The Evolution of Language


You can also listen to the interview (MP3 16.3MB).


Visit this channel at La Trobe on iTunesU.

Related podcasts

This podcast is part of the series Celebrating Charles Darwin. Please view related podcasts below:


Matt Smith:

Welcome to the La Trobe University podcast. I will be your host, Matt Smith, and I am here today with Dr. David Bradley from the Linguistics Program of La Trobe University. Thank you for joining me, David.

Dr. David Bradley:

Thank you for coming.

Matt Smith:

Now, you're here today to talk about the evolution of language.

Dr. David Bradley:

Yes, that's right, the process through time whereby the physiology of so-called speech organs evolved in the direction of producing speech more efficiently.

So if we look at the physiology of chimpanzees, they have various characteristics that make it impossible for them to make a full range of sounds like we make in modern human language.

We then look at the more recent hominids like Neanderthal and they've started to evolve in the direction of more modern human homo sapiens-like vocal tracts.

For example, the larynx, this box in the neck that regulates air going in and out of the lungs, has moved down in the neck by about 6 cms. Chimpanzees have it right at the back of the mouth. Neanderthals had it about halfway down to where it is now. And we have it, as I say, about 6 cms. down from the back of the mouth.

This means that we have a much more manipulable vocal tract and that we can make a lot more sounds. The tongue for example, has become much smaller if comparing a chimpanzee tongue and a homo sapiens tongue. Human tongues are smaller and therefore more able to make a range of manoeuvres. So we have a physiology that enables us to make much more sounds.

Looking at the brain capacity, again, we have a gradual evolution from apes, like chimpanzees or gorillas who have brains that are about 500 cc or a bit less on average, through to modern humans who have a brain capacity on an average of about 1400 cc, so nearly three times as big and a gradual progression from different hominids.

So different hominids like Erectus had a medium-sized brain, Neanderthal had a fairly large brain, and homo sapiens has, not only on average the largest brain of all the hominids, but also we can see from impressions on the inside of skulls that the Neanderthal brain was not as complex.

It didn't have as many folds that got impressed on the skull. Whereas modern homo sapiens brains have evolved a very complex structure which is adapted from other functions, but which allows us to process very complex structures including, of course, language.

So we have evolved this physiology and also the cognitive capacity that came with a larger and more complex brain. And we've also evolved a more complex mode of socialization. We have much larger societies in which people interact with a large number of people.

In a chimpanzee community, you have a group of 10, 20, 50 mostly females and their children who interact in the way that chimpanzees interact, about 20% of the time consists of grooming. But they don't actually communicate in the same way as humans.

Matt Smith:

Is it possible that changes in culture might have had an influence on our physiology?

Dr. David Bradley:

Well, obviously, if you're evolving in the direction of having a more effective speech-producing mechanism, that can also have consequences which are negative. I'm sure you know about the appendix, which is a vestige that is counter-productive.

We have the wisdom teeth, which are a vestige of us having a much longer jaw. With the larynx that I have mentioned before, we often choke because it's easy to get food down the wrong passage.

With a chimpanzee or a Neanderthal location of the larynx, you can completely close off the air passage from the food passage and you can never choke. But with this evolved lower larynx, we have more problems.

So we have problems that have arisen, which therefore proves that evolution has to have been in the direction of producing speech, because it doesn't make sense any other way.

Matt Smith:

How has speech itself evolved since we've got more of ability to make different noises and produce different sounds?

Dr. David Bradley:

Well, as you know, we don't have any records of Neanderthal speech. But from vocal tract models, it's clear that Neanderthals for example, could not have made all the different sounds like ee and oo and ka and ga, essential building blocks of all modern homo sapiens languages. So the physiology evolved in the direction of producing those sounds.

And then also, as you were asking in your previous question, the societies' becoming more complex made it essential for communication to become more complex and probably with the initial development of homo sapiens with its large brain and modern physiology, languages must have exploded from relatively simple systems among other hominids to the complex systems we have now.

Matt Smith:

Is our physiology still evolving then?

Dr. David Bradley:

Well, I think progressively fewer and fewer people have wisdom teeth, which is evolution in a sense because it means that you're losing something that was essential to crush food but I think, in general, we're not evolving that much.

Evolution in the case of language tends to go in leaps. And if we don't have any documentation of most of those leaps, because we only have indirect evidence: the size of the vocal tracts, the impressions on the inside of the skull of a Neanderthal.

What we do have direct evidence about is experiments that have been made to try and train chimpanzees to use communication systems like human language.

Matt Smith:

I've heard of that. There have been limited success.

Dr. David Bradley:

There's a whole series of experiments that have been taken by place over the late 70 or 80 years.

Matt Smith:

There's Coco the Gorilla that can do sign language as well.

Dr. David Bradley:

There's a long history of, firstly, people who are trying to teach chimpanzees to use vocal language which is impossible, as I said, because of the physical limitations. Then since the 1960s, attempts to teach chimpanzees sign language. And these have had limited success.

When they were implemented intelligently by socializing a chimpanzee with a human family which communicated with that chimpanzee in sign language all the time, that chimpanzee eventually achieved about a 150-word vocabulary and syntax, complex sentences of up to three words.

The sad thing is that the funding for these experiments was withdrawn about 1975, because somebody with power and influence concluded that they weren't working. And then the chimpanzees that had in these projects were retired to a setting where they could live their lives out.

They had one chimpanzee, a baby chimpanzee, that was introduced to this group of chimpanzees who are using sign language because they were taught in human families. That chimpanzee was socialized by these female chimpanzees, and that baby chimpanzee learned sign language.

Matt Smith:

Effectively or just making signs?

Dr. David Bradley:

To the same extent as the other chimpanzees, that is 150 or so words, the level of complexity that a human achieves in about two to 2-1/2 years, but never beyond that.

Probably this is because of the cognitive restriction. You have only 500 cc of brain if you're a chimpanzee, as opposed to 1300-1400 if you're a modern adult human.

Matt Smith:

Did it extend along the lines of inventing their own terms?

Dr. David Bradley:

No, it wasn't productive. It also didn't show several of the other design features of human language, that is we have the ability to talk about things that are not happening now, displacement. We can talk about what happened yesterday, tomorrow. We can lie.

Another design feature, as you say, we can do new things. We can produce an infinite range of possible sentences, most of which have never been said before in the history of the language.

And the third thing, and that's probably the most important, is that unlike a sign language, which has a limited repertoire of, in this case as I said, about 150 signs, all modern languages have the ability to produce strings of sounds put together into words where the sound itself has nothing to do with the meaning of the word. So the sounds c, a, and t don't mean 'cat' except when you put them together. If we look then at the other things that monkeys can do, for example, experiments have shown that monkeys can produce and replicate strings of information. So if you show them a series of pictures then they can replicate that series of pictures. But unlike the human language, which has these sounds that we can string together and can produce them very quickly, you can produce a string of sounds, 20-30 sounds, a second if you try hard. Whereas if you're producing and keying a recognition of symbols, that can only be done at a much slower rate.

So the design feature of having these separate sounds combined into arbitrary combinations to produce words with meaning, makes it possible that we can communicate much more quickly as well as more effectively.

Matt Smith:

So what other aspects are there to the evolution of language?

Dr. David Bradley:

From the perspective of linguistics, we have another way of looking at the history of languages, that is starting from now and trying to trace things back. So firstly of course, we look at written records which, unfortunately, only go back 5,000-6,000 years.

For example of Ancient Egyptian, we have only 6,000 years of records. In the case of Chinese, we have 4,000 years of records. But in the case of many languages, the written records have only developed recently.

So we need to use other techniques. We use what we call the comparative method to put data from a wide range of languages together to try and determine whether they are related historically, that is derived from some earlier original one language. And also, to try and trace what kinds of contacts that might have been between different societies.

When for example, the horse was domesticated 4,500-5,000 years ago in Central Asia, this was a very valuable thing to have because once you have horses, you can get them to drag stuff around. You can ride them. You can go and terrorize your neighbours by riding in with your swords waving and in effect, you can become much more powerful.

So this kind of innovation spreads rapidly. And that kind of innovation is interesting because it leads to the spread of often the word for the thing that you are carrying around, the horse or the new crop or the new kind of metal, whether it's copper or iron, whatever it might be.

And so by using that kind of techniques, we can get back more than 6,000 years. In some cases, some linguists would even claim you can get back to 10,000, 12,000, 15,000 or even 20,000 years ago simply by comparing modern languages to see what kinds of genetic descent they might show, what kind of relationships it might be.

For example, people working on the languages of the Indo-European family which is most of the languages of Europe and of course, also some of the languages of India and Iran and Central Asia, have tried to reconstruct the society that must have been based on the language that was spoken by the ancestors of all these people. What kinds of things that they do? What kinds of animals did they domesticate?

What kinds of crops did they grow? And then which crops did they get from somebody else along with the word for that crop? Which animals were domesticated by somebody else and then passed along to the Indo-Europeans?

So what I personally work on is primarily the Sino-Tibetan languages in China and the Himalayan region. And in this case, we can see things that some things were indigenous to China. That is, for example, rice and several kinds of millet and buckwheat were first domesticated as crops in China or the Himalayas. And so the diffusion of these crops to other areas of the world had to be from there to elsewhere. And therefore also, very often, the words for things like rice or buckwheat might have come from a language of that region.

Matt Smith:

OK, is it a remnant word from a dead language?

Dr. David Bradley:

You can't really say that a reconstructed language is a dead language in the sense that we really have a record of it. It's a hypothetical construct that's been built up by comparing a wide range of modern languages.

So for example, if we look at word that was probably used for Setaria Italica, a particular kind of millet which was first domesticated in China about 8,500 years ago, we see that the word for this in modern Chinese reconstructs back to a word something like 'tsik' maybe 3,500-4,000 years ago in Ancient Chinese. And the same crop reconstructs back to a form something like 'tsap' in Burmese which, of course, only goes back 900 years in its written form. So that's had time to change quite a bit more from the original Sino-Tibetan form.

And then we have a Tibetan form, something like 'cree' which again, has undergone a long series of changes because Tibetan only started to be written about 1,300 years ago.

So the written data enables us, by comparing that and also comparing other words with the same meaning from related languages to reconstruct back to an earlier form which was probably something like 'tsik,' 'tsap' or something in between and which we find cognates for, that is words that are derived from this and all of other language of the Sino-Tibetan family.

And then also, often, words for that same crop would've gone with the crop when the seeds were handed over to the people next door, to the Mongols.

Matt Smith:

That sort of thing became more common in the last 1,000 years or so where you got place names that still remain the same. You've got raccoons and moccasins over in America and those are remnant words.

Dr. David Bradley:

Well, English is a particularly good at sucking up words, usually borrowing when you're talking about loan words. But when you take a loan word, you're not stealing anything from the other language. They still have their word.

You know the words that came into English from Old Norse when the Danes came to Northeastern England in the 8th, 9th century; the words that came into English from French when the Normans conquered England in 1066; the words that came into English from Arabic during the Renaissance; mathematics and many other areas of science were transmitted from the Arabs to the Europeans; all the Greek and Latin stuff; and as you say, the American Indian stuff; Japanese; during the colonialism in India, Indian words made it English, 'calico' and 'sob' and so forth.

And it's not enriching or impoverishing the language that is the source. It's just taking something that you get from another culture and using the word for it from them.

Another good example of that is 'corn' which we know was only grown in the Western Hemisphere until 500 years ago. There was no corn in Eurasia or Africa until the Spanish brought it back with them from Central America about 500 years ago. And yet it's spread so widely that now we think of it as a fairly basic crop.

In fact, it's very new and therefore, the words for corn have to be new words, as well. The English word for corn used to be 'any kind of grain' and now, it's been specialized to be one particular kind of grain, namely maize or as they say 'ma-is' is the botanical name.

And in other languages, similar sorts of things happen. You take an existing word, change its meaning and then you have a new word. But anyway, you can either borrow words or you can create new words from your own resources when you borrow something from someone else.

Matt Smith:

Can I ask you did Darwin at all look at languages and how languages evolved?

Dr. David Bradley:

I think he looked at animal communication systems but not in great detail. I don't think he would've been looking at the historical linguistics aspects for several reasons.

One is linguistics has not been really considered as a hard science until fairly recently. Another reason is, as I've said, linguistics can only realistically try to get back 10,000, 15,000 or Mattimally 20,000 years. And that's really just seconds in terms of human and general evolution So we can't really compare.

As I say, linguistics starts from now and looks back and tries to see what kinds of relationships there were, both genetic descent from the same original language and contact things that go into languages from other languages. Evolution is looking at a much greater timescale.

Matt Smith:

So in a sense, you're just continuing to evolve his theories, which is rather fitting.

Dr. David Bradley:

Oh, yes. And as I say, we obviously are the product of a long string of evolution that's led to us having a much more adaptable vocal tract that produces speech, and a much complex cognitive processing system that enables us to do things that as we have seen, chimpanzees just can't do.

Matt Smith:

Dr. David Bradley, thank you for your time today.

Dr. David Bradley:

Thank you.