A Discussion of Chomsky’s Universal Grammar as an Account for the Origin of Linguistic Representations

One of the most pressing questions in bringing to bear any notion of representation in language and thought is where do representations come from? According to Marr’s three levels of description, hypotheses about cognitive capacities at the first two levels (computational and representational/algorithmic) must not only accurately describe the system, but must also answer questions regarding processing, acquisition, and often puzzling universal patterns. This means a proper understanding of the mental phenomenon of language would have to answer the typological and universal question of why we see the patterns of variation and similarity across languages that we do, and not other logical possibilities. Furthermore, it would have to explain the real-time processing signature of language comprehension and production. Lastly, it would have to answer the puzzling question regarding language acquisition that asks ​​how the learner gets from primary data to a grammar for the language relatively quickly, in the absence of negative data, and mostly independently of individual variables. Chomsky, in a rejection of Skinner’s radical behaviorism and in support of his own mentalistic approach, attempts to answer these linguistic questions through his theory of universal grammar. In this paper I will discuss how Chomsky tries to resolve the pressing question of where our linguistic representations come from through the poverty of the stimulus (POS) argument and Meno’s Paradox (or what Chomsky calls Plato’s Problem) in his attempt to develop a complete theory of language that not only describes, but properly explains the nature and origins of our mentalistic representations.

Chomksy’s poverty of the stimulus argument points out that within their linguistic environments, children are not exposed to rich enough data to acquire every feature of their language, but still learn to speak (Aspects, 57-8). This argument counters the empiricist idea that language is learned entirely through experience as the POS claims that the sentences children hear while learning language do not contain the information required for a complete understanding of that language’s grammar. The POS as a linguistic and philosophical argument takes the following logical structure: The first premise is that the speech that children are exposed to is consistent with many possible grammars. Secondly, it is possible to define data, D, that would distinguish the target grammar from all other grammars that are consistent with the input. Thirdly however, D is missing from speech to children. Fourthly, children nevertheless acquire the target grammar. Therefore, Chomsky concludes, the right grammatical structure arises due to some linguistic property of the child.

Chomsky derives the POS from Plato’s Meno which discusses the seeming impossibility of inquiry where to inquire is to come to know something that has not been taught to you by someone else. Meno says if I ask myself a question I do not know the answer to, I will not know I discovered the correct answer even if I were to find it. Socrates adds that asking myself a question to which I already know the answer is also futile because I already have the knowledge. The first premise of Meno’s paradox is that a person cannot inquire into what they know because “since he knows it, there is no need to search” (Plato 80e). The second is that a person cannot inquire into what they don’t know because “how will you know that this is the thing you did not know?” (Plato 80d). Therefore, “a man cannot search either for what he knows or does not know” and inquiry is impossible. Socrates attempts to resolve Meno’s paradox using the theory of recollection. He argues that when one learns something new without being taught, he is recollecting innate knowledge. Chomsky translates Plato’s paradox of knowledge acquisition into a paradox of language acquisition in the form of the POS. Chomsky further translates Plato’s solution to the paradox of knowledge, that of preexisting innate knowledge which can be recollected and in this sense learned, into his own solution to the paradox of language acquisition in the form of his theory of universal grammar. 

Chomsky’s universal grammar (UG) is the linguistic theory of a biological component of the language faculty that contains innate constraints on what the grammar of a possible human language could be. The theory argues that as children receive linguistic stimuli during language acquisition, they adopt specific syntactic rules that conform to this innate UG and that raised under normal conditions, humans will always develop language with certain properties like distinguishing nouns from verbs or function words from content words due to an innate biologically determined language faculty that knows these rules (Aspects, 57-8). This property does not know vocabulary or all of the grammatical parameters that can vary freely among languages, such as if adjectives come before or after nouns. However, this is not a problem because children do receive this information from their native linguistic environments. Innate UG thus allows Chomsky to theoretically account for those linguistic representations which paradoxically are not taught to us and yet are learned by us and furthermore to account for universalities such as nouns and verbs appearing in all languages. 

While logically compelling, Chomsky’s UG theory hosts drawbacks which threaten to undermine its viability. Firstly, since UG theories are not falsifiable, many consider them to be pseudoscientific. The grammatical rules linguists posit are just observations about existing languages and not predictions about what is possible in a language and therefore UG cannot claim to contain the innate constraints on what the grammar of a possible human language could be. Some argue that languages are so diverse that such universality is refuted by abundant variation at all levels of linguistic organization. Furthermore, some see UG as an oversimplification of neurogenetics as no link from genes to grammar has been consistently mapped by scientists and UG has no coherent formulation. Lastly, UG’s claim to reside in our genome is in conflict with biology because ​​it cannot have evolved by accepted neo-Darwinian evolutionary principles. It seems as though the fast-changing nature of language would prevent genetic structures which change much more slowly from catching up. Instead, it is just as possible that seemingly arbitrary aspects of language structure come from learning and processing biases that derive from the structure of thought processes, pragmatics, perceptuo-motor factors, and cognitive limitations and not innate UG. 

While there are valid objections to Chomsky’s theory of UG, it nevertheless remains a prominent theory in linguistics for its ability to, if true, account for the puzzling questions of how it is we acquire certain linguistic representations in the absence of environmental information and why we observe universal linguistic representations among languages and not other representational structures constituted by different logical possibilities. This theory, while not universally accepted, makes strides in that it at least attempts to account for the nature of our linguistic representations and recognizes that any proper theory of mentalistic representations must face the tremendous difficulty of explaining where our representations come from. I propose that one approach to further research into where our linguistic representations come from and determining if Chomsky is correct would involve further mapping of the human genome. It would be specifically worthwhile to identify and study genes associated with areas of the brain known via lesion studies to be associated with language comprehension and production such as Broca’s area and Wernicke’s area. If we can find evidence of syntactic logic coded in these genes, we might have an empirical basis on which to support Chomsky’s UG theory which would contribute to the development of a complete account of linguistic representations in the mind. Until then, it is clear that at the very least, universal grammar cannot be discounted as a potentially correct understanding of the origins of certain linguistic representations.

Works Cited

Chomsky, Aspects of the Theory of Syntax

Chomsky, Review of Skinner’s Verbal Behavior

Marr, Vision, Ch.1, pp.8-31

Plato, Meno