Good morning everyone.

Thank you so much Haizhou for your

very kind introduction.

And I also want to thank the organising committee

for giving me this chance to comment,

interact with colleagues from so many different fields, from so many different areas.

Speech

is a fundamental defining characteristic of our species.

Sometimes

our species is called

Homo sapiens,

but I'm not sure we deserve the word sapiens.

More likely,

some people said, we should be called

Homo loquens (talking man),

because

we're talking all the time, we're using speech all the time.

Because it's so fundamental, it's also

extremely complex.

I'm reminded of the cartoon that

Ray Liu showed yesterday

of an ocean of knowledge and

a few islands popping out.

Our job

is to build bridges

among those islands.

An island of electronic engineering,

island of linguistics,

computer science,

psychology,

neuroscience.

Each one

gives us an important window on speech.

And we need to connect them together.

The theme of this

conference

is the Diversity

of Spoken Language.

So this is the outlining

of my remarks today.

I will begin by asking, How is speech possible?

and I'll say something about

African origins,

the diversity that we see today in genes

and in words,

then I will proceed to

present three

case studies

of sound patterns in motion.

Sounds

hardly ever

behave as individuals.

They always

configure in patterns and move in patterns.

And I will

discuss consonants.

How they moved

with respect to Grimm's Law.

I will talk about vowels and

the Great Vowel Shift

in the history of English.

And I will talk about something that's not

found in all languages. Consonants and vowels are found in

all languages.

But in several thousand languages of the World you find tones.

And I will say something about how tones move.

And if there's time, I'd like to say something about speech and music.

Two cultural universals.

Two

cultural aspects that's present in every group

of human beings.

And I will close off.

I saw this cartoon few months ago and I thought maybe it's relevant.

Here you have one dolphin talking to another.

Although humans make sounds with their mouths,

occasionally look at each other,

there is no solid evidence that they actually communicate among themselves.

If somebody came up to me and looked at me and spoken Tamil.

I wouldn't understand him.

Or

Swahili,

or Caucasian,

or Abkhaz.

So in that sense our species is unique.

There is no other species

that have such great systematic diversity

among system of communication.

Diversity of language

is something unique to our species.

How did this

diversity come about?

Well, here's one story.

Lori Lamel

yesterday also

showed this beautiful painting by

Bruegel.

It is based on the story of The Book of Genesis.

God was annoyed

that people

had the hubris and

the arrogance

of building

a tower so high to reach the sky.

So God said, Go to, let us go down,

and there confound their language,

so that they may not understand

one another's speech.

Actually people all over the world

ethnographies told us,

have a very similar accounts

of creations and origins and so on.

My purpose this morning is to look at

diversity

more from

evolutionary perspective.

The key concept

behind diversity

is innovation.

In biological evolution

every generation of giraffes

has a certain degree of variation

in how long their neck is.

Every generation of wolves

have a certain variation

in how long their canine teeth are for fighting.

So nature selects,

there's variation.

As you innovate,

every generation has its variation.

And nature selects.

Similarly language.

Every infant,

in trying to reconstruct a language based on the

sounds, based on the language context in which it's born,

innovates.

And these innovations

mostly go by the wayside.

Many of them

are selected culturally rather than biologically

and persevere.

And as these innovations build up, accumulate more and more,

pretty soon

you won't be able to talk to each other anymore.

So, in biological evolution the innovations build up.

There's no longer mutual

reproductive

fertility.

You get a new species

in speech.

As these innovations become numerous,

build up over generation after generation,

you lose mutual intelligibility

and you have a new language.

So, the sounds as I said earlier

are not individuals.

They form natural patterns

and together

they

move in these patterns.

As these

innovations take place,

the social values attached to them.

Many of you may have enjoyed this great movie My Fair Lady.

Henry Higgins in the movie was actually based on the real English phonetician called Henry

Sweet.

But in the movie of course they took various liberties.

Here's Henry Higgins saying,

An Englishman's way of speaking

absolutely classifies him.

The moment he talks

he makes some other Englishman

despise him.

One common language I'm afraid we'll never get.

Oh, why can't the English learn to set a good example to people whose English

is painful to your ears?

The Scotch and the Irish

leave you close to tears.

Maybe that's why Scotch want independence.

Anyway according to Higgins,

There even are places where English completely disappears.

In America, they happen't used it for years!

Each one of us has to make a whole series of personal decisions when we

talk.

When I went to Hong Kong,

I noticed that in the textbook

to say you

in Cantonese

it's nay

with an n.

You find n in most of the Chinese dialects. In Mandarin it's nee

So I was told it's nay.

But as I

walked around the campus,

I noticed the young people don't say nay they all say lay.

The textbook says woo is gho

whether ??real/rear or nasal.

But the young people don't say gho, they say who.

So should I say

nay?

Closer to my native dialect

or should I try to be fashionable and say lay?

Lay is kinda not my age.

So, there are all these personal decisions

and these decisions build up

for group solidarity,

for group distinctions, and so on, and on and on

over the generations

new languages arise.

But let's take a look much for the back in time.

When did speech get it's first shot?

When did speech

start its trajectory?

I think a good case can be made for

over three million years ago,

okay.

In the region in Ethiopia

a bunch of physical anthropologists

were very lucky

and discovered

a relatively complete fossil of a

young woman,

middle teens,

called Lucy.

By studying Lucy

in great detail,

her hip bones

her ankle bones, her scull and so on,

they determined

that Lucy walked.

Lucy had erect posture,

essentially as we do.

So Lucy had made the transition to bipedal posture.

And

this is the reconstruction

of

Lucy.

This by itself, of course, is only part of the evidence.

Another part of the evidence which is also quite solid

is that they found, around the same time not very far from

this region called Afar,

footprints

covered by volcanic ash, so it's nicely preserved.

And here's the great

anthropologist by the name Mary Leakey

measuring in great detail

a whole string of footprints

and these footprints proved beyond doubt that over three million years ago

our ancestors were walking.

What are the implications of walking?

Well, in this excellent book written by Daniel Lieberman

called The Story of the Human Body

he contrasts

the

skeleton structure of the modern human

with that of the chimpanzee

and indeed there are many differences, many adaptations.

For us

perhaps the most important

is the restructuring of the entire upper body

and the head.

Rather than hanging forward

and help by muscle

the head's resting squarely

on the

cervical vertebrae,

or the spinal column.

It's

relatively

effortless for me to stand here,

not for a chimp.

I can lock my knees,

my skull is resting, its weight is on the spinal column.

There are many implications

of this new posture.

And immediate one of course, it freed our hands.

We became

tool makers in a serious sense.

Another very important adaptation

is that because of the restructuring of our heads

our larynx

has descended.

This is from a book

by Fitch, professor at Vienna.

And it shows the descent of the larynx

as well as the hyoid bone,

here's the thyroid cartilage and so on.

Even more clearly,

Daniel Lieberman's book,

makes it

very obvious why this is important for speech.

In the chimpanzee for instance whereas there's essentially just one resonance tube

to produce vocal sounds,

in our case

with the larynx

lowered

here's the larynx with the larynx lowered,

we now have two tubes.

A tube

that's the mouth

and another tube that's the throat.

And with the agile movement of the tongue

you can form much greater variety of phonetic distinctions.

And this of course is also extremely important,

but it's useless to have all that hardware

unless we have the proper

control mechanisms

and then control mechanisms

right in the brain.

Usually we see pictures of the

hemispheres of the brain,

but here's a nice picture

of the base of the brain.

If we lifted up the head

and looked at the brain from bottom-up,

we'll see a picture something like this.

Here's the temporal lobe, here's the frontal lobe.

The point that I wanna make

is that most of the cranial nerves,

twelve pairs coming out from the base brain,

most of them

are involved in speech.

That's how important the speech is

for

our life.

Henry Lenneberg,

some years ago,

published this

very nice diagram

showing the various nerves.

Here's the facial nerve,

which is cranial nerve pair number seven.

Here is (a) which is the trigeminal nerve

and perhaps the most

noticeable one

is this one that starts from around here in the base stand,

goes down all the way,

curls around the aorta of the heart

and goes back up

to control the larynx.

This is of course an effect of ascending up.

So five hairs of the twelve pairs of cranial nerves are involved in

controlling speech,

jaw movement, lip movement, larynx, tongue and so on.

So, probably

given both the

software in the head,

hardware in the throat and the mouth,

by about a hundred thousand years ago

our life

took a very important turn.

We left Africa.

Leaving Africa

is a hypothesis

that even Darwin mentioned

in the,

1871 book, Descent of Man.

But it's only within recent decade that the evidence become so strong.

The earliest evidence came in the 1980's

with mitochondrial DNA. Mitochondrial is only materially transmitted

and then came a whole spate of studies based on y-chromosome,

which is paternally transmitted

and now

they can do a huge amount of study

based on

single-nucleotide polymorphisms.

And it's largely

these genetic studies that leads to this picture

which was published in PNAS

a couple years ago

by a group of a geneticists at Stanford University.

So the scenario,

the best knowledge that we have for now,

is that humans

originated in Africa.

The earliest fossils take back to about hundred and fifty thousand years ago.

And

various genetic mutations took place

at around the same time. For instance

we have a particular version

of a gene called FOXP2,

okay.

Our version of the FOXP2 evolved probably

also around that time.

A whole bunch of things happened around that time

and

emigration started.

First to Asia,

back to Europe,

down to Oceania,

across the Bering Strait into the Americas.

And very quickly

all the way down to the, tip of South America, Tierra Del Fuego.

So here is an example of a genetic study,

done essentially by the same group at Stanford,

showing

how the

SNP, single-nucleotide

polymorphisms are related.

So here's the root of the tree

and in constructing these phylogenetic trees

often you have to take

outgroup.

Outgroup in this case is the chimpanzee.

Chimpanzee diverged from us about six million years ago.

So,

the chimpanzee provided a good

routing of the tree.

These are all Africans

and then step by step populate in the world.

Here's Asia,

here's the northeast,

here is

South America

and here are the

Pacific Islands.

What about language?

Obviously one people left.

They brought their languages with them.

And another

Stanford scholar professor Joseph Greenberg, was

actually my teacher,

classified

all of the seven thousand

or so languages of the world

into these major

superfamilies.

So, for instance, again just as with people,

languages have the greatest diversity in Africa.

A very successful

family of languages

is called Eurasiatic.

Europe .. Asia. Eurasiatic.

And within this family

there's one sub-family that's

extremely

dominant. That's called Indo-European.

Indo-European because at one end you have India

at the other end

you have Europe extending all the way to Iceland.

So English

is Eurasiatic,

because English

is West Germanic. Germanic .. Indo-European ..

Eurasiatic.

Going up the family tree.

Another language here .. Tamil.

Tamil is not related to any of the Eurasiatic languages.

Tamil is a member of Dravidian family.

Earliest forms of Tamil actually

were found

much more

northwestern.

Where it's not spoken.

It was founded in the Indus Valley.

With the coming of the Indo-Europeans

they were push further and further south, so that

Tamil is now in this small region here.

Malay as we saw in the last slide

is an Oceanian language

that

has a relatively recent history,

but what about Chinese?

Chinese belongs to a family called Sino-Tibetan.

Chinese is related to Tibetan.

Well, recognized

over a hundred years ago.

And

more recently

people like Greenberg say Chinese

is Dene-Caucasian.

Chinese is here in red,

but there are spots of red

all over the world.

According to several very eminent linguists, Greenberg's one of them,

Sergei Starostin in Moscow is another one, okay,

say all these languages are related.

Chinese is related

not only to Tibetan,

but also to the Yeniseian languages.

But not only to the Yeniseian languages,

but all the way across Bering Strait

the Athabaskan languages.

Including for instance

Navajo.

So,

together with genes, archaeology and so on

we are getting a better notion

of what we come from

and speech plays a very important role here.

Given all these languages,

how do we classify them?

Well, one way to classify is very straightforward, you know, where is it spoken?

Geographical.

This gives us a geographical distribution.

Another way is

typological.

What kind of structure does it have?

Does it have a gender system?

Is indirect object before the direct object?

Or is the other way?

Does it have tongues?

These are typological features.

So, this is a second way of classifying languages.

And the third way of classifying languages,

what is it related to?

Is Chinese related to Tibetan?

Is Tamil related to Telugu?

Which it is.

So on.

Sometimes

genetic or

historical relationship

and typological relationship

classified differently.

English for instance, as I said a few minutes ago,

is the Germanic language.

But it doesn't look anything like German.

English does not have chen (suffix),

a gender system.

German has three genders: masculine, feminine, neuter.

In English the primary

order for declarative sentences .. you put the verb in the middle,

I saw John.

In most of other Germanic languages, including German,

in declarative sentences

the verb is at the end.

All these structural differences.

So, when we want to use corpora of languages

to

use cross-linguistically

it's primarily the typological

classification that's important.

Not the genetic relation that's important.

Well, let's ask the question,

How did all this

diversity arise?,

Where did all this diversity come from?

For instance

here's a distribution by

area. This is the geographical distribution.

This was shown by an earlier speaker too.

According to the Ethnologue which is

essentially

missionary organisation where the team of thousands of linguists

travelling far-flung places of the world.

With one of their missions to translate The Bible

and in this process

they live there, they know the region well, they can report on the language and

according to the

Ethnologue there are seven thousand or so languages.

We shouldn't take these numbers very seriously. They are very good

rough guide.

For instance

when I went to Hong Kong, I couldn't understand the

word of the Chinese that they spoke there .. Cantonese.

Cantonese and Mandarin are supposed to be Chinese.

One language.

If this is so, then the

dozens of dialects in China only count once, even though they're mutually unintelligible.

On the other hand one year I was teaching in Stockholm.

Bought myself a little Volvo,

drove across from Stockholm to Oslo.

Swedish .. Norwegian.

No problem at all communicating.

Yet there we have two languages, okay.

So these numbers

just can't be taken too seriously,

but it's good to have kind of as a

guideline.

Was it

also this complicated at the beginning when we first heard language?

What does this diversity come from?

Well, linguists have two ways of talking about the origin.

Either

monogenesis,

all of the languages in the world

came from a single source

or at the beginning

there were various tribes in different regions

of the world maybe,

of Africa maybe,

and each of them came upon the idea of inventing language.

And some years back

I collaborated with the

mathematician, friend of mine at Berkeley, David Freedman.

And we did some statistical modeling

and

our thought

that is

much more likely,

given you have

a large number of sites,

so much more likely

the language arose

polygenetically.

This means, if this is right,

that some diversity was there

right at the beginning.

And this is not completely unreasonable because

you think about the invention of fire

that's had terrific consequences.

That was polygenetically invented.

Many people invented them independently.

You think about pottery.

Partly another major

prehistoric invention.

Independent.

More polygenesis.

You think about written language.

Many cultures independently invented written language.

So, spoken language, you know, not in the nuanced

powerful sense of spoken language now.

Spoken language

hundred thousand years ago

could have been invented polygenetically.

One person who took the monogenetic

approach

was Quentin Atkinson

who published this paper in Science few years back.

He

use the

corpus

world atlas of language structures which is maintained in Leipzig in Germany.

And looked at five hundred and four languages

in terms of their phoneme inventories.

How many consonants, how many vowels and so on.

And you can tell from the title of his paper

Phonemic Diversity Supports

a Serial Founder Effect Model

of Language Expansion.

In another words

he's trying to

construct the picture

kind of like this that we saw earlier for genes.

A serial founding effect.

He wants to do it

for languages and he does it

with this diagram

that you see on the

left side of the slide

from where you seti.

Without going into great detail of Atkinson's methodology,

I should mention that

his proposal was met with several very severe

criticisms

from three groups of scholars.

Some in Europe, some in America and so on.

And they are listed on the slide.

For instance

Cysouw

Dediu and Moran,

in the later issue of Science,

contested

both their databases

and their method

and came up with a very different picture

based on phoneme inventories.

Atkinson's

hypothesis would put

origin of language here, but actually there are many other sites.

In order to give us a clearer idea of how sound patterns actually move

I will now move to

three case studies.

Atkinson took the seven thousand languages of

global perspective, all the languages of the world.

I think perhaps we're not quite ready

to be that global.

It might be safer to stick with one family on which there is a very

detailed meticulous scholarship

that has been going on

since

the

eighteen century, latter part of the eighteenth century.

So, in Science 2004

we have a

picture of the Indo-European languages.

I'm not sure

whether you can catch everything

but there is not an English.

Traces to

Middle English of Chaucer,

Old English of King Alfred

and all the way up to Indo-European.

Indo-European also includes

Tocharian.

Which is an language that's become extinct.

You find remnants of it in western in China, in Xinjiang.

And you have

the

Indo-Iranian branch. Within the Indo-Iranian branch

you see Kurdish,

which is very much in the news these days. They're the ones fighting in Iraq.

So, Kurdish is actually an Iranian language belonging to the Proto-Indo-Iranian family.

This gives you an idea of the membership

doesn't tell us very much about the time scale.

In Nature in 2003

Russell Grey

and Atkinson

did an extensive

statistical analysis based on vocabulary.

They took almost two thousand five hundred words

from eighty seven languages,

all of from Indo-European and constructed this tree.

Along this tree

you see estimated time of death.

So, according to these people published in Nature in

2003

Indo-European

is about nine thousand years old.

As it gradually split

into

more language families

based on innovations,

okay.

A few thousand years ago

it

led to Germanic languages.

Eventually to English.

Those are

the,

the slide is kind of hard to see,

so I will take this section here

and blow it up

and we have this.

So, here we have English,

here we have German,

and we have the Romance languages,

Italian, Spanish,

and a separation here seems to be about

five thousand years ago.

What I'd like to

discuss now

is what other innovations

that led to just

this one little group.

How did

Indo-European

become Germanic

in a part of the world?

And this is how linguists operate, okay.

Here's some data.

Colin Renfrew is very esteemed archaeologist at Cambridge University

and he gave as a bunch of words.

The integers one to ten

in a bunch of languages.

English one, two, three, four, up to ten.

Japanese

hitotsu, futatsu, mittsu, yottsu.

Doesn't look anything like the rest of them.

So, Japanese

it's not an Indo-European language.

But the rest.

English

two

corresponds to Gothic.

Gothic is an extinct language but we know it because there's lot of textual material

preserved.

Gothic has twai.

Two ..

twai.

Why isn't the w pronounced in two

but is still written?

In Latin is duo.

In Greek

is duo,

in Sanskrit

it's dva.

In Russian is also dva.

Obviously there's similarity in there.

But is there something systematic that we can extract?

Yes, the systematic aspect is that

Germanic,

English and Gothic

T

corresponds

to other Indo-European languages'

D.

Let's take the next one.

Three

th is just a abbreviation

referring to the sound that Anne Cutler was talking about interdental fricative θ.

Three that's a interdental fricative.

So, here's a fricative, here's a fricative in Germanic

and these fricatives correspond to T here, T here,

T here,

systematically

as a pattern.

Without going through many more

this is the correspondence that we can extract

and

here's the great man

who formulated the law

Jacob Grimm.

Older brother of Wilhelm Grimm.

The brothers of the Grimm's Fairy Tales, okay. Jacob was also a great linguist.

In 1822

he published what's called the Grimm's Law.

He said

these

words

in Indo-European

pronounced with a

B(h)

D(h), G(h) type of sound,

became in Germanic

P,

D and G.

Sounds which were B, D, G in Indo-European

became P, T, K

in Germanic.

And sounds which where P, T, K in Germanic,

in Indo-European, sorry. Became F, θ, H

in Germanic.

Very regular.

We are able to look at some of these in comparison because

after the Norman conquest, about a thousand years ago,

English powered a lot of romance words.

French words, Latin words.

So we actually have

Latin and Romance words sitting side by side with the native Germanic words.

So take a word like

ped,

the pedal

on a bicycle.

It's foot.

So the P corresponds to the F.

P corresponds to the F.

The pad, the D

corresponds to the T.

D corresponds to the T.

So, the point that I wanna make with this slide,

I better move faster,

is that

it's a whole natural class that's displaced.

It's displaced not only as total patterns

but it's usually displaced

by single features.

As you go from here to here

you're going from

voice task ??

to voice ??.

As you go from here to here

you're going from voiced to unvoiced.

As you go from here to here you're going from stop to fricative and so

on.

It's

beautifully symmetric.

So, let's get closer to home.

Here's the sound change that took place.

It sided shortly after Chaucer's death

and it kept on going after Shakespeare.

It's called the Great Vowel Shift.

It's been studied in ??Greatstepft by Danish linguistic called

Otto Jespersen.

And some years back I also

followed up that study.

Why is it that we say

sanity,

but sane.

Why is it that in French

China's called Chine?

Chine versus China.

Christmas

versus Christ.

Why are there all these alternations?

It's because of the Great Vowel Shift.

And to hear one example.

The original vowel quality was actually a.

It used to be sanity.

But if you take away the suffix,

which protects this vowel,

the vowel itself raises.

So, it became sane.

So, we still say sanity

but we no longer say, he's insane.

We have to say insane.

As wind shift

we can say serenity

but we cannot say, this man is very serene.

We have to say serene

A has become E, E has become I. I can't go any higher.

So,

Christmas, the I,

has become Christ.

Chine

in French has become China in English, because of this vowel shift.

Again it's a systematic shift in patterns.

Okay, let's

spend a few minutes on

perhaps less familiar territory.

You build words not only with consonants and vowels

but also with tones.

People think of tones they think of Chinese

but there are thousands of tone languages in the world.

Here's a tone language called Trique,

in Mexico.

And in Trique

there are five levels of tones.

So, we have,

this is,

no,

gu du we ku.

Gu du we jo.

And so on.

Each with a very different meaning.

So, ku is bones.

Jo, ka, ?a, za.

Five different level tones.

With five different words.

But is it the case that

once you have a tone language always a tone language? No.

The paradigm tone language Chinese. People thought, Well, Chinese is a tone language.

Was it always a tone language? No.

Chinese became a tone language probably about

two thousand

twenty five hundred years ago.

And this is the result of

the loss of consonant clusters.

In Enghlish you have a lot of consonant clusters like

speech, play, spring, these all begin with consonant clusters.

But if you look at the Chinese dialects now, any dialect you want look at,

no consonant clusters.

So, how do we know that there was consonant clusters?

Chinese is not written alphabetically. How can you tell what it sounded like?

Well, it's true Chinese is not written alphabetically but a large part is written

phonetically.

It's not in terms of consonants and vowels. It's in terms of syllables.

So, for instance

here's,

I call these things sinograms

for Chinese characters. Here's a sinogram

that serves

as the phonetic

of this more complex sinogram.

Here's a sinogram

that serves as a phonetic.

Sinogram that serves as a phonetic and so on.

But then you ask me,

How is that phonetic?

They're different.

These are

spelled with a letter G but actually there's K sounds there.

Unvoiced un-aspirated velar stops,

okay.

These are all L.

Well, that's one of the

evidence

that they were consonant clusters.

If you take these words,

compare them with Tibetan.

Many languages in Southeast Asia

they also have constant clusters.

So, because these consonant clusters were lost,

tones arose.

Otherwise you'd just have too much ??.

And they arose in different ways.

So, here's Beijing,

Xiang, Xi'an, ??Hankou\Hong Kong.

and so on.

Each one

with the different number of tones.

In Mandarin they're four.

In Cantonese there are none.

So, this illustrates the four tones of

Mandarin.

This illustrates the nine tones of Hong Kong Cantonese.

Notice in Mandarin

because of the merger

??

??

??

have all become homophones.

Exactly the same pronunciation.

But in Cantonese

one is mid-level

??

??

??

three different tones

they kept.

And if you

did F0 analysis to extract the pitch.

This is actually.

my own voice

extracted from a PDP computer. Many of you may not even remember what PDP computers

are.

Back in 1973.

So, it is my voice saying

??

??

And here are the nine tones

from Cantonese.

You could plot these

on a

two-dimensional graph

using slope

and normalized height.

And you can see that the four tones in Mandarin

??

are quite evenly distributed.

On the other hand in Cantonese

it's kind of a mess.

There's a lot of overlapping

and this happens a lot whenever

a metropolitan centre

suddenly takes in a lot of influx of people

speak different dialects, different languages. The language will change faster.

So, as a result

in Cantonese tone two

and tone five

are merging.

Many of the words

are no longer distinct.

Tone three and tone six

are merging.

Many of them are no longer distinct.

I see I've ran out of time, Haizhou.

The last point on this tone part,

I'll just mentioned very quickly,

is that we saw a little bit earlier the vowel shift changed the vowels chasing

each other around, okay.

Low become middle, middle becomes high, high becomes

??deflunguised

Well, you can't say things in tones.

In Taiwanese

okay, here's

database on the Ministry of Education in Taiwan,

if you take each tones

and notice how they change.

They actually change in circle.

And detailed study of this

is available

in the literature.

The last point that

I wanted to discuss, but I can discuss with you individually,

is that many people think that language actually arose

from music.

The proto-language of the humankind

was actually singing.

Darwin himself

made many remarks to this

point.

So, in English, archaeologist Steven Mithen,

wrote the Singing Neanderthals

Origins of Music, Language, Mind, and Body.

And the linguist Fitch at University of Vienna

in his book Evolution of Language,

chapter fourteen

is all about

proto-language.

And here's a slide prepared by a graduate student at The Chinese University

showing

the

grammatical operators,

like drawing this type of trees,

applies very well

to music.

So, here's one of Bach's Menuets that

has

syntactic structures very much like

real sentences.

And this last slide shows that

actually

if you look at it just from the point of view region of interest with

fMRI

they're actually

distinct regions from music and language.

All this is very interesting, not too surprising

but we really need to go further

and look at the fiber tracks

not just the regions.

But how these regions are connected.

Okay, I have two more slides.

So, the roots of language then,

as a summary,

reach back over three million years,

when our remote ancestors transitioned

to bipedal posture,

restructuring the hands,

the vocal tract,

and the brain.

Speech

with its building blocks of syllables,

vowels,

consonants,

is a powerful vehicle for language,

emerged over a hundred thousand years ago.

Language and music are both universal to our species

and share evolutionary roots.

They have similar functions of communication,

similar principles of organisation.

Diversity in language is the cumulative product

of culturally selected innovations

made by numerous generations of speakers.

Spoken language has spawned

various auxiliary forms,

such as written language,

signed language, there is electronic media providing

additional windows for studying how we communicate.

And ever more powerful

technology of brain imaging and

computer analysis for spoken language and music

is already shedding much light

on our mind

and promises

to reveal

much more.

Thank you.

As much as we are a bit overtime I still welcome very short comments

after a mind-boggling talk that perhaps

bring back

speech and language

all the way to the roots.

If not

on behalf of Interspeech

we have a token of appreciation

that we would like to ask

the conference chair Haizhou Li to present to professor

William Wong.