Language: The Cultural Tool
Kerstin Hoge takes a trip into the heart of functionalism with a maverick iconoclast
This book is "a story about the joy of language". If it is unlikely to match the sales figures of Alex Comfort's The Joy of Sex, the allusion succeeds in further reinforcing Daniel Everett's image as something of a maverick iconoclast: a guitar-strumming linguist who prides himself on his distance from the ivory tower, trained in the art of survival hikes and used to self-reliance in the face of adversity and challenge. This is Everett's second book written for a non-specialist audience, and like his first, Don't Sleep, There Are Snakes: Life and Language in the Amazonian Jungle (2008), it draws heavily on his fieldwork and life among the Pirahã, an indigenous, semi-nomadic people of the Brazilian Amazon. While Don't Sleep, There Are Snakes combined ethnography with autobiography, charting Everett's loss of faith in Christianity (as well as in Chomskyan generative grammar), its successor seeks to popularise functionalist linguistics, an approach to language study that understands language structure to reflect language use. Functionalist linguists view language as a communicative tool, which we use to relate our experience and mental representation of the external world. Tools, whether linguistic or otherwise, can be understood only in relation to their function; and, what is more, they are cultural artefacts, created in response to the specific requirements of the society in which they originate.
In Everett's formulation, the functionalist conception of language can be summarised as "cognition + culture + communication", meaning that "each normal human being has a brain, belongs to a community with values, and needs to communicate, and the confluence of these states results in language". If culture can shape language structure, then language is "not prespecified in some instinct" and its acquisition is not based upon innate, language-specific structures and learning mechanisms. In other words, Everett's book is to be read as a treatise against Noam Chomsky's linguistic nativism, which postulates innate knowledge of language in the form of a set of abstract principles that guides us in the acquisition task, limits the number of possible grammars and accounts for the universal properties of human languages.
Everett assembles and summarises three types of extant arguments against nativism. First, he draws attention to the absence of clear physiological and genetic evidence in support of the proposal that there is an innate language faculty. Placing the burden of proof firmly on the nativist side, he argues that the fact that language functions appear to be widely distributed across the brain is problematic for the view that language is an autonomous mental organ, because such an organ ought to be associated with a genetically specified, fixed neural architecture. Distributed neural networks, however, may converge on local nodal processing zones, such as Broca's area in the left inferior frontal lobe, suggesting that language may be a distributed function "with some localizable aspects". Yet Everett is right that even if distributed neural networks converge on nodal foci, it has proved difficult to find a nodal region that is exclusively dedicated to language processing. Thus, brain-imaging studies have shown Broca's area to be activated not only in several linguistic tasks but also in the execution and observation of finger movements. And even if nodal regions associated exclusively with language could be identified, it still does not follow that these regions are innately specified, as they could result from learning or experience. Everett is also correct that the genetic evidence is more limited yet: FOXP2, prematurely hailed as a "language gene", but in fact a transcription factor that regulates the expression of other genes, has been found to be rapidly evolved in echo-locating bats, which calls into question the theory that the human FOXP2 variation is a unique language adaptation.
Second, Everett questions Chomsky's "poverty of the stimulus" argument (which holds that the output of the language-acquisition process is underdetermined by the input data) and suggests that "a language instinct...would be a costly investment for evolution to make if we already had the ability to learn languages via our pre-existent intelligence". He thus makes a case for functionalist linguistics as the more parsimonious theory from both an acquisitional and an evolutionary viewpoint, understanding language acquisition to rely solely on general learning mechanisms. The fronts in the debate on the role of linguistic input in childhood language acquisition, which has given rise to a vast literature, are hardened, and Everett's treatment here is a bit too tendentious to permit the reader a full appreciation of the empirical issues. For example, he confidently asserts that nobody has shown that cognitive explanations do not suffice to account for regularities in language acquisition, which shows a rather wilful disregard of a large body of nativist research.
Third, Everett challenges language universals as a source of empirical support for nativism, and claims that "human language is characterized by diversity rather than similarity". His position is informed by recent findings that there are fewer universal properties of language than was previously assumed, and that statistical universals of the type "if a language has verb-object word order, it will have prepositions (rather than postpositions)" are perhaps more aptly described as family-specific linkages, the result of coupled evolutionary change rather than robust universals that reflect innate biases.
While languages may turn out to be more diverse than assumed in nativist theorising, this does not mean that cross-linguistic differences in language structure result from cultural differences. Everett's defence of this position involves riveting linguistic and anthropological descriptions, but it is also beset with difficulties. A cornerstone of his argumentation is the lack of subordinate and relative clauses in the Pirahã language, which he derives from an "immediacy of experience" principle. According to this cultural principle, Pirahã utterances must be evidentially warranted, marked as "either experienced or as witnessed by someone alive during the lifetime of the speaker". Evidential licensing is encoded by means of a suffix on the verb. Crucially, the suffix can warrant only the verb and the "nouns identified in the action of the verb (the subject, object, and indirect object)". Complex sentences typically describe two distinct events (for example, He saw that the snake bit her involves an event of seeing and an event of biting) and hence the embedded clause cannot be licensed by the evidential suffix on the main verb, with the consequence that Pirahã grammar simply does not allow embedded clauses. For Everett, the cultural importance that the Pirahã place on the immediacy of experience has seeped into their grammar.
Ignoring the controversial question of whether the Pirahã truly lack embedded clauses, the problem with this account is that the idea that an evidential suffix on the verb can identify only its nominal arguments has limited explanatory power. Not all complex sentences involve more than one event. For example, in I see the snake that is green, the relative clause does not describe a distinct event, and the immediacy of experience principle will not apply to rule out these apparently non-existent syntactic structures. Also problematic is Everett's suggestion that it might not always be possible to trace the cultural influence that is responsible for a linguistic feature, given that "cultures may change or intermingle". To allow linguistic analysis to invoke cultural explanation in the absence of overt cultural evidence yields an unconstrained framework and ultimately renders the hypothesis that culture shapes language unfalsifiable.
This said, Language: A Cultural Tool is a very good read. Paired with a book by a similarly straight-talking linguist of a different persuasion, such as Steven Pinker's The Language Instinct: The New Science of Language and Mind, it will provide a most lively introduction to different approaches and debates in contemporary linguistics.
Daniel Everett's first love was music. He joined the school band aged 9 and at one time played every brass instrument. Throughout school he was in many bands, and he still plays guitar every day.
But having studied koine (New Testament) Greek as an undergraduate, he found a new love in linguistics after his first class in the subject at the University of Oklahoma in 1976. Less than two decades later, he was studying and living among the Pirahã people of the Amazon Basin.
Despite initial concerns for the health of his family - all of whom contracted malaria, with the exception of his youngest son, who was not even a year old when they arrived at the Pirahã village - Everett found the privilege and challenge of the experience awe-inspiring.
Incongruously perhaps, he says his favourite place to live is Manchester. Having been professor of phonetics and phonology at the University of Manchester from 2000 to 2006, he fondly recalls the city's "great restaurants, especially on the 'curry mile', and my favourite jazz club ever, Matt & Phred's".
In 2010, Everett became dean of arts and sciences at Bentley University in Massachusetts. This week, he will return to his home town, Holtville, California, for a gig with his old band, Offbeats and Rocs. It will the first time the bandmates have played together since they were teenagers.
Language: The Cultural Tool
By Daniel Everett
Profile Books, 320pp, £14.99
Published 22 March 2012
Kerstin Hoge is university lecturer in German linguistics, University of Oxford, and review editor of the Journal of Linguistics.