Skip to main content

Verified by Psychology Today


Nim Chimpsky and Noam Chomsky

Why language began with words.

Many people feel comforted and understood by animals. In that sense, they communicate with them. What they can't do, however, is talk with an animal. Whereas all animals, ourselves included, communicate, only humans use language to do so. That makes language special. To appreciate how, consider what it means to know language:

  • All humans can name things. No animal can.
  • All humans converse, taking turns as speakers and listeners. No animal can.
  • All humans can create sentences. No animal can.

These features of language are universal. What's less clear is when and how language originated. There are two schools of thought about that. The biblical view, which I won't consider here, states that language was simply created. The more modern view is that language evolved.

How language evolved is a problem that biologists, philosophers, psychologists, and linguists have debated since Darwin. There’s still no agreement about what features of language were selected from animal communication. Animal signals vary but the number a species produces is few, rarely more than a dozen. Animal signals are also unlearned and are not conversational. Until recently, these gaps between animal communication and language appeared to be too large to bridge.

This blog will describe a new approach to the evolution of language. It differs from that of Noam Chomsky, the world’s most prominent linguist, whose focus is on grammar. Instead of grammar, I will focus on words. From an evolutionary point of view, words are as different from any form of animal communication as they are from any grammar. They are therefore the first inevitable step towards language.

Duncan Rawlinson on Wikimedia Commons
Noam Chomsky, 2004
Source: Duncan Rawlinson on Wikimedia Commons

Most linguists, including Chomsky, agree that words evolved before grammar. Chomsky has nevertheless spent most of his career trying to discover a "universal grammar" that can generate any of the more than 6000 languages people speak. That is an admirable goal, but he appears to have minimized the fact that words are necessary to produce what he regarded as the quintessential feature of language: the ability to create an innumerably large number of meanings from a finite number of words. Without words, those meanings can’t be created.

Chomsky recently applied the theory of evolution to language by postulating a mutation that produced universal grammar about 100,000 years ago. Mutations, along with natural selection, are the two basic mechanisms of evolution. There is, however, a glaring problem with Chomsky’s hypothesis. A specific mutation for grammar would be, by orders of magnitude, the largest single mutation that ever occurred.

A more serious problem is that a mutation for grammar cannot account for words. To do so, it would be necessary to postulate more than 6000 mutations, one for each of the languages that people use. That would be quite a stretch. We are still left with the puzzle, how can natural selection account for words?

I explore that gap between animal communication and language in a forthcoming book, Why Chimpanzees Can’t Learn Language and Only Humans Can. The reasons chimpanzees can't learn language and humans can are two sides of the same coin. Chimpanzees can't learn that things have names; humans can.

Ironically, the discovery that chimpanzees can't name things resulted from attempts by various psychologists to challenge Chomsky by showing that chimpanzees, our closest living relatives, could create sentences. To get around the articulatory limitations of the chimpanzee vocal tract, some projects, including my own, attempted to train chimpanzees to learn rudiments of American Sign Language, a natural language used by hundreds of thousands of deaf people. The subject of my project, Nim Chimpsky (shown with me in the inset), was raised and taught by a group of teachers of sign language in New York City until he was almost five. Other projects attempted to teach chimpanzees to learn language by using arbitrary visual symbols.

Courtesy of Susan Kuklin
Nim with Herbert Terrace, 1975
Source: Courtesy of Susan Kuklin

All of these projects failed. In the case of sign language, analyses of videotapes of chimpanzees signing with their teachers revealed that the teachers inadvertently prompted the signs they anticipated the chimpanzee would make. Those prompts occurred a fraction of a second before the chimpanzee signed. My research showed that signs that seemed spontaneous were, in fact, cued by Nim's teachers. I explain this in an earlier book, Nim: A Chimpanzee Who Learned Sign Language.

In the case of visual symbols, sequences that the chimpanzees produced could be explained as rote memorization, like sequences used to enter a password when obtaining cash from an ATM. What those sequences have in common is that they were motivated by reward. They are clearly not conversational. Requests for rewards constitute a minuscule portion of human vocabulary. If such requests were all that an infant learned, she would never learn language. The moral of the “ape language” experiments is that it is futile to teach a chimpanzee to produce sentences if it can’t even learn to use words.

My book also describes some intriguing clues about the origin of words that resulted from recent discoveries by paleoanthropologists and by developmental psychologists. From their study of fossils, paleoanthropologists have suggested that a recent ancestor, Homo erectus, produced the first words. Research on infant development has identified non-verbal emotional and cognitive communication between an infant and her parents that are necessary precursors of her first words. As I will describe in a forthcoming post, such relations are uniquely human.

Language has been an embarrassment for the theory of evolution for more than 150 years. Now that we have simplified that problem by identifying the skills needed for learning language, we can begin to see those skills evolved from animal communication. I will pursue that topic in my next post, which will discuss the difference between animal signals and words.


Berwick, R. C., & Chomsky, N. (2016). Why Only us. Cambridge, MA: MIT Press. ISBN-13:026533492

Terrace, H. S. (2019). Why Chimpanzees Can’t Talk & Only Humans Can. Columbia University Press. New York, NY. ISBN-13: 9780231171106

Terrace, H.S. (1987). Nim: The Chimpanzee Who Learned Sign Language. Columbia University Press. New York, NY. ISBN-13:978-06713420413

More from Herbert S. Terrace Ph.D.
More from Psychology Today