El sueo de las aves

Translator: Gisela Giardino
Reviewer: Sebastian Betti

In the early 19th century,
doctor Real founded a model asylum.

To start populating it
he brings five inmates

from Santa Fe to Buenos Aires,
on a caravan,

through the deserted pampas.

This is a novel.
It’s “The Clouds” by Juan Jose Saer.

Towards the end, a massive fire
interrupts the journey.

And they all end up
taking refuge in a lagoon:

The psychiatrist, the madmen,
and a band of animals of all kinds.

Each in their world.

Real in it up to his neck.

He looks at the cows, horses,
birds and thinks

“how hard it is to think like an animal”.

to imagine what’s going on inside them,

how inaccessible all that is
to our reason.

And in that scene Saer makes Real say
a phrase that hit me:

“It’s easier for us to calculate
the movement of a remote star

than to imagine
the thoughts of a pigeon.”

My name is Gabriel Mindlin,
I’m a physicist.

My job was once to study

the chaotic wobble
of one of Saturn’s satellites.

And now I’m researching what goes on
in the mind and body

of a bird when they sing.

I say that Saer was right.

The brain of birds is something
that neurosciences study in depth.

Half of birds are not born
knowing how to sing.

They need to learn it from others.

And that’s why it is
an ideal system to study

what happens to a brain during learning.

When a bird sings different areas
of the brain light up,

and the challenge is
to understand the code

neurons communicate with.

That is, what neurons light up
and in what order.

It is well known that sleep

plays a very important role in learning.

And that’s why neuronal activity
is measured

not only as a bird sings
but also in they sleep.

And we found something really cute:

from time to time, they the same areas
of the brain activate during sleep,

that during the day generate singing.

Of course there’s a mystery here,

because if the instructions are played
but the bird doesn’t sing,

what’s going on?

When a bird sings,

the neurons instructions go down
to the respiratory system

and the vocal apparatus,

where they become sound.

When the bird generates the signals

but it doesn’t sing, it’s because
other neurons are inhibiting them.

So the bird can keep
a calm breathing rhythm at night.

The bird dreams of singing, but quietly.

(Bird singing)

My friend Franz Goller
from the University of Utah,

developed a method for measuring
the activity of those tiny muscles

that control singing.

Those of the respiratory muscles

that send the necessarily air
to generate sound.

And those of the vocal apparatus
that modify the properties of singing.

Collaborating with him
I developed mathematical models,

based on physics,

that allows us to translate
that muscle information in sound.

So now we can, besides
measuring muscle activity,

predict the associated sound.

To show you what I’m talking about,
this is how a real birdsong sounds

(Bird birdsong)

and this is a synthetic song.

(Synthetic birdsong)

It’s so realistic that some birds
mistake it for their own singing.

And this allows us to study
some rather funny problems.

Because we can change
the parameters of models

and generate a variety
synthetic birdsongs.

This allows us to study
what makes a birdsong sexy,

because many species use
birdsongs to court their partner.

In the case of female canaries,
for example, we explore

if more than the size of a candidate,
they’re interested in their ability

to sing certain difficult syllables
very quickly.

Three years ago, I received
an email from my friend, Franz.

You could tell by the tone that
the guy was euphoric.

He told me he’d been
reviewing muscle data

of a bird that had been
connected at night.

And he found that, from time to time,

very similar signals appeared

to those he measures when the bird sings.

Only the microphone
had not recorded any sound.

That is, the sleeping bird had moved
the muscles of the vocal apparatus.

And this is weird because it means

that instructions that were inhibited

and didn’t get to the respiratory system,
did get [to the vocal apparatus].

That is, when birds sleep they move
the muscles of the vocal apparatus.

We both knew what that meant.

We don’t have the code
that allows us to get

from neuronal activity to sound.

But we do have a way
to translate into sound

muscle activity
from the vocal apparatus.

They’re my mathematical models.

While reading Franz’s e-mail
my heart was racing.

Because we had hours of recording
of sleeping bird muscles

and we had the models ready
to synthesize sound.

So maybe we had
at your fingertips

the ability to associate sounds

to those brain activities
generated by sleeping birds.

Would it be that simple?

Would it be a matter of putting this data
in my models and listen to a dream?

The next day I arrived at the lab
trying to stay calm.

My desk is in a common space.

I share it with colleagues and interns.

And that day, like every day,
everyone was doing their own thing.

I didn’t tell them what I was up to,
not to spoil it.

First, I put together this data
that I share with Franz

and I looked for the sounds associated
to that muscle activity.

I prepared an audio track,
I made some mate, I put on my headphones,

and I started to listen

not the silence of the lab
on the night the data was taken,

but what was happening
in another dimension.

What was going on inside the mind
of a sleeping bird.

And I heard things like this.

(Synthetic birdsong)

That’s when I started laughing.

I was laughing, elated, happy.

They asked me what was going on with me
and I wouldn’t answer.

I laughed again.

I knew I was living one of those moments
you will treasure forever.

In fact, that night I got home
really moved.

I thought that
we had come a lot closer

to a species that’s so distant from ours.

One those it’s so hard to know
what’s going on inside them,

like my dear Saer says.

I also knew that this opened
many lines of research.

The frontier between sleep
and wakefulness seems impassable.

How much can we remember from a dream?

How much can we tell of what
happened to us in a dream?

It’s a little bit in the nature
of dreams to be elusive.

This result opens a path
to have an objective record

of what happens in a dream.

We can cross the border
between wakefulness and sleep

and come up with something
very concrete: a song.

All this work was very motivating.

And made of our lab a benchmark
on these topics.

However, if you ask me,
nothing compares to that feeling I had

when, in a little corner
at Ciudad Universitaria,

nature gave me the privilege
of being the first guy in history

to listen to the dream of a bird.