the preeminent behaviorist of the 20th century. Although my dissertation was about discrimination learning by pigeons, Skinner’s book, Verbal Behavior, inspired my interest in language. I was also aware of Chomsky’s criticism of that book and of other behaviorists who challenged Chomsky by starting ape language projects. As mentioned earlier, I eventually began my own project, only to discover that a chimpanzee’s signing was an artifact of its teachers’ prompts.
It took me many years to discover some positive implications of Project Nim. Knowing that chimpanzees couldn’t name things didn’t help me to understand how language evolved. Eventually, I realized that the root of the problem was Chomsky’s insistence that understanding the evolution of language required us to understand the evolution of grammar.
Although Chomsky devoted most of his career to discovering the nature of “universal grammar,” a grammar that could generate sentences in the more than 6000 languages that people speak, he had nothing to say about the origin of words. Without words, people can’t create “an infinite number of meanings from a finite set of words”, a feature that Chomsky emphasized was the essence of language. By focusing on naming, I believe that for the first time we are on fertile ground for finally unraveling the mystery of how language began, both in our ancestors and in human development.
I recommend a new reader turn to the chapter on the first year of infancy. Most people who spend time with infants (including their parents) think they are just playing and having fun. That’s of course true but virtually no one appreciates the structure and function of that play and how it inexorably leads to language.
Peter Hobson, a British psychoanalyst, has provided the best description I know of regarding the path to language: “Those psychologists who believe that humankind became unique by acquiring language are not altogether wrong. But they are not altogether right, either. Before language, there was something else more basic, in a way more primitive, and with unequalled power in its formative potential that propelled us into language. Something that could evolve in tiny steps, but suddenly gave rise to the thinking processes that revolutionized mental life. (…) That something else was social engagement with each other. The links that can join one person’s mind with the mind of someone else—especially, to begin with, emotional links—are the very links that draw us into thought.”
Those links begin with the practice of cradling and the opportunities cradling provides for social engagement. Among primates, only humans cradle their infants because new-born infants are the least developed. The volume of an infant’s brain is approximately 25 percent of its adult size; in chimpanzees, it’s 45 percent. Similarly, the human skeletal system is poorly developed. As a result, an infant cannot locomote and must be cradled for six months.
An important benefit of cradling is the proximity of the infant’s and parent’s eyes, allowing them to share each other’s affect and gaze, one of many quirks of evolution that laid the groundwork for language. In compensation, as it were, for the infant’s lack of mobility, infant and parent can observe and anticipate each other’s behavior to an amazing degree during cradling. How many times have you heard, and likely said yourself, when watching a baby in their parent’s arms, look at how the baby’s eyes are soaking up knowledge-– or words to that effect? Look how she watches other people. She’s like a sponge.
Developmental psychologists have shown that those interactions are rhythmic. Parent and infant take turns in expressing affect and in engaging in non-verbal auditory behavior (whimpering, gurgling, and so on). An important feature of those interactions is that they are conversational. After a mother smiles, raises her eyebrows, makes a sound, shakes her hands, and so on, so does the infant. Long before an infant utters her first words, she’s engaging in proto-conversations of affect with her mother and others.
While playing with an infant, it is commonplace for a parent to engage the infant’s interest in an object by looking at that object, waiting for the infant to gaze at it, and then look back at them and smile. Such examples of joint attention provide the first instance in which an infant and another person share the contents of their minds, in this example, knowing that each one saw a particular object.
There are two major implications I would like the reader to draw from my book. The first is the importance of non-verbal experiences that an infant shares with their parents. The other is how to overcome the weakness of Chomsky’s theory of the evolution of language.
Now that intersubjectivity and joint attention have been well documented by developmental psychologists, we need to learn more about their antecedents. For example, the renowned anthropologist Sarah Hrdy, has suggested that Homo erectus benefited from cooperation instilled by collective breeding, the practice of sharing the care of infants with relatives. Unlike a chimpanzee mother, who won’t allow anyone to approach a newborn infant for six months, there is evidence that Homo erectus’ infants were raised by “alloparents” in addition to their own mothers. Satisfying alloparents is presumed to strengthen intersubjectivity which, in turn, facilitates cooperation.
Recent advances in technology allow researchers to detect the focus of attention of a parent and their infant over long intervals of time. Such data will, for the first time, allow investigators to measure joint attention precisely in a variety of situations.
Chomsky’s prominence as a linguist is based on his concept of a Universal Grammar that can generate any of the languages that people speak. Those models have transformed linguistics and have contributed significantly to cognitive psychology. Chomsky’s anti-behaviorist stance has served him well in developing models of grammar. The same cannot be said for his treatment of words. Although Chomsky believes that grammar is innate and that it resulted from a mutation, the same cannot be said of words. Words have obvious behavioral origins, origins that are clearly social. The challenge is to determine those origins. If I were starting out as a graduate student and needed a field of inquiry to study, that would be my focus.
Given our current technological sophistication, I anticipate important discoveries about how language not only began but how it has also thrived. Words provide the glue that allows for and preserves learning, intelligence, knowledge, invention, discovery, understanding, wisdom, and love.
Herbert Terrace is a Professor of Psychology and Psychiatry at Columbia University. He has a BA & MA from Cornell University and a PhD from Harvard University. He began teaching at Columbia in 1961 and held visiting positions at the University of Sussex and Oxford University. He was awarded fellowships by the Guggenheim and Fulbright foundations and All Souls College (Oxford University). He is the author of Nim (1979) and Why Chimpanzees Can’t Learn Language and Only Humans Can (2019) and co-editor (with Janet Metcalfe) of The Missing Link in Cognition (2010) and Agency & Joint Attention (2013). Since 1961 his research has been funded by NIMH, NSF, and the James McDonnell foundations.