Inside the Fiercest Debate in Linguistics | Otherwords

- If your image of a linguist is of a mild-mannered bookworm

quietly pouring over a dusty tome in a secluded study,

then you've probably never seen linguists argue

over a controversial topic

because let me tell you,

it can get spicy.

And when it comes to controversial linguistic debates,

few are as heated as the theory of universal grammar,

commonly known as UG.

The arguments on either side can be mind-numbingly academic,

but don't let that fool you.

This is a passionate conflict

with each side casting their leaders as epic heroes

and their opponents as stodgy relics

or narcissistic charlatans.

Perhaps that's because behind all the technical jargon,

universal grammar is suggesting something so fundamental

about the human experience

that like religion or politics,

it can feel like a threat or a confirmation

of one's entire worldview.

I'm Dr. Erica Brozovsky, and this is "Otherwords."

(bright whimsical music)

- [Announcer] "Otherwords."

- There are a lot of characters in this story,

but few would disagree

that the most central is Noam Chomsky,

the linguist and political theorist

who's arguably had more impact on the field

than any other academic.

Chomsky developed the ideas of universal grammar

in the 1950s and 1960s,

proposing a theoretical system of syntactic rules

that are uniform across all languages

and innate to the human mind.

In other words, despite the seeming diversity of languages

around the world,

everyone is born with the same mental capacity for grammar,

which is a product of evolution by natural selection.

Chomsky famously suggested that if an alien visited earth,

they think we all spoke the same language

with minor regional dialects.

This makes some intuitive sense.

There are, after all, many grammatical features

that are found across completely unrelated languages

like the distinction between nouns, verbs, and prepositions,

or the use of affixes to adjust meaning.

According to Chomsky,

UG could explain what he called the poverty of the stimulus.

Generative grammatist like Chomsky have long believed

that beneath even the most seemingly simple sentences

lurks a deep grammar.

Of such startling computational complexity,

it seems impossible

that young children could master it so quickly,

and with so few examples to learn from,

unless the fundamental rules were already hardwired

into their brains.

Over the following decades,

he refined his theory of universal grammar

into something called principles and parameters,

which suggested that all languages adhere

to a set of concrete principles

and variation amongst languages

is due to certain parameters being switched on or off,

depending on culture, necessity, or random chance.

For instance, all sentences must have a subject,

but where that subject is placed in the sentence

is a parameter that varies by language.

Chomsky's theories revolutionized the field

and became the backbone of linguistic analysis

for a generation,

but there were still fundamental questions

that UG struggled to answer.

For one, if evolution is a slow, gradual process,

exactly how does grammar evolve?

Noun's first then verbs.

What part of the brain is responsible for it?

And since the human genome has been largely unchanged

for tens of thousands of years,

why would a system of expression complex enough

to describe particle physics

develop amongst the community of simple hunter-gatherers?

Since European languages tended to be overrepresented

in linguistic analyses,

there were also fears of distorted judgments

on what qualified as a universal characteristic.

English grammar, for example, relies heavily on word order,

so some accused Chomskians, as they're sometimes called,

of overemphasizing it when studying other languages.

Sure enough, as more diverse data rolled in,

proponents of UG had to repeatedly revise

their lists of principles and parameters

to account for unexpected grammatical features

of languages like Tagalog, Mandarin, and Warlpiri.

Towards the end of the 20th century,

alternative theories that challenged UG

started to gain traction

like usage-based linguistics, cognitive linguistics,

and constructive grammar.

Researchers like Michael Tomasello and Joan Bybee argue

that childhood acquisition of language

didn't require some mysterious grammar module

in the brain,

but could be explained

by more general cognitive abilities

like pattern recognition,

frequency gauging and memorization.

Tomasello pointed out the most children's early speech

consists of short lists of memorized chunks like gimme

or want juice

that are repeated for purely functional purposes.

As the vocabulary of chunks grows,

they may notice patterns or statistical frequencies

that allow them to create their own novel configurations.

And when you think about it,

adult speech isn't all that different.

Outside of literature and academic lectures,

most of our day-to-day conversations

are made up of pretty rote phrases

that you've probably heard thousands of times before

with little substitutions and alterations here and there.

Do we really think that our brains

are doing computational gymnastics

under the hood every time we make small talk?

Some even go so far as to suggest

that it wasn't until the advent of writing

that humans develop the kind of deep grammar

that Chomsky and other generative grammatist

spend their lives analyzing.

Meanwhile, advancements in brain imaging suggested

that there was no centralized language organ,

but that the act of speaking

was spread across many different parts of the brain,

much like other cognitive functions,

and despite certain genetic abnormalities

having an impact on language,

genetic research has failed to find any gene

or collection of genes specifically responsible for grammar.

But by this time, UG had become academically entrenched

and the majority of linguists were resisted

to scrapping the paradigm

they had been using their whole careers.

Chomsky himself was known to describe his critics

as liars and frauds,

and yet around the turn of a century

in response to new research and thinking,

he refined his theory of UG further

into something much simpler

and more explicable by evolution.

According to this new theory,

at some point in human history,

perhaps 50 to 100 thousand years ago,

the human brain evolved

a unique cognitive ability he called merge.

This allowed us to mentally combine two separate thoughts

into a set which we could simultaneously perceived

as individual pieces and a united whole.

That set could subsequently be used as an element

in another set and so on and so on

theoretically without end.

This creates a hierarchical thought structure

that would've been useful for things like toolmaking,

cooperation, and understanding our environment.

Language, however, is linear.

Our mouths are only capable of saying one word at a time,

so in order to translate these hierarchical thoughts

into linear sentences,

merge developed into grammar.

Essentially, Chomsky distilled

all his principles and parameters

into one universal characteristic of language, recursion,

the nesting of parts of speech into other parts of speech.

For instance, I linearly speak the sentence,

Amanda said that while you are out,

the green vase you got from your mom

fell off the shelf and broke,

but the meaning of the sentence is hierarchical

with recursive grammar as the intermediary.

According to merge, recursion was a foundation of grammar.

Its true universal element found in all human languages.

In some ways, it was a remarkable concession to his critics.

Not only did he abandon many of his previous claims,

he even acknowledged that the mental skill behind grammar

may have evolved as a general cognitive ability

before later being repurposed for language.

But for some opponents of UG,

even merge was too much of an assumption.

In 2005, linguist Daniel Everett

claimed to have found a language without recursion.

According to him, the Piraha,

an indigenous tribe he'd been living with off and on

since the 1970s,

did not use embedded clauses.

In English, we can say, Henry says it's raining,

with the clause it's raining embed within Henry says X,

but this is supposedly forbidden in Piraha.

Perhaps even more controversial

was the reason Everett offered for this idiosyncrasy

that the Piraha don't talk about

anything they haven't personally witnessed,

a trait he called the immediacy of experience principle.

Also, supposedly why they lacked stories, deities,

and origin myths.

They could say it's raining,

but Henry's opinions are pure hearsay,

so not worth talking about.

Everett's claims created a firestorm.

Many linguists accused him of misunderstanding UG or Piraha

or both.

Some even suggested

that Piraha were feeding him a nonsense language.

Chomsky called him a charlatan,

and some even suggested his conclusions were racist

for suggesting that the Piraha

were incapable of abstract thought.

On the less hyperbolic side,

a trio of Chomsky and linguists wrote an extensive response

to Everett's findings

that persuasively argue

that Piraha does in fact have clause embedding

and cast doubts on the logic of his immediacy

of experience principle.

Everett still has his defenders

including a glowing biography by novelist Tom Wolfe,

but his claims were not the final nail in UG's coffin

this some hoped for.

Even the advent of large language models

have only given both sides more ammunition.

Opponents of UG say it's proof

that a general intelligence can master grammar

without a specially designed language organ.

While defenders of UG point out

that AI models need to analyze billions of pages of text

in order to do so,

which seems to bolster

Chomsky's poverty of the stimulus theory.

Today, despite all the hoopla,

universal grammar is still accepted

and taught at linguistic departments around the world.

Surely some of that endurance is due to academic inertia,

but it does have several big things going for it.

For one, despite its brain melting complexity,

Chomsky's generative grammar

remains a cohesive, elegant system

that successfully predicts linguistic phenomena

in a range of diverse languages.

While alternative theories have certainly gained momentum

over the last several decades,

no one theory is as detailed or comprehensive.

Also, the uniqueness of language to our species

and the universality of it

across all members of our species

continues to convince many

that there must be something biological behind it.

Michael Tomasello argues

that the game of chess also has many unique structures,

but no one believes that this uniqueness

requires an innate chess playing module.

However, very few people can become chess masters,

and the vast majority

of us need direct instruction to learn it.

But barring some kind of genetic abnormality,

every human child becomes a master of grammar

before they can even tie their shoes.

Language is an enormous part of what makes us human.

It might be just an emergent technology like bows and arrows

or a side effect of other cognitive functions,

but it's hard to think of another phenotype

in the animal kingdom

that is as essential to a species identity

without a distinct biological root.

Fish swim, birds fly, humans talk.

Perhaps the debate over UG seems so passionately intractable

because its implications go beyond linguistics

to the very essence of human nature.

Maybe some people like to think

that our behavior is dictated by our biology

while others recoil from the idea that our conversations

are just repetitions of canned phrases.

Are we products of our genes or our environment,

nature or nurture?

Perhaps we'll never know,

but we can at least take pride in the fact

that what makes the question so difficult to answer

is also what makes us so special.

The astounding complexity of our brains and our culture.

If your image of a linguist is of a (speaking gibberish).

Chomsky developed the ideas of (speaking gibberish).

Language, however, is,

however, I can't say that wrong today.

However, here we go.

Also. (speaking gibberish)