Nova (1974–…): Season 45, Episode 101 - NOVA Wonders: What Are Animals Saying? - full transcript

Are you wondering how healthy the food you are eating is? Check it -
What do you wonder about?

The unknown.

What our place
in the universe is.

Artificial intelligence.


Look at this, what's this?


An egg.

Your brain.

Life on a faraway planet.

"NOVA Wonders" investigating
the biggest mysteries.

We have no idea
what's going on there.

These planets in the middle

we think are
in the habitable zone.

And making incredible

Trying to understand

their behavior, their life,
everything that goes on here.

Building an artificial

is going to be the crowning
achievement of humanity.

We're three scientists

exploring the frontiers
of human knowledge.

I'm a neuroscientist

and I study
the biology of memory.

I'm a computer scientist

and I build technology

that can read human emotions.

And I'm a mathematician,

using big data to understand
our modern world.

And we're tackling
the biggest questions...

- Dark energy?
- Dark energy?

Of life...

There's all of these microbes,

and we just don't know
what they are.

And the cosmos.


On this episode...

animals made all kinds of noise.

But what does it mean?

What they're telling
the female is,

"I'm free of parasites."

Can we crack
these mysterious codes?

The gesture means

"Travel with me."

His vocabulary

was like a two-and-a-half year
old human child.

This is Dr. Doolittle's dream.

"NOVA Wonders"

"What Are Animals Saying"...
Right now.


All around us
are alien tongues...

and they don't come from space.

From whale songs,
and wolf howls,

to birds chirping
and dolphins clicking...

The animal world is filled
with mysterious conversations.

Could we ever tap in?

We like to think that language
sets us apart from the beasts.

But are we really
all that special?

Today scientists are starting
to decode those communications,

discovering that
we might not be alone.

I'm Andre Fenton.

I'm Rana El Kaliouby.

I'm Talithia Williams,
and in this episode,

"NOVA Wonders"...
"What are Animals Saying?"

And what does it say about us?


Come on.

When I'm working with Zach...

Lie down.

It seems like magic.

Zach, that'll do.
Zach, here.

Experiencing that kind
of connection with a dog...

Come by.
Lie down.

You're in sync, it's like
any other connection

you have with someone.

Lie down.

Where you really get along,
and you're like,

"Oh wow, that was
an amazing conversation."

Robin Queen is a linguist

and competitive sheep herder
from the University of Michigan,

and like many of us,
she likes to think

she can talk with her dog.


I think we are,

as humans, we're fascinated
by the idea,

the Dr. Dolittle idea.

We want that to be true
in some way

When I started working
with the dogs,

I was shocked
at what they could do.

Don't lose it.

Lie down.

And for border collies
like Zack,

herding sheep
is just the beginning.

Get spider.

They appear to have

a sophisticated understanding
of language.

Get spider, bring it over here.

Their ability to learn
human words

are almost unlimited.

It seems to be, you know,

every year a new dog comes
by with a larger vocabulary.

A glance at YouTube
proves the point.

Tessa, get pumpkin.

Tessa can recognize
four different toys.

Good girl, bring it over here.


Get pig.

And that's puppies' play
compared to Gable...

He knows 150.

Good boy.

Chase, find meow, find meow.

But all of these pooches pale

in comparison to Chaser...

There's meow!
Come here!

I say throw me that ball.

Who has been proclaimed

the world's smartest dog.

Chaser was taught
the individual names

of over a thousand objects,

and that's really pretty cool

because it starts to try
and get at that question

of how special are humans?

Chaser can understand hundreds
of words,

but all she can do
is say "ruff."

She can't actually say

any of those words back.

Now, what would
be really impressive

is when Chaser starts saying,
"You go get the bunny."

Then I'd be impressed.

Find roach, find roach.

But as impressive
as Chaser's feats are,

do they qualify as...



Since the dawn of history,

we've imagined animals
to be like us.

I'm a horse, not a guinea pig.

Our stories are filled
with talking creatures.

But what's the reality?

Well, to answer that question,

let's talk about language.

To scientists,
it's a learned set of symbols

that can be combined
into infinite meanings.

Just consider these...

Dog bites man.

Man bites dog.

By just changing the order
of a couple words,

the meaning of the message
is completely different.

Well, this is the cornerstone
of human language.

It allows us to tell stories,
write poetry,

negotiate contracts,
whisper sweet nothings.

The question is,
is this skill unique to us?

There are two chimpanzees
in Rome

who have brought a new "twist"
in communication

between animal and human...

For over half a century,

scientists have been working
with our closest relatives

to answer that question.

Viki was adopted
when her own mother

couldn't feed her anymore.

At first, scientists
tried teaching language

to chimps by raising them
like children.

Perhaps one of the most famous
of these was Viki.

She loves all
the attention and affection,

and she loves everyone.

Do this, Viki.

Not only is this approach

now considered unethical,

it didn't work.

After seven years
of intense training,

she could barely
utter four words.

"Mama," Viki.

So, scientists switched
to sign language.

If you watch Koko closely,

she's learning
to put her fingertips

to her mouth to sign "eat."

Apes like Koko the gorilla

hinted that apes have
some ability for language.

Koko proved an adept student.

Everyone was amazed at how well

the little gorilla
was catching on.

Let's do just a little
bit more work

and then we'll get a whole bunch
of surprises, okay?

But it wasn't until this guy
came along

that researchers discovered

exactly how impressive
that ability was.

In the world of ape cognition,

Kanzi is, you know,
Elvis Presley.

All right, Kanzi, come on,
tell me what this is,

what's this a picture of?


Very good!

Instead of sign language,

Kanzi learned these.

They're called lexigrams,

abstract symbols
that represent words.

What's this?
Look at this, what's this?

An egg, very good.
That's an egg, good job!

And Kanzi has learned
over 400 of them.

Good, keep going!

Good job, Kanz!

Plastic bag.

Good, Kanzi, good stuff!

Amazingly, he started
teaching himself

this skill as a baby
over 30 years ago.

I want you to go put the onions

in your hot food.

Kanzi might have been

around 11 or 12 at the time.

His vocabulary,
spoken English vocabulary,

was assessed in comparison to

a two-and-a-half-year-old
human child.

Now, I want you
to take this spoon

and put it on top of the bucket.

Can you do that for me?

Can you put it on top
of that bucket?

Watching him today at 37,

it would appear
his language ability

goes well beyond vocabulary.

Can you put it
on top of the bucket?

Very good, that's a good job.

So I can ask Kanzi,

"Hey, Kanzi, can you put
the blanket on your head?"

And then I can ask Kanzi,
"Can you sit on the blanket?"

Can you put it on your head?

Nice job, Kanzi.

Very good, very, very good.

Can you... can you
put the blanket on the cube?

Yeah, now can you sit
on top of the blanket?

No, sit with your bottom on top.

Good job, Kanz,

that's it!

Very good!

What he's obviously doing
in that context

is understanding
not only the individual words,

but the order
in which they're arranged.

I think that's a big deal
because that is

one of the foundational elements

for human spoken language.

Kanzi's a huge deal.

The studies with Kanzi
and with the other apes like him

allowed us to get a window

into what great apes
might be capable of

in terms of learning our world
and our communication.

Science has gained a whole lot

from apes like Kanzi.

But, in all likelihood,

Kanzi will sort of
be the last of his kind.

Moving forward,
I think our approach

is shifted to one where
we're starting to focus more

on what the animals
are doing with one another.

Cat gorilla have visit,
Koko love.

Today, the questions

scientists are asking

is not whether animals
can learn our language...

Could you take
my shoe off, please?

But if we can learn theirs.

It's very interesting
that humans,

who are very caught up
in our own intelligence,

when they wanted to understand

whether other animals
had a language like ours,

the best thing
we can think of was,

can we teach that other species
to speak our language?

What you need to do

if you want to understand
animal communication

is leave
our own language behind.

Try as much as you can to become

more like an animal
and just not think in words

and see what they're seeing

and understand
what they're feeling,

what they're, what they're
communicating about.

And when you do that,
you discover a whole new world

hiding in plain sight.

When I first came
to the rainforest,

this was an alien world for me.

I had no clue what to do,
how to be,

how to move around in here.

Budongo Forest, western Uganda.

Cat Hobaiter
is setting off for work.

Trying to understand
their communication

means understanding
their behavior, their life,

everything that goes on here.

I'm the luckiest person
in the world

because I get paid to run around
rainforests with wild chimps.

I love this.

She's spent over ten years
studying these chimps.

It's been a pretty
incredible thing

to be able to watch

some of these chimps
from the day they were born

until adulthood.

And I get to see
the little detail,

the soap opera of their life.

I'm an outside observer,

but I've been here for so long

I feel a part
of the family, sometimes.

In the process,

she has unearthed a hidden form
of communication.

Those scratches,
shaking of trees?

To Cat, they aren't
random motions.

They're part
of an elaborate code,

a secret language of chimps.

All of these gestures

are a part
of chimpanzee communication

and they grow up with them.

To humans, it might
seem really subtle,

like a tiny little push,
or a tiny little pull,

and that's really hard
for us to see,

but I think to the chimps

it's very obvious
what's going on.

So she's sitting down,

looking up at her daughters,
and she's giving a big scratch,

so she's ready to go.

That's Harriet,
and that scratch,

it's not because she has fleas.

It's actually a signal
to her daughter Harmony.

And the little one's
coming down now.

Well, that scratch
has two meanings...

One of them is "groom me,"

and the other one
is "let's travel together."

The loud scratch
got her to come down.

They're all going to go

down the tree,

and that's them
leaving together.


To the untrained eye,

the gestures
don't look like much.

Only after hundreds of days
and even more nights

poring over 4,000 hours of video

did Cat start
to put the pieces together.

So, in this case

the gesture is
a big loud scratch,

but here it means
"travel with me,"

"travel together," so...

The reason she thinks

Harriet's scratch means
"let's travel"

is because she has seen

the same action and response
dozens of times before.

So, you've got Klaus,

the little young male chimp,
and his mom, Kalima.

And he's ready to go,
he wants to travel,

so he gives this big scratch

and then he comes around
the back of his mum, climbs on,

and they travel away together.

I sometimes look at this
and I wonder

if I'm seeing things,
if it's really there,

if I'm... you know,
if it's all kind of

in my imagination,

and it's not until
you're watching the videos

over and over and realizing

that you see
this little movement,

but afterwards, every time
you see that little movement,

the other chimp does something
that you start to think,

"Oh, there might be
something in there."

Like this one.

If you look carefully,

you can see the mother chimp
raising her foot.

In this case,
the gesture is a foot present,

and it means,
"Climb on me."

This is a hard one to see.

You've got the mother, Jenny,
walking down the transect,

and her little boy, James,

who she'd like
to have climb on her back.

What she does is, she stops,

she lifts her foot up,

and she looks back
over her shoulder at him,

so you know
that she's waiting for him.

She's waiting for him
to give the response

that she's looking for,
and, in this case, he climbs on,

and they travel away together.

But it's not enough

to just see the same gestures
over and over.

She needs to see some evidence

of a back-and-forth,
a conversation.

In this case, um...

He wants her to come
and be groomed by him,

so he's going to give

these big scratches...

And he's waiting for a response,

so, that didn't work,

she didn't do what he wanted,
she didn't do anything,

so he here gives
a little object shake,

and he gives the scratch again,

so he's combining
those two gestures,

but still nothing from her,

she's just not interested
at the moment,

so he's giving
a really exaggerated version.

It's like a back and forth
between the two of them.

Big scratch, object shake,

"Come on, I want to groom you,
come over here."

That seems
to have done the trick

because she comes down

and they start grooming.

The reason I know this

is an intentional gesture,

and not just a chimp
shaking a branch in the forest,

is because he gives it,
and he waits for that response.

And when he doesn't
get what he wants,

he gives it again, he persists.

But once he does get
what he wants,

then he stops.

And it's the same
as human conversations

and communications.

After you've passed me
the thing I'm asking for,

then I don't keep on
asking for it.


Cat has come up with

over 60 different gestures

with more than 19
different meanings.


Groom me.

And we're still picking up,
possibly, you know,

finding new ones
all of the time.

Move closer.

I think in terms of an

system of translation...


We probably have the most
meanings translated here.

Let's go.

Let's be friends.

And that's certainly compared

to a lot of other animal systems
of communication,

it's much richer,
it gives us much more detail

than we've been able
to find elsewhere.

Let's have sex.

It's easy for us to want
to focus in on language.

You know, we're quite
self-obsessed as a species,

we want to know what is it

that might be special
or different about ourselves.

But what the chimps
have going on here

is their own incredible,
rich world of communication.

And it would be really easy

just to focus in
on their vocalizations.

But the-the real subtlety

and texture
and all of those rich meanings

that we see
in the gestures every day

would be lost.

I think as humans,
we're so language-centered

that sometimes
it's easy to forget

that language is not
the only way we communicate.

Language is just one of
the channels of communication

that we have as humans,

and also not to think that

language is the only model
for animal communication.

In fact, we humans communicate

with dozens of different
expressions and gestures.

When you think about it,

animals have all this
and so much more.

Everywhere you look,

you find elaborate systems
of non-vocal communication.

From elephant body language...

to honeybees telling
their buddies

how far and where to fly.

Even the simplest of creatures
seem to have a lot to say.

Just consider these guys...

Habronattus formosos...

Jumping spiders.

Spiders have some of the most
unusual communication systems.

It really is kind of like
this fascinating puzzle,

and you use as much imagination
as possible

to sort of crack this language.


Damian Elias from UC Berkeley

spends his life
listening to arachnids.

Stay there...
There you are, come on.

Working with spiders
is not a normal occupation.

You might be surprised
how much this tiny creature

has to say... especially
when it comes to love.

Because these females

only mate once
in their lifetime,

they need to make
that choice count.

That decision better be
as informed as possible.


So you want to know

what spiders are saying,

you need to be
a bit of a voyeur.

Oh my god, I've probably spent

tens of thousands of hours
watching spiders have sex

and hearing spiders have sex

and thinking about
spiders having sex.

So he increases
the thumps as he gets closer,

Now he's really inching up
close to her...

It's like you're
trying to decode

some type of alien language.

It's only with technology
that we have now

that we can even try to

really decode what's going on.

So, what I'm going to do now is

I'm going to make
a female decoy.

I take a euthanized female,

and then take a pin
that has a small drop of beeswax

and then I lower that pin
onto the female.

And now using this decoy,
we can have the male court it.

Here we have
our courtship arena,

which is essentially
a needlepoint frame

and female pantyhose

with some pieces
of reflective tape on them.

I'm going to place
our female decoy

into this little rig

where it's hooked up
to this kind of pulley system,

and with it we can put her
in a lifelike posture

so she can fool a male,

and so how we record
these displays

is using a laser vibrometer.

So, this is what
you see right here.

The laser vibrometer converts
vibration into sound.

Spiders don't have ears,

and so they can't
detect airborne sound.

Instead, spiders detect
vibrations with their feet.

With the stage set,

it's showtime.

Think of it as a five-act
song and dance routine.

Act one...

So, he's doing
a sidling display there,

they're really exposing

a lot of the ornaments
that they have on their face.

And those oftentimes
are species-specific.

Females need to know

that it isn't a predator
that's trying to eat them.

And so you have

this very safe display
that starts very far away,

and as soon as he gets
close to the female,

he will start to do the
introductory display.

Now, the "singing" kicks in.

There... there, there

is the introductory display.

Essentially, it's like,
you know, "Listen to this,

"now I'm going to
start to tell you

a bunch of information
about myself."

And so, now he's going
to go through

a series of different signals.

First, the scrape.

Right there, that's a scrape.

The parasite load is really tied
to how loud the scrapes are.

By counting the parasites

on over a hundred spiders,
Damian found that

the louder the scrape,
the fewer the parasites.

So, what they're
telling the female is that

"I'm healthy,
I'm free of parasites."

Next, the thump.

The thumps are probably

to kind of make sure

to maintain
the female's attention.

And once they've snagged that,

time to put on the moves.

So, the third leg displays

serve to draw attention
to these ornaments

that are on the third legs,

and they kind of like,
shake them around.

And, depending on
how bright they are,

Damian thinks that these
tell the female

about his past.

By being able to correlate

the brightness
of these ornaments

and the quality of their food
when they were younger,

you can say
they're talking about

developmental history
and their feeding history.


And now for the finale.

And we have buzzes,
these long tonal signals

that inform the female
about the male's size.

So, the louder it is,
the deeper it is,

the more the female wants them.

It gets more and more intense.


He might destroy the female.

Okay, yep... okay, all right,
stop, get off.

Get off, get off.

Enough rehearsal.

Now for the real date.

Now we're going
to use two live individuals.

So we can kind of track
what the males are doing

and how the females
are responding to them.

To make sure he's right about
these signals,

Damian has to see how

real live lady spiders respond.

You can see that the female now

is just like really
looking at the male,

really looking
at what he's trying to do.

For the guy,
the stakes are high.

They aren't just singing
for their supper,

they're singing to make sure
they don't become supper.

When females
are assessing males,

they're deciding on whether

they're a potential mate
or whether a potential meal.

So right now, he's buzzing,
so he's getting really close,

so he's really going to...
kind of like wrapping up,

trying to get this female
to mate with him.

Now he's really in

the dangerous parts
of the display.

So, it's getting faster
and faster,

now he's going
to make a copulation attempt.

Oh, the female right there
said, "No way."


It's hard not to sort of
feel sorry for that male.

And this is especially the case
if the male is really

trying his heart out,
and he gets eaten.

Then I just feel
absolutely terrible.


So, where do all these
complex signals come from?

It turns out
what spiders are doing

with all those thumps and buzzes
is completely innate.

They're born knowing
a fixed set of sounds.

What might we find if we looked
further up the food chain?

For starters, it's clear
what we humans do

is very different.

We aren't born knowing language.

Our brains learn it
by listening to others.

This skill
is called vocal learning.

And only a few
other animals have it...

Whales and dolphins...

elephants and seals...

some birds,

and bats.

Like us, these animals have
a flexible communication system.

Scientists think that
if we're ever gonna find

a communication system
like ours,

it's going to be
in one of these.

If you enter a bat cave,

you hear a cacophony.

Thousands of individuals

simultaneously shouting
at each other.

And you ask yourself,

"Are they just shouting
at each other,

or is there more to it?"

Yossi Yovel is a bat biologist

from Tel Aviv University.

Bats are probably
one of the most vocal

and most social mammals
on earth.

Since the moment we're here,

they haven't shut up
for a second, right?

What exactly is the purpose?

Maybe we can say something about
what exactly they are saying.

To understand how difficult
a quest this is,

watch what happens
when Yossi's team

records wild bats.

So this is a very sensitive
ultrasonic microphone.

Here on the screen, you can see

the vocalizations
already in real time.

Trouble is,
there's just too much noise.

Looking at this screen like this

and trying to interpret
what you see would be

just like standing,
you know, in the middle

of a crowd of 500 people
shouting at each other.

So back at the lab,
Yossi's team has created

a tightly controlled
mini bat colony.

Welcome to my bat cave.

So, this is a male.

As you can see,
bats are extremely cute.

They're... some people
would describe them

as small puppies
or small flying dog.

So this is our
controlled environment.

So if the cave we visited
was like a stadium

full of thousands
of individuals,

this is like your living room
with a few friends,

and I can put the camera there
and monitor the full situation.


So I'm going to put him
on this wall

and he will probably
fly to the dark.

But placing bats
in a controlled environment

is just the first step.

To crack the code of bats...

Or any animal for that matter...

Requires making a connection
between action and sound.

And that, as it turns out,

is a real pain
in the you-know-what.

Unfortunately, the only way
to go about it

is to go over a million hours

of videos and just annotate.

First, the action.

What are the bats doing?

Generally, they're annoying.

They're really squabbling,
kind of...

No personal space.

Because bats live
in close quarters,

they fight a lot.

So making a database

of "what's the fuss about"
is key.

This is a-a fight
over food, basically.

So the first bat is holding
a food item in its mouth,

and the second bat is coming
to try and steal it.

In this case,

I would enter the context
of this as fighting over food.

We see a female protesting

a mating attempt by a male,

and it's a failed
mating attempt.

So, in this case,

they're sleeping,
one bat wakes up.

He shifts a bit,

and he annoys
the other bat by him.

It's a painstaking process,

but one by one,

Yossi's team creates a database

of over 100,000 bat spats.

We did this for

several months around the clock,

not missing
a single vocalization.

When you listen
to these vocalizations,

they all sound the same,

but that's because
you have a human brain

and not a bat brain.

So they turn
to the next best thing...

A brain of the silicon variety.

We fed this huge dataset

into a machine
learning classifier,

and if there are differences,

the computer algorithm
will learn these differences.

That is, if there is in fact

any connection between
the sounds the bats are making

and what they're talking about,
the computer will find it.

And after months of work...

We have essentially built
a simple bat translator.

One that can translate

four different bat calls,

even ones it hasn't
heard before.

This one is a fight over food.

This one a female
saying something like,

"Not tonight, big fella."

And this one: "Trying to sleep
over here, knock it off."

You can now take

a new vocalization,

a vocalization that
I've just now recorded,

for example, in my colony.

You can feed it
into this algorithm,

and the classifier
now will tell you

what was the argument about
without observing it.

It's fantastic,

this is Dr. Dolittle's dream,
you know, come true,

when you've cracked the system
and you can tell

that these calls have these
very distinct meanings.

And though they've only decoded
a few bat calls,

it's a start.

Is it possible other animals

are communicating
something bigger?

Much bigger?


I think everybody

is used to hearing
these beautiful,

melodic, lovely songs
from humpback whales,

but it's not always nice
to listen to.

When anyone asks me

how pretty their songs are,

I'm like.

Ellen Garland
is a humpback whale expert

from the University
of St. Andrews.

I have always loved

being by the sea and on the sea.

when I was six years old,

I declared that I was going
to be working with whales.

Whales, like bats,

are vocal learners,
and their songs

are among the most complex forms
of animal communication.

A single song typically

is anywhere from five minutes
to half an hour

just for one song.

So, these guys sing
for hours and hours on end.

Like human music,
whale songs consist

of repeated phrases and themes
made up of individual units.

On average, there's
about 34 to probably 36

different sound types
that we recognize

within the humpback
song repertoire.

And we name them how they sound.

So, moans, groans,
grunts, whoops...

So, we'll call that a trumpet.

We have a lot of low frequency,
very grunty sound...

And sort of ascending shrieks,

I feel like that one
is definitely

going to come back to haunt me.

Which begs the question: why?

Why are whales making
such complex songs?

One clue might be that
only the males do the singing.

Humpback song is really
an acoustic peacock tail.

It's extremely
showy and complex.

They're obviously
communicating with each other.

You sort of want to understand
why they're doing that,

what they're trying to say.

To find out,

Garland embarked
on the world's first

mapping mission of whale song.

I was to analyze song
across the South Pacific region

to try and understand

what the songs were
in multiple populations

through multiple years.

Across the South Pacific,

there are
tens of thousands of whales

living in separate groups.

Until Ellen came along,

no one had ever
compared their songs.

There were so many songs.

I couldn't keep them
straight in my head,

so I started to draw them.

And then from there,
I could actually lay them down

on the floor
by population, by year.

Next, she color-coded the songs.

You can absolutely
tell the difference

between these song types

because they have lots of
different sounds in them

and it's the particular
arrangement of these sounds.

So, this is the blue song type.

Now, if we listen
to the dark red song...

As you can see,
this is completely different.

Scientists thought that
at any given moment,

each group only sang
its own tune.

Well, we thought for a long time

that all the males in an area
sing the same song,

but that it's different
when you go to different areas.

It's different in,
whatever, Hawaii from Tahiti.

So, we expected to find

that all the songs

within a year would be the same.

So I started analyzing

and I started
with the eastly population

of French Polynesia,

and there was some
interesting irregularities

in there, shall we say.

And I was like
"Hm, this seems strange."

Strange because in
French Polynesia in 2006,

not all the males
were singing the same song.

Sometimes the whales
were singing the red song

and sometimes the blue.

And then I went
to the next population over,

the Cook Islands.

And then I got to Tonga.

And then I got to New Caledonia,

and, of course, finally,
to East Australia,

there was sort of a disconnect.

The same songs kept turning up,

but in different places.

I talked with other researchers,

and they were like,
"Wow, I've seen that song type.

What is it doing over there
in that year?"


What was going on?

It wasn't until Ellen
mapped everything out over time

that a picture began to emerge.

Consider the blue song.

In 2002, it enters the charts
in East Australia.

In 2003, it's all the rage
in Tonga.

2004, it's a hit in Samoa.

And by 2005, it's number one
in the Cook Islands.

back in East Australia,

those trendsetters had picked up
a brand new tune.

All of the males

threw the current blue song
out the window,

and started singing
this dark red song type.

And then once
they were singing it,

it was then passed
to the next population over,

which is New Caledonia,

and all those males learned
this brand new song type,

and again and again
across the South Pacific,

so to Tonga, American Samoa,
the Cook Islands,

and finally to French Polynesia.

It's almost a game of telephone
across the South Pacific.

It was kind of like Beatlemania,
you know,

when the British invasion
came over

and transformed American music.

And this didn't
just happen once.

As Ellen dug deeper,

she found that this same thing
happened year after year.

And that was the really big
eureka moment.

The fact that we see
repertoires of song

shifting from one population
to another

across the Pacific
in humpback whales

shows that humpback whales
have cultural transmission.

That's a big deal,
because culture

was once thought
to be uniquely human.


No one knows
how these songs start,

but why would male whales

put so much effort
into switching them?

We think that it's something
to do with novelty.

A novel song makes you stand out

against the background
of singers around you.

You want to be able to stand out
to that female

and maybe you'll get
more matings.

But are they just sexy tunes?

Could there be any lyrics?

So, exact content in them,

what their message is,
that's still unknown.

Could we ever know

if any information
is being exchanged?


It's the same problem faced
by scientists at SETI

who listen to signals
from space,

hoping to find signs
of intelligent life.

SETI is really interested
in knowing

whether or not there are
other beings in the universe

that are intelligent.

And one of the ways to do that

is to quantify
and understand communication.

So why not start by trying
to decode

the "alien tongues"
right here on earth?

Looking at the stars and saying,

"Are we alone?"
I don't think is as useful

as looking at the millions

of other communication systems
that are nonhuman on earth,

and studying them
so that if and when

an extraterrestrial signal
is received,

we'll have a feel
for nonhuman communication.

Easier said than done.

We're faced with a big problem,

which is we don't have any idea

what the meanings
of the sounds are,

so we can't translate them,
we've got no Rosetta Stone,

we can't say this sound means
"fish" and that means "dog."

Think about it for a second.

Imagine you were an alien

peering down on Earth,
trying to decipher

what these odd creatures
had to say.

How would you know
what to listen to?

This music stuff?



When you think about it, we
humans are making lots of noise.

And only a fraction of it

contains information
we call language.

How would you be able
to pick out the right parts?

Well, this is where
the math comes in.

We're really looking for

a statistical fingerprint
for language.

Is there something about the way

that the sounds
have been put together

into a sequence

that is characteristic
of language?

Consider ours for a moment.

In 1945, linguist George Zipf
asked his students

to plot out the frequency
of each of the 264,430 words

used in James Joyce's Ulysses.

He drew a straight line
through it,

and it had a 45-degree,
minus-1 slope.

Oddly, the most frequent word

occurred exactly twice as often

as the second most
frequent word,

three times as often
as the third most frequent word,

and so on down the line.

In the logarithmic scale
that mathematicians use,

it looks like this.

So he thought,
"That's interesting,

what if I take another book?"

Darwin's Origin of Species.

Same thing.

What if I take a Chinese book?

Same thing.

Turns out, every human language
on the planet

follows this rule... from Swahili
to Arabic to Eskimo.

It's called Zipf's Law.

It suggested that
the structure of language

is fundamentally the same
across different languages.

So what about animals?

Brenda McGowan at UC Davis

and Laurance Doyle at SETI...

Yes, the Search for
Extraterrestrial Intelligence...

Wanted to find out.

So they decided to analyze

one of the most intelligent
animals we know... dolphins.

They communicate with

an elaborate repertoire
of whistles.

By categorizing whistles

into what we would call words,
if you will,

and I don't mean
that literally, but the idea

is to categorize signals
into types.

Brenda McGowan had collected
a bunch of signals

and gotten their frequency
of occurrence,

and one morning,
I got up, and decided,

"Well, I wonder
if this obeys Zipf's Law?"

And wouldn't you know...

It obeyed Zipf's Law.

So I went and had a cup of tea,

and then I went back
and did it again,

and it obeyed Zipf's Law.

I was pretty excited.

Because, I mean,
it could have been anything.

I mean, what's the probability

that you're going
to find something

that's a negative-one slope
in another species?

It's, you know,
not only exciting,

but it seems highly improbable.

It's one of
those moments in science

where you're going
"Wait a second,

"dolphins have
a communication system

with potential complexity
as complex as humans."

It doesn't measure meaning,

but it does measure
what they could be saying.

It doesn't necessarily mean

that dolphins have language.

It just means that they may have

a complex communication system
that functions like language.

Which brings to the question...
Do any animals have language?

People have set up language

as being really
the only remaining trait

that separates us
from all other animals.

The trouble is that language
cannot be simply binary,

it cannot be the case
that we have language

and no one else
has even a part of a language.

That goes against
everything we know

about how evolution works.

So, there must be a spectrum

of linguistic ability
among animals.


And in fact,
all the research today

is telling us how much we
share with animals.

Very good!

But a huge mystery remains.

Where does language come from?

Unlike our other features,
like opposable thumbs

or walking upright,
there are no fossils for speech.

The only way
to answer this question

is to dive deep
into the biology,

into our brains, our cells,

and the very genes
that make up you and me

and every creature on earth.

Could it be that we're not
as special as we think?

Many people have been assuming

that we're much more
different than animals

when it comes to language.

When we start
to realize the similarities,

then we start to learn
how we can get at this mystery

of where language came from.

This is the question that drives

Erich Jarvis
at Rockefeller University.

A formally trained dancer
from the Bronx,

he's long been fascinated
by language.

I felt like being trained
as a dancer

trained me
to become a scientist,

because both require
a lot of discipline,

hard work, creativity,
lots of failure

before you get success.

And in the past 29 years,

Erich has had a lot of success,

but the path to get there
was not easy.


I guess my story begins

being born here
in New York City.

We had what one might
consider a broken family.

My father, he eventually
became homeless,

and he was later
killed by a gang

who were killing
homeless people.

So I grew up
with a single mother,

we were not a wealthy family.

Culturally, we were wealthy.

I followed my mother's wisdom
of trying to do something

that has a positive impact
on society,

so I decided I'm going
to become a scientist.


I had to learn that
it is more difficult for me,

because I didn't
have much to compare to.

There wasn't anybody
in my family,

anybody in my friend circle,

anybody in my neighborhood
that I knew was a scientist.

Nonetheless, Erich forged ahead,

delving for answers
about the origin of language

in the brains of songbirds.

This mystery of where
language came from,

ten years ago,
we had very little clue,

But now, we're at the point

where we're starting
to understand

how language brain
pathways evolve,

and the underlying genes
that control that.


Little did he know

a huge clue would come
from a single family.

When we first heard about
the family,

it was the first time
that anybody had found

any genetic change that causes
something specific for speech.

Three generations
of the Kearney family

had difficulty speaking.

Analysis of the family's DNA

led to a gene called FOXp2.

Humans with a mutation
in the FOXp2 gene,

who are otherwise normal,

have trouble
making complex sounds.

They can do "ka-ka-ka-ka,"

but they have trouble producing
complex syllables,

like "condition."

Songbirds also have
a FOXp2 gene.

And when Erich inserted
the same mutation into them,

they too had trouble.

Then the birds
can't imitate properly,

just like in humans.

Even though we're separated
by 300 million years

from a common ancestor,
a gene became used

for a similar purpose in humans
and vocal learning in birds.

Turns out,
all animals have a FOXp2 gene,

but it was assumed that

it only affected communication
in vocal learners.

But if this were true,

why would all animals
have the gene?

Erich wondered if its effect
on communication

could be more profound.

So, he decided to try
the same experiment

in a species that doesn't learn
its vocalizations: mice.

They don't just squeak.

They sing.

When pitched down
to the human hearing range,

actually sound
like songbird songs.

It's amazing.

And like many songbirds,

the males sing
to impress the ladies.

Usually when you put
female with the male,

he produces these complex,
very modulated syllables.

We call them the sexy songs.

But unlike songbirds,

mice are born
knowing their songs.

Our assumption was that mice
are vocal non-learners,

so putting this human mutation
that causes a speech deficit

shouldn't do anything
to their vocal behavior.


If the FOXp2 mutation
does affect mice,

that would mean
the roots of human language

spread well beyond
a handful of vocal learners.

So above the cage here

is a microphone that detects
in the ultrasonic range.

To find out, you need to take
twin mice like these...

Identical in every way
except the mutation.

First, the normal mouse.

I'm going to go ahead
and put him in a cage now

and see how he responds
to this female,

and I'm going to expect since
he doesn't have the mutation,

that he's going to produce
more complex songs to her.

So, here we go.


There he goes.

That's a complex syllable type.

There he goes, see?

So like we have
these pitch jumps here,

from here to here, here to here,

and then these
long syllables like this

followed by a short one.

This is what a normal animal
should be singing.


Now for his brother,

the mouse carrying
the same mutant version

of the gene that affects speech
in humans and songbirds.

Okay, so now I'm going
to take his brother,

who has the FOXp2 mutation,

and I'm going to put him
in the cage.

So our question is
will his mutation

affect his ability
to produce song, and if so, how?

Ah, here he goes, here he goes.

These are more simple syllables.


Here you go... he's singing.

So this guy,
he's behaving normally,

but he doesn't seem
to want to produce

these more complex sequences
as we've seen in his brother.


This female, she's like, "Eh."

So, what you see here

are sonograms of the sounds

that these mice are producing,

and what kind of almost
looks obvious here,

this is the complex song

that the wild type mice sing
to the female.

You take the FOXp2 mice
with the mutation,

instead of doing this,
they do this.

The simple song,

where they have these
simple syllables,

not the same as what you're
seeing in the wild type mice.

So I'm actually even struck more

about the stark contrast
that I'm seeing

in these two brothers,

one that doesn't have
the mutation and one that does.

Everything else about them
is the same.


What it means,
according to Erich,

is that the roots
of human language run deeper

than we previously thought.

Even in a species that's born
knowing its vocal repertoire,

FOXp2 appears
to affect the ability

to make complex sounds.

And it suggests that it's not
a black or white world

of the haves and the have-nots,
it's a continuum.

And it brings us closer
to these other animals,

in our abilities,
in our cognition, in our speech.

I'm not saying we're the same,
mice and humans aren't the same,

we're more advanced,

but we're closer
than what people realize.


Language is like
the last barrier

that we seem to hold
as being truly unique,

so we really sort of have
to change our way of thinking

about what I would
call a continuum

between other animals
and humans.

If we only think about
human language,

we're only focusing on
what might be shared

between human language

and communication
in other species,

we could be missing so much
of what other species do.

I think we've discovered enough

and had enough surprises

to be absolutely sure that we've
just scratched the surface.

And there is
this amazingly complex

and wonderful world to explore,

which should keep generations
of biologists

and psychologists busy
into the future.