The link between evolution and language | Richard Dawkins

0:06You're listening to TED Talks Daily, where we bring you new ideas to spark your curiosity every day. I'm your host, Elise Hu. Across species, communication shapes literally everything in the world, according to the legendary evolutionary biologist Richard Dawkins. In this fascinating conversation with linguist and TED guest curator John McWhorter, Richard dissects why communication is the thread that ties us all together and how our genes may be the ones pulling the strings.

0:41so Richard let's just get right started as people say at the beginning of interviews like this I want to explore and I think you do too the fact that if linguist me reads the selfish gene I'm always thinking oh this is just like language and you have an interest in how the evolution of language is similar to the evolution of creatures. And so just to explore the parallels, the first thing that I've always wanted to ask you is about excess. And so in any language, any language, the language specifies more things than it needs to. If you speak that language, you think of it as normal, but it isn't.

1:21And so for example, if you learn about the future tense in English, You're taught that we say will. I will buy you some socks. But if you think about it, that's not a sentence. When would you say, I will buy you some socks? That would only be at the culmination of an argument. Really, more likely would be, I'm going to buy you some socks. Or you could say, I buy you some socks tomorrow, which gives it a sense of event. Or I shall buy you some socks, which doesn't really mean anything, but yet it's another way of saying it. Our future tends overdoes it. I've often thought so much of language really doesn't need to be there.

1:59No language needs to be as picky about the future as English. Do creatures, do plant, plantures, do they overdo it in the same way? And if so, why? I suppose you could say that poetry is something extravagant and overdone. And if I think about the nearest approach to poetry in wild creatures, it might be something like A Peacock's Tail, where it's far from utilitarian. The idea is that the male is attempting to seduce a female, and there's massive overkill there. There's too much. It's too much. When Charles Darwin proposed his theory of sexual selection, his co-discoverer of natural selection, Alfred Russell Wallace, didn't like it, because he wanted it to be more utilitarian.

2:51and Darwin was happy to say simply it's female whim. Female simply like this extravagant thing. This argument wasn't really settled until the 1930s when R.A. Fisher pointed out mathematically, well intuitively mathematically, that if you make certain assumptions, there is a runaway process, an exponential runaway process as natural selection works upon female taste, genes of female taste getting more and more extravagant, male tails get more extravagant, and this runs away exponentially to produce this ridiculous, extravagant advertisement. There are other theories, but I think that's perhaps the nearest approach you get to what you want.

3:44And the only thing that could stop it is if the tail became so large that it would interfere with. Eventually, utilitarian considerations will interfere. I sometimes wonder whether, I think Jeffrey Miller has suggested this, that it might be that human poetry, epic poems, recitals, singing, is a form of sexual selection, and that humans might have evolved the capacity to advertise to the opposite sex by being virtuoso poets, virtuoso reciters of tribal epic poems, something of that sort, like a peacock's tail. That's a pleasing idea, especially for language people. Yeah, I have a related question then.

4:33A lot of what is in a language is dragged along, like to use an aging analogy, tin cans dragged along behind a car after a wedding. I've never actually seen that outside of a cartoon, but it must have happened at some point, probably in the 50s, but that. And so, so much of language is that. For example, in the English language, we'll say something like to dribble. Now, that comes from drip. So, a drip is just dwip, dwip, and then dribble is dwip, like that. Now, you think, well, okay, that's just that one. But then imagine you are nipping on something. So you just go like a wolf nipping. But then nibble is kind of like.

5:20And so that all means. And we use it in words like a giggle and spackle and sparkle. One spark is you get shocked from a fuse that goes out. Sparkle is like that. Now, we can't make any new words with ol anymore. You cannot walk ol across the stage. It's long dead. Only some ancient Germanic person would have been able to use it. So, it's dead. Now, the analogy to this, I think, with DNA is junk DNA. Is it true that there is DNA dragged along like that that used to mean something but didn't, or is that analogy off? That is true. The phrase junk DNA is rather overused in different ways, but I think what you're talking about is what are called pseudogenes, where they're just dead vestiges, actually.

6:13It's a little bit different than what you're dribbled. It's a dead vestige. And the most interesting example of this, it's kind of fossil that's gone, is the sense of smell. As you know, humans have a very poor sense of smell compared to many other mammals. The interesting thing is that we still have the genes that would make us have that kind of sense of smell. They've just been turned off. So we have a whole repertoire of ancient mammal smell genes. They've just simply been switched off. They're vestiges. They're pseudo-genes. I sort of feel if only they could be turned back on again, we would experience wonderful exotic perfumes that we cannot imagine and I could imagine the sort of wine connoisseurs would go far beyond the inklings of lead pencil mixed with blackberry in the satisfying finish that kind of thing I mean that would go way way beyond if only we had all these jeans turned on Yes, it would.

7:23But these genes can be turned back. I just thought of an example. This truly is spontaneous, and that's not my usual. Ol, wordle. And so that's not a word, but now it is. We're reusing that suffix, and so it's a cute little wordy thing. And so you can kind of bring it alive, and maybe there'll be other things ending in ol nowadays, such as your tax assessmental or something like that. Have you tried hardle, by the way? What's hardle? Hardle. Oh, it's wonderful. It's just much more difficult. It's hardler than wordpress. Yes, yes. That is perfect. Anyway, I have another G. Professor Dawkins question, which is this.

8:03I had been present when you said that there was a way of telling a difference, this is getting into my thing, between a language and a dialect. People are always asking a linguist, what's a dialect? What's a language? How can you tell? Is Scots a dialect of English or is Scots another language? And you said, and I thought this was very clever, that you know that someone's speaking a dialect of something else rather than a separate language if when you speak the dialect, the native speaker either laughs at you or is insulted. Yes. They don't see you as doing something important. Whereas if you're speaking Japanese, they're not going to laugh at you, at least when you're there.

8:41They are honored that you're trying. And so that. Is there an equivalent to that in telling a difference between species and subspecies? Yes. What were donkeys and humans? If I were to go into a Glasgow pub and say hoots mom the new, I would probably get thumped. But if I were to go into an Amsterdam pub and attempt to speak Dutch, they would be all over me. They love you, right. And that's the difference between a dialect and a language. Well, there are two species of frog in North America, microhyla carolinensis on the east side and microhyla olivacea on the west side and olivacea has a higher pitched call so that's olivacea with a high pitched call and carolinensis as a lower pitched call there's not much difference between them now you might expect that if it was like you might expect a climb that you would gradually get, as you look from, these are very closely related species.

9:46As you go from west to east, the call would get gradually deeper and deeper and deeper. That's what you might expect. The reverse happens. It's a reverse Klein. As you get into a zone of overlap, the Olivacea call goes up and the Cerenensis call goes down so that these creatures are exaggerating the difference between them. So in the zone of overlap, the difference between them is exaggerated compared to where they don't overlap. This is equivalent to my going into the Glasgow pub and attempting to do a Scott's dialect and being hit for... And being jumped. And I think it's probably true. I mean, Dovzansky had this theory that when new species form, when an ancestral species diverges to form new subspecies and then new species.

10:46There is an intermediate and interregnum when natural selection favours anything to differentiate them, anything to make them sound different or look different or smell different, to exaggerate the difference. And so there is a kind of acceleration of the speciation process. There are many people who say the same thing about human speech, that people deliberately try to distinguish themselves from speakers of other dialects for social identification. Yes, we're not one of those. I frankly suspect that a lot of that has to do with maybe a small collection of words. I am skeptical that people change their accent on purpose because it's so deeply seated.

11:30However, there are other linguists who would counter me on that. And certainly it's true of words that you try to use different words. Yes, yeah. Yeah. So here is another question for you. Natural selection means that creatures' changes are all or maybe mostly due to survival. And so the longer neck of a giraffe presumably was so that it could reach higher. that helped in terms of, you know, getting food. And then there's the okapi, which is a giraffe that didn't happen to do that. Now, of course, in language, the way I mainly think of it is that most change in a language, how you get from Old English, Beowulf, Middle English, Chaucer, Modern English, this, is not survival.

12:22You know, Old English was fine the way it was, Middle English sounded perfectly normal to Chaucer, this to us feels like the only way English should really be, and that's the way it's always been. It's drift. It's just that things happen, and you don't know it in your lifetime, and next thing you know, it's a different language. Now, there's also drift in natural selection, right? Well, drift is in a way the opposite of natural selection. Evolutionary change can come about through drift, where there's no selective force, there's no advantage in this gene or that gene. It's just random random drift.

12:57And that's probably a very important force in evolution. It's not the force that produces interesting things. It's not the force that produces adaptation, that produces better wings, better legs, better voices. But especially if you look at a molecular level, if you look at evolution molecularly, what you see is that changes are neutral. They have no... It's rather like changing the font on your word processor from Geneva to Times New Roman. The meaning is the same, but the actual letters are different. So that's an extreme example of drift, and it's very important. But selection is the interesting part of evolution.

13:41And what I wonder is, really putting a question back to you now, is it all drift in language evolution, or is there a, we could call it memetic selection, where if you think of words as memes, then using the word meme in the proper sense, by the way, not the so-called internet meme, is there a sense in which natural selection, not natural selection, a form of selection favors certain... What about the great vowel shift, for example? Is it possible that... Meaning, why the word, I made a hat. Why is it spelled like that? It says made Why do we say made? Because the a became e over time No one noticed But that's called the great vowel shift Okay, so suppose one vowel shifted for some reason Then that might have caused a necessity for other vowels to shift In order to disambiguate Because there might have been confusion resulting from the first vowel shift Could the whole series, a cascade of vowel shifts, have followed for functional reasons?

14:55Disambiguation. It's a great idea. Folks, imagine the vowels. Do any of you have chinchillas? Of course. So, you put them all in a cage, and they're always kind of moving all around each other. Chinchillas are like vowels. The vowels in your mouth are always kind of moving around. And yes, there is a such thing as a chain shift where it almost seems like the language is trying to keep things clear. That's a weak force, but just as often the vowels fall together and create homonyms that give you trouble. Or a very typical language thing. I'm sitting here, I'm flying by the seat of my pants, and so unfortunately I'm going to have to go into a weed.

15:32Here's the weed. There are languages spoken in northeast Siberia. I'm not even going to give their name, but that's where it is. And think about in English. So we have what? what, where, why, when. All of them begin with what. You kind of like it that way. It makes them easier to learn all of those question words. In these northeastern Siberia languages, you have a series of words like that 5 ,000 years ago that all begin with k, and they all feel alike. The k dropped off because the first sound in a word might drop off like you say, what's that, instead of what's that. It all dropped off. And so that means that in those languages now, instead of what, where, when, why, it's ut, air, and I, like that.

16:12They're not as clear, they're not as learnable, they're all screwed up. That kind of thing happens every bit as much as things that help make things clearer in language, and nobody minds, because by the time a child grows up with it and realizes that there's something off about it, it's too late to learn anything else. And so I feel like language change is maybe random-er than creature change. Yes, what I don't understand is why North Americans don't differentiate between can and can? Because we don't have to. You do, because cannot. They're the exact opposite of each other. But context takes care of so much, and so we just suffer.

16:49Okay, well, I have to agree to differ about that. I want to ask you, I want to challenge something. Okay.

16:59Proto-Indo-European, supposed to be the ancestor of a very large number of languages, all European languages except Basque and several Indian languages. The music for it is... That's proto-Indo-European. It's in what is now Ukraine. Continue. Well, in evolutionary biology, it is a fact that if you take any two animals you like, like, for example, you and a kangaroo, there is a single individual animal, which is the most recent common ancestor of you and a kangaroo and that's an individual, I literally mean an individual there was a mother animal, looked a bit like a shrew or something like that which had two children playing together in the dust and one of those children became the ancestor of humans and the other of those children became the ancestor of kangaroos that is literally true an individual Now, I suspect that linguists suffer from, well, you've heard the phrase physics envy, academic disciplines envy physicists.

18:10Well, I suspect that linguists envy biologists this ability to trace back to a single ancestor. I don't think there was such a thing as Proto-Indo-European in that sense. I suspect that it was never a, well, of course there was, but the idea that there was a single language which then spread by some kind of mysterious influence all over Europe and the Indian subcontinent, it would have been a hybrid of some sort, as I believe we could say English is, because English is a mixture of Germanic and Romance languages. I bet Proto-Indo-European was a hybrid in the same way, and it's biology envy that makes linguists try to trace back to a single...

19:14We do have that envy. I wake up with it daily. But you are correct about proto-Indo-European. We'll never know exactly, but no doubt it was shot through with probably words and grammar of other languages. My best guess would be languages spoken today in the Caucasus mountains, for one. Certainly, there's no reason why not, because the people were always conquering each other and mating and mixing together. It had to happen. And so, no, there was no pure proto-Indo-European. And even, I'll give you even more, it's becoming clearer that, you know, Indo-European is many, many languages, several dozen languages, it's becoming clearer that they trace back to at least two different dialects of Proto-Indo-European.

19:59I won't go further than that. And it's clear that there's that. And all languages are hybrids to an extent. However, I would say that there would theoretically have been some one Proto-Indo-European language that was mixed that became the other ones. and there was constant mixture, kind of like, if I may, the way you talk about the life of a gene, which is snaking through all sorts of various creatures such that it's hard to say exactly, and tell me if I'm getting this wrong, that there was this proto-kangaroo and this proto-marsupial and then a proto-placental, and then there was just one shrew.

20:42My impression was that maybe there were other shrews that contributed genes to both of those strains as things went through. In previous generations, all that I said was that you can trace back to a single common ancestor. Admittedly, you can go back before that, and then, of course, all genes will come in from everywhere. By the way, in bacteria, it's quite different. In bacteria, it's much more like, I think, of language, where you really do have cut and paste going on all the time between bacterial genomes. We can assume there's always been mixture. The pure language would have been spoken for about 10 minutes, probably somewhere in Africa, above what is now the Sahara, maybe in East Africa.

21:25After that, once you have more than one, there would have been hybridity between the two. There was always mixture as things went along. So basically, yes. And you know what? Richard, I think that we have reached the limit of our time, which is unfortunate because I could do this until I expired. But we have limited time. And so thank you very much for being here to allow me to live out one of my life's fantasies. This was wonderful. Thank you very much.

22:01That was Richard Dawkins in conversation with John McWhorter at TED 2025. If you're curious about TED's curation, find out more at ted.com slash curation guidelines. And that's it for today. TED Talks Daily is part of the TED Audio Collective. This talk was fact-checked by the TED Research Team and produced and edited by our team, Martha Estefanos, Oliver Friedman, Brian Green, Lucy Little, and Tonsika Sungmarnivong. This episode was mixed by Christopher Faze-Bogan. Additional support from Emma Taubner and Daniela Balarezo. I'm Elise Hu. I'll be back tomorrow with a fresh idea for your feed. Thanks for listening.

22:43you