Is Language an Instinct?

And other myths.

Posted Dec 19, 2014

 Vyvyan Evans
Source: Source: Vyvyan Evans

In my recent book, The Language Myth, I investigate one of the dominant themes that has preoccupied the study of language for the last 50 years or so: whether the rudiments of the human capacity for grammar—central to language—are innate. This idea originated with the research of the American linguist and philosopher, Noam Chomsky, beginning in the 1950s, and gathering momentum from the 1960s onwards. The idea, in essence, is that human infants are born equipped with a species-specific Universal Grammar—a genetic pre-specification for grammatical knowledge, that ‘switches on’ at an early point in the process of acquiring their mother tongue; and this being the case, it takes much of the pain out of language learning. From this perspective, human infants acquire language because they come with a hard-wired knowledge of aspects of grammar—although there is no meaningful consensus on what these aspects might amount to, even after over 40 years of looking. This enables a child, so the party-line claims, to ‘pick up’ their native language. I presented a very partial, thumb-nail sketch of just some of the relevant issues in a short popular science essay, published in Aeon magazine, here. And, I have discussed the issues further in a full length radio interview, available to be listened to here.

The Language Myth
The Language Myth (Cover design by Zoe Naylor; reproduced with permission)

In a series of recent posts, summarised here, a number of distinguished linguists, who broadly adhere to Chomsky’s proposition that there is an innate Universal Grammar, suggest that I have either misrepresented the claim(s) associated with the research programme surrounding this hypothesis, and/or misunderstood it; and, in three specific cases that they draw attention to, that I have supported my arguments using findings which they claim to have been refuted—they appear, at least in one case, when discussing what is known in the jargon as Specific Language Impairment, to be referring to the short Aeon essay, rather than the fuller discussion in the book.

The Language Myth is written for a general audience—not specifically professional linguists—and takes the form of an evidence-based rebuttal of aspects of the world-view developed in the popular, best-selling books written by Professor Steven Pinker of Harvard University. Indeed, Pinker’s first popular book, The Language Instinct, published back in 1994, provides my book with its title, albeit, with a twist: The Language Myth plays on Pinker’s book title, which I cast as the eponymous ‘language myth’. Indeed, claiming language to be an instinct is self-evidently a myth, as first pointed out by psychologist Michael Tomasello in 1995—see his book review here.

But importantly, The Language Myth directly takes on what I see as the larger theoretical and ideological world-view of what I have elsewhere dubbed ‘rationalist’ language science. While my target is the presentation in Pinker’s various books, it necessarily encompasses more than just the research programme initiated by Chomsky and his co-workers.

It also addresses fundamental issues and questions in cognitive science more generally, and the range of Anglo-American linguists, psychologists and philosophers of the second half of the twentieth century who helped shape it. For instance, I consider the nature of concepts, our ‘building-blocks’ of thought—and whether these might be innate, in some meaningful sense—the relationship between language and the communication systems of other species; whether language, and the mind more generally, might consist of distinct, and enshrined neurological systems—sometimes referred to as ‘modules’—which evolved independently of one another, for a specific mental function; whether the human mind has its own innate mental operating system—sometimes referred to as ‘Mentalese’, or our Language of Thought; and whether language can, in some shape or form, influence habitual patterns of thought—sometimes referred to as the Principle of Linguistic Relativity, famously proposed by Benjamin Lee Whorf (and not to be confused with the straw-man argument for linguistic determinism—the idea that thought is not possible without language; thought clearly is possible without language, as we know from research on pre-linguistic infants, adults who have suffered language loss—known as ‘aphasia’—as well as studies on other species, who have often sophisticated conceptual capacities, in the absence of language; Whorf explicitly argued against linguistic determinism).

The rationalist world-view boils down to the claim that the linguistic and cognitive capacities of humans must ultimately, and at least in outline, be biologically pre-programmed: that there’s no other way, ultimately, to account for what appears to be unique to our species. In The Language Myth, I argue that there are six component ‘sub-myths’ that make up, and mutually inform and sustain this particular stance. I dub them ‘myths’, because they were proposed, in most cases, before any real evidence for or against was available. And since evidence has become available, most objective commentators would be hard-pressed to say that any of these ‘myths’ have much in the way of clear-cut evidence to support them—I take a slightly stronger position, of course; my assessment is that there is almost no credible evidence. So, here are the six:

Myth #1: Human language is unrelated to animal communication systems.
The myth maintains that language is the preserve of humans, and humans alone; it cannot be compared to anything found amongst non-humans, and is unrelated to any non-human communicative capability. And the myth reinforces a view that there is an immense divide that separates human language from the communicative systems of other species. And more generally, it separates humans from all other species. But recent findings on the way other species communicate, from apes to whales, from vervets to starlings, increasingly suggest that such a view may overstate the divide that separates human language and non-human communicative systems. Indeed, many of the characteristics exhibited by human language are found, to varying degrees, across a broad spectrum of animal communication systems. In point of fact, we can learn more about human language, and what makes it special, by seeking to understand how it relates to and is derived from the communication systems of other species. This suggests that although human language is qualitatively different, it is related to other non-human communication systems.

Myth #2: There are absolute language universals.
Rationalist linguistics proposes that human babies enter the world pre-equipped to learn language. Language emerges effortlessly and automatically. And this is because we are all born with a Universal Grammar: a pre-specification for certain aspects of grammar; whatever the ultimate form of these putative ‘universals’ might be—a universal being a feature of grammar that is, at least in principle, capable of being shared by all languages. Moreover, as all languages are assumed to derive from this Universal Grammar, the study of a single language can reveal its design—an explicit claim made by Chomsky in his published writing. In other words, despite having different sound systems and vocabularies, all languages are basically like English. Hence, a theoretical linguist, aiming to study this innate Universal Grammar, doesn’t, in fact, need to learn or study any of the exotic languages out there—we need only focus on English, which contains the answers to how all other languages work. But like the myth that language is unrelated to animal forms of communication, the myth of language universals is contradicted by the evidence. I argue, in the book, that language emerges and diversifies in and during specific instances of language use.

Myth #3: Language is innate.
No one disputes that human children come into the world biologically prepared for language—from speech production apparatus, to information processing capacity, to memory storage, we are neurobiologically equipped to acquire spoken or signed language in a way no other species is. But the issue under the microscope is this: the rationalist linguistics world-view proposes that a special kind of knowledge—grammatical knowledge—must be present at birth. Linguistic knowledge—a Universal Grammar that all humans are born with—is hard-wired into the micro-circuitry of the human brain. The view that language is innate is, in a number of respects, highly attractive; at a stroke, it solves the problem of trying to account for how children acquire language without receiving negative feedback, from their parents and caregivers, when they make mistakes—it has been widely reported that parents, for the most part, don’t systematically correct errors children make as they acquire language. And children can and do acquire their mother tongue without correction of any sort. Moreover, children have acquired spoken language before they begin formal schooling: children are not taught spoken language, they just acquire it, seemingly automatically. But, such a strong view arguably eliminates the need for much in the way of learning—apart from the relatively trivial task of learning the words of whatever language it is we end up speaking. The fundamentals of grammar, common to all languages, are, at least in some pre-specified form, present in our brains prior to birth, so the language myth contends. But as I argue in the book, a large body of evidence now shows these specific assumptions to be incorrect.

Myth #4: Language a distinct module of the mind.
In western thought there has been a venerable tradition in which the mind has been conceived in terms of distinct faculties. With the advent of cognitive science in the 1950s, the digital computer became the analogy of choice for the human mind. While the idea that the mind is a computer has been a central and highly influential heuristic in cognitive science, the radical proposal, that the mind, like the computer, is also modular, was made by philosopher of mind Jerry Fodor. In a now classic book, Modularity of Mind, published in 1983 whose reverberations are felt to this day, Fodor proposed that language is the paradigmatic example of a mental module. And this view, from the perspective of rationalist linguistics makes perfect sense. According to Fodor, a mental module is realised in dedicated neural architecture. It copes with a specific and restricted type of information, and is impervious to the workings of other modules. As a consequence, a module can be selectively impaired, resulting in the breakdown in the behaviour associated with the module. And as a module deals with a specific type of information, the module will emerge at the particular point during the life cycle when it is needed. Hence, a mental module, in developmental terms, follows a characteristic schedule. The notion that the mind is modular might, on the face of it, make intuitive sense. In our everyday lives we associate component parts of artefacts with specific functions. The principle of modularity of design is both a practical and sensible approach to the manufacture not just of computers but many, many aspects of everyday commodities, from cars to children’s toys. However, the evidence, as I argue in the book, provides very little grounds for thinking that language is a module of mind, or indeed that the mind is modular.

Myth #5: There is a universal Mentalese.
The language myth contends that meaning in natural languages, such as English, Japanese or whatever, derives, ultimately, from a universal language of thought: Mentalese. Mentalese is the mind’s internal or private language, and makes thought possible. It is universal in the sense that all humans are born with it. It is language-like, consisting of symbols, which can be combined by rules of mental syntax. Without Mentalese we could not learn the meanings of words in any given language—spoken or signed. But as I show in the book, Mentalese assumes a view of mind that is wrong-headed: it assumes that human minds are computer-like. It also suffers from a number of other difficulties, which make this supposition deeply problematic.

Myth #6: Language does not influence (habitual patterns of) thought.
While everyone accepts that language affects thought in the sense that we use language to argue, persuade, convince, seduce and so on, according to the myth, thought is, in principle, independent. The myth contends that the Principle of Linguistic Relativity—that systematic patterns in grammatical and semantic representations across languages influences corresponding differences in patterns of thought across communities—is utterly wrong. As I show in the book, not only does Pinker, and other rationalists mischaracterise the thesis of linguistic relativity—that the language we speak influences how we habitually think, categorise and perceive the world—he is also wrong in another way. Despite Pinker’s assertion to the contrary, there is now a significant amount of scientific evidence suggesting that, in point of fact, the linguistic patterning of our native tongue has indelible and habitual consequences for how we perceive the world. Of course, the question then arises as to how significant, in terms of influencing individual and cultural world-views, one takes this evidence to be. In a recent book, The Language Hoax, its author, John McWhorter, plays down the significance of the relativistic effects of different languages on the minds of distinct communities of language users. While I disagree with McWhorter’s position—and his review of the relevant evidence is at best partial—given the sophisticated methodologies that now exist for directly and indirectly investigating brain function during routine cognitive and perceptual processing, any objective commentator would be hard-pressed to deny the relativistic influence of language and non-linguistic aspects of mental function.

Ultimately, whether or not one accepts the general argument I make, in The Language Myth, boils down to one’s ideological as well as one’s theoretical commitments. Academic research, like any other human endeavour, inhabits a socio-cultural niche. And ideas arise from assumptions, and principles, sometimes explicitly rehearsed, sometimes not, cocooned within the institutional milieu that helps give them life, and sustain them. In terms of the specifically Chomskyan element(s) of the rationalist world-view that I argue against, my view is that perhaps most damaging of all, has been the insistence that the study of language can be separated into two distinct realms: ‘competence’—our internal, and mental knowledge of language—and ‘performance’—the way in which we use language. Chomsky’s position is that performance arises from competence—given his assumption that fundamental aspects of competence—our Universal Grammar—is, in some sense, present at birth. Hence, competence, rather then performance constitutes the proper object of study for language science. But I, and a great many other linguists, believe that the evidence now very clearly shows this perspective to be wrong-headed: our knowledge of language, so-called ‘competence’, in fact arises from use, from ‘performance’. And Chomsky’s logical error, as I characterise is, has held the field of (Anglo-American) linguistics back for too long.

My rationale for writing The Language Myth, and debunking the world-view presented in Pinker’s popular writing was the following. Pinker’s popular presentations of rationalist cognitive science, at least amongst undergraduate and beginning graduate students, and the informed lay audience, is arguably better known than the work of Chomsky, Fodor and the other leading lights of rationalist cognitive science. And his characterisation—whether one likes, or not, the analogy of language as an ‘instinct’, that Pinker coined—of language and the mind as, ultimately, biological constructions, is widely believed. Many of the standard textbooks, used in the stellar universities across the English-speaking world, promote Pinker’s works as essential readings. Moreover, they portray the sorts of arguments he promotes as established fact. Things are really not that clear-cut. At the very least, the (popularisation of the) rationalist world-view is on very shaky ground indeed. I, of course, didn’t write The Language Myth for committed rationalists; I don't pretend to be able to convince them--it appears, to me at least, that in the case of many such colleagues, their commitment is ideological, rather than being based on an objective and critical evaluation and appreciation of the voluminous evidence. And of course, while they may accuse me of being partial and/or prone to misunderstanding in my presentation, as I show in The Language Myth, the same accusation must then be applied to Pinker, but with several greater degrees of magnitude!

In my next few posts, I’ll be examining some of the evidence, for and against, each of the component myths that make up the rationalist world-view. And in so doing, I’ll also address some of the criticisms raised by Chomksyan colleagues who have objected to my portrayal of things. Whatever one thinks on these issues, these are fascinating times in the study of language and the mind, and an exciting time to be an academic linguist. And my advice to all objective and curious-minded people is to read The Language Myth, and make your own mind up. Some representative and high-profile reviews of the book are below, to give you a flavour of what’s in store.

Book review in The New Scientist 18 Oct 2014
Book review in the Times Higher Education 13 Nov 2014