1080. Linguist Andrew Cheng explains why people’s accents shift over time, especially when they move—and how YouTubers make perfect data subjects. If you've ever cringed at your old voice recordings, this one’s for you.
1080. Linguist Andrew Cheng explains why people’s accents shift over time, especially when they move—and how YouTubers make perfect data subjects. If you've ever cringed at your old voice recordings, this one’s for you.
Andrew Cheng is a professor of linguistics at the University of Hawaii. You can find him on Bluesky at LinguistAndrew.
NEW: Sign up for my AP Style webinar on June 12: bit.ly/4k1XmpI
Use the code MACMIL for $50 off.
🔗 Share your familect recording in a WhatsApp chat.
🔗 Watch my LinkedIn Learning writing courses.
🔗 Subscribe to the newsletter.
🔗 Take our advertising survey.
🔗 Get the edited transcript.
🔗 Get Grammar Girl books.
🔗 Join Grammarpalooza. Get ad-free and bonus episodes at Apple Podcasts or Subtext. Learn more about the difference.
| HOST: Mignon Fogarty
| VOICEMAIL: 833-214-GIRL (833-214-4475).
| Grammar Girl is part of the Quick and Dirty Tips podcast network.
| Theme music by Catherine Rannus.
| Grammar Girl Social Media: YouTube. TikTok. Facebook.Threads. Instagram. LinkedIn. Mastodon. Bluesky.
Grammar Girl, I'm Mignon Fogarty, and just a heads-up that today's show was originally released a couple of months ago as a bonus segment for people who support the show, the Grammarpaloozians.
If you like what we do every week and want to support the show — and get ad-free podcasts and bonus episodes like this right away — you can become a Grammarpaloozian! Just visit quickanddirtytips.com/bonus to learn more.
And now, on to the episode.
MIGNON: Greetings, Grammarpaloozians. I am here with Andrew Cheng, a professor of linguistics at the University of Hawaii in Honolulu. We just finished our conversation about language in movies, which was just fascinating. If you haven't heard the main show yet, head over there and listen to that. It was great, but Andrew has done also a study. He's looked at some YouTubers and how their dialects have changed over time and when they moved, and so we're gonna talk about that today. Andrew Cheng, welcome to the Grammarpalooza segment of the Grammar Girl Podcast.
ANDREW: Thank you! It's so exciting and exclusive.
MIGNON: Yes, very exclusive. There's a velvet rope at the — digital velvet rope somehow.
This paper that you did is about second dialect acquisition, which sounds very academic, but it's actually really interesting. So can you explain what you mean by that?
ANDREW: Yeah, it just means that you maybe speak one dialect of a language, and then at some point in your life, you speak a — you acquire a second one, a different one. Dialects here refer to anything you might consider as a very main or well-known dialect of a language. There's Australian English, and there's British English, and there's American English, and Singaporean English, right? So if you grow up speaking one of those dialects and then you move to another country where they speak another dialect, you could either keep your own dialect or you could shift the way you pronounce things and start speaking in that second dialect as an adult. So it's really very interesting to what extent people shift their dialects, if they do it consciously or unconsciously, and whether we can really track the changes in pronunciation over time.
MIGNON: This is a backing up, higher-level question, but how would you define the difference between a dialect and a language?
ANDREW: That is — that's the million-dollar question for linguists. As a sociolinguist, there's actually a third term that I like to use, and that's just language variety. And that is a very neutral term that just refers to anything that's different. You can have variety A, variety B, variety C. Whether it's a dialect or language is actually a political question.
There's a saying that everyone uses in linguistics, which is that a language is just a dialect that has an army and a navy. What that means is that the difference between what you call a language and a dialect is dependent on who's labeling it and how much political power that they have. So we've got various languages that will say, "I'm a language because I'm associated with this country."
But then we have the linguists who look at those languages spoken in those two countries and notice that they're actually really similar to one another — mutually intelligible. If someone speaking language X can completely understand almost everything that someone speaking language Y is saying, we could consider them dialects, actually, of the same language, but don't tell that to the politicians.
Don't say that to the people who are a little bit more nationalist and have that pride: "My country, my language."
MIGNON: I love that you brought up that quote because it's the epigraph in my book. I can't quite reach it, but it's ... "The Grammar Daily," which is one of my books ... it's the opening quotation.
ANDREW: Yeah, I love that.
MIGNON: Yeah, it's a classic. It's really good.
So, YouTubers, I imagine this is a relatively new academic endeavor to study YouTubers.
So can you tell me how you came upon the idea of looking at these people and how they talk and why YouTube?
ANDREW: I am going to age myself here or date myself. I grew up in an era when YouTubers were a really massively influential source of entertainment — like the early days of YouTube vlogging, where you had early celebrities like KevJumba or Nigahiga, and especially, speaking from the Asian American perspective.
This was a time in the mid-2000s when you wouldn't see a lot of Asian Americans in mainstream media, but with the advent of this new social media website, YouTube, like suddenly anyone could become viral, and anyone could become really popular just by posting whatever they wanted for the whole world to see.
And so I got very interested in that. Then I went to college, and I went to grad school, and my life got in the way. Suddenly, 10, 15 years later, I'm revisiting some of these videos and wondering, "Huh. I wonder how these people have changed." I grew up with some of these YouTube vloggers.
Do they still do the same thing? Are they still making the same content? The answer is no, not at all. But some of them are still making videos all the time and still talking, and I just was very curious about their voices. I then took a seminar in graduate school that was about language change across a lifetime, and this is really interesting because it connects to all sorts of ongoing linguistic questions, like theoretical questions such as: Is there a critical period after which it's impossible to learn a second language?
Does someone's accent fossilize in place at a certain age, and then you can't change it for the rest of your life? Or once people are in their older adulthood, like 70s, 80s, 90 years old, what kind of changes happen there? So thinking about these questions and then combining it with my personal interest in YouTube vloggers, I was like, "Hey, it's really hard to do a longitudinal study of real people because you need to ask one person to come back to your laboratory and record them every year for 10 years."
That's hard to do, but YouTubers have been recording themselves daily, weekly, whatever, for a really long time sometimes. And if we just analyze their voices from their videos going back however many years, you instantly have a longitudinal dataset that can be used to analyze any changes in their voice over time.
MIGNON: Oh, that's fascinating. You could do the same thing with podcasters.
ANDREW: I absolutely could. How long have you been doing Grammar Girl?
MIGNON: Eighteen years.
ANDREW: Eighteen years. All right, so you can send me a video from 2007?
MIGNON: Six.
ANDREW: 2006? And I can compare it to your voice now, and we'll see what's changed.
MIGNON: Not much. I don't think. I haven't really moved. So, it seemed like in your study, one of the key elements was that the people had physically moved.
ANDREW: Yeah, so one of the YouTubers that I followed moved from Hawaii to Nevada. So Hawaii has a very particular dialect of English that we call Hawaiian English or Hawaii English. It sounds quite different from other varieties of English. And then this person moved to Nevada, which has more of a general West Coast, western type of English.
Then the other YouTuber is originally from upstate New York, grew up in Boston, has a very particular, maybe a very familiar accent to those who've — I don't know — watched "Good Will Hunting" or whatever, Ben Affleck, Boston-type speech.
So think of that Boston "pahck your car" and "Hahvahd yahd" type dialect. You get a little bit of that, but then she moved to Los Angeles, as many viral YouTube stars do. I'm thinking about two vloggers whose accents started on opposite sides of the United States, and then they moved to the West Coast area.
To what extent did the hallmarks of their accents or their dialects from their respective places change to fit the mold of what's more common in California or the West Coast?
MIGNON: And what did you find?
ANDREW: I found a pretty interesting ... Okay, so first of all, the answer was not 100%.
It's not the case that anyone would just completely wholesale leave their old dialect behind and then start speaking like a Valley Girl Californian. But every vowel system — or every person's accent — is what defines an accent is composed of lots of different parts. So you can talk about the vowels, you can talk about the consonants, you can talk about pitch.
And so if I just broke down each dialect into its component parts, you could see that there were some shifts, like a few vowels here and there that would move and others that would not.
The example from the Bostonian was her "ah" vowel, like the one in "Boston," right? When we think of the typical Boston stereotypical accent, we say, "Boston, Hahvahd yahd," that "ah" sound, which in California — I'm a Californian, so mine would be more like "aw, Boston." And we do see that movement from "ah" to "aw" over time for that speaker.
But none of —
MIGNON: The vowels? Linguists often talk about the vowels. Why not the consonants?
ANDREW: I think it's because vowels are a little bit easier to analyze for me, and vowels tend to be like the locus of where dialectal variation
occurs in English. Like when you talk about all the different — if you just list out all the phonetic sound differences between every dialect of English, mostly it has to do with their vowels. That's not to say that there aren't consonant differences, but those are just...
MIGNON: I'm wondering why that is. Is it because of the way we make vowels in our body?
ANDREW: That could be it. Vowels are really slippery creatures. Yeah, the difference between an "E" and an "IH" and an "OO" and an "AH" just has to do with where your tongue is, and it's not like a — it's like the difference between when you play piano keys and you've got, you press this key and it makes that sound versus playing on a violin where you can slide your finger up and down and go in between notes all the time. So vowels can be slippery like those on a violin. It's really easy for them to change. It's really easy for them to shift slightly, subtly, in any direction.
MIGNON: Interesting.
ANDREW: But changes in consonants have to be a little bit more concrete.
MIGNON: Fascinating. And I'm not familiar with the Hawaiian accent. What was changing there from Hawaii to the West Coast?
ANDREW: One hallmark of the Hawaiian accent is a lot of what we call monophthongized vowels. And so that's a vowel that maybe sounds more like a vowel that undergoes a little bit of change, like saying "uu," and you've got that little off-glide at the end, "uu."
And in Hawaiian English, that becomes monophthongized; it becomes more stable, so just "uu." Actually, maybe a better example would be "o" in the word "boat." We're gonna go on a boat. My Californian English and a very standard American English pronunciation is "O." You've got a little off-glide; you can look at my lips, "O, boat." But for Hawaii English, we're more likely to hear "boat," especially for those who are very local and grew up in Hawaii.
So we've got that vowel monophthongization that happens. We also have, ah, another example that I write about in the paper is the vowel "a." For my particular dialect, Californian English, if I say the word "can," "C A N, can," the presence of the "N" sound will change the nature or the quality of the "a" before it, so it becomes more like "can" versus "can," and "can."
Very subtle difference; hopefully, you can hear it.
MIGNON: Yeah, I can.
ANDREW: For Hawaii English, we don't really have that, so it's more "can" all the way. Actually, sometimes more "can." So what I found for our Hawaii YouTube vlogger was after moving to Nevada and just being surrounded by more standard or West Coast American English rather than Hawaii English, there were some aspects of his vowels that became less monophthongized, less Hawaiian over time, and more like standard American English.
And I'm always using scare quotes when I talk about standards, because who is — just like who is to determine what is a language versus a dialect, who gets to say what is standard and what is not standard? It's just whoever has power, right?
So for both of these YouTube vloggers, I'm noticing that it was just little parts of their vowel systems that really subtly changed.
Maybe it's to the extent that a person who's really paying attention could notice, but in other areas, like, some of these other changes were so subtle that I would be surprised if anyone really noticed the change. And it really speaks to this idea that a lot of linguistic change and linguistic convergence — like when we change the way we speak — sometimes it's so subtle that we don't even notice until years down the road.
And then you listen to your past self and think, "Whoa, that's literally what happened with one of these vloggers," the one from Boston, where she, in one of her videos in the late 2010s, was like, "I listened to my early YouTube vlogs and cringed at my accent," and she makes commentary on how, "Wow, I sounded so Boston back then," and now she doesn’t.
So she is aware of that change. But there's many times, like during the change is happening, it's very hard to notice.
MIGNON: And why do you — I know that we tend to talk like our friends and the people that we're around. Is it as simple as that? Why do people change the way they speak, or is there more to it than that when someone moves like that?
ANDREW: A sociolinguist would probably say that a lot of it has to do with prestige and how positively you evaluate people who speak a certain way. This could be people that you talk to all the time or could be like what you hear on the radio. You don't have to necessarily be in conversation with somebody to pick up verbal tics in the way that they speak.
And you also don't have to necessarily like somebody to also maybe subtly pick up on ways that they speak, too. A lot of what we call convergence, linguistic convergence, happens subconsciously, but it can be affected by your evaluation of that particular accent or dialect that you hear.
MIGNON: And I can see the pros of studying people who have been making YouTube videos for years and years because you've got all that data. I imagine, though, that there must be some disadvantages.
ANDREW: Yeah, as an experimental linguist, one thing I like to have is a lot of control. Not control in a "you must do what I say" way, but control as in, like, the ideal situation for recording vowels or consonants is just to have someone say the same
word in the same
tone, in the same place, like the same recording with the same equipment, once a year for 10 years.
If we could really control, tightly control, all the other contexts of sound, that would be the best way to really verify if the sound, if the change that we're observing has to do with change in the voice rather than a change in other surrounding contexts. But you can't do that when you're using found data.
The same thing happens for any social scientist who wants to look at a Twitter corpus or anyone who's just studying, I don't know, comments on a Reddit thread of some kind. You can't really control who is giving you the information, why they're giving you the information, or who people really are.
So there's a lot of unknowns in the data. That could be one disadvantage. I think that's the main one.
ANDREW: And then there's the audience effect, which has to do with people speaking differently depending on who they're talking to. So the way that I'm speaking now to you, or to the imagined future audience for the podcast, is definitely different from the way I talk to my students when I give a lecture, and that is definitely different from the way I talk to my partner or to my friends, and that's definitely different from the way I talk to my family.
Even disregarding different languages that I would use with each of these people, even if I was using English, the tone, the intonation, the stress patterns, my pitch—all that's going to be quite different. And so when we're thinking about the audience of the people that I studied for this project, perhaps the way that they were speaking early on in their careers is quite different before they became famous versus after they became famous and their audience grew by a thousand times. Or maybe in the different genres of videos that they made. Some videos being straight-up monologues, me talking to my camera, versus others being a conversation or others being some kind of recorded
skit that was very popular in the 2000s and 2010s. All those things need to be taken into consideration and can't be really controlled for as well when you're doing this kind of found data analysis.
MIGNON: Yeah. Yeah. I imagine people hear a big difference in my voice when … because I do two shows a week and one is scripted and one is an interview conversation like this. And I know I speak very differently in those two situations. Yeah.
ANDREW: I would love to take a look at the phonetics of that if you're curious.
MIGNON: It's all public.
So yeah, would you do a study on YouTubers again like that? Do the pros outweigh the cons?
ANDREW: I think that I would. I would love to continue on in this vein of linguistic studies of popular media. And I think it's mostly, obviously from my standpoint as an academic, the theoretical question of what drives second dialect acquisition is quite interesting, but I'm also always on the hunt for ways to get more students and more just laypeople engaged in linguistics.
And if I can do that by analyzing the voices of celebrities or YouTubers or movies, anything that gets people drawn into, wow, what is linguistics? I consider that a win, which is why I do what I do.
MIGNON: Wonderful. That is so interesting. We're going to wrap up. We always ask for book recommendations. It's time for the book recommendation segment. Pretend we are your best friends and what books would you tell us to read?
ANDREW: Oh my gosh, Mignon, have you read "Babel"?
MIGNON: Yes. Oh my gosh. I have.
ANDREW: R. F. Kuang. It is so good. Now, people were recommending that book to me before it was even published, and I put off reading it for two years for who knows what reason. But the minute I picked it up and opened it, I was like, "Oh, okay." It’s about linguists. It's an epic fantasy about linguists in an alternate universe where linguistics is magic.
MIGNON: Magical linguists.
ANDREW: I just thought this book was written for me, and it totally was. It's so good.
Highly recommend anyone who likes fantasy or is interested in languages or just wants to fall into this rabbit hole of a really well-constructed fantasy world. From a more academic perspective, if you've got those in the audience, a lot of the trajectory of my academic career was changed after I read a history book called "The Making of Asian America." It's by Erica Lee, published in 2015. It's very, it’s like a textbook, but it's written extremely engagingly, and it just covers every community in the United States that could be defined as Asian American.
Who are they? When did they get here? Where did they come from? And what are some really important pieces of their history and how they contributed to American history that, you know, maybe you've never heard of or maybe they don't teach in schools very often? So that was one that, after I read that, I just thought, "Oh, there's so much more to my identity as an Asian American and the fabric of American culture than I ever thought or I ever imagined."
MIGNON: Amazing. That sounds like one of the books. It sounds like a book you might have been assigned for school, but then you would keep forever because it was so interesting and useful.
ANDREW: Yeah. And I sadly never took an Asian American Studies course, so I wasn't assigned it, but I'm so glad that my friend recommended it to me, so I'm recommending it to you and your audience.
MIGNON: Wonderful. Thank you. And is there a third?
ANDREW: And the last one, yeah, the last one is "Parable of the Sower," by Octavia Butler. So this is a sci-fi classic, and I'm just recommending it now because it seemed so weirdly prescient.
It is about a dystopia, and I'm not going to comment too much on the current state of the world, but I do think that if you read "Parable of the Sower," you might wonder, was this really written 22 years ago? No, 32 years ago? Or was it written last year?
MIGNON: Yeah, I've been seeing a lot of people recommend that book. I actually checked it out from the library, and then it expired before I got to reading it. So ...
ANDREW: I highly recommend it, and it also has a sequel called "The Parable of the Talents," and both of them are very short, very readable, and a fascinating look into the genius mind of a fantastic sci-fi author, Octavia Butler.
MIGNON: Those are wonderful. Thank you so much, Andrew Cheng from the University of Hawaii in Honolulu. Thank you so much for being here. Where can people find you?
ANDREW: I had so much fun, thank you. You can find me on Bluesky, LinguistAndrew. Also on Twitter/X, and I do have an academic website. If you just want to Google my name and University of Hawaii and linguistics, you'll probably find me.
MIGNON: And that's Andrew Cheng, C H
ANDREW: C H E N G, yeah.
MIGNON: Awesome. Thank you so much for joining us today.
ANDREW: You're very welcome. Thank you so much for having me.
I hope you enjoyed that. And if you want to hear Andrew talk about language in movies like "Parasite" and "Anora," you can find that episode with him on your feed; it's episode 1060.
Finally, thanks to all the Grammarpaloozians who support the show! These bonus episodes happen because of you. If you'd like to help support the show, you can sign up right on the show listing in Apple Podcasts or learn more about other ways, including getting text messages with fun tidbits from me at quickanddirtytips.com/bonus.
That's all. Thanks for listening.