Review of the book How Language Began: The Story of Humanity’s Greatest Invention:
In grad school, I remember hearing about Daniel Everett as a controversial and somewhat heterodox figure in the world of linguistics, but until now I had never read any of his work.
Everett’s controversial claim is that a lot of the structural and especially syntactic features of human languages that are commonly thought to be universal, and without which language would be unimaginable, are actually not universal and not fundamental at all.
He bases this claim on his observations of the languages spoken in the Brazilian Amazon, especially Pirahã. Everett claims that Pirahã lacks hierarchical and recursive structures, yet is still a complete human language and perfectly adequate for communication. This contradicts the Chomskyan idea that language is at its core an innate, computational cognitive system that manipulates and combines symbols into larger structures. Instead, Everett claims that the essence of language consists in the use of symbols (i.e. words) for communication, and that this naturally leads to a whole host of auxiliary features and behaviors, of which syntactic communication is just one.
The big discontinuity
In this book, Everett tackles a big question: How did language begin? Did it evolve, was it invented, or some combination of the two?
It’s a particularly hard question because our nearest relatives, the great apes, don’t seem to have a communication system that could serve as a “protolanguage” or direct precursor of human language. Chimpanzees, bonobos, and gorillas do communicate with each other with sounds and gestures, but they don’t seem to have any direct equivalent of the words or phonological and syntactic structures found in human languages. Although they are capable of learning to communicate with symbols if trained by researches, non-human primates cannot speak using their vocal tract (larynx, tongue, lips, etc.) the same way we do.
So, how did our hominin ancestors start talking? In the Chomskyan paradigm, language originated in Homo sapiens as a result of biological evolution. The standard argument to support this conclusion is indirect and somewhat subtle, but (simplifying) if our linguistic abilities surpass what we would be able to do merely with general-purpose learning and cognition, they must depend on innate cognitive adaptations; therefore, language could not have been invented, even if the details of how the innate linguistic endowment is implemented in our genomes and brains, and how this evolutionary event occurred, are still highly speculative. And because Homo sapiens show larger brains, changes in the vocal tract anatomy, and more complex behavior than previous hominins, it is thought to have occurred relatively late in human evolution.
Everett argues for a different story. He thinks language was invented, probably almost 2 million years ago, by Homo erectus. According to his theory, erectus were the first to invent symbols. This, in turn, created a context where the brain and body could coevolve with the nascent languages: an example of the Baldwin effect, in which individuals who were better talkers had an advantage given the cultural and behavioral context, and the languages themselves also changed over time to meet the needs and abilities of their users.
It all started with symbols
According to Everett, Homo erectus had already reached a level of behavioral complexity 1-2 million years ago that suggests they probably had highly developed cultures and communication systems. Although most non-stone tools don’t leave any direct evidence, there are indications that erectus could do a lot more than hit two rocks together. They migrated all over the world and colonized islands, which suggests they must have built rafts or canoes, and probably planned their migrations, hunts, and fishing expeditions.
First, these hunter-gatherer communities would have invented symbols: in order to coordinate their collective activities, they would have made specific sounds (probably accompanied by body movements) to refer to objects, actions, and relations. At this point, Everett’s big idea starts to kick in: once you have symbols, fully-fledged human languages would have emerged gradually as a result of the interactions between speakers, their bodies, cultures and communicative needs, the structure of information in their environment and implicit in social interaction. Further refinements made by Homo sapiens would have been the result of a coevolutionary process that presupposes an existing linguistic environment.
In the rest of the book, Everett elaborates how the existence of a symbolic communication system within a cultural context might have led to the emergence of a wide range of linguistic phenomena of interest to linguists. For example, phonetics with vowels, consonants, and syllables might have emerged because this is the most efficient way to transmit information in a way that is easily perceptible by the auditory system. Phonological systems might have converged on “double articulation” of phonemes (small meaningless elements) that combine to form words (larger, meaningful elements) because of the necessity to produce a large set of easily distinguishable symbols. Grammar with hierarchical phrases and sentences might have emerged as a result of demands on working memory and to help distinguish the different roles a symbol can play in an utterance (e.g. topic, comment, subject, object). From the start, speech would have been accompanied by gesture because our interconnected brain tends to recruit the whole body and its neural representation in the service of complex tasks. The first language users would also have resolved ambiguities using shared background knowledge and cultural assumptions, just like we do.
We may never know exactly when and how language originated. Nevertheless, attempts to address this question can help us interrogate our concepts and assumptions about what language is and how it works. Whether or not you think Everett is onto something here, we shouldn’t limit ourselves by dogmatic theoretical beliefs in this area, given how much there remains to be discovered, and, probably, re-conceptualized about language.