Is Grammar Innate? by Mark Lowe
This article aims to solve an EFL puzzle through philosophical analysis. The puzzle is Chomsky’s theory of innate grammar: is it true or not? The answer to this puzzle has important consequences for language teaching it tells us whether we should teach language cognitively through understanding, or whether we should teach it intuitively through helping students to pick up language as young children do. To help solve this puzzle, I draw on the ideas of four philosophers: Ayer, Popper, Wittgenstein and Searle.
Let us start with a recap of Chomsky’s theory:
‘The fact that all normal children can readily acquire the language of the community in which they grow up, without special instructors and on the basis of very imperfect and degenerate stimuli, and further that children can learn certain sorts of language such as are exemplified by natural human languages, but cannot learn other sorts of logically possible languages, provides overwhelming evidence that each normal child contains in some unknown way in his or her brain an innate language acquisition device (LAD), and this LAD consists at least in part of a set of deep unconscious grammar rules’. (Searle: The Rediscovery of the Mind, Chapter 10)
This theory has very significant implications for language teachers. If grammar is innate, it develops naturally, like the limbs on a body or the petals on a flower. There is therefore no need to teach grammar. Our task as teachers is simply to provide the right conditions and the right diet, and the grammar will grow of its own accord, like teeth. The innate grammar theory provides theoretical support for teaching methods based on acquisition rather than learning. It provides a rationale for the so-called Natural Approach and for other humanistic methods. It has long been widely accepted in our profession as true: it has long been part of the received wisdom of language teaching and applied linguistics.
But in the last few years, many people have begun to have reservations about the theory. Some reservations are on practical grounds (eg the methods derived from the theory do not seem to work very well, and many excellent teachers find other methods more effective). Other reservations are on theoretical grounds (eg the theory is confused – the evidence does not seem to hold water, etc). Bad jokes have gone the rounds. ‘You can’t find Chomsky’s books in the library? Try the fiction section.’ As a teacher and director of studies, I want to be sure that the methods I use, and the methods I encourage my teachers to use, are coherent and based on sound theory. I look to philosophy to clarify the issues and to distinguish valid from invalid reasoning. I look to philosophy to help determine whether grammar is innate or not.
Philosophers have raised several objections to innate mental phenomena. The debate is as old as Plato and Aristotle. Locke and Hume argued strongly against: Kant argued interestingly in favour, with his mental categories. In our time, three powerful objections have been expressed. The first stems from positivist philosophy: what is true must be verifiable, or it must in principle be falsifiable (in Popper’s variant of positivism). The second derives from Wittgenstein’s language philosophy: what is true must be free of distortions caused by language muddles, and consistent with our understanding of the nature of language and the mind. The third comes from the philosophy of John Searle: a theory of language must not only be free of linguistic distortions, but must also be consistent with what we know about the working of the brain. Let us consider these objections in turn.
In Language, Truth and Logic, A.J.Ayer distilled the Vienna Circle’s positivist doctrine on scientific method: a scientific theory is valid if and only if it can be verified by empirical evidence. In his The Logic of Scientific Discovery, Karl Popper proposed a generally accepted variant of this doctrine: a hypothesis is valid if and only if it can in principle be falsified by empirical evidence. On both counts the innate grammar theory is invalid, because there is no evidence that can either verify or falsify it – because there is no test that could prove the existence of non-existence of innate grammar rules in the brain. Since the theory does not follow the criteria for a scientific hypothesis, it is therefore not a valid theory. Positivists maintain that it is a metaphysical chimera.
In addition, and as John Searle (among others) argues, there is abundant evidence to support alternative theories of language development. For instance, in his classic study of child language development Learning How to Mean, Halliday does not need to posit innate grammar to explain what happens when a child develops language. His young son Nigel starts with one-word utterances- (Mum, cock-a-doodle-do, jam), moves on to two-word utterances (more jam,go walk) and then produces three-word utterances (Let’s go walk, and I want Bartok – Nigel’s enchantingly Hallidayan way of demanding to hear music). Halliday explains the stages of Nigel’s language development by the functional need to communicate, by imitation of and extrapolation from models he hears, by his parents’ encouragement, and by what is known of the workings of the human brain. The innate grammar hypothesis is not cited: it would be redundant. In his Introduction to Functional Grammar, Halliday offers a detailed theory of language based on the same principles. Again, no innate grammar is required to explain language or language development in this study. Language develops not in accordance with alleged abstract rules of grammar hard-wired into the brain, but through pragmatic responses to practical needs. Language development follows the same principles as evolution: no separate edifice of theory is needed to explain it.
The innate grammar hypothesis also begs the question. In other words, it rephrases a mystery (how do children acquire languages?) as a solution (children learn languages by means of a black box called the LAD). We do not know how the black box works: we have no understanding of what goes on inside it. Genuine scientific theories explain the unknown in terms of the known. Either what is large is explained in terms of what is small (for instance, matter in terms of atoms and molecules, or brain activity in terms of synapses and connections), or a mystery is explained in terms of a generally accepted theory, such as gravity or evolution. The innate grammar hypothesis does neither. The ‘solution’ is as mysterious as the puzzle it seeks to explain, and is therefore not a genuine solution.
Let us summarise so far. Positivist philosophers hold that the innate grammar hypothesis cannot be proved by evidence, that the available evidence can better be explained by alternative theories, and that the theory begs the question. It is therefore unproven.
Let us now examine Wittgenstein’s analysis of the issues involved. There are two central points. The first is technical, concerned with following a rule. The second is general, related to Wittgenstein’s view of the nature of language. We consider these points in turn.
Wittgenstein thought long and hard about the central claim that is made by the innate grammar hypothesis: namely, that we follow unconscious mental rules, including the rules of grammar. The conclusions that follow are based on the sections of Philosophical Investigations devoted to following a rule (paragraphs 185 – 242) and from Hacker and Baker’s Commentary on the Philosophical Investigations (vol. 2).
The conclusions may be summarized by saying that Wittgenstein’s analysis of what it means to follow a rule reveals deep confusions in the theory of innate grammar. That theory posits hidden rules that are innate and ‘known’ implicitly, even though nobody knows them explicitly. These ‘rules’ are supposed to be encoded in the brain, so people follow rules of which they have never heard and which they could not understand if they knew them.
Here are three key problems identified by Wittgenstein in the theory of rule-following:
1. The theory is not conclusive. Alternative theories fit the facts equally well, if not better. There is thus no convincing proof that makes the innate hypothesis preferable to other hypotheses. (Halliday’s theory is one example)
2. The theory is confused and it leaves key questions unanswered. Are these rules merely part of the description of how language works? Or are they part of the software of the biological computer in our brains? How exactly do we have knowledge of these rules, which though innate and tacit, are not accessible to conscious understanding? Where do these rules come from, anyway? Did God put them there as a sign of special favour to human beings, as it were? (as some theologians have maintained) What is the ontological status of these rules? And so on. The theory is entangled in a net of unanswered and unanswerable puzzles and confusions.
3. Expressibility. How are these rules ‘realised’ in the brain? If a child’s ability to master a language presupposes the child tacitly ‘knowing’ a complex array of grammatical rules, how are these rules represented to the child’s mind? Are they written in DNA letters? Can there be rules at all, which have never been and can never be expressed?
‘These questions are merely conceptual confusions dressed up as theoretical problems’, write Hacker and Baker. They derive from muddles about the status of ‘rules’. A proper understanding of rules and rule-following, Wittgenstein argues, depends on the following principles:
If Wittgenstein’s analysis of following a rule is valid (and few would question it today), then Chomsky’s theory is invalid. Wittgenstein’s analysis leads to the conclusion that the innate grammar hypothesis is nothing but a linguistic muddle masquerading as a theory
Let us now consider Wittgenstein’s general theory of language and whether it is consistent with the innate grammar hypothesis. The view of language presented in his Philosophical Investigations (and elsewhere in the many posthumously published collections of Wittgenstein’s notes) is permeated with the idea that language evolved in order to meet pragmatic human needs, and was not generated by abstract forms in the brain. This view is argued all the more persuasively because Wittgenstein had earlier espoused the opposite theory. In his early Tractatus, language was thought to reflect the logical form of the world: in the later Philosophical Investigations, language was thought to reflect human needs. There are suggestive parallels between Chomsky’s innate grammar theory and the Tractatus theory of language: both posit idealized abstract forms that in some sense generate language. The reasoning which caused Wittgenstein to repudiate the Tractatus view also leads to the repudiation of the innate grammar hypothesis. Language simply does not work that way. ‘The meaning of a word is its use’, to quote Wittgenstein’s well-known aphorism. Language reflects practical use, not abstract form. So Wittgenstein’s mature philosophy of language is not consistent with the innate grammar hypothesis, and Wittgenstein’s philosophy in general finds much that is unsatisfactory and confused in the theory.
In the next section we see how the American philosopher John Searle has developed Wittgenstein’s ideas into a comprehensive and coherent theory of language and mind, and how Chomsky’s ideas look in the light of this theory.
John Searle, professor of philosophy at Berkeley, University of California, has made many important contributions to our understanding of language and the mind. His first book, Speech Acts (1969), provided a framework for the profound and brilliant, but often rather unsystematic insights of his Oxford philosophy teacher, J.L.Austin, and of Wittgenstein. Intentionality (1983) explored the relationship between language and the outside world it seeks to understand and to mediate. The Construction of Social Reality (1991) relates language to man’s social life, and especially to the many organizations and institutions man has created, such as governments, marriage, financial systems, law, sport and games – and how language makes such activities and systems possible. The Rediscovery of the Mind (1994) both summarises his previous work and adds insights from neurophysiology into the workings of the brain. These themes are further developed in recent books such as Consciousness and Language (2002) and Mind (2004). Searle writes with matchless vigour and clarity, and his books are a joy to read. His contributions to our understanding of language and the mind are unsurpassed among thinkers active today.
Searle’s theory of mind starts from an analysis of the language that we commonly use to talk about mental activity. Most of our problems in talking about the mind stem, Seale believes, from the fact that this language is saturated with the notion that there are two separate domains in our heads: mental and physical, mind and matter, eternal soul and mortal body. This duality derives from ancient and primitive philosophico-religious ideas about the mind, which have deeply influenced the way our language has evolved. Searle believes that the language generally used today to discuss the mind still contains these fossils, which lead to underlying confusions, and which muddle our thinking and obscure our picture of reality. The reality is that our minds are made of the same stuff as our bodies: the mind is a biological organ like any other. Searle’s philosophy of mind seeks to explain language and the mind in terms of neuro-physiological structures such as synapses and connections, and to eliminate metaphysical ghosts from the machine. This is the scientific foundation on which Searle’s more detailed studies are built.
Searle has discussed Chomsky’s theories in many books, articles and reviews. The analysis that follows is taken from The Rediscovery of the Mind and from Intentionality.
In The Rediscovery of the Mind Searle writes: ‘Chomsky claims that innate, unconscious rules cause verbal behaviour. In other words, there is a cause/effect relationship between ‘rule’ and language. But studies of neuro-physiology indicate that language is caused not by ‘deep unconscious rules’ but by neuro-physiological structures that have no resemblance to the patterns of language at all. The brain’s hardware produces patterns, but these patterns are not causally related to language produced by humans: they merely delineate the possible forms that human languages can take.’
Searle goes on to argue that such evidence as exists for ‘innate grammar’ and the LAD that supposedly contains it can be explained more satisfactorily and more simply by an alternative hypothesis. The so-called Language Acquisition Device is not a discrete part of the brain. Rather, the whole brain is a Language Acquisition Device. Language is a complex affair, requiring the participation of every part of the brain to master. Moreover, the brain is a biological organ like any other, except that it is more complex. The physical components ands structures of the brain – its synapses and connections – account for the sorts of languages that can be learned by human beings. There is no need to postulate, in addition, rules of innate grammar. Such a postulation is redundant and incoherent.
Searle also analyses the notion of ‘unconscious’ employed in the theory of ‘unconscious rules of grammar’. He finds that the notion is confused and the cause of theoretical muddles. The term ‘unconscious’, as used in Freud’s theory of the structure of the mind, relates to what is repressed, not to what is in principle unavailable to conscious knowledge. Searle adopts the term ‘non-conscious’ to denote mental and physical biological processes that cannot be known consciously, such as the systems of digestion, balance-while-walking, the workings of the brain’s synapses, and so on. He thus avoids ontological confusions about the status of the mental and physical constraints that are caused by the structure of the brain. They are simply part of our biological systems.
A further basic point relevant to the theory of innate grammar is made in Searle’s book Intentionality. According to that work, the whole Chomskyan focus on the inner linguistic workings of the brain is back to front. We understand language rightly not by peering into the entrails of the brain, but by studying how language helps us to deal with life: to predict, to organize, to make sense of things, to make things happen in coherent ways, to handle conflict and to love another person. We can better understand how language really works, Searle maintains, by seeing how it helps us to live than by looking for imaginary rules in the head. This view is similar to Wittgenstein’s and to Halliday’s: it is essentially a functional rather than a systemic view of language.
Chomsky’s theory is ‘a stunning mistake’, to quite Searle’s words. Where we are unable to find meaningful conscious processes, we posit meaningful unconscious processes, even deep unconscious ones. But such processes do not exist. They are a meaningless chimera.
The reasoning that leads to the rejection of the innate grammar hypothesis also leads to the rejection of those other bastions of the Chomskyan edifice: deep structures and universal grammar. They are myths. The whole Chomskyan castle in the air begins to disintegrate.
Let us conclude. Positivist philosophy, Wittgenstein’s language analysis, and Searle’s philosophy of language and the mind all reach the same conclusion: Chomsky’s innate grammar theory is not valid. So? In the words of David Hume: ‘if a theory cannot be measured, and if there is no evidence for it, consign it to the flames.’
If the innate grammar theory is invalid, then language teaching methods that depend on it are also invalid. If grammar has to be mastered through use and practice and understanding - and is not innate - then language teaching methods that teach grammar through use and practice and cognitive understanding are valid. And that is the main practical conclusion of this enquiry into the question: ‘is grammar innate?’ Grammar is not innate: it has to be learned. Sound language teaching methods start here.
(a version of this article was first published in Modern English Teacher in July 2004)