Is Language an Instinct?—Response to Vyvyan Evans (part 1)

Nov 27, 2015 by


About a year ago, Aeon published an article by Vyvyan Evans, titled “Real talk”, which was recently brought to my attention by Polish linguist Piotr Gąsiorowski. Evans’s article in Aeon summarizes some of the key ideas in his book, The Language Myth: Why Language Is Not an Instinct. Upon its publication, I was planning to lay out a detailed response, also in a book format; however, that project has not been completed, chiefly for reasons of time and other obligations, so I would like to publish sections from the manuscript in the form of a mini-series of posts on this blog. As is true of most of my posts here, it is aimed primarily at a broader audience with little or no background in linguistics.

The gist of Evans’ article (and book) is summarized nicely in the introductory sentences: “For decades, the idea of a language instinct has dominated linguistics. It is simple, powerful and completely wrong.” But not so fast. The theory that Evans sets out to criticize does not focus on the “language as instinct” idea as such. That phrase was coined by Steven Pinker as the title of a book first published in 1994 that has since become a classic. But the theory that Evans criticizes, which I shall call “Universal Grammar theory” (or UG theory, for short), took shape decades earlier and, to the best of my knowledge, neither Chomsky nor other linguists working within that theoretical framework ever used the phrase “language instinct” as a serious technical term.

Pinker’s book was addressed to a broader audience, and—like any good writer—he took into account that audience’s (presumed) common knowledge. The assumption against which Pinker focused his argument is that language is a cultural invention, like metalworking or the wheel. This idea that language is something that people “figured out how to do” may seem ridiculous to a professional linguist, but it is still far more commonly believed that Evans seems to realize, even more than 20 years after the publication of Pinker’s book. I find that this notion is deeply rooted among my adult education students all the time. One of the reasons why this assumption is so entrenched, I think, is because many non-linguists often equate language with writing. Not too long ago I had to explain to my students that (Biblical) Hebrew and Yiddish are two completely different languages that have less in common than English and Russian (i.e. many words from the former in the latter, but little grammatical similarity). I later realized that the confusion had to do with the fact that Hebrew and Yiddish look similar as they are written with the same characters. Another question I often get, which also comes from the same conflation of language and writing, is why writing post-dates spoken language by so long (roughly 95% of the time, spoken language existed without any writing). The answer, of course, has to do with the “cultural invention” nature of writing—but not of language—which required a certain level of civilizational development.

Taking into account that assumption, widely-held by non-specialists, Pinker’s essential point is clarified: language is not a human invention in the sense of the wheel or even writing, but a biological fact about our species. In this laymen’s sense of “instinct”, what Chomsky, Pinker, and other proponents of UG theory say about language fits quite well: language is thought of as an inherent human inclination, much like spiders’ weaving of webs. This is also roughly how Evans defines an instinct in the Aeon article: “an inborn disposition towards certain kinds of adaptive behaviour”. In what sense language (or language use) can be viewed as an adaptive behavior is an interesting question, which I will address in more detail in a later post.

But if we are to consider a stricter technical sense of “instinct”, as it is understood by biologists, there is a clincher: as Evans points out, instinctive “behaviour has to emerge without training”. He writes: “A fledging spider doesn’t need to see a master at work in order to ‘get’ web-spinning: spiders just do spin webs when they are ready, no instruction required.” Similarly, baby kangaroos climb into their mothers’ pouches upon being born. Sea turtles, newly hatched on a beach, move toward the ocean. Honeybees communicate the direction of a food source by “dancing” a certain way. Nobody teaches baby kangaroos, sea turtles, or honeybees to do these things and yet they do them. Thus, calling language an “instinct” seemingly suggests that language is a purely biological phenomenon—all nature and no nurture!

Critics like Evans easily show that this view is incorrect, stressing the socio-cultural component of language. Stories of so-called feral children, either lost by accident or purposefully deprived of human contact by cruel parents, show that without exposure to human language within a certain time-window known as the “critical period”, children never develop their language ability. Evans describes

“the appalling story of Genie, a girl in the US whose father kept her in a locked room until she was discovered in 1970, at the age of 13. The general lesson from these unfortunate individuals is that, without exposure to a normal human milieu, a child just won’t pick up a language at all.”

How well such feral children can pick up language after being brought into normal communicative environment (or even with specialized training) depends on the age at which they are “discovered”. In one well-documented case, a girl, known in the literature as “Isabelle”, had been hidden in the attic by a deranged mother and never spoken to, until she was discovered at age 6. At that time, Isabelle had no speech and her cognitive development was estimated as that of a 2-year-old child. But within a year she was able to catch up with other 7-year-olds, even in her linguistic development. A much sadder case is that of “Chelsea”, a partially-deaf woman incorrectly diagnosed as suffering from a mental disorder, who was “discovered” at age 31 and fitted with hearing aids. Despite concerted efforts to teach her to speak, all Chelsea ever managed was to learn a large vocabulary; her “grammar”, the ways she put these words together, remained virtually non-existent. Her “sentences” were strings of words like “Breakfast eating girl” or “Banana the eat”—resembling more the utterances produced by “Nim Chimpsky”, a chimpanzee supposedly taught to speak American Sign Language (ASL), than those of a typical two-year-old. Exposure to human language at the right time is thus vital to the development of language. “Language is not something that emerges automatically, and effortlessly”, Evans writes in The Language Myth (p. 3). As a result, how can it be reasonably regarded an instinct?

It should be noted, however, that many behaviors that certain species are biologically predisposed for require some exposure to either environmental or social factors. For example, the communicative “dance” of bees is adjusted to the latitude at which they live: the latitude determines the position of the sun above the horizon, one of the elements that define the “dance”. More similar to the “learning” required for human language to be fully realized is the way that birds “learn” their species’ song. Studies conducted with singing bird species, such as finches and warblers, show that when newly-laid eggs are incubated in soundproof chambers and the hatchlings are hand-reared in individual and acoustic isolation, they can sing only a very simple tune, representing “the inborn component of the song”. The more intricate “embroidery” of the song has to be learned, or at least acquired via exposure to the singing of other male finches. According to an article in British Library, “there are also species [such as the Indian hill mynah] whose song (and indeed whose entire vocabulary) is handed down culturally, generation by generation”. Should such behavior be called “instinctual” or should there be a different term for it? Either way, the claim of UG theory is that human language fits into the same category as bird singing: some “software” needed for its development is innate but exposure to other members of the species is necessary for its full realization.

There is another obvious yet significant difference between human language and other instinctual animal behavior: within the same species, instinctual behavior is essentially uniform. All spiders in the same species weave webs the same way. All bees in the same species encode the information about the direction of food source using the same “dance moves” (but different species do different “dances”). All birds in the same species produce their species-specific song. But humans obviously do not all speak the same way. Babies pick up the language spoken around them: as Evans writes, “a child born in Tokyo learns to speak Japanese while one born in London picks up English”. Children adopted across linguistic barriers speak the language of their adoptive family rather than their biological parents: a child of, say, Vietnamese parents adopted by English-speaking Americans at a young age will learn to speak English just as well as her American-born peers. In fact, if she ever wants to learn her “biological mother tongue”, especially after the critical period is over, she will probably have as much trouble as any non‑Vietnamese person. Simply put, we are not born to speak any particular language.

But generative linguists, advocating UG theory, have never claimed that we are. The expression “language instinct” refers to a predisposition for human language in general, not for any specific language. It might be more useful to think of individual languages as variations on a common theme, with Universal Grammar as that motif, much like different recipes for chicken soup are variations on a common theme. Sure enough, Scottish cock-a-leekie, Thai tom kha gai, and Eastern European matzo ball soup have little enough in common in terms of ingredients, flavor, texture, and color—just as the Scottish Gaelic, Thai, and Yiddish languages each have a sound and a feel of its own. But if we examine chicken soup recipes carefully enough, we will find common ingredients: chicken and water are certain to show up in every recipe, as do other types of ingredients, such as aromatics (whether leek, galangal, or garlic), vegetables of some sort, and perhaps some doughy element, like noodles or matzo balls. The procedures for making the soup will also be different but the same types of techniques (e.g. boiling the soup) will be used in different recipes. We could even write a Universal Chicken Soup recipe, one that will contain not all the details of how any particular chicken soup may be made, but a lot of blanks to be filled in for specific recipes. Universal Grammar, similarly, is not a detailed how-to manual for language, but rather a fill-in-the-blanks “recipe” of how human language works. In other words, Universal Grammar is not a denial of linguistic diversity, but a common denominator of the diverse expression of human language—the similarities that are often obscured by the more conspicuous differences across languages.

In the following post, I will address the issue of how linguistic diversity is addressed within UG theory, as well as Evans’ claim that “absolute grammatical ‘universals’ common to every one of them […] is not what we have discovered”.


Related Posts

Subscribe For Updates

We would love to have you back on Languages Of The World in the future. If you would like to receive updates of our newest posts, feel free to do so using any of your favorite methods below: