[Music]

This is KZSU Stanford.

[Music]

In title opinions coming to you from the womb of the great cosmic serpent,

wherein our rebels begin.

In the beginning was the word, yet the word did not come first.

There is always an earlier, more sacred, and a rural hour to the dawn.

And there is always something that comes before the beginning, before birth, before the coagulations of meaning.

Here's a quote to get us underway.

By the time a newborn emerges into the cool air, its mind flares with recognition when it hears the language that flowed outside the flesh walls that cradled it.

In the womb, infants have absorbed the clique clique of the rhythms of their mother tongue, its cadences and rising falling motifs, its predilection for hurting consonants together,

or for spacing them out as sparse counterpoints to the airiness of vowels.

It is then, and in the months to follow, that an infant and her native language come to possess each other.

Before naming, before meaning, before there is even a notion that sound bears the burden of meaning.

Those sentences come from a recently published book titled Lingua File, "A Life of Language Love," whose author joins me today to talk about the connections between language and love, language and music, language and self, and of course language and meaning.

Julie Cetavi is an acclaimed writer and linguist whose work ranges from the scientific to the literary and everything in between.

I'll mention two of her previous books, memory speaks on losing and reclaiming language and self, and an introduction to cycle linguistics.

Cetavi has taught linguistics and psychology of Brown University and the University of Calgary, and now she joins us here on entitled opinions.

Julie, welcome to the program.

Thank you, Robert. It's a pleasure to be here.

Your new book, Lingua File, has three parts, childhood, maturity and loss. I'm especially fascinated by what you have to say about childhood, the way we're born into language and what takes place before meaning, before meaning being the title of the first section of your first chapter.

Human language is a mysterious phenomenon to say the least and above all when it comes to our induction into our native tongues.

In your book, you write about the fluid nature of our initial experiences with language. You ask, I quote, "How do infants intuit the shapes of words immerse as they are in language in its liquid form?"

So here's a question to start us off. How much do we really know about language in its liquid form given that none of us has a reliable memory of floating in that amniotic fluid, if I may call it that?

That's a wonderful question. And I think it really speaks to the desire to interrogate that mysterious part of ourselves that we can't access through memory.

But I would add that even our conscious awareness of language as adults has limits as well. There are many aspects of language that are simply not that available to our conscious awareness.

And that's one of the things that really drew me to the language of science is the fact that this opened a door into the inner self, if you will.

So yeah, let's visit that period of time before memory that is so mysterious and so profound.

So how do we know what infants know about their language? I think it starts as with all knowledge with careful observation. So I'm going to tell you about the first minute of my daughter's life.

When she came out into the world, a nurse held her and I spoke and immediately her head turned towards me. I could see on her face that there was some searching and some deep attention that was triggered by my voice.

And I suspect that observations like that are what have fueled scientists to ask the question of what is known to a child to a newborn baby, what is it that stands out in their experience.

And the answer to that is by interrogating where they press their attention, what captures their attention.

As you know, a newborn baby doesn't have many actions that we can use to respond to us. We certainly can't ask it for its introspection, but nor does it have many behaviors.

But there are some behaviors that reflect attention and one of these is sucking, sucking is a measure of arousal and arousal is connected to how deeply you're paying attention to something.

So we know, for example, that if we put pacifiers into the mouths of newborns and these pacifiers are wired up to electrical devices that measure the rate of sucking.

And we played over a loudspeaker the sounds of the infants mother tongue that they have heard a sum of while in the womb, they will suck at a faster rate than if they hear a foreign language that is dissimilar from their mother tongue.

So that tells us that they have learned something in the womb that there's some familiarity that they have with their mother tongue.

That connection between the infant and the mother is primordial also a new account of the bond that language creates between, I mean, it's a pre linguistic bond to be sure, but that infant and newly outside of the womb recognizes the sound of of its own matrix in that sense.

Yes, that's a beautiful way of putting it. Yeah, absolutely.

And there's this trauma of birth. It has to be a trauma of separation. And to hear that voice is connection and language is connection. In fact, I think the.

The earliest meaning of logos in Greek was really that of relations relational as a bond. It's a connection. So love plays a really crucial role in your understanding of how language acquisition takes place and what the purposes of language subsequently are in terms of creating connections between people and with community and so forth.

But it really begins with this primordial bond between infant and mothers, right? Do I get you right? Yes, yes, you've articulated it beautifully. And I really wanted to highlight that in my book is that from the very beginnings, even before birth, our exposure to language is in the context of these fierce, fierce attachments that we have to our caregivers, or even to the body within which we began our existence.

And I think that shapes many aspects of language throughout our lives all the way until the moment of death.

So the language, as you mentioned, one, you know, the infant can recognize a language, even though it doesn't understand a word of what is being spoken. In fact, doesn't even understand that a language is made up of words.

What is it that the infant hears when it hears? Let's call it the native, you know, it's native language, namely the language that his mother was speaking while it was in the womb.

That changes very significantly over the first months of infant's life. So in the womb, it can only hear things that are muffled through the walls of, you know, all of the body parts that it's passing through.

And it's picking up primarily rhythmic properties, patterns of stress and timing that are typical of each language. After birth, the baby can hear much more fine-grained aspects of speech. So the individual consonants and their fine-grained distinctions.

It comes to learn which sounds are common in their language and which sounds are not particularly useful for signaling contrasts.

So it adapts its perceptual system to be specific to the sounds of the mother tongue, so to speak. And it also learns to identify boundaries of chunks of sound that kind of grouped together.

So even before a child understands the meanings of words, it's able to identify that certain syllables cluster together and can be separated from the speech stream as units.

And that probably serves as a very useful anchor upon which to later on tether meaning to.

So Julia, it sounds like you're pointing there to the connection, the profound connection that language has with music.

And here, I'm really quite taken with what you write early on in your book about. You say that over time, language becomes less fluid as more and more clumps of sound solidify.

These offer themselves as vessels for meaning and the strands of sound and meaning begin to braid together at subsequent stage.

But there was a time in each of our lives when we experienced language as sheer music.

We felt its rhythm in our bodies, we recognized its patterns and motifs.

We could be soothe by its repetition then surprised by its departure from pattern, but we did not understand it.

We might live, I think, much as birds do.

The only go on to claim that strangely enough, when it comes to language at least.

Our species has a lot more in common with birds than with our own primate ancestors.

Well, it has certain things in common for sure.

And one of those is that birds like humans are exquisite vocal learners.

And apes are terrible at this.

Apes are very bad at reproducing sound. They're a much better at reproducing science with their hands, for example.

So researchers have been far more successful teaching chimpanzees and gorillas sign language than spoken language.

Birds, on the other hand, do learn to make very complex and highly structured dreams of sound, much like speeches.

And here, I'm referring specifically to birdsong and not just the isolated calls that birds also make.

And what I think is fascinating is that we don't have any evidence that there is meaning that is conveyed in those streams of sound in the way that language uses speech to combine units of meaning together.

And of course, as humans, we also produce meaningless streams of structured sound in the form of music.

That also is something that's universal to humans.

Well, some biologists might claim that birdsong has, if not a semantic meaning it has a purpose which is, you know, mating or signaling, I like your suggestion better is that it's a way of self manifestation.

It's an announcement that I am here. I sing, therefore, I am.

Yeah, admittedly, that's somewhat speculative because I have even less insight into the mind of a bird than I do into the mind of a newborn.

But yeah, there is just no evidence that they're gathering together to form laws or hold seminars or anything of the sort of thing that we do with speech.

Right. I'm interested also because of my long, ongoing engagement with John Batista Bico, the 18th century Italian theorist who wrote a book called The New Science, which delves into the origins of human culture.

And also of language and I'm going to read a few axioms that he has, and see what you make of them because when we're talking about infancy, he also speaks about the infancy of the human race.

He says, "Mutes make themselves understood by gestures or objects that have natural relations with the idea they signify."

"Mutes utter formless sounds by singing and stammers by singing teach their tongues to pronounce."

Another axiom, "Men vent great passions by breaking into song as we observe in the most grief-stricken and the most joyful."

And from these axioms, Bico writes, "It follows that the founders of the human nations having wandered in the wild state of dumb beasts were inexpressive save under the impulse of violent passions and formed their first languages by singing."

This is a theory of the origins of language. Now, from one point of view, there's nothing more feckless than what are the origins of language because it's always a matter of speculation.

However, the primacy of music, I think there's a lot of empirical evidence to suggest that language does arise from this musical, let's say, configuration.

And we still hear it even when you're speaking or I'm speaking, and as you claim, the infant can understand the music of the maternal tongue, so forth.

Here, I wanted to draw attention to what he says about the importance of gesture.

Before speech, I gather that you think that gesture is hugely important for children as they're learning to speak.

Well, you know, can I back up a little bit and comment on some aspects of what you just said?

I love that quote very, very much, but I'm also struck by the fact that we humans have in parallel the systems of music, which as you describe can give expression to those passions that we just need to burst forth with.

We still have access to that, if you will, meaningless aspect of structured sound, even as we've developed language and speech, so the two are not serving exactly the same purpose clearly.

And gesture, I think it's the fact that gesture can be recruited for language if speech is not available and that it can serve, as far as we know, as well as speech for communicating, suggests that language is not tethered to speech, that to some extent it can be separated from sound and, you know, the qualities of music.

I think this is something that's actually quite profound about language that it has a relationship to music, but it also has some independence from music.

I think that the times when we most want to use language to give expression to those vast passions, I would say, come in the form of poetry, where maybe we return somewhat to the musicality of language, where poems pay special attention to the rhythms of language and the sound patterns of language, the repetitions.

So I think poetry in a sense is where those two threads, the need to convey meaning and the need to be expressive through the musicality of sound really come together.

Exactly, and poetry is the famous sound and sense. Yes, conjugation that you have in poetry. And you know, you were saying that language and music can be separated, they're also united.

And when Vico says that men vent great passions by breaking into song as we observe in the most grief-stricken and the most joyful, I think it's interesting that grief, there are great tragedies, esculus, for example, in the Persians where

After the tragedy, there are moments in the manuscript thing where you have nothing but a string of vowel sounds, and the commentators, they don't know what to make of it, they think that there's been a corruption, but in a sense, when a passion gets too intense, then it just becomes a kind of vocalization, non-Semitic, it's the expression of grief.

I know that Jean-Jacques Rousseau, who was also an 18th century theorist, he also believed in the musical origins of the language, and of course, he, you know, being French and being a romantic, he thought that it was joy, the other candidate.

And he imagines the first young man and the young woman around a well, prior to, you know, semantic language, and that there would be some kind of vocalization of love, where the content of the melody, it would be just a melodic phrase would be, "Mm-m-wah, love me."

But that would be communicated sheerly through the music of it, not through semantic meaning.

And birds to get back to these lovely creatures seem to do something similar. They vocalize to their young in a way that, where that function seems to be to strengthen the bond and the response between chick and parent bird is just a way of linking together through vocalization.

Yeah, that's what's so intriguing to me, Julie, is that the role that love plays in the human equation between mother and child, and also as you go on to point out that the child then wants to align with other minds.

And where there are these other minds around it, and that it also is dimly aware that language is a way of aligning its mind or, you know, with other minds, because there is this primordial need to connect to relate, and that love is a word for that.

And yet, we're not the only species that love's intensely, it's, you know, there are all sorts of pack animals, and you're talking about birds and, you know, the mother and the chick, and so forth.

But somehow, when it comes to human beings, this kind of natural love feeds our capacity for communicating with one another linguistically.

Yeah, exactly. So can I relate an experiment that I just read about to you and ask you what you make of it?

So here is an experiment that was done with cats, domestic cats.

It follows up on a technique that is used to study infants typically about 14 to 16 months of age, and looks at how readily they map words onto objects or images.

So the scenario is that a cat looks at an image, and here's a word at the same time, so a picture of, say, a cloud, and here's meadow.

And this paired sound and image goes on for a number of times. And then, at one point, the sound switches, and we hear another word, bopu.

The cat does a double take and stares longer at the screen, registering that there's something unusual about this.

In experiments with human infants have been taken to, you have shown similar things, and this has been taken as evidence that the child has mapped the connection between the word and the image.

Now, in this experiment, it was shown that cats actually required far fewer presentations of these sound word pairs than human infants did.

That'd be curious, what do you make of that?

Well, I would have to reflect on it further to give you a deeper answer about my first reaction is to imagine that a cat's sensitivity to sound is related to all the exquisite sensors it has that enabled it to identify the mouse or some other sort of thing in its environment.

And that any sort of anomaly or the switching of these sounds, these vocal sounds would alert its attention, because they are so much more attuned to the sonorous environment in which they live.

But that is just spontaneous response.

That's a very reasonable hypothesis, but I'd like to offer a different hypothesis.

And one that suggests that the child is slower to make the mapping because it has a more sophisticated understanding of words and what they do.

And what has been shown in this experiment is simply the learning of an association between a sound and an image.

But a child understands that sounds have different functions and sometimes they are used as signs that is that they stand in for something else.

And sometimes they have some other purpose and sometimes signs are uttered in the vicinity of objects, but not to refer to that particular object.

So, for example, an adult might in looking at an object with a child say, "Oh, wow!"

Now, wow is not the name of that object. And the child understands that there are other things that you can utter in the vicinity of an object that don't necessarily link the meaning that there isn't that sign relationship to the object.

So, what children do is actually look for positive evidence that a sign has been uttered with intent.

And if we put a child in a more natural setting, they're playing on the floor with a variety of objects with a parent.

And the parent says, "Oh, look, a Modi!"

A child might be looking at a particular object, but doesn't immediately assume that Modi refers to the object that she's looking at.

She shifts her gaze to the parent and tracks the object of the parent's attention and learns quite readily often after only one presentation of that sound object pair that the parent intended to refer to the object in the parent's focus of attention.

So, that's a much more sophisticated way of mapping sounds to objects than is displayed in this particular experiment.

Now, we haven't done that second version of the experiment with cats. We don't know to what extent they're able to track the object of adult's attention.

But I think that there's an awful lot more going on than simple association of the sound with the presentation of an object.

And that really speaks to the drive to align minds. What the child is doing immediately is trying to discern the intent of the adult upon uttering this sound.

What did they want to do? Did they want to comment? Did they want to name it? What was the purpose of this utterance?

And I can't resist bringing insane Augustine into this discussion because Augustine was the theorist of, you know, for Augustine, everything was motivated by desire.

And of course, for him, it has to always end up being love of God. Anything, you know, less or other than love of God was not the genuine, most genuine form of love.

But his understanding of language acquisition seems very much in line with what you were just describing because he, for Augustine language acquisition is really also motivated by desire.

And it's a desire to find out what other, what's going on in the world around them on the one hand to get into the minds of the adults, but even maybe more importantly, is be able to appropriate the means of communication of one's own desires.

And usually they were egotistical, or craven desires that they just want more food than their rivals sibling or something.

Very much like you do in your book, when you imagine what you were like when you're a child and, you know, on your mother's knee trying to learn language.

Augustine imagines that he learned to speak by observing these preformal forms of communication by the adults who surrounded him when he was still speechless.

And he says, and what they meant was made plain by the gestures of their bodies by a kind of natural language, common to all nations, which expresses itself through changes of countenance, glances of the eye, gestures, and intonations which indicate a disposition and attitude.

Just as an infant, I began to exchange with those about me the verbal signs by which we express our wishes and advance deeper into the stormy fellowship of human life.

I love that line about the stormy fellowship of human life that there's so much packed into that line, but yes, it's really want to get into it.

Yes, wait to get into that story.

Yes, exactly, exactly, but if I can, that is, you know, Augustine was such an astute observer, and I think what he's describing in that passage is also surprisingly something that is common to all humans, but that we don't necessarily share with our close eyes.

So the ability to read each other's state of mind through gesture, eye gaze, expression is something that apes appear to not be particularly good at, and other animals appear to be much better at.

It's been shown, for instance, that chimpanzees do not understand pointing as a communicative gesture.

If you take a piece of food, put it in one of two containers and scratch, you know, switch around the container so it's not visible where the food is, and then you point to the one that has the food in it.

The chimpanzee won't recognize that as a signal to say, try this one, it will choose a random, whereas a dog will understand the gesture of pointing and will try to, you know, get into the container that has been pointed at.

So a simple elements of, you know, simple cues that we have access to that allow us to align minds and to discern each other's intent as an opening for language is not something that's universally shared even among some of our closest animal relatives, but maybe distributed elsewhere in the animal kingdom.

Yeah, yeah. And of course, this phenomenon of meaning where the child for the most part lives within a condition where it only partially comprehends what's going to take place around it.

For the most part, it's not comprehending, but it knows that there's something going on that it's meaningful. Yes.

And it feels excluded and wants to enter into it. That also reminds me of another passage in Augustine in that in this case, but when he gives the example of someone an adult in this case encountering a dead word.

He used the examples of Temetuma, a word that had become obsolete. And even though the adult does not know the meaning of the word knows that the word is meaningful.

And therefore wants to understand what word means. And he says that if anyone therefore applies himself with ardent diligence to know and inflame with this zeal continues the search can he be said to be without love.

What then does he love for certainly something cannot be loved unless it is known, nor does he love those three syllables that he already knows, but suppose he were to love them for this reason because he knows that they signify something.

And I think the child knows that a lot of things signify things. Yes, they signify it, but they're in the very drift of meaning.

They are even when language is still a liquid form that they haven't extracted units from I really, really wish that I could invite a gust into a dinner party of language scientists. I think he would be fascinated and they would be fascinated by him.

But the scenario that you describe is actually something that has been shown with three month old infants where they are, they hear a word and then they see a sequence of images, things that belong to a similar category so they might see a cat, a cow, a dog, a horse, and then something that violates that category, an airplane.

And if a word has been played accompanying these images, they show a shift of attention when they see something of a different category.

That suggests that what the word has done is to their attention to notice some of the commonalities between these unfamiliar objects to notice something about them.

You don't get the same effect if you play a simple tone, for instance, something non linguistic.

So there's something about the presence of that word that is instigating a form of attention that would otherwise not be present.

To me, I always feel that there's a very strong link between attention and love. So when I hear that passage by a gust and that's what I think of as that the love that's expressed there is that desire to know the attention being pressed on something in the world.

And I know that you've thought about the difference between how we learn to speak and artificial intelligence. You say there's a huge difference that what's missing, or let's say the big difference is perhaps that there's no such love that motivates artificial intelligence.

And it's the way it learns language and with no sense of the mind of another or that languages is a means of expressing once will. Can you say a little bit more about the difference between language acquisition along humans and artificial intelligence?

So I think that the way that an artificial intelligence system works is very similar to that experiment with the cats that I described. It's a pairing of image to sound to the extent that sound or the language is paired with anything at all.

There's just some raw stimulus that is associated with that. And then the patterns of statistical regularities that are present within the images and within the linguistic system.

But, you know, as I suggest, the small child is doing something much more rich and complex than the cat has been demonstrated to be able to do.

I think this is why children can learn language on the basis of far less input than a large language model can, you know, it essentially needs to slurp up almost the entirety of the internet in order to have a reliable grammatical linguistic system.

And I think it's because we have these, well, first of all, we live in bodies and have sensory organs so that we, we create meaning from our physical environment and then can overlay that into language as well.

And there's very much this element of connecting to other humans and aligning minds and discerning intent. And large language model simply cannot do that. They mimic language and they will generate language as if they had states of mind and intentions and goals and desires.

But they do not. And I think that that's what makes them so confusing for us humans to interact with.

Because they lack the driving impulse of our own psychology, which is motivates. They are not motivated the way the child is motivated to break into the fellow stormy fellowship of humanity. So, that's right.

Well, Julie, we essentially covered only the, you know, the beginning of your book. So on another occasion, maybe you can come back on to entitled opinions and we can talk about, you know, what happens later in life.

Well, nothing but please be more Robert. Well, it's been a real pleasure. I remind our listeners. We've been speaking with Julie said to be a cycle linguist. Can I call you a cycle linguist. Yes. Yes, you can cycle linguist from Calgary and thanks again for coming on Julie and I am Robert Harris and for entitled opinions. Thanks for listening.

Well, I'm up here in this room. I'm looking all around. Well, I'm looking at my belly button window. And I see a whole lot of friends.

And I wonder if they don't want me around.

What seems to be the fuss out there. Just what seems to be the hate.

'Cause you know if you just don't want me this time around. Yeah, I'll be glad to go back to spirit.

And even take along the rest before coming down the shoot again.

Man, I sure remember the last time.

So if you don't want me now, make up your mind aware when if you don't want me now.

Give a take, you've got 200 days. 'Cause I ain't coming down this way too much more again.

I know what they got pills for ills and thrills and even spills, but I think you're just a little too late.

So I'm coming down into this world that is regardless of love and hate.

And I'm going to sit up in your bed, mama, and just to grin right in your face.

And that I'm going to eat up all your chocolates and say I hope I'm not too late.

So if there's any questions, make up your mind. 'Cause you've got to ever take.

The questions ain't your mind. Get it for take. You've only got 200 days.

And I'm looking at my belly button window and I swear I'll see nothing but a lot of friends.

And I won't live the one way around.