Words are combined into phrases and sentences in a dazzling array of patterns, collectively referred to as syntax. The complexity of syntax has long confounded researchers. Consider, for example, the previous sentence. There are all sorts of patterns in the order of the words of that sentence, patterns that are familiar to you and me and other speakers of English. Those patterns are critical to the transmission of meaning and to how we think as we create sentences. It was no coincidence that I put “complexity” after “the,” or “syntax” after “of,” or “researchers” after “confounded,” to cite just three examples of many in that sentence alone. You and I know that “researchers” should follow the main verb of this particular sentence, in this case “confounded.” If I put that word somewhere else it would change the sentence’s meaning or make it confusing. And we know that articles like “the” should precede nouns, as should prepositions like “of.” These and other patterns, sometimes referred to as “rules” as though they represented inviolable edicts voted on by a committee, help to give English sentences a predictable ordering of words. It is this predictable ordering that is usually referred to when linguists talk about a language’s syntax.
Without syntax, it would seem, statements could not be understood, because they would be transferred from speaker to hearer in a jumbled mess of words. This is, it turns out, a bit of an oversimplification since a number of the world’s languages do not have rule-governed word order to the extent that English does. Still, let us stick with the oversimplification for now, because it hints at something meaningful about speech. Many languages, like English, tend to put the subject in front of the verb, and the verb in front of the object, as in “The syntax confounded the researchers” or “Sergio kicked Neymar.” In other languages another order may hold, for instance the object may precede the verb, as in “Sergio Neymar kicked.” In fact, the latter sort of ordering seems to be more common than the former across the world’s languages. Most languages have default word orders: Strong conventions determine how units of meaning are encoded sequentially, even iconic words and these conventions help make language intelligible. Still, the conventions can be exceedingly complex and take kids and adult language learners years to learn. Here is an illustration of a particular word order that helps convey meaning in English: (1) Sergio kicked Neymar and ran away.
This is a very straightforward sentence, but note that you must be familiar with a convention to understand who did the kicking (Sergio) and also who did the running away (again, Sergio). I do not need to say “Sergio kicked Neymar and Sergio ran away” for you to interpret the sentence. A word-order convention of English lets you know that, Because Sergio came first in the sentence, he is also the (omitted) person who did the running away. This convention does not exist in all languages. In some Amazonian languages, the equivalent sentence would mean that Neymar, the one being kicked, ran away. The point here is simply that there are countless “rules” about things like this in English syntax, and as English speakers we often fail to appreciate how many such conventions we must be aware of to convey our thoughts and comprehend those of others. Similarly, speakers of other languages must be familiar with an incredible array of distinct word-order conventions.
Syntactic conventions can be exceedingly complex, and any given language contains so many of them that linguists have long wondered how individuals can learn them. To many linguists in the twentieth century, learning language was largely about learning syntactic rules like those related to who did the second action in a sentence such as the one in example (1), rules that helped people produce and decipher sentences. Various theoretical models were put forth to offer frameworks for understanding how human syntax, with all its complexity, is even possible. Some of the models suggested that humans are genetically hardwired to decipher syntactic patterns, predisposed to make sense out of the stream of words they begin hearing at birth. These models focused on the complex syntactic rules of languages, especially a few well-studied languages like English. They suggested that learning a language was primarily a process of learning two components of the language: its dictionary or, more technically, its lexicon (which consists of all the units like words, prefixes, and suffixes, and the meaning of all the units), and its grammar. The grammar consists of the rules that allow people to put the lexicon’s units of meaning into predictable orders so as to construct even larger units of meaning. An increasing number of linguists now think this “dictionary and grammar” model of language was misguided. According to them there is no real distinction between words and sentences, as odd as that claim may seem, and no material distinction between a dictionary and grammar. We will return to this point toward the end of the chapter. The dictionary-and- grammar view of language benefited from the rise, beginning in the mid-twentieth century, of the theory spread by linguist Noam Chomsky. His theory became the dominant paradigm for the study of syntax during the latter part of the twentieth century and continues to influence scholars in some circles today.
One of the key distinguishing features of human languages, according to the Chomskyan paradigm, is a syntactic feature known as recursion. Recursion refers to the use of a structure within another structure of that kind, for instance when you place one clause inside another. If you are not sure about what a clause is, perhaps this helps: A clause consists of a subject and a predicate. All sentences have at least one clause, but some sentences have two or more clauses. In the following sentence a clause is inserted into another clause to form a coherent thought.
(2) Neymar knows [that he is a great dribbler].
In this example, the bracketed clause serves as the object of the whole sentence. It is a clause serving a function within a larger clause. We can also embed clauses recursively in the middle of each other, as in the following example.
(3) Neymar, [who likes to beat defenders [who think they can stop him]], placed the ball in the bottom corner of the goal.
Chomsky and others suggested that the ability to recursively combine clauses like these is at the core of human speech, implying that it was a key characteristic shared by all human languages. Countless studies have been published on recursive phenomena like embedded clauses, as boring as that might sound. Most of the relevant studies were on English and other European languages, though many were also on unrelated languages. In the last fifteen years or so, however, some linguists have made prominent suggestions that recursion is not so fundamental to syntax, and language more broadly, since embedded clauses like those evident in the previous two examples seem to be lacking in some languages. In 2009, linguists Stephen Levinson and Nick Evans pointed out that, judging from the data, syntactic recursion is not actually found in all languages. Part of the evidence they relied on comes from the famous case of the Pirahã, an Amazonian language I have already discussed. My father published a series of papers around fifteen years ago that described the absence of evidence for recursion in Pirahã (among other things), contradicting the claims of Chomsky and others regarding the proposed universality of recursion in the world’s languages. Instead, the language does not seem to allow for clauses to be placed inside each other but only placed next to each other. Recursive structures equivalent to “Sergio kicked the boy who ran” have yet to be documented in the language. Instead the Pirahã equivalent of this sentence would be something like “Sergio kicked the boy. The boy ran.”
One of the key distinguishing features of human languages, according to the Chomskyan paradigm, is a syntactic feature known as recursion.To someone outside linguistics, this may not seem to be a particularly controversial claim—namely that a language in the Amazon places clauses next to each other instead of embedding one inside the other. To many linguists, however, it became a hot topic. Some expressed skepticism that such a language could exist. To others, like Levinson and Evans, it was not an implausible claim given the extreme diversity of the world’s languages and given that claims of “linguistic universals” always seem to fall apart as the number of languages considered grows and grows. As it stands, no outsider has been able to show that Pirahã speakers use recursion of the kind predicted by Chomsky and colleagues. Some might counter that this is simply because only a handful of people have actually learned Pirahã, as it is exceedingly difficult for outsiders to learn, by all accounts, and so our data on the topic are limited. From this perspective, maybe we just have not come across recursion in the language yet despite the hundreds of hours of recordings. To date, anyhow, no clear evidence for recursion in Pirahã has been offered.
The topic of recursion in Pirahã became popular both inside and outside academia, with articles published in venues as diverse as Language and the New Yorker on the topic. As linguist and syntactician Geoffrey Pullum has noted, part of what was lost in the discussion of Pirahã was the fact that it is not the only language that undermines the notion that recursion is a fundamental feature of syntax. In a 2020 article, Pullum surveyed a series of studies of non-WEIRD languages from the last few decades, all of which appear to lack recursion. Let us consider some of the examples discussed by Pullum.
Pullum notes that Ken Hale, a linguist who studied Australian languages, observed back in the mid-1970s that clauses in the Warlpiri language are not embedded inside each other but rather just loosely adjoined. This appears to be the case in other Australian languages as well. Around the time that Hale was making his observations, a linguist named Des Derbyshire published a series of findings on the Amazonian language Hixkaryana. Derbyshire noted the absence of recursion in the language. Instead, he said, the syntax of the language relies on similar strategies as the syntax of Pirahã. Derbyshire, who had spent decades as a missionary working with and living among the Hixkaryana, found that Hixkaryana clauses are simply placed next to each other rather than inside each other.5 A much more widely spoken language in which recursion does not seem to be relied upon is conversational Indonesian, one of the languages with the most native speakers. Linguist Robert Englebretson made this point in a comprehensive grammar of Indonesian. Indonesian is not the last example that could be offered, but it illustrates the point that some non-Amazonian and non-Australian languages may not rely on recursion during conversation. It is very difficult to prove the complete absence or impossibility of recursion in a given language. What we can observe, however, is that there are multiple languages that have been documented in the last few decades for which we have no clear evidence of recursion. This should likely give us pause prior to considering recursion a key feature of syntax and, more broadly, to considering recursion critical to how we think in order to speak. Another factor that might give us pause is that recursively embedded clauses are relatively infrequent in speech, even in the many languages like English that allow them.
All the discoveries of languages that lack recursion date to the latter part of the twentieth century but have received much more intense scrutiny in the first part of the twenty-first Century in the wake of the findings from Pirahã and in the subsequent well-publicized debates on recursion. Note that none of the exceptional languages are Indo-European. This is not surprising in the sense that the languages of many regions are so distinct when compared to the languages of better-documented language families. Many of the languages indigenous to Amazonia are completely unrelated to each other, let alone those of other regions. The findings on languages with “unusual” syntax offer another case in which the extent of diversity of speech, like other facets of human behavior and thought, has been underestimated because of a research bias on people and languages in WEIRD populations. Even today, the vast majority of syntactic research is conducted on the languages of speakers of a handful of WEIRD languages. Like many claims about universals in human psychology that may be called into question by examining very distinct populations worldwide, claims about universals in syntax tend to face challenges once a truly representative sample of the world’s languages is considered. Evans and Levinson suggest that all linguistic universals that have been proposed, including recursion, do not hold up to closer inspection. Conversely, many very complex phenomena that do not surface in WEIRD languages and which have never been proposas universals are actually very common in many languages of regions like Amazonia. To scholars like Evans, Levinson, and many others like myself, the common perception of certain linguistic “universals” originated at least in part in the biased sample of languages that served as the basis for linguistic inquiry during much of the twentieth century. As that bias waned with the continued documentation of unrelated languages worldwide, the evidence for universals uncoincidentally began to wane as well.
__________________________________
This excerpt is adapted from Caleb Everett’s book: A Myriad of Tongues: How Languages Reveal Differences in How We Think, used with permission from the publisher, Harvard University Press. Copyright © 2023 Caleb Everett.
