Transcript of AE 1089 - Interview: Why People May Discriminate Against Foreign Accents with Dr. Shiri Lev-Ari

G'day, you mob. Pete here, welcome to this episode of Aussie English. Today I have an amazing episode for you guys. I have the pleasure of interviewing Dr Shiri Lev-Ari, now Dr Shiri Lev-Ari is a cognitive psychologist studying language and real-life social networks. She has a master's degree in cognitive studies of language from Tel Aviv University, a PhD in psychology at the University of Chicago.

She's done postdoctoral research at the University of Chicago and the Laboratory of Cognitive and Psycho Linguistic Sciences of OM in France, and she's now a lecturer at Royal Holloway at the University of London. So, we have an amazing talk today, guys.

We chat about her recent publication that looked at how native speakers of a language discriminate against foreign accents and deem them as less trustworthy, and how exposing individuals to foreign languages actually helps increasing this perceived trust.

We talk about her work looking at social networks and how they affect language, so this is social networks, not as in Facebook or YouTube or anything like that, but your friends, your family, the people that you interact with at work, everything like that when it comes to speaking and using a language. We talk about the importance of diversity and variety of a languages speakers in your network for improving your proficiency in a language.

We talk about how slang words develop and spread throughout social networks, how our brains perceive foreign sounds in a language that isn't our native language, and whether you can ever train your brain to be able to hear these subtle differences between similar sounds that are foreign to you.

The speed of language evolution in the modern era with our extreme connectivity and the fact that language is now recorded extensively, and we also talk about the effect of colonisation on language evolution and how pidgin languages emerge and then rapidly turn into fully fledged Creole languages with the subsequent generations of children who end up speaking this language.

Anyway, it's a great episode. I know you're going to enjoy it. Let's get into it.

Shiri Lev-Ari, welcome to the podcast.

Happy to be here.

Yeah, thank you. So, I got onto you recently because I was reading The Conversation, which is an online publication that I really, really enjoy. And you had an article in there. "Here's why people might discriminate against foreign accents, new research".

So, guys obviously go and check out this article, go give it a read and give it a share. But I then obviously wanted to learn more about the author and get you on to the podcast to talk about the findings of this study that you've done. Can you give us a bit of background on you, how you ended up working in the area that you do, doing the research that you do?

Sure. So, I study language, but I studied language from really a social perspective. So, I really care about how this social environment influences, how we learn language, how we use language, how the language then also influences our social behaviour.

And this, I have to say, is not my main line of research. My main line of research actually looks at social networks and how that influences language use. But I was working- When I started working on this project I was just working about interactions between native and non-native speakers and how the expectations we have about non-native speakers influence, how we process their language and things like that.

And then I started hearing about this research that shows that when something is really hard to process, you believe it less. And I was working, like I said, about interactions between native and non-native speakers, and we know that it's harder to process the speech of non-native speakers.

So, I thought, oh, what? Like, you know, what would be the implications for interactions between native and non-native speakers? And would that lead people to really believe non-native speakers less? So, that was really the start of this project.

Awesome. And did you have, like a sort of history here? Because by the sounds of it, I think, and your name, because I've seen it written, I assume you're Israeli and you've learnt English as a second language to the, you know, an incredible proficiency.

Did that have anything to do with you being interested in this field of sort of social networks and linguistics as well? Having, I assume, passed through that, that kind of thing in your life, too?

I think that was, I mean, potentially it was part of what led me to that. I mean, I was just interested in language in general, and then I was trying to think about what type of expectations people bring to interactions with other people. And just when I was trying to think for examples of cases where people come with expectations, maybe as a non-native speaker, one of the things, I thought, okay.

People definitely have expectations about non-native speakers when they interact, about the fact that they're going to have lower proficiency, that they're going to be hard to understand, like all these kind of things. So, it might have led me in that way. But- Yeah, but my research- I mean, it might have led me a little bit, but my research is more general, that I just care about language and society.

Brilliant. So, how did you set this study up and answer this question of truthfulness versus, I guess, untruthfulness when it comes to a person's accent and how they're perceived by other people?

Yes. Actually, I should say so, I'll just mention that this is actually based on two studies, one that I've done a while ago, and I should say that this latest study was actually also the master's thesis of Kasia Galka. So, she's the first author. She, like I said, she did a masters with me. So, the first study from quite a while ago, actually, we tried to look just at the basic to see, can we find that people believe non-native speakers less.

So, the way to do that was really just to play people different trivia statements. So, some of the trivia statements were true, some were false. So, it would be things like "ants don't sleep", so that it is that it was all about like nature, planets, things like that that are not culture specific.

And then people just heard, like I said, different statements and I heard them from, some of them are from native speakers and some of them from non-native speakers. And the non-native speakers had Korean accent, Italian accent, Polish accent, Turkish accent. So, a variety of accents, some of them mouths, some of them heavy. And yeah, like every person heard, both native and non-native.

But of course, if someone heard something in a native accent like "ants don't sleep". Someone else heard like "ants don't sleep" in an Italian accent and someone else heard "ants don't sleep" in a Korean accent and so forth. And basically, what we found is just that people rated, and basically, I should have said, everyone was told two things that are really important, that they're goal is to see how good they are at assessing knowledge.

So, it was really not about- So, they're thinking that we're trying to see how good they are, basically knowing if true statements are true or false. And another thing that is really important, we didn't want it to be about prejudice.

We really wanted it to be about difficulty of processing. So, we told people that these speakers are just reading out sentences that we gave them, so they're not coming up with their own sentences. It's not about whether you believe them or not. They reading out what we gave them, they don't even know if it's true or not.

And to really make it believable, we also made participants record sentences before they started, and we told them we're going to play them for future participants. So, they really think it's just people reading aloud, it doesn't matter who the speaker is and still people rated that the trivia statements as less likely to be true if they heard them with an accent compared to if they heard them in their native speech.

So, that was the original study. So, it really showed that, yeah, people believe non-native speakers less. And even when it's not even what the non-native speaker says, they're just messengers. And what we did now in the new study, we also wanted to see if we can reduce the bias.

So- But to find like more evidence that it's about difficulty, but also to reduce the bias. This time we chose just one foreign accent. So, this time we did it in the UK and there are quite a lot of Polish immigrants in the UK, so we decided to focus on Polish accent.

So, what we did is basically- So, first of all, the main task was the same. People heard trivia statements, some of them were in English like native English accent, some of them were Polish accented English, and they just needed to rate them. And first of all, I should say everyone on the task again believe the statements more when they were native speech compared to Polish speech.

But we also had the manipulation to see if we can reduce the bias and the way we do that is before they started doing the task, they just heard different stories and some of the people heard the stories in just native British accent, and some of them heard it in Polish accent. Now, the thing is that we know that exposure to speech makes it easier to process it, so the more you hear foreign accent, the less it is hard to understand it.

So, really, experience is the best way to get used to their accent. So, like I said, people just listen like, you know, they listen to stories and some comprehension questions completely unrelated to the previous statements, not the same topics.

But they just got used to the Polish accent or not if they just heard the native British speakers, and then they did the trivia task. And what we found is that those who previously were exposed to the Polish accented stories showed smaller bias. They still showed bias.

Everyone showed bias regardless of the position, but we managed to reduce it, so they were less biased than those who didn't get any exposure. And I should say the entire exposure was 10-15 minutes. So, hopefully, if you have more meaningful exposure, maybe it can reduce it more.

Yeah, that's astonishing. And I guess you, for context, you did this with English obviously being the main language. Would you expect to replicate across effectively every single language on Earth? Or would the cultural background of certain languages and where they are geographically and how much influence they have from other cultures influence that?

Like if you are say, I don't know, an Amazonian tribe that has never really come to contact with anyone else besides maybe surrounding tribes, would you expect it to be even greater the measurement of, I guess, distrust or what they would rank it as being untrue because, you know, it's the history there?

Yeah so, I focussed on discrimination, that is not about prejudice. Discrimination that is about prejudice also exist; we know that accent is one of the best marker of group memberships. This even goes to the Bible. You can find over there like this quote about, Shibboleth.

So, basically according to how you pronounce shibboleth versus siboleth, it means whether you are a group member or not. Really, there's like- Like there's a segment there that says it's actually quite troubling because they were kind of standing on the border.

And if people mispronounce the word, they were killed because that shows the out group members. Basically, it's very old accent discrimination, really not new, really not about globalised word versus not. I actually hope that globalised word would reduce it. But it's very old. But I'm looking really at the difficulty of processing. Now, the difficulty of processing is really just about the basic fact of how we process language.

So, the idea is that people who have- What is foreign accent? Foreign accent is that you pronounce words in ways that are not normative because they're influenced by your other language. And because they're not normative they don't fit with what people expect when they listen to language. They can also call ambiguity because sometimes the wrong sound is produced.

It's always harder, and it's going to be harder regardless of whether you're in a small ethnic group in the Amazonas or whether you're in Japan or United States or Australia or India, you're still going to find that people are harder to understand because you're not used to the way they pronounce things.

And in addition to the non-use, they're just like, you know, built in ambiguities and uncertainties just because, you know, people collapse over sounds that are different to you. So, it's just always going to be harder. So, to that degree it should be true everywhere. Now it's true that in addition to that, there's also prejudice, and the prejudice can either enhance it or reduce it depending on the culture you live in.

But the fact that something is harder to process is always going to be true. I think one potential case if there are so many immigrants from someplace such that, you know, you already- Like I said, exposure helps. You're so used to hearing that accent because it's everywhere, because there's such, I don't know, intense contact between speakers of the different languages that you are already used to hearing that accent.

What would happen- Would you expect the same thing to happen with, say, really difficult to understand English accents?

So, despite you obviously being the same native speakers of the same language you have, say Glaswegian English, which for the average person who has never had any kind of exposure to it, it will sound like a completely different language and take them quite a while to get used to the dramatic differences in pronunciation.

I remember I watched a- There's a clip about one of the, I think he's from Glasgow and he has that accent, obviously. And he's a politician in the UK and he's trying to ask a question of someone else who's another politician who I think has a background from New Zealand.

And he gets up and he's like, I'm going to need you to repeat. And the guy repeats himself. And then again, he's just like, I'm sorry, I- It must be my background, but I don't understand you. Can you please submit it in writing?

And so, would you expect the same thing to happen there despite being the same, you know, speaking the same language as another person, but if you have two different accents and you haven't had exposure to them, you would see exactly the same thing with trust there?

Yeah, so a couple of people thought about it and tried to look at it, one of them looked at it with regional accents in Spain, and one of them actually looked specifically in the UK and they compared for an accent, so it was Mandarin Chinese. And regional accents, which was actually Scottish. I don't remember if it was from Glasgow or where it was, but I remember it was some variety of Scottish accent.

And yeah, it was surprising that they didn't find- So, for example, the person who looked at both Mandarin Chinese and Scottish one, she did replicate the fact that people believe less something when it was said in like, you know, Mandarin Chinese accent, but not when it was with a Scottish accent. I don't remember enough who were the participants, so I don't know how much exposure.

She's located in Edinburgh, so I don't know if some of the participants already had some exposure to Scottish accent. But at least, for example, with the Spanish study, they just assumed that maybe the difficulty of processing regional accent, even though it's there, it might not be as big as the difficulty of processing foreign accent.

I'm not really sure, like people did think that it should happen as well...

Yeah.

...So far, we don't have evidence for the same thing with regional accent.

Could it also be that I guess people recognise that accent and they're like, well, they're from my country effectively, right? They're from the United Kingdom, so they give them more trust despite maybe having processing difficulties. Because they sort of see, oh, in-group, you know, in-group despite maybe slightly different accents.

Yeah, I mean, it could be that there's at least less prejudice in terms of actually processing. I don't know, because like I said, my main interest is how people, even people that are not really prejudiced would discriminate just because even if you're not prejudiced, you're going to have a harder time processing, so it should happen regardless.

I think part of the reason, but I really don't know, is that foreign accent has several properties that regional accents don't. For example, foreign accent, some people say there's more variability, so it's not as consistent.

So, it's easier to get used to regional accent because it's very consistent, you just this is how they produce those things, you can get his state, whereas with a foreign accent, they might produce it in a whole myriad of different ways that makes it harder... (both talking) ...It depends...

Sorry, you mean if you were like to get ten French people who spoke English, they would all potentially have subtleties and differences as opposed to a consistent set of pronunciation rules in English.

Precisely, that's one of the arguments that people make. I don't actually know if it's been tested, but it's been said often. Another thing is that foreign accent often leads to collapsing over several sounds because they don't exist in your language. So, for example, you don't know what "the" is, so you're just going to say, 'sir', but you're also going to say 'sir' for 'sir.

So, you're basically or, you know, you can't make different vowel contrasts, you just choose one vowel for all these vowels. Usually in regional accents, it depends, but for the most part, you still maintain the contrast you just produce them differently, so you move all vowels to something different.

So, again, there's more information there. So, it could be that all these things just make it easier to understand regional accent. But like I said, I think more research needs to be done. But so far, people didn't find any evidence for a similar effect with regional accents.

So, what are the broader implications of this most recent study that you've done? Should it be something that, you know, if migrants were reading this article and they'd be like, oh my God, discrimination because of foreign accents.

Do I need to worry about improving my pronunciation or changing things? Is it the kind of thing that they should take that on board? Or is this the kind of thing that maybe governments and native speakers in groups need to at least be aware of and try and implement changes themselves?

Yeah, I would go more with the second one like rather than the former. And the reason is really that, I really think that non-English speakers already have a burden, right? They already need to put so much effort into communicating, I really think that societies should realise that there is discrimination that is built in that.

Even if they're not prejudiced, of course, there's also I'm not saying that there's no prejudice in the world, but even regardless of prejudice, there's going to be discrimination because of that, and we need to take that into account and try to find ways to solve it.

And yeah, the fact that we find that exposure helps, that means we need to think- So, it depends on what we're looking at, but for example, if you're looking at people that we know will have to deal with a lot of non-native speakers, maybe we want to have built in training that is about exposure to non-native speakers.

Maybe we also want to think about more diverse workplaces where people would interact regularly with non-native speakers, and that actually should help both with prejudice and accent comprehension. So, it's like a win-win situation. So, I really think this should be the direction of how we make it easier for native speakers to get used to foreign accent and find it easy to understand it.

Have you noticed a change with the BBC by any chance over in the UK, with the representation of people or journalists or people that are shown on TV? Because here in Australia, we have the ABC, the Australian broadcasting- I've forgotten the thing, but they- I've noticed in probably the last three years they've really gone down the diversity rabbit hole of just trying to make sure that all of their-

The people they interview, the people that they have doing the work, that are doing the journalistic stuff, they're from all sorts of backgrounds, they are migrants, they are people from Ireland or Scotland. And so, initially- My initial reaction was, oh my God, we're going to lose kind of what the Australian accent is in terms of the, you know, the ABC and what it should be showing migrants.

But then I thought, actually, this is probably better if we think about it from the population of the, you know, born here, Australians with the Australian accent, because they're going to be getting so much more exposure to not just different cultures and everything and people of different races, but all these different ways of pronouncing different words and these different accents.

So, yeah, I guess my first question is, does that help? And the second one is, have you noticed that locally in Britain?

So, yeah, I do think it should help a lot. I think that that's fabulous, that they do that. I mean, it's good for many reasons from representation that you just want to see more people like you and what the country is actually like. But also, yeah, it should help with getting used to accent.

It's usually better to listen to not just one speaker per accent, several speakers per accent that helps you understand better what just about the speaker versus the accent. But yeah, I think this is the way to go. I can't really say whether that's true over here. I have to say I don't watch- I don't really consume news via television, so...

She's too busy doing research.

Well, I mean, there are just other outlets to consume, so.

No, all good. I thought I would ask. I would like to also ask you about social networks and their importance for people learning foreign languages because I was watching, I think, a lecture that you had on YouTube about this, and it seemed really interesting.

I think you brought up that the word like "great" may actually have different meanings, depending on the person who's saying it when they're responding to say, how was the movie? If someone says, oh, yeah, great. Or someone says, yeah, great! You know, they may have different meanings and the importance of having lots and lots and lots of different people so that you can sort of learn that vocabulary and the nuances there.

So, yeah, can you tell us a bit about social networks and their importance or their significance when it comes to languages and linguistics?

Yeah, sure. So, basically, the idea is that we learn language from our environment, right, whether it's our first language or our second language, we always learn it from the environment. So, it really matters what our environment is. And a lot of my research shows that the more diverse our environment is, the better we are at basically learning a language.

And that's actually even true for native speakers, not just non-native speakers. And I can really take the example that you said about "great" because I think, yeah, that's a nice one from a couple of my studies and I the idea over there was to look at how people really differ in how they use language. Right.

So, if you ask someone, oh, how was the film last night and they say, great. For some people, "great" would be, really amazing. I enjoyed it so much. And for some people, "great" will be kind of like a synonym of okay, great, it was nice, fine. And like just people differ in how they use that.

And what I was interested in those couple of studies was to try to see whether, like I said in general, I find that having diverse social networks or having more people, and it's actually also true, although I don't necessarily study that, that different people are also going to be different from each other.

You're going to get more exposed to how people use language, so you'll be better at understanding it. So, you'll understand that, oh, the "great" of that person, because of their speech style and the other words that they used probably means really good. But the great of that person, considering the other words that they used in their speech style probably means not so amazing.

And I did a couple of studies to really look at that, so I really just- This was even native speakers. Of course, this is going to be even more true with non-native speakers. But I was trying to show that even native speakers, if they keep on having a large and diverse social network, they'll be better at languages in their own native language.

So, the first study was really, like, very simple. I asked people about their social network, and I should say social network, I just mean people you interact with. They don't have to be your friends, they could be your colleagues, your boss, your clients, your family, like, you know, just people you regularly interact with. So, I asked them like, you know, I measured how many people they regularly interact with.

And then I showed them restaurant reviews. And what I asked them, and these were restaurant reviews that I collected beforehand. And I just asked them to guess, like not guess, but estimate how many stars the person gave the restaurant from one to five.

Very clever.

Yeah. And basically, I find that people that had larger social networks were better at doing that.

Wow. Really?

Yeah. Is this a four or a five? Or is this a three or a four? So, that was really, really nice. Of course, the problem with that is that you don't know the direction of the correlation. So, is it the fact that because you have a larger social network, you're also better at understanding people? Or is it that you have a larger social network because you're really good at understanding people?

True.

We don't know which one causes the other one. So, I led and also did an actual experiment that tested that. And in that experiment, what I did is initially I, again, collected this time reviews for products, specifically chairs from different people. So, I showed them dozens of chairs, didn't need it to write reviews. And then I did a trick of basically, I wanted to make it about language learning.

So, I replaced some of the words with novel words. So, the person who reads the review doesn't really understand it. It says things like, "this chair is normal." And you don't know that "normal" was beforehand "horrible". But I changed horrible to normal. So, that way, people need to learn the new words, but the new words really function like the old words. So, basically, I replaced "horrible, bad, okay, good and great".

So, you know the entire spectrum just with new words. And the point is, like I said before, and people use these words differently. And then I had new people come and basically listen to, well, read actually the reviews from the previous people. But I made a trick of basically manipulating they're social network. So, to iterate- So, everyone read 40 reviews with the same amount of exposure to these words.

But those 40 reviews came from just two reviewers or eight reviewers. So, they either read like a lot from each person or just a little bit, but for many people. And then I tested how well they understood novel reviews that I gave them from other reviewers.

And basically, I found that if you were exposed to, you know, the input from multiple people, you were better. So, they had the same amount of exposure, but just because it came from multiple people, they could learn these words better.

So, even though I did it, like, you know, the same sense that was said it's true for both native and non-native speech. That means that also as a second language learner, if you keep on interacting just with one or two people that are native speakers, it's not going to make you as good at learning how to use words compared to if you interact with multiple people.

Is there an upper limit to that? Like if I were- I speak Portuguese with- My wife's Brazilian, so I speak Portuguese at home. If I were to get on, say, you know, a language learning website like italki or something and got a thousand different tutors and had a single lesson with each of them once. Would that be as effective as, say, having or ineffective compared to, say, having a thousand lessons with 10 tutors or 50 tutors?

Like is there an upper limit where if you were to just- You need time with the same people on a regular basis to be able to absorb, get those repetitions of the use of those words to be able to really ingrain that context? Or is it more just about broad-spectrum exposure to as much as possible?

Yeah, I would assume that- Well, I guess now people understand the concept better, so logarithmic rather than linear. So, I do think that, you know, the difference between two and one is very big. The difference between like, you know, three and two is still big, but not as big as two and one. Like I would say, I didn't necessarily test that, but that would be my hypothesis...

Yeah.

...If like, you know, ten is much better than one, but the difference with twenty is still better than ten. But the increase is not as big as ten to one. And yeah, it could be that a 100 versus 110, probably going to make a small difference. Like 200 might still be better than 100. But like, you know, you'll need- It beco- Like, you know, the addition of every speaker becomes smaller, the more speakers you already interact with.

Because the idea is that if you want to be exposed to multiple ways of using language...

Yeah.

...At some point, if you have a large enough network, you kind of sampled all the different ways that people use language. So, just encountering a few more people, they're still going to use the language like, you know, probably like people you already encountered. So, it might not be as useful to meet more and more.

So, what are the sort of roles of social networks, too, for developing new words? Right. I think I was watching you, and you were talking a little bit about that in this lecture. And I was thinking Australian English is known as probably the dialect of English that has the largest amount of slang that we make up. So, we have something like 5,000 diminutive slang words that we use.

And I think the equivalent, you know, for other dialects is usually around 1,000, maybe 2,000. We have some ridiculous amount. So, a good example of a word that I think escaped Australia and went wide was "selfie". Right. So, we came up with that, I guess someone somewhere.

So, what is the role of social networks and how do these words kind of end up developing and then escaping? Is there something, too, that allows certain words to escape and just become widespread and accepted by other people? Or yeah.

Yeah, there is actually some research on that, not so much mine, but there is the idea of in general, if you look at community size, then first of all, the more people there are, the more words that there'll be invented. It doesn't mean that all of them will spread, but more people will invent words. And there is something that shows- I'm trying to remember the details of that study.

I feel bad when I'm not citing the person who found this. But the idea was that in general, I do think that I found that in larger communities, the words are more maintained, and you don't lose words as much. So, in smaller communities, words are much more likely to go away. I don't want to say too much. I'm not- I don't remember that well. But the idea is that, yes.

Things like the structure of the networks, it's both going to be how big it is. It's going to be how connected it is. Because, you know, like different communities can be very interconnected, where only your friends also talk to each other and their friends, friends talk to each other.

Or it can be quite sparse that you have friends that don't know at all and don't interact with your other friends and so forth. So, all these things really influence both how information spreads, whether it can be maintained, whether it can reach the entire community or go away.

So, all these things matter, but I don't remember enough of the details to keep going through that.

No, no, no. That's all good. I'm just curious. These are sort of just questions that come to mind, and I'm always really fascinated with, you know, Australian English, obviously, and why it has so much more slang than other dialects.

And I think a big part of it is that our culture is very- It's kind of really tied in with our culture to be sort of inventive with language. And I think that's why we create these words and end up maintaining them and using them a lot more.

Whereas I think if you were to go to say Great Britain, from again, the limited understanding that I have as someone who's like genetically British but has never been there.

I feel like there would probably be the opposite kind of pull for, you know, Received Pronunciation speakers where they would probably be like, don't use slang, you know, stay away from it a little bit and maintain that certain culture or the way that you appear to other people.

But it's always something interesting that I wonder about. Do you think that social media has had a massive effect on those sorts of things, too? Where it allows words to be created and spread rapidly, but it potentially can also just lead to them dying out because there's so much new uptake, you know, that can just spread around the globe overnight.

Yeah, it is true, and things now like, I think that a lot of- So, like language change happens- Can happen much more quickly because information like words really and, you know, people also change the grammar all the time, and all these things really change very quickly. Will it also lead to dying more quickly? That's an interesting question. I really don't know. I think that it depends.

It depends whether some of the things that you're inventing are not just replacing the previous ones, right. They're adding more shades of meaning, and if they're not replacing the other ones, then there's no reason for the other ones to go away. You're just going to have a larger and larger vocabulary.

But of course, at some point you can't maintain all of it, but- Actually, yes, I mean, I definitely think it's going to accelerate language change. I don't know. I think it's a really interesting question to ask whether it would also lead towards dying more quickly.

I'm not sure because you can say that because nowadays written evidence for it everywhere, like, you know, it's there, you can see it and encounter it, whereas you know, species ephemeral like you lose it. Whereas over here you can also find a tweet from two years ago, so- I'm not sure it's an interesting question. I don't know.

What would you expect to with like, say, language evolution, with the internet and with the fact that we can just save everything now? Because I imagine that lang- Like again, from someone who's not a linguist, I imagine that language was evolving so much more rapidly when you were, say, an isolated group in the Amazon.

Even today, those groups probably have language, I would imagine that changes really quickly because it's not written down. And so, you know, you would imagine what you pass on is liable to just have new words and then your kids make new words. It's not like you get to keep reading literature that was written down decades or 100 years ago and be like, these are the words that we use.

Here's the dictionary, and there's no deviating from this. We can add to it, but we don't really change any of it and say this no longer means this.

So, do you think with the internet and with social media and the fact that we can save everything today, if you were to fast forward, say, 200, 300, even a 1,000 or 2,000 years, you would be like the pronunciation may differ, but it's going to be recognisable? You- You know, English writing isn't going to suddenly be character based like Chinese.

So, I do think, I don't know if it's the internet, I definitely think that starting to write down things probably slowed language evolution. Yes, precisely the reasons that you're saying. Now we know this is the word. This is like, you know, this is sort of how it's written. So, I definitely agree that probably the invention of writing really slowed down.

Obviously, we cannot really- Since language- We cannot study how quickly language changed before writing was invented because we don't have any access to that. But I think that's true. But I think, sorry, you had another point that I forgot that I wanted to answer, too.

About pronunciation changing or?

Yes. So, that's actually- I'm not sure, so- So, one thing that is interesting is that small groups, what I find in my studies. So, I should actually say that how I conduct some of my studies of language evolution. So, I basically get people to create language in the lab.

So, I really show them normal things that they've never seen, and they need to interact about them with new people, but are not allowed to use any language they know, they create gibberish words. And as they create gibberish word and then... (inaudible) ...for several hours, they actually created a language.

And when I say they created a language, I really mean that, you know, they start, first of all, having a vocabulary that everyone agrees on. But they also start creating grammar and morphology, like they really start creating a language out of talking in gibberish. But I manipulate the type of social structure that I have.

So, sometimes I have very small groups, sometimes they have large groups. I manipulate how interconnected they are. And I find out actually small groups are much more likely to do really weird things. So, in a sense, all large groups sort of look the same. If you get a large enough group, they're always by the end reach sort of the same type of complex grammar at the same time, like, you know, they're going to be very similar to each other.

It's kind of like the law of big numbers, like you're going to something that is like easy for everyone, and it fits their- Like, you know, everyone. But with small groups, you can be really influenced by like one weird invention of one person and everyone goes with it. And because it's a small group, they can all, like, accommodate each other...

That's so interesting because that's like evolution, right? If you have a small population, mutations can arise and then they diverge rapidly. Whereas if you have a really large population, it happens a lot more slowly and you're not going to have something. Yeah.

Precisely. So, yes. So, in that case, I don't know. But like, you know, pronunciation can go in multiple ways just because, like I said in small groups, and especially if we tend to think that society now lives in larger groups compared to in earlier times...

Yeah.

...You know, I would actually assume a lot of variability back there with each group going in some weird direction. And I was like, you know, and now, you know, because language is, yeah, is used by larger groups, for the most part, they converge more.

But yeah, like I said, there's a lot of speculation there. But definitely we've seen all the experiments, really small groups can end up with languages very different from each other in like for good or bad, like, you know, just very different.

Yeah, I always wonder. I think I was learning about Australian English and the pronunciation changes since being developed because we had this really weird- The reason we have this, well, the accent that we do to some extent is because we had all of these convicts come over in the late 1700s, right, when we were colonised, and we had a bunch of people from London.

But then the convicts tended to be people from, you know, the lower socio-economic parts of society from Scotland, from Ireland. And you had them all in a very isolated place with only maybe 1 to 2,000 people and then their children that were born in the colony were the ones who had their own accent. So, they noticed within the first generation, they were like, shit.

All these kids are developing their own pronunciation as opposed to copying their parents. They're like, we are much more influenced by our peers than we are by our parents. And so, that was how we ended up with this rapid change in pronunciation from British English.

I didn't know that about the history, but that's really interesting and it makes sense. So, I would say it's probably- It's not so much, probably they ignored the parents' pronunciation, but it probably tried to also find some way that also, I wouldn't say, averages across them.

Yeah.

...You know, takes over not only from your own parents' accents, but the other accents, that's something that makes sense. So, in general, this is something that, I don't know if you want me to talk about it, but the whole idea of how language evolves.

So, this idea how you move from kind of a pidgin to a creole, so pidgin is when you start having two groups of speakers of different languages start interacting. So, for example, you might find it in trade relations, in colonialism. So, really intense interaction by people who don't speak the same language.

Like they're good example, I think would be the pidgin that's spoken in Papua New Guinea, right. Because they'll have all their native languages, but then they have the mixed with English really stripped back with grammar, and it's very specific, kind of like interaction-based language that's used.

I don't know enough about the situation in Papua New Guinea, but that sounds very much like the way it is. As you said, it tends to have, yeah, not much grammar. And you know, it's communicative, but it's not, yeah, it's not fully grammatical and it's not the way we usually like, you know, languages as they usually are.

But what happens once you usually start having children, the children need to acquire their language and what the children do, they actually create the grammar. And it's not like they sit down and say, I'm going to create the grammar, but they need to find some structure to something, so they do take from their parents, but they have to do more than that.

And then actually, you see that with each generation you have more grammar. That's basically the new learners, and it becomes it's called the creole. So, actual like, you know, languages that do look just like other languages in terms of like, you know, having, yeah, like, you know, more grammatical structure.

And so, yeah, that happens often, you can find it in many places, a lot in Africa because of like, you know, colonialism or in like, you know...

The Caribbean, right?

Yes. So, it's very common. Some people say that that's how Swahili started. Like, you know, arbitrator's and Bantu languages. So, that's how you got created Swahili. But yeah, they become real languages, but by the learners.

It's the infants that need to learn it. The parents don't really have such a structured language and they create it. And yeah, it could happen with accent as well because you need to find something that, yeah, allows you to bridge the gap.

Do you have children, if I can ask?

No, I don't.

Because I was going to ask if you taught them, if you had them, both Israeli and English. I've got young children, one is two and a half, and he speaks both Portuguese and English, but he mixes them, and it's really interesting to see him create new words for stuff that he doesn't know.

So, when we were- A good example is like in Portuguese pássaro is bird, right. And so, we were trying to teach him these words, you know, there's a bird, there's a bird. And I had no idea how much we'd be talking about fricking birds to kids, but you see them everywhere, and so the kids are always like, what's that? What's that? But he just came up with his own, oh, that's kaka. That's kaka. Like, there's a kaka there.

And so, now, even now I refer to birds quite often with him as, oh, look, kaka. Like there's a Kaka. Do children- Are they the ones that kind of are the evolutionary pump when it comes to language and language evolution? Is it mainly- Is that where all the change tends to happen?

If you were to look at the demographics of a population that speak a certain language, would you imagine that it's the young children or the adolescents that are the ones creating that diversity that leads to languages changing?

Yeah. So, actually, for many, many decades, that's what people in linguistics and psychology thought. So, they thought that it's really the children who changed their language. And like I said, there is quite a lot of evidence to that if you think about things like creole or the example that you just gave me the accents in Australia.

But I think that actually over the last few decades, people understood that it's actually not just children, so adolescents have a big part, but even adults. If- I probably if I could find some recording of you from 10 years ago, 15 years ago, you'd probably noticed that you don't actually quite talk the same way.

Yeah, I was wondering that.

There's like this really, really nice study by Jonathan Harrington and what he did, he looked at the Christmas announcements, like the Christmas speech that the Queen makes every year...

Yeah, yeah.

And he tracked it for, like, you know, many years, they basically found that her pronunciation changed over the years. So, basically, the Queen no longer actually speaks the Queen's English.

I was wondering that, I've got grandparents who are in their nineties, and I always think, did they speak so kind of- They almost speak like received pronunciation Queen's English kind of English. And I always wondered, what did they sound like as teenagers, you know? Did they really sound like this?

And I guess the best thing about being a podcaster, if I can maintain this until I'm 90, there'll be like an almost daily, you know, series to look at and be like, how much have I actually changed? I imagine, too, its vocab too, and certain phrases and expressions.

And I've noticed that they'll come and go, I'll be having conversations with someone, and they'll use something and something subconsciously is like, oh, that's really good.

And then all of a sudden, I'm saying, ostensibly all the time, you know. So, it is really funny how that obviously, that it just evolves and changes over time throughout your life, right.

I think that the best way to see it, or at least in my experience, if you sometimes meet people who were born in your country but left, I don't know some time ago...

Yeah.

...So, you suddenly talk to them, and it's not just their slang, slang is part of it, but even outside they use like expressions. It's like, yeah, I guess we used it in the eighties or like, you know... (inaudible) ...But they still say, because it's really just the way you phrase things really changes all the time.

So, I guess what I'm saying is that we all change language, both adults and children because we don't keep the same way of talking throughout our lives.

An amazing example of that for you to potentially go down the rabbit hole of is Australia and the Italian migrants that came in. I think about the 50s and the 60s to Melbourne, and there's a street that we have in Melbourne called Lygon Street, which has all of these Italian restaurants and pubs and cafes and everything on it that have been there for that long.

And it's really funny, you'll talk to them and they're like, we're the authentic Italians, you know, with the authentic food and the way that we speak. And I'll meet Italians who've come yesterday to Australia, and they'll be like, these guys are out of a time capsule, man. You'll talk to them and it's like, I'm talking to my grandfather.

But their kids have learnt that Italian. And so, it's this really funny thing where it's almost like they've been taken- They've been frozen in time and taken away, and then- They don't go back to Italy often, you know, they'll just grow up in Australia, but they speak Italian. And so, you have this weird thing of when they interact with Italians, they're like, both of them think each other is weird, you know.

I would just say that probably the Italian speakers on the street also don't quite speak the way they actually talked back then. So, they also changed, it just didn't change in the same direction as the people in Italy. They just change on their own trajectory.

Did- Sort of finishing up because I know I've kept you quite a while. What's the sort of history with Hebrew and Israel? Because you obviously had a lot of Jewish people from all over Europe and, you know, the continent, they're coming into a single area.

Did you have a really interesting, you know, sort of melting pot of accents and dialects when that happened? And has that led to a sort of really, really diverse language that you now have? Or is it more sort of homogenised?

In terms of accent, you mean?

Yeah, and vocab and everything, I guess, too. Because I imagine that was really different depending on where you were from originally.

Yeah, although I would say that usually people who came to Israel didn't speak Hebrew, so they learned Hebrew in Israel.

Oh, okay. Really? Sorry, forgive my ignorance. Forgive my ignorance.

Yeah, it's just that Hebrew was not used as a spoken language. Sometimes it was only used like, surgically like, you know, people used it when praying. But that was about the only context that people used it outside of Israel. And I'm not even sure that everyone actually prayed in Hebrew rather than in translation, but I'm not sure about that.

But really, they couldn't speak Hebrew, only when they came to Israel they learned Hebrew. So, you don't sense they learned the same version, which doesn't mean- And I think there's quite less variation in Israel compared to other countries.

I think potentially because it's small, potentially because we actually move a lot. I think it's quite uncommon to find someone who grew up and spent their entire life in the same place. You move an hour away, but an hour away in Israel is like half of the country away.

So, like that's in- You know, I think we don't have as much geographical variation. We do have socio-economic variation, it does exist, but that, you know, that always- Yeah. So, that kind of thing does exist in a tiny bit of ethnic variation as well, just because- So, it's not so much that people learn different versions of Hebrew, but if you learn Hebrew, and beforehand, you know, it just like non-native speakers.

Like we said, when they learn a new language, they bring their knowledge of the previous language into the new language. So, if you previously spoke Arabic, you're going to sound different than if you previously spoke Russian or Polish. So, in that sense, there's some variation. But yeah, it's also, yeah, unclear to what degree it's also maintained in coming generations.

Some markers completely- Like that there were ethnic markers seem to disappear for the most part. But yeah, we have a lot of socio-economic variation, but not so much yet. But that's less about- Like I said, because they only learn Hebrew when they come to Israel.

Yeah, it's interesting. Australia's kind of the same, if you were to come to Australia and drive around, you wouldn't really see geographical differences in terms of pronunciation. There's a few little things like slightly different, someone from Queensland might say school instead of school.

So, there's that kind of thing, but it's much more related to whether you're in the city or you're in rural environments where you might have a broader sort of more nasally and more diphthongs kind of accent. And so, we compare that to Britain, and I've met people- There's a guy that I met from Yorkshire in Australia, and he was like, yeah, for the first 40 years of my life, I didn't go more than 30 kilometres from where I grew up.

And I was just like... (shock) That's just so foreign to me, it blew my mind. Anyway, last question is based on your research, what sort of advice would you have for people coming to Australia as, you know, migrants that are learning English as a foreign language based on your sort of knowledge about social networks, how can they best kind of use that in order to improve their language or at least speed up the process of learning English?

I ask because I know a lot of people who come here and then end up living and interacting with people from their home country, and I'm always like, try and limit that. Anyway.

Yeah, I mean, I don't know if I would say limit that, but just make sure that you also have interactions with native speakers. I think there is a lot of support you can get from interacting with people, you know, who are also migrants and experience the same thing. But yeah, I think you just want to have a diverse network.

So, make sure that you interact with several different types of people who are native speakers. So, both like different people, but also preferably of different ages, different like, you know, socio-economic backgrounds.

Like, you know, the more you can get diversity in your input, the more you can really get used to being able to understand new speakers and get a better idea of, you know, what the English in the new place is. So yeah, that would be my recommendation.

Would it also be with foreign accents as well? Does that also have an important place?

Yeah. Do you mean in order to- So, yeah, if you want to get used to also a foreign accent, then yes. Interact with a lot of foreign speakers of that accent and you'll become better. So, there is a lot of literature that really shows that. The more people you interact with, the better you get idea of that accent.

Brilliant. Well, Shiri, thank you so much for coming on. Where can people find out more about you and your work and what you do?

Yeah, so I have a website, but I don't know. I guess, if you google my name and social networks, I'm going to assume that that's going to come up. So, yeah...

It does. It's the first thing to come up.

Okay, good. So, yeah. So, all my publications over there are available for free, for anyone to read and, yeah, feel free to contact me.

Brilliant. Thank you so much again.

Thank you for having me. Bye.