It’s easy to take language for granted. Pretty much every human in the world can learn a language, so how special can it really be? But if you take a step back and look at the wider view, the fact that humans are uniquely able to develop complicated languages is astounding. And despite decades of study, there’s no consensus on why exactly language is possible. But there’s one contender that has been dividing the linguistics community: Universal Grammar.
In this episode of Multilinguish, we look at the theory of Universal Grammar, which posits that humans are born with a genetic predisposition for languages. We explain what exactly that means, and explore why some linguists are so adamantly against it.
Multilinguish: Are We Born With Language?
In the first part of the episode, Thomas Devlin (me!) and Babbel’s Senior Learning Experience Designer Jennifer Dorman define what Universal Grammar really is. Which is easier said than done. Then we talk about Noam Chomsky, who’s the linguist behind this theory. We close out the half by looking at studies that have been used as evidence for Universal Grammar, including the wug test.
In the second part, Thomas and Jennifer return to talk about all the arguments against Universal Grammar. This includes Michael Tomasello’s usage-based approach, which is one of the most popular alternative theories that tries to explain why children are able to learn languages so easily. Lastly, we talk about why Universal Grammar matters, and what it means for humanity that we all have some innate ability to acquire any language in the world.
This episode was produced by me, Thomas Devlin, and edited by Brian Rosado. Special thanks to Jennifer Dorman for her help and insight on this episode. Jen Jordan is our executive producer. Our logo was designed by Ally Zhao.
What Exactly Is Universal Grammar, And Has Anyone Seen It? | Frontiers In Psychology
B.F. Skinner: The Writer And His Definition Of Verbal Behavior | ABAI
A Review Of B.F. Skinner’s Verbal Behavior | Noam Chomsky
Brain Mechanisms In Early Language Acquisition | Neuron
Why Only Us?: Language And Evolution | MIT Press
Constructing A Language | Harvard University Press
Thomas Devlin: From the language app, Babbel, this is Multilinguish. I’m producer Thomas Devlin. If you’re listening to this, you were probably once a baby. And as a baby, you somehow were able to gain the ability to understand and produce language. It’s something that everyone does and so it’s easy to take for granted.
But it’s at the heart of one of the most contentious questions about language. Why is it that humans, unlike every other creature on earth, are able to learn complex languages? In this episode of Multilinguish, we look at Noam Chomsky, universal grammar, and the argument that has torn the linguistics community apart for decades.
Universal grammar is one of the most important concepts in linguistics, but defining it can be difficult. To help me understand universal grammar better, I talked to one of Babbel’s senior learner experience designers, Jennifer Dorman.
Jennifer Dorman: A universal grammar, more or less says that grammar is innate, or languages is innate. That people are born with this innate, biologically programmed set of grammatical categories. Things like nouns and verbs, word order, that sort of thing. And this innateness hypothesis more or less means that humans are unique in that they are genetically wired for language acquisition.
Thomas Devlin: At its simplest, universal grammar is simply a theory that says that humans are born specifically with the ability to acquire languages. The name can seem a little misleading at first, because usually when we think about grammar, we’re thinking about the set of rules that make up a language. English grammar, for example, has rules for what order we can put words in and how we conjugate them. So, is universal grammar related to these other grammars then?
Jennifer Dorman: Well, in some ways, yes. So, we are definitely talking about some very specific things that we would all understand as grammar. So for instance, the fact that languages have different types of words. So, what we call grammatical categories, like a noun, a verb, a modifier, like an adjective or an adverb.
But I should add, even for people who very much believe in this theory of universal grammar, there isn’t necessarily consensus as to what that base set of grammar elements are baked into our biological code. So, there are some people that say there are maybe a handful of principles that are generalized across all languages, and others who say maybe there’s only one.
Thomas Devlin: So, already, we’ve run into a disagreement on what universal grammar actually is. But to really get at the heart of this topic, we have to talk about the man who is at the very center of pretty much every discussion about it, Noam Chomsky.
Jennifer Dorman: So, Noam Chomsky is one of the few household names in linguistics. You might know him from lots of other things he does on the cultural or social activism side. But realistically speaking, he is absolutely one of the most pivotal characters in the world of linguistics.
Some might say, one of the most divisive characters in linguistics. But as we talk about universal grammar today, you’ll probably see why Noam Chomsky is associated with, well, one of the larger controversies in the science of linguistics.
Thomas Devlin: Chomsky has been working in linguistics for decades and outside of the world of academia, he’s probably most famous for his stances against the Vietnam and Iraq wars. But back in the 1950s, before he became very politically involved, he was a linguistic student at the Massachusetts Institute of Technology. And at the time, there was a different prevailing theory about how children learn languages.
Jennifer Dorman: In the 1950s, the major idea for how children actually begin speaking and understanding this language, was behaviorism. And the guy you would look to for that, not the only linguist to talk about this, but probably the most famous, was B.F. Skinner. And his sort of seminal work, which was published in, I think, 1957 was called Verbal Behavior.
And essentially, he promoted this idea of a blank slate because behaviorism, which you might know a little bit from a set of dogs and bells and salivating. So Pavlov’s behaviorism conditioning studies. But B.F. Skinner essentially, said that there’s no innate information necessary for language acquisition.
Thomas Devlin: If you’ve ever taken a psych 101 course, the names B.F. Skinner and Ivan Pavlov probably ring a bell. No pun intended. Their behaviorist approach to language learning meant that children don’t need an innate ability to learn a language. They just listen to the world around them and are taught how to speak through a system of positive and negative feedback.
When a child says something with proper grammar, for example, a parent might encourage them, so positive feedback. And if the child is wrong, then the parent might correct them, which is negative feedback. In 1959, Noam Chomsky wrote a review of Skinner’s Verbal Behavior and basically tore it apart.
And in this review, he started putting forward the ideas that would one day become universal grammar. And as far as reviews of academic books go, this one really rocked the boat. Chomsky, who up until then was barely known, was taking on Skinner, whose work was massively influential.
Over the next few decades, Chomsky developed these ideas more and more. But as we’ve already mentioned, there’s disagreement over what even makes up universal grammar. So how do you prove whether it’s true? Well, there are two main kinds of proof. The first is observational.
Jennifer Dorman: Some of the observational elements that led Chomsky to formulate this theory, is just observing way infants across the world in learning every possible language known to man, how they tend to acquire that language. So, not just the order of the types of skills that we see emerging, but also, the time span of when those things emerged.
Thomas Devlin: If you were to entirely buy into the behaviorist approach to language learning, a child’s environment would have a huge impact on their language ability. And while not every baby develops at the exact same rate, the average across the world is remarkably similar.
A baby can learn Hindi as quickly as it can learn English, or any other language on earth for that matter. Plus babies learn language much faster than they learn most other skills. And based on these facts alone, it can seem like babies must be born with an innate ability to learn a language. Another observational argument that Chomsky has used, is something called the poverty of the stimulus.
Jennifer Dorman: What children produce, they produce things that would be nearly inexplicable when you consider that, if we think about the fact that if children were born as like a blank slate, they would have to hear examples of all different types of language. They would have to know what was right, what was wrong. Things would have to be reinforced, for them to actually produce the kind of language structures that they can.
But yet, generally speaking, most caregivers are not providing explicit instructions. Saying, “Yes, little Tommy, this is correct. And this is incorrect. And you should say it this way.” Or, “Don’t say it that way.” There just simply isn’t enough raw material in a child’s sensory input to account for what they are able to produce with relative ease.
Thomas Devlin: While all of this observational data is great, what science really relies on to prove and disprove theories, is empirical data. To prove universal grammar, you need studies that show the hard numbers about how children learn languages.
Jennifer Dorman: Empirical tests are a little bit more challenging, but there have been some really interesting studies in laboratories. And, yes, with infants and children. In a very safe, of course, and very ethical way, that seemed to potentially support universal grammar.
The emergence of sounds that infants produce tends to be fairly standard across languages, even when the language they hear around them does not actually have those sounds in their language. So there’s no phonological inventory. But yet, when we look at how infants, so infants from around eight months, 10 months, how they can recognize certain sounds in laboratory settings.
So for instance, the L and the R sound. So, la and ra. There have been extensive studies with American and Japanese infants, or also American and Korean infants, to see whether or not they are able to distinguish between these two sounds. And so, up until about 10 months, both Japanese and American infants are able to reliably distinguish between la and ra.
They start to lose that a little bit after 10 months, where the Japanese infants start to become, in a sense, a little less accurate in identifying those different sounds. And by about 12 months, the Japanese infants are really having a very hard time identifying difference. Whereas, American infants are getting more accurate.
What’s really interesting about this is that these L and R sounds are not part of the sound base of Japanese as a language. So you’re hearing… So think about it. This eight month old child in Japan is able to distinguish these two sounds that they don’t hear in their natural environment.
The fact that they get a little bit less accurate while American infants got more accurate, is interesting because, of course, in English we hear L and R as two different sounds. So, lands and rand are two different words in English. So that’s maybe some empirical evidence. And another test that I really like is the Wug Test. And is a Wug Test something you’re familiar with, Thomas?
Thomas Devlin: The Wug Test is possibly one of the most famous linguistic experiments of all time. And it’s very simple. A group of children were shown an image of a blob-like creature and told this is a wug. Then the image would change so there were two blob-like creatures, and the children would be told now there is another one. There are two of them, there are two, blank. And then the children would be asked to fill in this blank, which they almost always did with the correct answer, wugs.
What this proves is that children, aren’t only repeating what they’ve heard before, and that they can create new words using rules that they’ve never been explicitly taught. This goes back to the poverty of the stimulus mentioned before.
If children were born with a blank slate, this shouldn’t be possible. The fact that they can then, is evidence for universal grammar. But not everyone buys this. And after the break, we’ll look at universal grammar’s many enemies, what other theories are out there, and what this linguistic theory really means for us.
Steph Koyfman: Hey, there. It’s Steph. Multilinguish is brought to you by Babbel, the language app. Our marketing team wants you to know that we offer an app that teaches you 14 languages. From Spanish, French, and Italian to Portuguese, Russian, and more, Babbel’s app is created by real language teachers and experts.
You learn how to have conversations in real life situations, like flirting with a stranger, or meeting your significant other’s in-laws. We’re offering multi-language listeners 50% off a three month subscription. New customers can get this offer by visiting babbel.com/podcast. That’s B-A-B-B-E-L.com/podcast. Now, back to the show.
Thomas Devlin: Welcome back. Before the break, we defined universal grammar and looked at some of the strongest arguments for its existence. But as we’ve mentioned, this theory is extremely divisive in the linguistic community. And one of the biggest problems with the theory is that it is kind of impossible to prove it’s false.
Jennifer Dorman: A good theory in science in general should be falsifiable. And there are a lot of people say that this is just not the case with universal grammar, because we can’t definitively prove that it’s there.
And by the way, I mentioned a few studies, empirical studies that may support it. There are also plenty of studies that we could look to to say, “No, that’s not the case.” For instance, studies where children have been able to produce grammatical errors. Or I shouldn’t say been able to. Did produce grammatical errors that, in theory, universal grammar should preclude.
Thomas Devlin: And no matter your personal thoughts on universal grammar, this is not a good thing. It’s hard to tout a theory that can never be tested. There have been certain people who claim to disprove universal grammar though.
At one point in the history of the theory, Noam Chomsky wrote that he had finally figured out the one element of grammar that is shared by all languages, recursion. Recursion basically allows speakers to embed smaller structures into larger structures. One example in English would be making a noun possessive. You can say the phrase, “Mike’s car,” but you can also say, “Mike’s daughter’s car,” or, “Mike’s daughter’s boyfriend’s car.” It’s this kind of recursion that allows language to expand infinitely.
When Chomsky defined universal grammar this way though, it became more open to attack. Another linguist, Daniel Everett, took advantage of this. He studied a language called Pirahã, which is spoken by an indigenous population in Brazil. Everett claimed that this language had no recursion at all. And therefore, universal grammar must be wrong.
This direct rebuff of Chomsky made a splash when Everett’s initial paper was published in 2005, but it didn’t completely destroy the theory. And that’s because again, universal grammar is hard to falsify, because when flaws are pointed out, the theory can change to address those flaws.
As recently as 2016, Chomsky published a book that refined his idea of what universal grammar is. So rather than focus on various attacks on universal grammar, and there are many, it might be more helpful to look at alternative theories. And one of the most popular theories right now is Tomasello’s usage-based approach.
Jennifer Dorman: So essentially, the usage-based approach is an alternative to language acquisition. So an alternative to universal grammar. And it essentially is encapsulated by understanding that language emerges through use. The structure, so the grammar, emerges through use. And that children rely on general cognitive skills.
So it’s not that the children are born with a blank slate per se, but they’re not born with any sort of innate understanding of language. Rather they’re born with some innate cognitive skills that allow them to learn lots of things, one of those being language.
Thomas Devlin: In many ways, the usage-based approach is very similar to universal grammar. It argues that children have remarkable ability to learn a language, but instead of saying that it comes from an inherent language gene, Tomasello says it comes from other cognitive abilities children have.
One strong piece of evidence for this is how the brain handles language. If universal grammar were real, that would mean that there is probably one specific part of the brain that is handling all language, but that’s just not the case. The brain is very complex and many different parts come into play when a person is listening and speaking. And if you want to learn more about that, you should check out our episode from earlier this season on the topic called, your brain on language.
But the usage-based approach also has its flaws. It doesn’t explain, for example, why children are so good at avoiding mistakes when learning a language. There just seems to be something else going on to explain human kinds remarkable ability to learn languages. And we could spend all day looking into more theories and studies. But at this point you’re probably wondering, does this matter to me?
Jennifer Dorman: Pretty much not. I mean, I shouldn’t say that definitively because I have heard some linguists talk a little bit about, “Well, if universal grammar is accurate, then we should explicitly teach grammar in the second language.” By that, I mean, at Babbel, when we structure our language courses, we don’t just say, “Okay, here. Here’s your Spanish course. Here’s your Turkish course.” We say, “Okay, here’s your Spanish course for your American English learner. Here’s your Turkish course for your German native speaker.” We localize it.
And that actually probably is a little bit in line with this idea of universal grammar because localizing language instruction, especially for adults usually means that we try to take a look at that first language, your native language. And see what in that first language can we build upon? What are some things that might be blockers or cause a lot of confusion?
So for instance, as a native English speaker, I mean, we don’t have grammatical gender. I mean not really in English anymore. So when I learned German or when I learn French or Polish, and these are languages that have grammatical gender, I need to be explicitly taught what that means. Whereas, if I’m a native French speaker and I’m learning Spanish, for instance, this doesn’t matter so much because I already have this concept of what grammatical gender is.
Thomas Devlin: When it gets down to it. If a big paper came out tomorrow definitively proving universal grammar is true or not, your life probably wouldn’t change. But there is something about the concept that is pretty remarkable, no matter who you are.
Jennifer Dorman: But it is an interesting cognitive exercise to think about, “Huh, let’s take a step back. How do infants and children do that?” Because we just take it as a matter of course. We look at children and we think, “Wow, that’s fantastic. It’s amazing what they can understand at a young age, at a time when they are still eating dirt and crawling.” And yet, somehow they can still infer communicative intent, sometimes just by looking at our eyes or a couple of words.
Thomas Devlin: Language is amazing. The fact that we can communicate complex ideas and emotions just by opening our mouths and emitting noise, is in many ways the most remarkable part of being human. And we really still don’t understand how it works. Whether universal grammar is true or not, Noam Chomsky and his theory have changed how we understand language forever.
Even those that don’t agree with Chomsky, still think that there is something inside of us, some cognitive ability, that allows us to create language. And from there, that allows us to create civilization and human society. This could seem like it’s some silly academic argument that doesn’t really matter to anyone, but it’s really about who we are, how we got here and what universal traits, bind humans across the world together.
Multilinguish is a production of the language app, Babbel. This episode was produced by me, Thomas Devlin. Editing and sound design by Brian Rosado. And a very big thanks to Jennifer Dorman, whose insight was invaluable to this episode. You can read about today’s episode topic and more on Babbel Magazine. Just visit B-A-B-B-E-L.com/magazine. Say hi on social media by finding us at Babbel USA. Finally, please rate and review this podcast. We really appreciate it.