Why does everyone say OOO! when they see someone fall down? Why do we say YUM when we feed a baby? And what’s the deal with fillers like UM?
For this episode we’re talking about non-lexical vocalisations with Dr Eleonora Beier and Dr Emily Hofstetter.
Also: linguists are diving into Grambank, a database with detailed information about grammatical features in over 2,500 languages. With its release, we’re talking to project leaders Dr Russell Gray and our own Dr Hedvig Skirgård.
Listen to this episode
Patreon supporters
Huge thanks to all our great patrons! Your support means a lot to us. Special thanks to our patrons at the Supporter level:
- Iztin
- Termy
- Elías
- Matt
- Whitney
- Helen
- Jack
- PharaohKatt
- LordMortis
- gramaryen
- Larry
- Kristofer
- Andy
- James
- Nigel
- Meredith
- Kate
- Nasrin
- Joanna
- Ayesha
- Steele
- Margareth
- Manú
- Rodger
- Rhian
- Colleen
- Ignacio
- Sonic Snejhog
- Kevin
- Jeff
- Andy from Logophilius
- Stan
- Kathy
- Rach
- Cheyenne
- Felicity
- Amir
- Canny Archer
- O Tim
- Alyssa
- Chris
- Laurie
- Aengry Balls
and our newest patrons at the Listener level: Benedict and Terry
and at the Friend level: Boo88
FauxFox also bumped their pledge up to the Listener level
and Tadhg doubled their pledge!
Become a Patreon supporter yourself and get access to bonus episodes and more!
Become a Patron!Show notes
Grambank
https://grambank.clld.org
About Grambank | Max Planck Institute for Evolutionary Anthropology
https://www.eva.mpg.de/linguistic-and-cultural-evolution/research/grambank/
World’s largest grammar database reveals accelerating loss of language diversity
https://www.colorado.edu/today/2023/04/19/worlds-largest-grammar-database-reveals-accelerating-loss-language-diversity
The rise and fall of linguistic diversity at Eurovision
https://globalvoices.org/2023/05/01/the-rise-and-fall-of-linguistic-diversity-at-eurovision/
Dr Beier’s website
https://eleonorajbeier.wixsite.com/ejbeier
Do disfluencies increase with age? Evidence from a sequential corpus study of disfluencies
https://www.researchgate.net/publication/369557791_Do_disfluencies_increase_with_age_Evidence_from_a_sequential_corpus_study_of_disfluencies
Jennifer E. Arnold, Maria Fagnano & Michael K. Tanenhaus | Disfluencies Signal Theee, Um, New Information
https://link.springer.com/article/10.1023/A:1021980931292
Leelo Keevallik, Emily Hofstetter, Ann Weatherall, & Sally Wiggins: Sounding others’ sensations in interaction
https://doi.org/10.1080/0163853X.2023.2165027
Emily Hofstetter, Leelo Keevallik: Prosody is used for real-time exercising of other bodies
https://doi.org/10.1016/j.langcom.2022.11.002
Emily Hofstetter, Leelo Keevallik, and Agnes Löfgren: Suspending Syntax: Bodily Strain and Progressivity in Talk
https://www.frontiersin.org/articles/10.3389/fcomm.2021.663307/full
Emily Hofstetter’s YouTube channel
https://www.youtube.com/@emdoesca4249
NON-LEXICAL VOCALIZATIONS: Vocal practices for achieving embodied coordination
https://nonlexicalvocalizations.com
Here’s why some LGBTQ youth are now embracing the nonbinary pronoun ‘it/its’
https://www.yahoo.com/lifestyle/heres-why-some-lgbtq-youth-are-embracing-non-binary-pronoun-it-its-223331366.html
It/Its pronouns are no less valid than any other pronoun set
https://themicheab.medium.com/it-its-pronouns-are-no-less-valid-than-any-other-pronoun-set-ce8ce0c7de96
Diversity of Nonbinary Youth | The Trevor Project
https://www.thetrevorproject.org/research-briefs/diversity-of-nonbinary-youth/
How Women End Up on the “Glass Cliff”
https://hbr.org/2011/01/how-women-end-up-on-the-glass-cliff
The Glass Ceiling: Definition, History, Effects, and Examples
https://www.investopedia.com/terms/g/glass-ceiling.asp
100 Women: ‘Why I invented the glass ceiling phrase’
https://www.bbc.com/news/world-42026266
The invisible danger of the ‘glass cliff’
https://www.bbc.com/future/article/20220204-the-danger-of-the-glass-cliff-for-women-and-people-of-colour
Paul Smith, Peter Caputi, Nadia Crittenden: A maze of metaphors around glass ceilings
https://ro.uow.edu.au/cgi/viewcontent.cgi?referer=&httpsredir=1&article=4091&context=hbspapers
Transcript
[Transcript provided by SpeechDocs Podcast Transcription]
DANIEL: Nora, I feel myself doing more UMs and UHs because you’re doing it.
DR ELEONORA BEIER (hereafter NORA): [LAUGHS]
DANIEL: That’s odd.
NORA: Yeah. I mean, then you start becoming hyperaware of it.
DANIEL: [CHUCKLES] Yeah, I suppose.
NORA: Um…
[BECAUSE LANGUAGE THEME]
DANIEL: Hello, and welcome to Because Language, a show about linguistics, the science of language. I’m Daniel Midgley. With me now: she’s the proper linguist on the show, and that’s because she got her degree in Proper Linguistology. It’s Hedvig Skirgård.
HEDVIG: I like linguistology. I think it’s a great name. We should change the whole discipline name to that.
DANIEL: Right! Because if you study art history, you don’t become an artist, you become an art historian. Eh?
HEDVIG: Mhm!
DANIEL: We study about linguistics. And occasionally languages.
HEDVIG: We study about languages. We study linguistics. Right?
DANIEL: Languology. I propose Languology.
HEDVIG: Languology I’m pretty open to.
DANIEL: Okay.
HEDVIG: In general, I believe that there’s so many debates about, like, Oxford commas and singular THEY and whatever, and people have all these… you got to pick sillier things and then choose those hills to die on. You know?
DANIEL: Sillier things?
HEDVIG: Like languagology instead of linguistics.
DANIEL: Right. GRAMMATICALISATION versus GRAMMATISATION versus…
HEDVIG: No, that’s too important. That’s… people actually argue about that. You got to pick something sillier and then argue about that. And by that, point out that the discussions they’re having are maybe not that important.
DANIEL: I still feel like that’s pretty silly.
HEDVIG: No.
DANIEL: Anyway.
HEDVIG: That’s what I mean. That’s the point.
DANIEL: Okay. You know what? You can always go sillier. That is true.
HEDVIG: Yeah! You can always go sillier. And when people are being silly, you should bring them down to earth by just pointing out, like: Hey, this is not that important. We’re all going to die. You know, I saw someone on TikTok say, “You should start every meeting with, “We’re all going to die. Nothing’s that important.”
DANIEL: Death is coming for us all. I’m not going to get bent about the Oxford comma. I’m going to get bent about the Cambridge hyphen. No! The Cambridge semi-hyphen. Yeah. Now we’re getting very silly. Well, Hedvig, I’m glad that you’re here with me to get silly on this episode because you’re doing a lot for this one.
HEDVIG: Uh, yeah! I’ve invited my boss on, which is scary and fun. So, we look forward to a fun segment with Dr… so he’s technically Director Professor Dr Russell Gray. I think he has all those titles, and Germans love to use all of them, but I just know him as Russell.
DANIEL: Oh, okay. I better use them then.
HEDVIG: No, because he’s from New Zealand. He’s kiwi, he’s not going to get hung up on…
DANIEL: Oh. Well, that’s all right, then.
HEDVIG: Yeah. But anyway, Dr Russell Gray, my boss and project leader for Grambank is joining us to talk about Grambank. And then at some point, we’re also going to talk about Eurovision.
DANIEL: Mm-hmm. This is our annual Eurovision episode.
HEDVIG: Those are my jobs for this episode. What are your jobs?
DANIEL: Ben’s not here, so let’s do it. My job is, well, our usual Words of the Week, which I think actually you did a bunch of those too. I’m feeling a bit garfunkelish. Hm.
HEDVIG: Really? Did I?
DANIEL: Yeah, you did, actually.
HEDVIG: Oh, yeah. It’s because… I’ll tell you, Daniel, what explains how many Words of the Week I’ve added, it’s that I’ve realised that, A: I can use Discord on my phone.
DANIEL: Oh. Good.
HEDVIG: So, whenever I have a bright idea, I can immediately go to our patron-only Discord, which you can join by becoming a Patreon, and you can hear all my crazy ideas.
DANIEL: Yeah, you’ve been dropping those in the show-ideas channel, haven’t you?
HEDVIG: That’s the right channel, right?
DANIEL: Yeah, I do that too.
HEDVIG: Okay, good. Anyway, so I have those thoughts, and then my day goes on and I completely forget about them. And then, Daniel notes them down and writes them in our show notes. And then, he goes…
DANIEL: Yes, I do.
HEDVIG: “Oh, Hedvig! This Word of the Week.” And I go: Oh, yeah, I saw that on TikTok two weeks ago. Uhhh… [CHUCKLES]
DANIEL: “Oh, that made it in? Oh, shit.” All right, good. Ben is not here with us this time, unfortunately. He is occupied but we’re going to get him again the next time around. What my job is, is I’m going to be talking about not words, but things that aren’t words. For example, when everybody’s watching someone on stage and there’s a blunder where they fall down, they have a big spill, and everybody in the audience goes, OHHH! Why do we do that?
HEDVIG: That is weird. You’re right.
DANIEL: When we feed a baby, why do we say YUM, even though we’re not eating anything? And what’s the deal with fillers like UM and UH? Eh? You ever thought about those not words?
HEDVIG: I have. And if you pay attention to different people, you’ll notice that people have preferences. I know a guy from Australia, his favorite filler is, YOU KNOW.
DANIEL: Oh, yeah.
HEDVIG: Oh, yeah. Yeah.
DANIEL: It’s really obvious when somebody uses it a lot and I have to edit them.
HEDVIG: I feel for you.
DANIEL: Yep. No, no. It’s good.
HEDVIG: What’s mine?
DANIEL: You don’t… er. Uh. I listen to you a lot, Hedvig, and I haven’t noticed anything really intrusive.
HEDVIG: [LAUGHS]
DANIEL: You just do things normally. I think we use LIKE a lot.
HEDVIG: Massive cop-out.
DANIEL: No.
HEDVIG: But I don’t use it as a floor-holding filler, do I?
DANIEL: No, I would have said something years ago if there were anything that was super, super intrusive.
HEDVIG: You’re just a coward.
DANIEL: But I am going to be talking… what if I just don’t have anything? What if I just like the way you talk? [FAWNING VOICE] ~I just like the way you talk. You just sound good. I think you’re grouse.~ I’m going to be talking to two scholars who have done work on the non-lexical things that we say, that is, things that aren’t words. Non-lexical vocalisations like “hmm.” We’re going to be talking to Dr Eleonora Beier of the University of California, Davis. Also to Dr Emily Hofstetter of Linköping University. And we’re going to find out more about how these seemingly peripheral nonword items can actually tell us a lot about how we use language, how children learn language, and what are we doing when we combine our minds. So, that’s going to be for this episode.
HEDVIG: That’s really cool. Actually, when I first started doing linguistics, that’s what I thought linguistics was mainly about. And then, [LAUGHS] I discovered that there was quite a little amount of research on it.
DANIEL: It’s true.
HEDVIG: And that it was hard to do research on it. So, I’m very excited to hear what they’re going to say.
DANIEL: We’re digging into it. We’re rolling up our sleeves.
HEDVIG: Sweet.
DANIEL: And you know what? We’re also doing something very special on our next episode as a bonus for patrons. It’s very unusual. We’ve never done this before. We have arranged a kind of mini conference for our episode…
HEDVIG: Nice!
DANIEL: …because we found out that there was some work that we like from Dr Caitlin Green, Dr Rikker Dockum, and Dr Aris Clemons. They’ve all published their work in the oxford collection on inclusion in linguistics and also in decolonising linguistics, asking questions like, “How do we make linguistics a bigger tent? How do we make linguistics a better discipline for everybody?” And we’ve asked them, and they have agreed to present their work on a special live episode, and all of our patrons are invited.
HEDVIG: Ooh! That is such a good idea. Who had that idea? It wasn’t me, but I kind of want to steal it.
DANIEL: It was me.
HEDVIG: It was you! It’s such a good idea. It’s so much fun.
DANIEL: Good idea, Hedvig. You did great. That was such a good idea you had.
HEDVIG: [LAUGHS] No no no, it’s fully Daniel’s idea. I’m just being cheeky. But that is actually a really good idea. I’m very glad we’re doing it.
DANIEL: Yeah. I’m just so interested in this. And they had presentations. We’re all going to pile into a big old Zoom room. Everybody. Patrons can ask questions. It’s going to be super. So, keep watching your Patreon account if you are a patron. If you’re not a patron, why don’t you become one? You’ll get all kinds of bonuses like live episodes, bonus episodes, shoutouts, mailouts. And, of course, you can hang out with us on our Discord server which, you know, I mean, it’s kind of fun sometimes, it’s not too bad.
HEDVIG: I just upload nice cat pictures, so that’s pretty good.
DANIEL: Hedvig uploads cat pictures where they’re pretending to be plants.
HEDVIG: Yeah.
DANIEL: So, that’s patreon.com/becauselangpod. Our first big piece of news is that Grambank has been released. It’s a database with a lot of information about many different languages in the world, what kind of behaviour they show. Hedvig, you are the project coordinator on Grambank, so I want to ask you about it. But we also have a very special guest with us, the project leader, the eminent Dr Russell Gray of the Max Planck Institute for evolutionary anthropology. Russell, thanks for coming on and telling us about Grambank.
DR RUSSELL GRAY: Thank you. Thank you very much, Daniel.
DANIEL: All right, both of you. I’m going to ask you questions. First up, tell us how Grambank came to be.
RUSSELL: Well, it was a long time in the making. We’ve been working on it for over 10 years. It goes back to an idea that Quentin Atkinson and I had, I think, in 2012. And we thought, “Wouldn’t it be wonderful to have a global database with all the information you ever want to know about the world’s languages?” And we put in a grant for that, and obviously that was ridiculously ambitious, and we didn’t have the resources to do it, but at least it got us started. And Grambank is one realisation of that vision we had a long, long time ago.
DANIEL: Okay, now, if I am a linguist, what am I going to be using Grambank to do?
RUSSELL: If you want to find out about grammatical variation, the standard things that are covered when people write grammars or sketch grammars, grammar is for you, but it’s not all things to all people. In the early phases, and Hedvig will remember this very well, spent a [HEDVIG CHUCKLES] lot of time discussing vigorously about which features to include in Grambank. So, it now contains 195 features that range from word order, verbal systems, case marking, gender marking, things like that. But it won’t contain… it can’t contain every possible grammatical feature that any linguist on the planet might be interested. So, we excluded things that were very hard to code consistently or that were sort of only found in a small number of languages, a small number of areas, because you want it to have a large set of data. We wanted things that could be compared across most of the languages of the world.
DANIEL: Okay, so for example, if I want to know if a language has dual and plural pronouns, is that going to be in there?
HEDVIG: That one’s in there, yeah.
DANIEL: Okay, what else?
HEDVIG: We have lots of questions, and like Russell said, we can’t be everything to everyone. So, I don’t doubt that if our listeners go in and look at grambank.clld.org, you might be like, “Oh, but my favourite thing that I love about verbs is not in there.” And that’s all right, that’s okay, we’re one dataset, but we have a lot of questions. So, 195 features actually cover a lot of topics. We cover things like comparison, like, “Hedvig is more giggly than Daniel” is a kind of comparison. And we have questions about that. We have questions about negation, tense and aspects, and loads of fun things. The Grambank website is up. Anyone can go there and play around right now.
DANIEL: Cool. How many languages are we talking about here? The world has 7000-ish languages. How many have information included?
HEDVIG: So, 2,000…
RUSSELL: So, in the first… Go ahead, Hedvig.
[LAUGHTER]
HEDVIG: This is funny because Russell and I both have been sitting through some interviews and said the same things.
[LAUGHTER]
HEDVIG: So now, we’re competing. The first version has 2,430 languages in it and some dialects of some languages.
RUSSELL: The aim is to cover every language that there’s a grammar or sketch grammar for. So, we hope to cover, I guess, up to about 3,500 languages. And our coverage at the moment, it’s pretty good in terms of covering all the major language families, all the major kind of world regions, and say if you compared it with previous projects that are a bit similar, like the World Atlas of Language Structures, one of the problems with WALS is if you think of a matrix of languages with all the features, it’s 85% empty. It has a lot of missing data. Whereas, I think, Grambank’s only 24% empty in terms of filling out that matrix. So, we’ve tried to maximise the coverage and get, as far as possible, sort of an unbiased sample of the grammatical diversity of the world’s languages. A lot of previous work has focused far too much on languages of Europe and we went out of our way to make sure that we didn’t do that.
DANIEL: What about signed languages? I know that those are just different in lots of ways, but is there any way to incorporate them?
RUSSELL: We have two, right, Hedvig?
HEDVIG: Yes, so we have, I believe it’s Finnish and Japanese sign languages in there. This was a collaboration with Hannah Lutzenberger, who used to be at Radboud and now is a postdoc somewhere else, I’m sorry, I forget exactly where, but she works on sign language typology and she filled in two languages. But it’s hard because we have questions like, “Do you use tone to mark blah, blah, blah?” What is tone in a sign language? Is it a super segmental feature, like raising your eyebrows? Is that tone? So, some features make sense, some features are harder to translate directly from typology on spoken languages to sign languages. And there are efforts like Ulrike Zeshan and her group and other groups are trying to do sign language typology starting from sign language principles. And in Grambank, we did start with thinking mainly about spoken languages, so not all features are going to fit really well. We have had a lot of interest. So, besides the languages we coded, we’ve also been contacted by people who said they want to contribute more sign languages.
DANIEL: That’s cool.
HEDVIG: Yeah! Yeah. And we’re open to that. We’re very happy. We love sign languages. We just know that not all features are going to fit very well.
DANIEL: That’s fair.
RUSSELL: Perhaps, this is a good point to say that this is just the first release and it’s an evolving project and we’ll be adding more languages. But also in Grambank 2.0, we will move from covering just sort of core grammatical features, if you like, to adding in other data sets that are perhaps regionally more specific or perhaps we could add in more systematically sign languages. So, the idea is that we cover these 195 features in a common way, but it also connects in little future versions as a repository for other datasets that will be linked into the Grambank ones. So, I think it’s the start. It’s certainly not the final, final product.
DANIEL: Well, it’s exciting to see it expand. I’m curious about language relatedness. When languages do some of the same things, like they have the same kind of word order or they have gender on their nouns or whatever, there could be a couple of explanations. One is that the languages are related, and they both do that thing because they both come from a parent language that did that thing. But another is that the people who speak those languages have been hanging out in the same area for such a long time, and they influence each other because they had to communicate, and that’s why they do the same things. Does Grambank tell us anything about genetic relationship versus areal similarity?
RUSSELL: Absolutely. And that was one of the questions we tried to answer in the initial Grambank publication. And really, much to my surprise, we found that there was a lot of really strong genealogical signal. As you point out, for a long time, linguists have debated the relative importance of genealogical inheritance versus geographic diffusion. And typically, people will point to certain features and tell a certain kind of narrative. But there’d been no global quantitative evaluation. And really, I was actually pretty skeptical about the degree of genealogical signal in these features because, for a start, often there’s a limited number of possibilities, so you can get parallel evolution, and there’s well-known examples of areal diffusion. But overwhelmingly against over all sorts of domains of grammar, we found that genealogy substantially more important than geography, which surprised me.
DANIEL: Wow. And is that because, do you think when people are in contact, they’re just good at switching? When they stop talking to those other people, they go back to talking how they were going to talk? They don’t change very much, and they don’t influence each other? Or maybe all the influence is in vocabulary. I don’t know.
RUSSELL: I think partly it’s just that people haven’t really quantitatively assessed it. And often, linguists are more interested in the differences, the things that don’t go with general things. I think it’s partly because the examples have been cherry-picked in the past. I mean, certainly areal diffusion really does happen. But overall, what we did was to fit the 195 grammatical features on a global language tree, which was a sort of summary of what we currently know about language relatedness. It combines all the classification evidence in Glottolog with Bayesian phylogenies, and with other bits of information to produce a sort of, if you like, draft global genealogy for the world’s languages. And when you fit the data on that kind of way, there was really striking genealogical signal.
HEDVIG: And we should say as well that what we did was, like, what Russell described, is that we took our grammatical features and tested for each of them if genealogy or space was explaining more, that’s not the same as comparing to how much more genealogy plays to lexicon, right?
RUSSELL: Yes, exactly.
HEDVIG: So like, it could be that lexicon is still more predicted by genealogy than grammar is. But the thing was that what was surprised us is that space was having such a small effect.
DANIEL: Amazing. I want to ask you about language endangerment. Has this project of Grambank told you much about language endangerment so far?
HEDVIG: Well, we know, and as many of our listeners of our show know, that there are a lot of languages in the world and a lot of them are under threat of no longer being spoken or transmitted in any way to the next generation. And as we’re heading into the United Nations Decade of Indigenous Languages, we’re as a discipline thinking more and hard about what this means for people and communities and what it means for research and all the impacts it has. We’re losing this window into the great diversity of the human mind. And we should also be clear, as we are on this show, about the causes. So, the causes are primarily linked to colonialism and the effects of colonialism. But specifically in our paper, we wanted to ask, besides the tremendous loss that this is to the communities, what do we lose as scientists? What do we lose as researchers? And we found that these languages that are currently classified as under threat, if they do end up going away, if they do fall asleep, then we lose a disproportionate amount of knowledge about what it means to be human. The languages that remain are not as diverse as all the languages that are alive today.
RUSSELL: And I think if I could just add to that, the thing that the Grambank analysis showed, I think most linguists are aware of the language endangerment situation, and most linguists are aware that diversity gives us a window into human history, human culture, and human cognition, that once it’s gone, it’s gone. And hence, the vital importance of language documentation works. But I think the thing that the Grambank paper showed is that the loss is going to be very uneven. That in some areas of the world, like North America and Northern Australia, the situation is really quite dire. What we did was use a kind of measure from ecology that had been used to measure ecological diversity, and we applied that to grammatical diversity. And the scenario is really quite bleak in some regions, for example, Australia and North America.
DANIEL: Mm. So, we found out from Grambank a bit about language relatedness, language endangerment. What other projects are you aware of that people are using this database for?
RUSSELL: Lots. [LAUGHTER] Really, this is just the first thing. To name a few, just this week, a group of us have been working on a paper testing claims about language universals. The idea that the human mind constrains the range of possibilities. So, we’ve used the Grambank data to test 191 putative linguistic universals. And I’m really quite delighted with the results that we’re just writing up at the moment.
HEDVIG: And to maybe illustrate for some of our listeners who have been through and through the generative episode and things, we’re talking about universals here in what might be known as like the functional or typological sense. So, for example, all languages that have a verb at the end have their post positions, so like the “boat on” instead of “on the boat” and not the other way around. And these can be just statistical. They don’t have to be always true. It’s not the same as Universal Grammar, but it’s related.
RUSSELL: Yeah, exactly. Yes.
DANIEL: There’s a tendency.
RUSSELL: Exactly. So, when I say universals, that’s just shorthand for statistical tendencies. And yeah, we use very powerful Bayesian phylogenic methods to test this. And because we’ve got this large sample of languages, we have a lot of statistical power to detect these statistical tendencies. And I’m not sure if I want to give away the answer yet. Maybe people will just have to…
HEDVIG: No, let’s keep them…
DANIEL: Okay. That’s cool.
HEDVIG: Don’t do spoilers.
DANIEL: I would love to know, though, are there any just weird little things that tend to travel together that you just wouldn’t expect at all? When we see this one thing, we also see this other weird thing, and we weren’t expecting that to happen.
RUSSELL: I think it should have to wait. I’m sorry.
DANIEL: Aw, man, okay.
RUSSELL: Maybe as a teaser, I’ll just say that you could say the glass is half full or the glass is half empty. I’m the glass is half full. I think about half of the putative linguistic universals, these statistical tendencies survive our analyses when we control for spatial and genealogical nonindependence. So, we have to wait and see until we publish it which ones survive the tests and which ones don’t.
DANIEL: [CHUCKLES] I’ll be looking forward to that.
HEDVIG: Good teaser, good job.
DANIEL: Yeah. Such a tease. Oh, my god.
RUSSELL: Yeah, just to keep teasing. One more. Another one is to test claims about the linguistic niche hypothesis, the idea that societies of strangers have more complex languages than societies that are large and have lots of second language speakers. And again, because we have this global sample, we can test that claim about how the sociolinguistic environment might shape patterns of grammatical complexity and we have the papers currently under review. Fingers crossed about that.
And again, I’m not going to tell you what we find. So, if people are interested, they should come along to a workshop we’re having in September here in Leipzig on the 13th and 14th September, where we’ll be showcasing, discussing, and evaluating lots of these Grambank-related projects.
HEDVIG: I was going to say, if you like things like this and you log onto our website and you find a language that you like and you find that we have missing information or that you doubt our coding or something like that, you can get in touch with us. We’re still, like Russell said, going to be publishing new versions. So, we are open to people contributing and saying, “You got this thing wrong,” or, “Here’s some information that you didn’t have.” So, if you do that on our website, and you go under Contact, you can find out how to do that. And we’ll also have a workshop in September with limited amount of spots where we’ll guide people through coding and adding and revising our coding. Which I’m really looking forward to.
RUSSELL: That just reminded me. We’ve really forgotten to mention a really important thing about Grambank, and that is about the coders and that you coordinate in the system. And I think this is again something that’s quite different about Grambank. Rather than just reflecting the work of one or two people or collating a whole lot of individuals’ work. What we did with Grambank is create this global network of young linguists who are coding grammars. But we did it in what I think is a really interesting way, in a way that produces much more reliable results, that each of these coders is connected via GitHub to a feature patron. So, every feature in Grambank, and there’s 195 features, has an expert that advises on the coding of that feature. And so, what typically happens is that the coders are working away, but they’ll encounter sort of problematic edge cases and then they’ll raise a query on GitHub and the feature patrons will kind of advise, and there’ll be some iterated discussion. And in this way, I think the judgments through that iterated discussion become a lot more reliable. So, we really leverage this global community and this interactive discussion to, I think, produce as reliable of results as we possibly can. So, thanks to all our wonderful team of coders and to our feature patrons, and to Hedvig who coordinated all of this.
DANIEL: Wow, it’s like you’ve got all these language experts filling down the column and then crossing that, you’ve got the feature experts working their way across on all the different languages. That’s really cool.
RUSSELL: And I think another good feature is rather than just having the decision there, actually if people are interested in, say, the coding of one particular thing in one particular language, they can potentially burrow in. And we’ve kept a record of all that scholarly discussion on GitHub, so you can see the discussion that went on behind the datapoint.
DANIEL: What’s wonderful to me about being a linguist is that I get to see individual cases of language use, but I also love getting to see across the broad spectrum of a lot of time or a lot of different languages. It must just be amazing having it all before you like a gigantic landscape. Does it feel that way?
HEDVIG: …um?
DANIEL: What’s it been like working on this?
HEDVIG: It’s been hard.
RUSSELL: Yes, I was going to say exactly the same thing.
HEDVIG: I don’t want to lie.
RUSSELL: Coordinating a manuscript with 105 authors, that’s not easy. Coordinating a manuscript with 105 linguists, that’s even harder. And this has been…
HEDVIG: [LAUGHS] We’re not that bad.
DANIEL: Yes, you are.
[LAUGHTER]
HEDVIG: Okay, all right.
RUSSELL: People do like to argue. It has been hard work. And I think the job, it’s not finished yet. I think the challenge now is to keep going and cover all three and a half thousand languages that we could potentially have grammars or sketch grammars, to add new sets of features, and to answer interesting questions with the data. So, I think it’s started. It’s nowhere near finished yet.
DANIEL: What a fantastic project. Thank you for doing it, Dr Russell Gray and Dr Hedvig Skirgård. Thank you for coming and talking to me, us, each other about this project. And where can people find it?
HEDVIG: grambank.clld.org
RUSSELL: Thank you. And just a final thought. Look out in the future for Numeralbank and Parabank.
HEDVIG: Ah, yes!
RUSSELL: They’re coming too.
DANIEL: Whoa.
HEDVIG: We have so many banks. We’re the bank department.
DANIEL: You’re the bank bank.
HEDVIG: We should start minting money. We have a bank-bank.
DANIEL: Hold on, go back. Numeralbank, about?
HEDVIG: Guess. [LAUGHS]
RUSSELL: Numeral systems of the world. So, we have a project when we are writing this up at the moment, where we have data from over 5000 languages of the world characterising the diversity of the numeral systems. It’s part of our QUANTA ERC Synergy Project. So, that’s coming. I hope that will be out by the end of the year, hopefully.
DANIEL: And Parabank?
RUSSELL: Parabank, a database. This is a project in collaboration with researchers at ANU, Nick Evans and Kyla Quinn. And it’s about pronoun paradigms and it’s also linked to kinship paradigms.
DANIEL: Oh, man.
RUSSELL: They’re working on getting that out, I think, also by the end of this year.
HEDVIG: And we’ve already released Lexibank as well. So, Lexibank was released first. That’s Lexicon and Clicks as well, which is connections between Lexicon led by Mattis List and his team. And then, we have Grambank that we released now. And then we’re going to release more banks. We’re a fountain of banks.
DANIEL: Oh, my god. See, I was trying to drive to a conclusion, but you just opened it up there because now I have more questions. [LAUGHTER] We’re not going to get to them today. We’re going to have to come back and do some more. Wow.
HEDVIG: Yeah.
RUSSELL: Thanks. All language knowledge in the entire world. Amazing.
[LAUGHTER]
DANIEL: Thank you both for being here.
RUSSELL: Thanks, Daniel.
HEDVIG: We’re nothing if not megalomaniacs.
DANIEL: Now, it’s time for our annual Eurovision report with Hedvig. You watched Eurovision this year?
HEDVIG: I watched the Eurovision grand final 2023. Yes, I did. Did you?
DANIEL: No, I did not. And that’s okay. I’m happy not to watch it and then I listen to you talk about it. That’s how I like to participate. But I do have a question. You didn’t watch the semifinals?
HEDVIG: No.
DANIEL: Is the semifinals just them performing the same song again?
HEDVIG: So, the semifinals happen before the grand final and they are exactly as the name suggests. So, there was, I think, 15 in the first semifinal, 16 entries in the second one, and a certain amount of winners from those qualify for the grand final.
DANIEL: Okay. And they just do the same song each time?
HEDVIG: They do the same song. Yep.
DANIEL: Okay, cool. If you got one song, you do it.
HEDVIG: Yeah. I, as a rule, tend to just watch the grand final because I like to hear the songs for the first time and be surprised. A lot of other Eurovision heads not only, like, watch the music videos they put out before the semifinals, then they watch the semifinals. But I only watch the grand final and I know I have some other people in our Patreon group who are the same. It was really fun this year. There were a lot of great songs. So, as a reminder for those who don’t tune in, Eurovision is a song contest. It’s a contest for performers and songwriters. You have to produce original songs. People compete country wise. It was started after the Second World War as a way of unifying Europe. It’s called Eurovision, but it has since expanded to include countries next to Europe, specifically in the Middle East and in North Africa. So, like Morocco and Israel and stuff like that. Countries like that participate every now and then. And did you know, Daniel, that SBS in Australia bought the rights for Asiavision?
DANIEL: Okay. So, they’re doing that.
HEDVIG: No, they’ve decided to drop it. I just looked it up earlier.
DANIEL: Oh, no. Is it going to go to like Seven or something?
HEDVIG: No, they just don’t think Asiavision is going to happen. It’s too hard to organise, too many time zones, too many… they’re just going to drop the ball.
DANIEL: So, there was going to be an Asiavision and then they’re like blehp.
HEDVIG: Exactly. And there was also an attempt at doing a US version called American Song Contest. They did do it with Kelly Clarkson and Snoop Dogg as presenters. But, and I quote one of the critique reviews on Wikipedia, “Americans just don’t have enough self-deprecating camp to pull off Eurovision.”
DANIEL: Yeah. That is part of the appeal, isn’t it?
HEDVIG: Yeah. You have to be able to be silly and have fun and make fun of yourself. And a friend of mine watched American Song Contest and she said that Americans are way too earnest.
DANIEL: Just taking it too seriously. You know what?
HEDVIG: Yeah.
DANIEL: Actually, I would not watch Asiavision, but I would listen to you talking about it.
HEDVIG: Okay. Well, Asiavision is not happening, and US Vision is not happening. So, the only thing that’s happening is that more countries are joining Eurovision. And Australia was in it yet again with, I thought…
DANIEL: Yay.
HEDVIG: …very great entries.
DANIEL: From Perth!
HEDVIG: From Perth? Oh, I didn’t know that.
DANIEL: Voyager’s from Perth.
HEDVIG: That’s amazing.
DANIEL: Yeah, this was our year.
HEDVIG: They were really good. I really enjoyed them.
DANIEL: They came in 9th!
HEDVIG: And they were really fun in the Green Room. They came pretty decent.
DANIEL: They just seemed super chill and genuine and nice. I was glad to see them.
HEDVIG: They’re really cute. They were in the middle of eating their lunchboxes when they got some points and everyone zoomed in on them and they had like food in their mouth. I thought it was really funny. [DANIEL LAUGHS] Anyway, Australians are always really fun at Eurovision. I hope New Zealand would participate and other countries. Let’s just all pile into Eurovision, apparently nothing else is working.
DANIEL: Awesome.
HEDVIG: But anyway! Let’s get back to linguistics.
DANIEL: Bringing the world together.
HEDVIG: Okay, we’re a linguistic show.
DANIEL: Back to linguistics. Here we go. Language.
HEDVIG: So, Eurovision happened. It was in Liverpool in the UK. Ukraine won last year. But Ukraine can’t host Eurovision.
DANIEL: No, it wouldn’t be a good idea, would it?
HEDVIG: Because they’re at war with Russia. They have been invaded by Russia, and they cannot host a massive international event. Liverpool was the chosen candidate for the sort of proxy host country… city, and they did a lot of things to include Ukraine in the show. So, they said some things in Ukrainian every now and then, they had a lot of Ukrainian flags. I’ve also been told by people who have been on the streets of Liverpool that there are a lot of restaurants who are offering Ukrainian food and menus in Ukrainian, and there are a lot of Ukrainian flags around. And I didn’t know this. One of the reasons that Liverpool was chosen as the Grand Final is because there’s actually a significant Ukrainian immigrant population in Liverpool.
DANIEL: Ah. So, this really was an attempt to make it, like, Little Ukraine, attempting to bring them in. That’s great.
HEDVIG: Yeah, exactly. So, in terms of languages, so for the Grand Final, there were 26 entries, and 13 out of those were not entirely in English.
DANIEL: Not entirely. So, there were lots of countries where they’re not English-speaking countries, not officially, anyway. And yet, they used English in the song.
HEDVIG: Right. So, for example, Poland… Blanka from Poland had a song called Solo, which was entirely in English. English is not an official language of Poland.
DANIEL: Yep. Now, Diego on our Discord has linked us to a report on globalvoices.org by Langoly that says lots of acts are using English as a way of maybe trying to have more global success, because English has this prominence as it does.
HEDVIG: Yeah.
DANIEL: Do you think that’s successful?
HEDVIG: Well, I think what happened in the beginning of Eurovision is that you had to sing in an official language of your country. So, UK could sing in English, Ireland could sing in English. I think Malta or something like that were allowed to sing in English as well, but everyone else sung in the official language of their country. And it was felt that gave an unnecessary advantage to Ireland and UK and arguably, specifically Ireland, because Ireland actually won the most. UK doesn’t win that often. And that it was unfair because a lot of people know English and could understand what the songs were about and were therefore more likely to engage and connect with the song and vote for it. Whereas if they heard another language, they would be like, “This is a funky tune, and I like the chorus, but I don’t get it.” So, maybe you wouldn’t vote for them. However, that’s going back and forth, because in Eurovision, it’s very popular to showcase your country. So, very often people show up in sort of some sort of national dress or dress of Indigenous people from that country, or in particular, if it is actually an Indigenous representative. So, in the grand final this year, there were 26 entries, and 13 of those were in not English. And that’s actually quite a high, how do you say, from the semifinals to the grand final, almost all the songs that were not in English made it through. Only Romania didn’t make it through.
DANIEL: Oh!
HEDVIG: So, if you sang in not English in the semifinal, you were more likely to go to the grand final.
DANIEL: So, if they’re trying to use English as a way of gaining success, it’s not working as far as getting through to the final.
HEDVIG: I mean, still 26 minus 13 is… oh! 13. [CHUCKLES] I’m smart.
DANIEL: It is.
HEDVIG: So, still half of the songs in the Grand Final were in English.
DANIEL: Yeah. That’s still a lot. Jeez. Okay.
HEDVIG: That’s still a lot for one language. So, you didn’t have a bad chance. It’s just you had a better chance if you sang in not English. Both Romania and Moldova sang in Romanian because Romanian is the official language of Moldova. Romania didn’t make it into the grand final, but Moldova did.
DANIEL: That happened last year too. Didn’t Moldova use Romanian last year? I seemed to remember that.
HEDVIG: Yeah, I’m pretty sure they did. I’m pretty sure you’re right.
DANIEL: Now, Sweden won. Spoiler alert. Lena on our Discord reckons that Sweden won because it’s been 50 years exactly since ABBA won with Waterloo. Am I right? It was Waterloo, wasn’t it?
HEDVIG: It was Waterloo. I was defeated, you won the war.
DANIEL: What do you think of that theory?
HEDVIG: I think that theory holds a lot of water. And here’s also where I’m going to get a bit nerdy, which is the grand final has different ways of voting and counting votes every year. It used to be only jury votes, so you could understand this as committee votes. So, every country had a committee and then they had a number of points, and that’s where you hear the “La France! Douze points!” That’s that. And then recently, they’ve decided to incorporate more popular voting, so now it’s split. So, there’s jury vote and popular vote.
DANIEL: Nice. And is that changing things?
HEDVIG: It is changing things. I think that it should only be popular vote. I think expert committees are BS.
DANIEL: Wow. Radical. Okay.
HEDVIG: It’s music.
DANIEL: You firebrand!
HEDVIG: Wait, how is this radical? It’s music. [DANIEL LAUGHS] We should let people vote. It’s aesthetics. It’s not like, “Who made the best bridge that can hold the most weight?” It’s not an objective contest. Anyway.
DANIEL: I can see a case for that. I can also see a case for a split between the critics — the experts — and the people. I can see it that way.
HEDVIG: If we’re going to have this contest and it’s going to be about self-deprecating camp, I don’t think we should have, like, taste police in there. You know what I mean?
DANIEL: [LAUGHS] Wow.
HEDVIG: It’s supposed to be silly. It’s supposed to be a chance for artists no one’s ever heard of to make a splash.
DANIEL: Yes.
HEDVIG: Anyway, I was going to answer your question, which was about the conspiracy theory about why Sweden won. As the Swedish person, I feel like I need to respond to that.
DANIEL: Please, please.
HEDVIG: So, it is indeed true that next year will be 50 years since ABBA won with Waterloo, which gives Sweden a great opportunity to have a very, very, very ABBA-themed Eurovision contest.
DANIEL: Oh, that’s fun.
HEDVIG: And the jury vote was the deciding factor in Loreen from Sweden winning. The popular vote by quite a large margin, went to Finland, Käärijä, with the song Cha Cha Cha, which was in Finnish and which I voted for twice.
DANIEL: Did you!
HEDVIG: Yes. I must say that there are a lot of people on the internet throwing shit at Loreen, which is really bad taste. And also, Finland’s Käärijä has said, “Please don’t,” because she just did her song and did her best and then voting happened. It’s not like… But I do have to say I think it’s also really unnecessary to say that Loreen’s song was very bad because I don’t think it was very bad. I don’t think it was the best song in the contest. I voted for Finland, but I don’t think it was terrible. I understand that people liked it. I understand that it got a lot of votes. I think definitely should have made it to the top five. Personally, I don’t think it should have won, but I think it’s… Anyway, don’t throw hate at Loreen, but we do need to have a frank discussion about these committees because they seem to be a bit out of line.
DANIEL: Okay.
HEDVIG: Now, let’s see if I have anything else to say about the language. Oh, there’s been a lot of faff on the internet about the BBC sign language translator who translated into British Sign Language. Reminder for everyone:…
DANIEL: I noticed that.
HEDVIG: …there is not one sign language, there are many. And the one that you might have seen video clips off online is translating into specifically British Sign Language and specifically for the song Cha Cha Cha by Finland. It looks quite fun, the signs that he’s doing. So, people really enjoyed it.
DANIEL: It sounds like one of those things where everyone sees a very expressive interpreter and they think, “Oh, my goodness. I had not realised that signed languages could be expressive. This interpreter stole the show,” blah, blah, blah. And it’s like: uh, well…
HEDVIG: But that’s also because what they’re doing right, is they’re performing the rhythm and the vibe of the song along with the semantic word content. And specifically for songs that are as upbeat and silly as Käärijä’s Cha Cha Cha, it looks pretty fun. And also, if you look on stage and you see the performance, the performance is also very fun, like the one that the singer is doing. So, it’s not a surprise that the sign language one is similar in style.
DANIEL: So, some sign language interpreters made a splash. What about Spanish? In years past, we have found that Spanish is sometimes thrown into songs…
HEDVIG: Yes.
DANIEL: …even if the performer doesn’t have a Spanish language background, just because Spanish has associations of fun times and party times. Is that still happening this year?
HEDVIG: It’s not really happening, perhaps with the exception of …Finland, again. We’re talking a lot about Finland, but that is because they were, in my opinion, the best. So, their song was called Cha Cha Cha. And they did have dancers on stage that looked like they were like the cha cha dance. Like they were dressed to dance that.
DANIEL: Appropriately.
HEDVIG: And as far as I know, that is a Spanish or at least Latin song, this style.
DANIEL: Mhm. It is.
HEDVIG: So, that’s the closest I could see through to our theory of Spanish being the party language. Other than that, I didn’t see any other evidence, actually.
DANIEL: What about minority languages? Were there any songs performed in a minority language of that country?
HEDVIG: No. So, the Czech entry was in English, Ukrainian, Czech, and Bulgarian. Those are all quite large languages. And I don’t know if Ukrainian and Bulgarian is a minority language of Czechia. It might be. But other than that, no. Everyone else, like…
DANIEL: Nobody was making a point that way.
HEDVIG: No, not that I saw, no.
DANIEL: Were there any other language things you noticed that surprised you?
HEDVIG: Um… not really. Maybe the Croatian one, they made a point of… their song is called, and I’m going to pronounce this wrong, Mama ŠČ. It’s S with a little tiny V on top, and then C with a tiny V on top and I don’t know what that’s about.
DANIEL: Oh! It’s the [ʃʧ] sound. I know that Russian has that sound. <Щ>.
HEDVIG: Yeah.
DANIEL: Mm. It must be in that language too.
HEDVIG: It must be in that language too. And a lot of the commentators and everyone were making fun of, or just were making fun of themselves, having trouble pronouncing it. I think the least surprising language thing that happened in Eurovision this year was that France performed in French. They did another… I think, Édith Piaf nostalgia, but a little bit electro. What’s another one? Oh, Slovenia’s song is technically… it’s in Slovenian, but the song is called Carpe Diem, which is Latin.
DANIEL: It’s true.
HEDVIG: Meeh?
DANIEL: Okay.
HEDVIG: Eeh meeh? That’s maybe something?
DANIEL: Well known enough to be borrowed in lots of languages, perhaps.
HEDVIG: Yeah, exactly. Exactly.
DANIEL: Mm. All right. Well, this has been our Eurovision language report. [HEDVIG LAUGHS] Hedvig, thank you for bringing that to us. I really appreciate it.
HEDVIG: You’re welcome! And, Daniel, will you just watch Eurovision sometime?
DANIEL: Next year. Hey, is there a way we can watch it live together?
HEDVIG: There is, but you’ll have to be up at like 04:00 AM.
DANIEL: I’ll do it, and I’ll do it with our patrons. That’s what we’ll do, if you’ll warn me in advance.
HEDVIG: Okay! Oh, my god. Okay.
DANIEL: We’ll make it a thing.
HEDVIG: I had no idea you’d be up for that. Eurovision is best viewed in a large group. Like, you should have a party and people dress up and people have wigs on.
DANIEL: I know, but I can’t get anybody up at 4am, so that’s not going to happen. But I will have it open on my laptop, and then I will also have a Discord window open, and I will comment with everybody who wants to watch with me. That’s as close as I get to a party.
HEDVIG: I should say: because it’s going to be in Sweden next year, I am going to try and get tickets, so I might be trying to get to one of the semifinals or something.
DANIEL: You’re going to be there! That’s going to be so cool!
HEDVIG: I’m going to try.
DANIEL: Okay, you’ll be holding down the fort at the actual performance, and I’ll be holding down the fort on Discord. Okay. Watch for it, people. And finally, our final news item. Hedvig, I want you to take a guess. As we get older, do we say UH and UM more? What do you think?
HEDVIG: More.
DANIEL: More? Any idea why?
HEDVIG: I think [CHUCKLES] that as you grow older, it gets harder to think.
DANIEL: Picking her words carefully. Yes.
HEDVIG: Cognition gets harder, and it’s hard to plan sentences, and you do fillers in order to give yourself some stalling time to think of what you’re going to say.
DANIEL: Are you right? Surprisingly, no. But…
HEDVIG: What?
DANIEL: …to find out why… I know! This is a surprise. I got to talk to Dr Eleonora Beier. She’s a postdoctoral scholar working at the University of California Davis, Center for Mind and Brain. And she came up with a very unusual and very cool method to find out. Because, as you could imagine, it would take a super long time to study people across the lifespan and see if they start saying UM and UH more as they age.
HEDVIG: Oh, the same people. Yeah, sure.
DANIEL: The same people. So, she studied videos of celebrities doing interviews at different points in their lives!
HEDVIG: That’s smart, because you have a sort of controlled environment, because every time it’s an interview. It’s not natural conversation. But if you’re comparing an interview with an interview, you’re comparing apples and apples.
DANIEL: Precisely. I thought that was so cool. So, when I found out about this research, I had to ask Nora about it. So, I started off by asking her about UM and UH, which we usually think of as fillers. And I asked, is that really the best way to think of them, as fillers?
NORA: Um, yeah. So those could be called non-lexical fillers, meaning that you’re trying to fill in a pause with something that’s not a word, so UH or UM are not words. Um, but you can also think of lexical fillers, where you say things like, YOU KNOW, or WELL, that are also not super meaningful, but you’re also trying to buy yourself time by saying those words.
DANIEL: Well, that’s what I would have said. A few years ago, if you’d asked me, “What do you do when you’re saying UH and UM?” And I would have said, “Oh, it’s a traffic director, because you’ve got an air pocket in the brain. You don’t know what to say yet, but you don’t want to give away your turn. So, you say, UM, UM, UM UM UM until you know what to say.” But lately we’ve been seeing a lot of work that suggests that UH and UM do a little bit more. Uh, for example, have you noticed anything like that? I’ve got one.
NORA: Um. Um. [CHUCKLES]
DANIEL: Um. [CHUCKLES]
[LAUGHTER]
DANIEL: Argh!
NORA: Well, why don’t you go ahead with what you’ve seen?
DANIEL: I will. Okay. So, um, hyperawareness of UH and UM. There was an experiment where they gave people sentences that contain certain words, and sometimes they would hear, “My sister broke her, UM, leg,” or, “My sister broke her leg.” And then, they would go back after they’d heard all the sentences, and they would say, “Did you hear the word LEG in any of these sentences?” And if they had heard UM before LEG, they remembered LEG better. Yeah. And also, children, when they have a made-up word like DAX, and they hear, “Look at the, um, dax,” and then they see two screens, one with a familiar object and another with an unfamiliar object, and they hear the UM, “Look at the, um, dax,” which they’ve never heard before, they’ll look at the unfamiliar object if they hear UM because I guess they imagine that the adult is going through more a cognitive hard time. So, UM and UH and actually can help memory, and it can help guide interpretation. And that was when I stopped cutting UH and UM from the show because I thought, “Well, it’s there for a reason.”
NORA: Yeah, exactly. I mean, basically, um, these fillers and these different types of disfluencies are just this very normal part of our speech, right? Sometimes, when people do different types of speech training, they’re told to try to remove them and reduce them. But it turns out that, first of all, most people tend to not notice disfluencies unless they’re just very, very obvious, but we tend to filter them out. And it actually makes the speech sound more natural. And actually, for example, some of these voice AI softwares like Alexa or Siri sometimes sneakily introduce disfluencies to sound more natural.
DANIEL: Argh! I don’t want that.
[LAUGHTER]
DANIEL: I think you should act like a robot. You act robotty. You’re a robot.
NORA: Exactly. And then it turns out, like you said, that they’re not just filtered out or, you know, help us sound natural, but they can actually be useful for the listeners. So basically, because we tend to assume that if someone is pausing or is saying UH to buy more time, it makes us assume that it’s because they’re trying to retrieve a kind of important word. Maybe it’s a word that’s less typical, um, and that’s why it’s taking you a little bit longer to remember how to say it. Um, and therefore, it could be very important for the listener to understand it and pay close attention to that word. Um, and so, yeah, that could be one of the reasons why we see better memory for those words because it’s like a little signal that saying, “Hey, I’m about to say something really important here, so pay attention.”
DANIEL: Ooh. Okay. I’ll get ready. Yeah, that’s cool. Okay. So what attracted me to your work here was a study that you’ve done looking at, uh, disfluencies, UH and UM, but not just among a group of people, but among like, certain people, famous people who have had interviews at different points in their life and studying whether they start saying UH and UM more or less as they age. How did you get onto this? This is such an intriguing design.
NORA: Well, the honest answer is this was a project that my PhD advisor several years ago kind of pitched to me, both to me and my colleague, Nene Chantavarin. So, we kind of worked on this together. Um, and yeah, we were interested in looking at whether people become more disfluent with age. Uh, and the reason why this seemed really interesting is that there’s a lot of review papers that kind of state this as fact or seem to be pretty consistently saying, “Yeah, we have a lot of evidence that people become more disfluent with age.”
DANIEL: Well, it makes sense because as you get older, you slow down a little bit, you have a little more trouble coming up with the words. And so, it seems likely that you’d throw an UM or an UH in there just to hold your turn while you’re mentally cranking through. I can see why everybody would think that.
NORA: Yeah, um, and that’s basically the theory for that, um, is that with age, there’s lots of cognitive processes that slow down. Uh, it might be harder to retrieve the words that you want to say. Uh, you might have even more vocabulary to go through. Like with age, you accumulate more words, more knowledge, and so you have to go through more of them to figure out what’s exactly the word you want to say.
DANIEL: Ah, yeah. Makes sense.
NORA: Yeah. There’s a lot of reasons why you would expect people to be more disfluent.
DANIEL: Ah, I hear a BUT coming.
NORA: But [CHUCKLES] when we really looked at the actual studies that, uh, were looking at disfluencies over age, uh, it turned out that there was a lot of inconsistencies. Uh, first of all, so we’ve been mostly talking about UHs and UMs, so those would be filled pauses. But there’s other types of disfluencies, different ways to kind of, um, categorise our fluency. And two other types that we were interested in are repeated words. So, if, uh, for example, if I say, um…
DANIEL: If, if, if.
NORA: [LAUGHS] If I said like, uh, “I went to the-the house down the street,” I repeated the word THE. And then, another that were interested in is repairs, which is where you say the wrong word, and then you correct yourself. So, for example, if I say, um, “I went to the house… I mean, the mansion down the street.”
DANIEL: Yeah yeah yeah.
NORA: Um, you’re switching a word. And so, different studies were, uh, looking at either all of these or only some of them or looking at other things like silent pauses. Uh, and basically, they were all finding different things. Some were finding that people as they age produce more UHs and UMs, but they don’t repeat more words. And then, other studies would find that they repeat more words, but they don’t say more UHs and UMs. So, it was just very inconsistent.
DANIEL: Okay, now, before you tell me your results, I want to know who you looked at. Whose videos did you study?
NORA: [CHUCKLES] So, we had a sample of 91, uh, famous people, I believe.
DANIEL: Wow.
NORA: Yeah, we picked from actors, politicians, musicians, directors, TV anchors. Like, we tried to split it across different professions because we were afraid that, for example, if we’re only looking at politicians, maybe those people were, you know, super trained in the way they speak, and they’re a little more similar to each other. And so, we were trying to make it a little bit more varied.
DANIEL: Yeah.
NORA: Yeah, so for actors, were looking at people like Hugh Grant, uh, or Harrison Ford. Uh, we we’re looking athletes as well, like Magic Johnson. We’re looking at authors like, um, Kurt Vonnegut, Ray Bradbury, Stephen King. We’re looking at business leaders and politicians, uh, uh, musicians like Bob Dylan, uh, directors like David Lynch. There’s a long list.
DANIEL: Okay. Okay. And can you give me an idea of what kind of age range? What was the maximum… did you see Ian McKellen, like, when he was younger and then when he was a lot older? What was the biggest range you had, or a typical range?
NORA: Um… So, the the overall range of the whole, um, corpus that we created goes from, I think, 20 something to 94 years.
DANIEL: Yeah yeah yeah. Okay.
NORA: On average, the time span for each individual person that we had was, uh, 28 years. Quite a lot of variability. So, it went from just 13 years to 51 years for…
DANIEL: That’s a long time.
NORA: …one person. Yeah. We had a minimum three interviews, but sometimes up to five or six interviews, uh, that were at least seven years apart.
DANIEL: Okay. Wow. All right. Okay. Okay. I think I’m ready for some results. So, going in, I was ready to find that UMs and UHs increased as a person ages for the reasons that we mentioned. Harder to recall, more words to choose from, cognitive slowing down a little bit. What did you find?
NORA: So! [LAUGHS] We also expected all of that, and then in the end, we found the most consistent finding was that people slow down as they age. So overall, people become, uh, a little bit slower when they talk. But then in terms of the actual, uh, types of disfluencies that we talked about, we didn’t find any increase in the UHs and UMs.
DANIEL: Ooh.
NORA: And we also didn’t find an increase in repairs. So, saying the wrong word and changing it.
DANIEL: Wow.
NORA: Um… the only thing that seemed to increase with age, besides how fast they were talking, was, uh, the repeated words. So, people, as they age, tend to repeat their words a little bit more. Uh… but this was a very small effect, so I don’t want to draw super strong conclusions. So, I would say that really the most important, um, effect that we found was that people were slower as they aged.
DANIEL: Okay. But no real increase in UMs and UHs. That is super interesting.
NORA: Yeah.
DANIEL: Now, [SIGHS] I’ve got to ask here. These were interviews. They knew they were being recorded. And when people know they’re being recorded… I mean, we’re not aware of everything, and we can’t always control our UMs and UHs, but I kind of can. Like, I’ve been doing radio for a while, and when I’m on, I think I don’t do as many UMs and UHs. I kind of trained myself out of it. I’ve trained myself to be more fluid. Not everyone who is famous is going to have that kind of training. But could it be a function of just knowing that they were on camera?
NORA: For sure. Yeah. So that’s one of the big shortcomings of this, is that because we’re limited to these, uh, recorded interviews, and pretty much all of the people are somewhat familiar with being interviewed and being on camera, um, these are not “normal people.” They’re all kind of trained and have a lot of experience. So, it could be that this is, ou know, kind of a finding specific to this very narrow sample of people. And the reason why we have this limitation, of course, is that to do this type of study where you get the same person and look at their speech patterns over their age, if you did this, not with this type of corpus, you would take 80 years to do the study because you need to follow up with the same person over a long time. So, this is kind of why we were limited to these people.
DANIEL: I mean, my PhD might take 80 years, but it’s not like you have that much time. And that might not be a serious shortcoming either, because it’s not like these are all news readers or all actors. Some of these people are, like, sports figures or musos or people who aren’t really, really focused on verbal performance and haven’t gotten any training that way. So, I think that knowing that, it’s probably an acceptable trade off.
NORA: Yeah, yeah, exactly. And that’s why we were really focusing on not having just only actors, you know, but having more variety.
DANIEL: So, have you gotten any feedback on this from people who have done earlier studies? I’d be interested to see how you’ve changed the field with these new findings.
NORA: Um, I haven’t had any direct feedback since publishing. There definitely was a lot of feedback during the review process, um… We had a little bit of a hard time publishing this. I think the main reason is a lot of the times, if you have clear-cut results that fit your hypothesis, reviewers will give you an easier time publishing. And if you get more interesting results that don’t quite fit your initial hypothesis, people always come up with different ways you could have analysed the data and why you did things wrong. Um, so, we tried to incorporate as much feedback as possible and we definitely went through a lot of reviews. I think it really helped making the science better. Um, so, yeah.
DANIEL: That’s cool. All right, so now we know a little bit more than we did before about dysfluency. Has it made you look at dysfluency differently, or aging?
NORA: Well, I guess the main takeaway I took from all of this is that what I find the most interesting result in the end is that age itself is not very good at predicting disfluencies, um, but instead it’s more individual differences in how we speak. So even though we didn’t see a lot of differences with age, um, differences in disfluencies were related to other aspects that we measured. So for example, we were looking at how much variety of words people were using when they spoke and like, whether they used kind of more frequent words, like everyday words or kind of rare words. And those were the predictors that, um, seemed to be correlated with disfluencies. So, it’s not so much that you become more disfluent with age but with age, you might have changes in your speech patterns, you might have more varied vocabulary. And then, if you have a more varied vocabulary, you might be more likely to be disfluent because you’re taking longer to sort through your large vocabulary to find the words you need. So, there’s really a lot of variability.
DANIEL: Wow. Just one more thing about your work. There’s another project that you’ve been doing about — I’m going to quote here — “the neural dynamics of anticipatory attention.” And the reason I’m thinking about this is because we had an interview earlier this year with Dr Mark Ellison about how our brains, when we hear conversation, they’re making predictions, they’re constantly, like, sensing patterns in words. And we often use things like intonation, and also pauses and UMs and UHs, I guess, to… manipulate is the wrong word, but to influence the attentional state of our listener. And so, when I saw this work of yours about anticipatory attention, I was really curious. So now, I’m ready [CHUCKLES] to pay attention. What is this work that you’ve done?
NORA: Yeah, so this is basically the work I’m doing right now. So, it’s kind of in progress, um, but it’s similar to what we were saying at the beginning with the study you mentioned where if people heard an UM, they were more likely to pay attention because they thought the next word would be really important and so then they could remember it better. Uh, and as you said, there’s other ways that people can predict when they should really pay attention and when they don’t necessarily have to. [LAUGHS] So, like you said, intonation, if I say, “This is the thing that’s important,” that pitch accent, it’s like this strong increase in pitch and intensity of my voice that is cueing you to pay attention.
DANIEL: LIKE is another one. You know. “I was like heading off to work and like this guy.” So, the LIKE is a focal marker, meaning, “The next thing I say is what I want you to focus on.”
NORA: Yeah, exactly. And you can use other cues, like you can say, “What I did yesterday is,” and so that’s basically a syntactic focus construction.
DANIEL: Ooh, cleft sentence. Yeah.
NORA: Yeah, exactly.
DANIEL: Okay, cool. Cool.
NORA: Basically, this current project is looking at how exactly do these different cues in the speech guide our attention. And the really cool aspect of this is that I’m thinking of it as a temporal prediction, which is a fancy word for predicting when something will happen.
DANIEL: Cool.
NORA: And so, we know that as we are listening to someone speak, we’re always predicting what words will come next. You’re probably familiar with the feeling of you’re trying to say something and the person that’s listening to you is filling in for you and saying the end of the sentence.
[CHUCKLES]
DANIEL: Yes. We finish each other’s…
NORA: …sentences. [CHUCKLES]
DANIEL: Sandwiches!
NORA: Or sandwiches, yeah. So, we’re always trying to predict what the next word will be. Uh, but what I’m trying to, uh, investigate now is not just how we predict what will happen next, but also when those words will happen. Those cues, like disfluencies or the focus, are just ways or little cues that tell us, “Oh, I need to pay attention now,” or, “Oh, actually what I’m going to hear now is not super important, I can pay attention later.” And we’re measuring that with EEG, which is a measure of neural activity.
DANIEL: Oh, okay. That’s where the neural dynamics come in, because you’re examining… [EXHALES] what’s the term? Is it event-related potential or… how does it work exactly?
NORA: Yeah, so the EEG measures this ongoing neural activity that are essentially neural oscillations. That’s why they’re called brainwaves. And, um, the ERPs, or event-related potentials are specific responses to… if you hear a specific word, for example, you will get a response to that word.
DANIEL: Yeah, it goes voom, like the wave goes really high.
NORA: Exactly. Uh. But you can also look not specifically tied to one point in time, you know, like to that specific word, but you can also look at how the brain waves are behaving over a longer period of time. And so that’s mostly how I’m going to be using them. It turns out that when we are really focused on some stimulus in our environment, like if I’m really focused on what you’re saying, our neural oscillations will have a certain pattern, versus if I’m kind of mind wandering and starting to think about what I’m going to do later for dinner, we start seeing a different pattern over the course of the EEG. Specifically, it’s what’s called alpha oscillations. Meaning that we start seeing a lot of oscillations at one particular frequency and that’s how we can tell if you’re paying attention or not paying attention.
DANIEL: Right. Wow, this stuff’s so interesting. Well, look, I want to hear more about this when you’ve gotten some more of this done, because I think that looking at how we manipulate attention… And also, I would love to know about how the speaker is… what’s going on neurally for the speaker as well as for the listener and looking at some some some dyadic interactions. I would just be fascinated by that.
NORA: Yeah, definitely.
DANIEL: Oh, man. We’ve been talking to Dr Eleonora Beier, postdoctoral scholar working at the UC Davis Center for Mind and Brain. Nora, thanks so much for hanging out and talking to me today. This has been super interesting.
NORA: All right, thanks so much for having me again.
DANIEL: And now, it’s time for another iteration of Related or Not, brought to us by our great patrons. In this game, I give you two words, you have to guess whether they’re etymologically related or if the similarity is merely coincidental. Diego gave us this one, and that means I can play, because I made my guess before I looked up the answer.
HEDVIG: I am going to introduce Ben into the conversation. I have gotten out my dice.
DANIEL: Okay.
HEDVIG: And I have a D6, a six-sided dice. And if it’s an even number, then he guesses that it’s related.
DANIEL: Okay, very good. Should we do him first? Go ahead.
HEDVIG: Yeah. Say the word, and then, uh, I’ll do him at some point. Yeah.
DANIEL: Okay, okay. Hedvig, do you remember when you came to Perth and we went to the museum, Boola Bardip, just outside there?
HEDVIG: Yep.
NORA: They had these little waterspouts built into the ground that spray water vapour every once in a while.
HEDVIG: Uh-huh. Kids love it.
NORA: My daughters were there as well, my two young ones. Now, the water comes up in kind of a MIST. A mist. And when my two daughters play in these water features, their clothes and their shoes get quite wet with all that mist. It makes them quite…
DANIEL AND HEDVIG: MOIST.
HEDVIG: I see where this is going.
DANIEL: So, the words for today are MIST and MOIST. Related or not? Ben, what do you say?
HEDVIG: Ben says related. I rolled an even number.
DANIEL: Okay. Very good.
HEDVIG: Okay. MIST and MOIST, I don’t think that they’re related.
DANIEL: Okay. Uh, any reason?
HEDVIG: I think that MOIST is related to a word for soft, and I think that MIST is… I don’t know, something else.
DANIEL: This is great. You’re so good at playing this game. I like hearing how… the alternate theory. Okay. Here’s what I wrote. I wrote: I could easily see how they could have come from the same source. You take MIST from maybe Old English, and then you add some kind of adjective inflection and get MOIST because, I mean, we used to change words by adding some kind of sound internally, like MAN and MEN. But upon further reflection, in this case, I don’t think that was how it worked. I think MIST was Old English, and I think MOIST would have been more… it looked kind of French to me, kind of more from Latin. So, I’m going to say no.
HEDVIG: Okay.
DANIEL: So, Ben says yes, we say no. [LAUGHS] Answer: it is no! Not related.
HEDVIG: Yes!
DANIEL: Yes. We beat Ben!
HEDVIG: We beat Ben.
DANIEL: We beat Random Ben.
HEDVIG: [LAUGHS] I’m laughing at the dice.
DANIEL: He’s going to love that. So, I was right. MIST is Old English, and Etymonline says it’s perhaps from a Proto-Indo-European *megh, which means to urinate. And of course, that made me think of micturate, which is to wee.
HEDVIG: Wait. What?
DANIEL: Yeah, to micturate. Yeah, it’s a word that never really caught on, but to micturate is to pee.
HEDVIG: Okay.
DANIEL: So, when it’s all foggy and misty, you say, “Man, it’s really pissing around here.”
HEDVIG: Yeah.
DANIEL: Which is strange because when I pee, it doesn’t come out in the form of vapour. It’s not that way. But, um, I digress.
HEDVIG: [LAUGHS]
DANIEL: And then MOIST, uh, I nailed it again. It comes from Old French, moiste, and possibly from Classical Latin, mucidus, which means moldy, which is where we get MUCUS. But then, we also have Latin… It’s also possibly Latin, musteum, which is a word for new wine. And that’s why we say that things smell kind of musty.
HEDVIG: Also, must is just what we call cloudy apple cider in Scandinavia.
DANIEL: Is it really‽ You call it MUST?
HEDVIG: Mm-hmm.
DANIEL: Fantastic.
HEDVIG: And it could get some alcohol in there by the fermentation process.
DANIEL: But it’s not necessarily alcoholic and probably wouldn’t be alcoholic, maybe just a little bit?
HEDVIG: Yeah, just a little bit.
DANIEL: Well, then that’s interesting because that preserves the sense of new wine. Not the fermented stuff, but just brand new, just basically fruit juice. Fascinating. Okay. So, MOIST is not related to MIST, but it is related to other murky words like MUCUS, MUSTY, MUST in Swedish, and even MUSKY, if something smells musky or has a certain kind of musk aura to it. So, not related.
HEDVIG: Interesting. The most important conclusion is that you and I won, and Ben lost. [LAUGHS MOCKINGLY]
DANIEL: Sorry, Ben. He can complain, but that’s what he gets for not showing up.
HEDVIG: Yeah, exactly.
DANIEL: Thanks to Diego for that puzzle. You can send us a pair of words if you like, hello@becauselanguage.com. You can also, if you have some cash lying around, sponsor an episode and we’ll mention you during the game.
[TRANSITIONAL MUSIC]
DANIEL: I’m talking to Dr Emily Hofstetter from the Department of Culture and Society at Linköping University. Emily, thanks for coming on and talking with me today.
DR EMILY HOFSTETTER: Oh, thank you so much for having me.
DANIEL: I ask this to everybody. Tell me something that you think is really, really cool about language.
EMILY: I think for me, it’s how fast and effortlessly we coordinate with other people. We can do all these really nuanced things and we do it all the time every day, and we don’t have to think about it. And it’s just like magic watching it happen. I’m always astonished at just, “Wow, that was a really great expression really. How did they get that done?” That is so much fun, to get to come into work and look at all the time. It’s amazing.
DANIEL: So, what exactly are you looking at when you’re looking at coordination? Because we’ve been talking a lot this year, this season of Because Language. Yes, we have seasons.
EMILY: [LAUGHS] Yes.
DANIEL: We’ve been talking a lot about how language, we use it to coordinate joint attention and joint intention and make sure that we’re on the same page and then we can accomplish stuff together. What does coordination look like for you?
EMILY: Um. Gosh, I mean, it’s sort of the bread and butter of everything that we do, is trying to come to some kind of place or situation where we can keep going on the basis that we have some kind of mutual understanding, but also just mutual task that we’re engaged in doing because everything that we do is coordinating. Even on really longitudinal scales where immediate interaction or immediate language isn’t being used, people make products for each other to use or whatever. But on an immediate scale, that’s everything from understanding how that other person is doing today and what it’s going to be like having an interaction with them, through to getting work done together, from doing construction to surgery, anything like that. Everything we do requires making sure that we can do it in the right timing with each other and the right same task, that we have some idea of what they’re about to do so that we can adjust what we’re about to do and keep going on that way. What I get to do is look at how the everyday things are remarkable, and so how everyday tasks that people are up to doing are actually really intricate detailed things that we’re just so practiced at doing that we don’t have to think about it.
DANIEL: I’m thinking of the way that my children just sort of rope in other children [EMILY LAUGHS] to play at the park and it’s just so effortless and they bump into each other and then they coordinate, and then they’re playing.
EMILY: Oh, yeah. And they have no problem just picking up and trying anything out. And I don’t know if it’s just that there’s less at stake for them, but they have no… they don’t have a lot of baggage about maybe the better way is conversation can go badly or awkward or anything. And so for them to say, yeah, we just get started. I mean, I guess often it’s the speed that astonishes me… and that can even happen when people are driving. People’s ability to anticipate where someone else is going to be on the road and the fact that there are as few accidents as there are. It’s really the same with verbal interaction that we don’t… um, the timing of when we do stuff is so incredible. This is something that maybe happens less with how we’re talking right now because of the format. But when you’re talking with somebody in person, there’s usually a lot of really precisely timed continuers, so things that people say like HMM and YEAH, and they come at exact moments in order to ratify what somebody has said but also prevent them from thinking, “Oh, I’m about to take a turn and take over talking now.” And the precise way that people time that is pretty neat.
DANIEL: Yeah, and then you record a show like this and because we’re on opposite sides of the world, there’s a one-third of a second pause and it just screws up the whole timing thing. I notice this when I’m editing.
EMILY: Yeah, exactly. Well, so I mean, that happens… People can do that when they’re on the phone. Anybody can sing with each other in real time, but if you do it over the phone, which I do a lot with my mother and my child right now, my mom is on the other end of the phone in another continent and we can’t sing the song at the same time. For one of us, it will sound like we’re in time and for the other person, it will sound like we’re not. And so, it’s almost impossible to talk or impossible to keep singing because you hear it with that slight delay and it sounds ridiculous. Kind of like how if you feed people’s talk back to them while they’re talking, it’s almost impossible to continue.
DANIEL: That’s right.
EMILY: So yeah, we do everything in order to have this incredible sense of timing, and it’s crazy.
DANIEL: Now, typically linguistics, as a lot of people have studied it, has focused just on example sentences on paper and then we look at the grammaticality of it. And if we study dialogue at all, we bring in a lot of assumptions, like one person speaks at a time. And obviously, this means we’re not getting the full picture. But enough people, I guess, have studied linguistics thinking, “Well, that’s close enough.”
EMILY: Mhm.
DANIEL: But what’s missing from the picture if we’re just looking at the syntax of sentences on paper like we do for a lot of linguistics?
EMILY: Yeah, there’s so much missing. It’s like we’ve cut out our tongues or we’ve cut out our lungs or something out of the process, as well as our faces and everything. We use so many different resources at the same time to be able to talk in real interactions. There’s a lot to do with prosody, obviously, so the phoneticians will be going: But there’s more than syntax! or whatnot. The way it sounds is important.
DANIEL: I was thinking of this as a dad because when I am trying to communicate that we’re keeping it light, I sometimes sing as a dad, as like, [IN A SINGING TONE] “We are going to the park, let’s get our shoes on.” And it’s a thing that I drop into and I think what I’m trying to communicate is, “I’m not being angry, there’s no stress, but let’s just get going,” you know, that kind of thing. And the way that I communicate, that really matters.
EMILY: Yeah, absolutely. And there’s so much about how we’re physically interacting and whether or not we’re looking at each other that kind of determines whether we expect someone to be talking. And there’s so much that we can do if we look at real sentences instead of ones that are abstracted, where we can look at how the syntax is interrupted or redone, that have to do with how we can coordinate and make sure that we have mutual understanding or that we’re doing the same task and things like that. So, you can do… just kind of how I’m doing here, I started a sentence and then I cut myself off. I do a heck of a lot of self-repairs where I cut myself off and start again. And so, those are really critical even when just looking at the textual output that people have in real interaction.
DANIEL: So, you study non-lexical vocalisations and that’s really what we’re talking about here today. And I guess non-lexical vocalisations are the expressive sounds that we make that have some kind of communicative power or not, but they’re not words. Am I getting that about right?
EMILY: Yeah, exactly. I mean, it’s kind of arbitrary where we distinguish between a word and not a word, both for any linguist, also for lexicographers and even for everyday people, what counts as a word comes up sometimes: ~Is that really a word?~ and whatnot. And so the stuff that we’re very consistently not calling words, we would call non-lexical vocalizations. But, yeah, so they’re sounds, they’re messier and more chaotic than what we typically think of as words.
DANIEL: Well, I notice them when I’m going through the transcripts for the show, like I want to get some of the things… the ARGH out, but I just don’t know how to spell them because they’re not conventionalised. So, I guess that’s what we’re talking about. And I was using a lot of them in getting ready for this talk with you, because you know… you’re looking through the material. You’re trying to write a show and write questions. I know that we’re going to talk more or less spontaneously here because that’s what we’re doing. But even so, when you’re prepping, you sort of lay out the information in ways you’re going to need. And sometimes, it is hard and it is frustrating to know how you want to lay things out. So, I’ve been sitting at the table going ARGH, AARGH or sighing like [DEEP SIGH] [EXHALES NASALLY] a nasal sort of thing.
EMILY: Yeah, exactly.
DANIEL: And it struck me that these noises that you make when you can’t find the words, I feel they just really go deep. They must just be part of some kind of ancient brain. Do you ever feel that way?
EMILY: I mean, I always joke that the reason I’m on this project is because mostly I just speak in weird sounds, just wander around the house going, “Nyeah,” [CHUCKLES] some kind of weird connection to my reptilian brain maybe. I often have difficulty finding the words and will just gladly start to vocalize to make up for it, which doesn’t really help anybody except me.
DANIEL: But does it? I mean, maybe it does.
EMILY: Well, maybe. Yeah. I mean, it certainly lets them know that I’m trying to think of something.
DANIEL: Okay. There you go.
EMILY: Potentially, the tone of where it might be going.
DANIEL: Well, that’s it. Right? We were talking just earlier in this episode to Dr Eleonora Beier about UH and UM. And I used to cut those out of the transcript, but now I don’t because they have meaning. They serve as — sorry to recapitulate, audience — they serve to focus our attention, they serve to show, “I’m having a hard time thinking of the thing.”
EMILY: Yes.
DANIEL: I mean, of course, they’re traffic directors, like always. Is there anything I’m missing? What else am I…
EMILY: Oh, yeah. Well, sorry not to say, “Oh, yes, you’re missing so much.” [DANIEL LAUGHS] No, I mean agreement, “Oh, yeah.”
DANIEL: Yes.
EMILY: To me, from the way I look at things, UH are ways to most of the time, to indicate that you’re holding a turn, that the turn that you’re speaking is not yet finished and you’re not wanting somebody else to jump in and take their turn.
DANIEL: Yep.
EMILY: Or sometimes, it can be one to sort of indicate, “Uh, I’m about to take a turn or something,” looking for an opportunity to get the other person to stop. But it can also be used to foreshadow that something you’re about to say might be, uh, potentially problematic or unspecific or something like that. Um…
DANIEL: Oo, you did that just there. What were you doing?
EMILY: That one was on purpose. Uh, I hang my head in shame. [LAUGHS]
DANIEL: No, what were you doing just then when you said, “Something that’s, uh, problematic”? Were you hesitating there? Like, “I don’t want to say problematic, but I’m going to say problematic.”
EMILY: Yeah, exactly. I don’t want to say that it’s… um… now that one was more genuine.
[LAUGHTER]
DANIEL: Stop it!
EMILY: What I was going to say was vague, I guess. I don’t mean problematic in the sense that someone’s going to get in serious trouble, but it might be. I mean, people do use this when talking about really sensitive topics though, and then talking about gender or racism or something like that. There’s a lot of UHs because they want to make sure they get the wording right. And it’s pretty hard when their automatic way of thinking about it, or talking about it in any case isn’t quite apposite to what they want to do. Um… So, they’re really fascinating, UHs and UMs. But you were talking about taking them out, I’ve started experimenting with AI transcription as a first stage in some of the transcription work that I’m doing, because I have a sizable corpus that I’m working with that normally I would just do everything by hand, but it’s becoming not practical. But most software that does automatic transcription automatically removes those, which… aah!
DANIEL: Yeah. And kind of and sort of as well, which isn’t relevant. But those are important. Those matter, leave them in!
EMILY: And we had a conference one time — because so many conferences went online during the pandemic — and the software that was being used to broadcast the talks was removing the sounds that we were trying to present about. [DANIEL LAUGHS] It was just hysterical. It was amazing. I mean, it was also very frustrating.
DANIEL: That’s terrible!
EMILY: I mean, no one could have anticipated this was going to be the thing, but, oh, wow. And that becomes a problem with all sorts of softwares designed to think of that stuff as junk language, which, you know, isn’t true.
DANIEL: No, it’s really not. Well, since we’re talking about AI, I’ve been thinking about embodiment like a lot. A lot. We have the bodies we do, and they feel things. And this can have an impact on how language works and the metaphors that we use. And I’ve been thinking about this as it pertains to… of course, everybody’s obsessing these days about large language models because they’re really amazing and really new and really odd. And we have bodies and computers don’t, and computers also don’t go, “Daargh,” “Guh,” that kind of thing. And so, it seems like I’ve been asking our guests what’s uniquely human about language, and I feel like non-lexical vocalisations might be it, you know?
EMILY: They’re definitely part of it. Whenever I think, “Oh, this is something that AI doesn’t do or can’t do,” it’s usually because no one has thought to put it there yet. Um.
DANIEL: Or they’re ripping it out.
EMILY: Yeah, exactly. I think one of the things that people react to when they see things, like a few years ago there was Google’s Voice Assistant that would make really realistic-sounding voice calls that were ostensibly spontaneously generated. And then also what ChatGPT and stuff does, whenever people freak out and go, “Wow, it’s amazing what they can do,” I’m usually sitting thinking: they’re just doing this kind of interactional practice, and so they’ve suddenly figured out through the language model that that’s what people do in that spot, and that’s why it sounds so amazing, but it’s actually very straightforward and it could have been programmed in a long time ago if people had known to look for it. So, ChatGPT does all of these repairs where it’s going, “Oh, I’m sorry, I’ll fix that for you.” And they’re very good at recognizing when the human counterpart is initiating a repair. When they want to fix something that the computer has said or said that’s not quite right or something, it’s very good at recognising so many different ways to initiate those adjustments. And that makes it seem really human. I’m actually working with someone in the department here. We just did a paper on a toy robot that does use non-lexical vocalizations, [DANIEL LAUGHS] which was cute, but it doesn’t always use them correctly. Um…
DANIEL: Like what?
EMILY: Well, um, when you say UM, one of the things… you have a very, very small amount of time afterwards, 0.1, 0.2 seconds, um, which is a time frame that computers can’t really respond in, um, yet. Uh. You should be talking, so if you go UH, you don’t have much time after that. Um… Or it takes a lot of extra interactional work to get more than that really small amount of time. And so, the robot was supposed to produce someone’s name. You would teach it your name and then it would repeat back the name and be like, “Oh, it’s nice to meet you,” or whatever. Um, and do a happy dance kind of thing. Um. And it would do this whole display where it tries to memorize your name and it thinks about it. And what’s really happening is this is a cover for the robot processing, um, the name learning. And then it goes, at the end of that sequence, “Uh,” and there’s this really long silence.
DANIEL: [LAUGHS]
EMILY: There’s no need for it there. They just programmed it to be way too long. And so, the participants constantly in the data are going, “Is it broken? What’s happened? Oh, no.” There’s just this look of panic on their face and they’re looking to the other people who know how to use the robot. They glance over to them as if to check. And then just as they’re about to give up on it, the robot goes, “It’s your name, right?” And so it’s just… it comes at the wrong time.
DANIEL: Nice try, computer.
EMILY: Yeah, it’s very sweet.
DANIEL: It is.
EMILY: So, yeah. And I mean, there was a wonderful video. Boston Dynamics had a video of a robot that could do all this lifting and things. And the original video is just with music or something like that. But someone redubbed it to have the robot swearing and complaining about its work. And then, it would lift this box in one section, it went [HEAVY GRUNT]. It was adorable because the robot in the original doesn’t do that. But people thought it was really important to add that if you were going to make a joke about the robot. So…
DANIEL: Yeah, true.
EMILY: Anything like that definitely makes it seem more human.
DANIEL: And you have done work on noises of strain.
EMILY: Yeah. [LAUGHS]
DANIEL: And what those communicate. I want to get to those. But first, I want to ask what are you trying to uncover by studying these non-lexical vocalisations like HUH and ARGH, MMM! and things like that? What are you trying to learn? What got you here?
EMILY: What got us here was wanting to know more about the mundane aspect of language. So, how people use it in everyday life and in all sorts of circumstances that don’t exist in a lab or don’t exist on paper. Um… And wanting to know more about how people coordinate, how people understand each other. And also, increasingly recently, how… uh… people display that they are having some kind of sensory experience or that they know that someone else is having some kind of sensory experience, um, from seeing to tasting to touching and so forth, um, as well as sort of energy and pain and strain. And so, how those things get communicated and made public in interaction was what we wanted to understand more about. And this was also something that was just really missing in a lot of linguistic work. And so, we were trying to fill the hole as well as understand what role these had in interaction and what role they had for people to understand each other better.
DANIEL: Okay. Well, then let’s get to some of the work that we’re talking about here. Ah, I watched a video once where there was a performance and somebody fell down. One of the performers fell down, and the entire audience went, “Ohh!” So, they were sounding out this… I guess you could say it was a spontaneous cry of alarm. Is that all that’s happening or is something else going on?
EMILY: I mean, Goffman… The sociologist, Erving Goffman, was one of the earlier people to talk a lot about these and he called them “response cries.” And he mostly talked about them as something that people did for themselves. So, when you trip, you “argh” or “oops” or something like that and that kind of displays to other people both that the thing happened, but also that you are aware that it happened and that it wasn’t intentional and that normally you would have control over your own body, but you just had a lapse in that ability or something.
DANIEL: Are we talking about OPE or WHOOPS, or things like that?
EMILY: Yeah, exactly. Although he also had a very small section… I mean, it’s an amazing paper. It’s one of the only linguistics papers I know to talk about straining when you’re having a bowel movement. Great. Thanks, Goffman!
DANIEL: Thanks, Erving. Which is not communicative or not intentionally. That one just happens during the natural course of events. I don’t think it’s intended to be communicational, is it? I mean, you’re alone, hopefully.
EMILY: Well, I think that’s the thing, is that when something happens and becomes public in any accessible way, it becomes something that is usable for communication, even if it came out without the intention to do so. And so, people…
DANIEL: So to speak.
EMILY: …can… yeah.
[LAUGHTER]
DANIEL: Sorry.
EMILY: Oh god. No no. [CHUCKLES]
DANIEL: We’re talking about pooping, folks.
EMILY: Yay.
DANIEL: We talk about poop all the time on the show, by the way.
EMILY: Oh, that’s good. That’s a relief. [LAUGHS] It happens when you talk about body straining. It gets there. Um.
DANIEL: That’s interesting. I mean… What I’m getting from this is that when there are things, even if they’re involuntary things, those are on the table now. They’re available.
EMILY: Yeah, exactly.
DANIEL: For interpretation.
EMILY: Precisely. And so, one of the things that I look at right now… I mean, I haven’t actually published it, but it’s my piece. One of the things I’ve got in progress is looking at how the breath apparatus that is involved in doing rock climbing motions or indeed many different kinds of strain motions. Whether or not it’s intentional, it regularly results in a specific sound path where it goes [STRAINED SOUND] or some kind of very small strain grunt and then often silence. And then when the body relaxes, there’s this some kind of [EXHALES] outbreath.
DANIEL: [EXHALES]
EMILY: Yeah, exactly. And this is the body releasing the tension in the torso, releasing the diaphragm, releasing the glottis and things like that. And that means that when you are watching someone climbing and spotting them, making them safe, you know when they are no longer straining and when they have their balance because they’ve had an outbreath. So, it’s one of the ways that people then take that nonintentional sound and turn it into something that is communicative and useful for what they’re doing.
DANIEL: Wow.
EMILY: Um… so yeah, I think a lot of nonintentional sounds become part of it. But then, how did we get onto this? I feel like I haven’t answered an original question that got us started here.
DANIEL: It was somebody falling off the stage and then everyone says, “Ohh!”
EMILY: Yeah. So then, when it comes into doing this, that was what were really surprised to start seeing a lot of is that people don’t just do these for themselves, they do it on other people’s behalf. Um… So, this is really common when feeding infants, is you’ll feed them, and the moment the spoon goes in the mouth, the parent or the caregiver goes, “Mmmmm.” And sort of makes the sound for the child. And then about a year later, the child starts doing it for themselves, um, but it starts with the parent doing it, even though the parent isn’t eating, they’re not doing anything at all, they’re just feeding the child. Um…
DANIEL: I feel like a lot of these things are actually kind of, like… we’re talking about non-lexical, but I feel they lexicalise because you’d think they’d be spontaneous, but actually they’re different. Like I would say, “Ow!” and a Spanish speaker would say “¡Ay!”
EMILY: Yeah, exactly.
DANIEL: So, it feels like this is almost lexicalising… It’s a mixture of spontaneous and lexical. Have you noticed this?
EMILY: Oh, for sure. There’s tons of conventionalization. I’m learning Swedish because I live in Sweden and my child speaks mostly Swedish and only a tiny bit of English. And weirdly, the non-lexicals are almost entirely in English, so they say ¢ and they say OOPS and stuff, but the one in Swedish is AJ and HOPPSAN. And so, it’s totally different sounds, even for very related languages. The word for when you’re learning the word, whether you call it a word or a non-lexical, the sound…
DANIEL: I’m going to give it a pass on this one.
EMILY: Yeah yeah! For when you get new information is… typically for new information. When somebody’s telling you something, they’re informing you, you often go, “oh” in English.
DANIEL: Oh!
EMILY: Yeah, exactly. And then in Swedish though, you don’t. You say “jaha,” which is very different and I have so much trouble doing that, so I kind of end up going “oha” or something ridiculous where I’m trying to transfer over and then if I try and switch back to English, I’m going, “Ah, ah,” or, “Oh.” Messing it up that way. So…
DANIEL: Gosh.
EMILY: They absolutely are conventionalized within languages. There are some that are quite broad. I don’t remember if… One of Nick Enfield’s papers, he’s on the “Is HUH a Universal Word?” paper that they have.
DANIEL: I was just going to mention HUH? because we talked about that back in the Talk the Talk days, almost a decade ago, where every language has something like HUH? because everyone needs that job, but it’s always that prosodic thing, like Huh? well, maybe not always, but just about always [INTONATIONAL UPRISE] “uh?” It’s got that intonational contour.
EMILY: Yes. So that’s what they found… this is a paper by Dingemanse et all, and they did a cross linguistic study, and almost all the languages they studied, there’s what, 30-odd of them? have some kind of usually nasalized back vowel, like, ah, uh, hah, something like that. Um… Except for Danish, which…
DANIEL: [SHOUTS] Oh, of course! it’s always…
DANIEL and EMILY: It’s always Danish.
[LAUGHTER]
DANIEL: What do they do?
EMILY: They swear up and down that the way to do this is to say “hva.” And “hva” also exists as an option in Swedish and other Scandinavian languages. But they swear up and down that there is no “huh,” and there definitely isn’t much of it in the data that they have.
DANIEL: Oh, my gosh. And it’s not [RISING] “hva?”; it’s [DROPPING] “hva”.
EMILY: Yeah, “hva.” [QUICK RISE] Hva. Hva!
DANIEL: Okay. Well, having that contour…
EMILY: So, they can do a bit of an upturn.
DANIEL: Okay, all right.
EMILY: Anyway, so… but it is very widespread anyway, so that’s one of the ones that seems to have quite a similar phonology in a lot of different language families.
DANIEL: Okay. And this is reminding me also of something that you uncovered in your work studying Pilates classes in Estonian, which is kind of an up and then a down pair.
EMILY: Yeah, exactly.
DANIEL: Could you tell me a little about that?
EMILY: Yeah. I was so delighted that when we published this, someone responded to us, “If you’ve been in an exercise class, you know what this is.”
[LAUGHTER]
DANIEL: What did the instructor do?
EMILY: So, anytime that they did… I mean, yeah, this is easiest to hear. So, in my head right now is playing the Estonian, but I cannot say Estonian words in a way that is at all accurate. So, I’m trying to go back into English to, “Go into the exercise and go out of the exercise.”
DANIEL: Yeah yeah yeah yeah yeah.
EMILY: If you’re going up, and your toes are up, “Up on your toes and down now again,” and something like that. And so, whenever they would go into some kind of trajectory of exercise, then it tends to go up a lot, and then it comes down from that afterwards to come out of it. And they use that to coordinate with the class. So, everybody is basically in time.
DANIEL: Well, that’s the eternal rhythm, isn’t it? Because that reminded me of nothing less than adjacency pairs, because I’ve done a lot of work with adjacency pairs and somebody asks a question and that takes the tension up a level, and then that tension needs to be… I don’t know if tension is the right word, but it needs to be resolved with answer. And it’s no coincidence that question intonation goes up? because then you’re putting it on a shelf. You can’t just leave it on the shelf. You got to bring it down. So, calling it an up and down pair, it just seems very natural. I think that’s just some kind of eternal thing that happens. You raise the tension, it’s got to be lowered. You go up and then you got to go down.
EMILY: Yeah, exactly. Yeah, I mena, it’s interesting to think of it as an adjacency pair because a lot of what we were thinking about was in terms of how this is used across participants to create synchrony as opposed to sequence. But definitely the up and then the down is itself a sequence. So, absolutely on that front that it’s a pair of sound. Absolutely.
DANIEL: Okay. So, what I’m getting here is that when we do these kinds of non-lexical vocalisations, they have a number of functions. They can serve to help everyone coordinate their actions, like in an exercise class. “Okay, move your leg this way. Now, move your leg that way.” It can also be in the case of YUM for the baby, “Mmm.” It’s like you’re taking that mental state that you imagine somebody else has and you’re putting it into… you’re doing it yourself as a way of… Oooh, sorry, I’m out of road. What am I doing when I say HMM to a baby?
EMILY: I mean, in a way, we’re really struggling with this ourselves, is that there’s not an obvious action. This isn’t an easily mappable kind of speech-act-sounding thing or like, a request or an offer or anything like that. It seems to be that there’s definitely a sense where they’re doing being together within an experience where they’re enacting a jointness of the thing that they’re doing and sensing together. But it also definitely instructs the person how they should be sensing it. So, the baby should be taking this to be tasty! What wonderful food I’ve prepared for you. This is what we do when we taste delicious food.
DANIEL: You do think this is delicious, don’t you? Bit passive aggressive, I got to say.
EMILY: Yeah, a little bit. [LAUGHTER] And with the Pilates class, it’s kind of the same, is that you can infuse the instructions with strain as you do them, like, “[GRUNT] Go up.” Or, “[STRESSED] Yeah.” And that also instructs them that there is a certain amount of effort that’s expected with what they’re doing.
DANIEL: It’s a way of saying we are all together experiencing this together, which is super language-y, because that’s what we do with language. That’s one thing we do, is we form and maintain social relations. But this could be like, “We are all doing this sound together,” or, “I’m doing with you.” Do you know what it reminds me of? Okay, this is going to be a weird reference, but you’re in Sweden, so I’m going to go there. Ari Aster’s film Midsommar.
EMILY: Oh, yeah.
DANIEL: Can you guess the scene I’m thinking of? Because I’m thinking of a scene…
EMILY: No, because I haven’t watched it! I’m a person who avoids horror films at all costs.
DANIEL: I normally am too. And yet, I have really enjoyed Aster’s films. I found out that I thought I was scared of horror, and I am when it involves… um… violation of the body. But when it’s smart suspense, I really like that. So, I like Ari Aster’s films. And so with Midsommar, there’s a scene where Dani, the protagonist, is super-duper sad and she’s crying. Something is not going right. And she’s in a village where there’s this group and it’s their ceremonies that they’re doing. (Heh-heh.) And all of the women her age are there with her, and they’re clustering around her, and she’s crying, and they’re crying, and they’re acting out her grief. And they’re all together, and she’s looking around wonderingly at them because they’re feeling what she’s feeling. And she’s never had that with anyone. So, when I think about using these vocalisations, I just kept flashing on that scene for some reason because it was a time when they were all very clear that we are all on the same page here. We are feeling your feelings and we are in a situation where we understand each other. That’s like building empathy.
EMILY: Absolutely. That’s a really good example, I think. There’s so many studies of, I don’t know, rituals but also just events that happened in different groups around the world where people have these very emotional experiences together. And historically, people have often really exoticized those as some kind of ~native practice~ or whatever [DANIEL LAUGHS] as opposed to just, we all do this and we do it in different contexts and in different ways. And so, there’s nothing exotic about it all. We’re doing it in an exercise class, or we’re doing it with our children. And that it can happen in really dramatic ways where we all have those really strong emotions of grief together, or it can happen in really quiet, subtle ways of simply enjoying the same food. So, there’s definitely a lot, I think, here in terms of how we manage to overcome any boundary that exists between our bodies and their ability to sense together by using different practices in the voice and in the body to show that we are, in fact, together, that we have the same experience.
DANIEL: So, it’s not just about intersubjectivity… And this is a word that I ran across in one of your papers, and it’s just right. There’s intersubjectivity, which we talked about with Dr Nick Enfield in, uh, the… /‖ ‖ ‖/ [TRYING TO REMEMBER] episode called Consequences of Language. But there’s also intercorporeality, which is just perfect.
EMILY: Yeah.
DANIEL: It’s just perfect. That’s what it is.
EMILY: Yeah, for sure. It’s a wonderful word. Merleau-Ponty, he’s the philosopher that got started with this word, as far as I understand it. And it’s a wonderful description of having bodies that are together and that are linked in some fashion. And there are different ways to accomplish that. I mean, in some versions of theory, this is just sort of the basic way that we exist as humans is that we “have” intercorporeality with others. Our bodies are always mutually sensing, because whenever we’re in the same place, we know where each other’s bodies are through our sense of perceptions or proprioception. And we can also touch each other and have these reflexive touching experiences where we are both touched and being touched and things. But it’s also something that we can highlight and accomplish by coordinating our bodies together in a specific way.
DANIEL: And, of course, we would do that as humans! Of course we would do it, because we have these soft, vulnerable bodies. We can’t run very fast. We don’t have an exoskeleton. So, if you can’t do that, if you can’t link together with other humans, then you’re very vulnerable. So, yeah, it would be really important to do that even before language got started. But after language got started, yeah, we could do it a lot better. So, I guess intercorporeality is another one of those things that language built off of, but then also helped.
EMILY: Yes. Yeah, exactly.
DANIEL: Wow. Now, I guess this takes us to the beginning of language then. I… uh… You know that in 1866, the Linguistic Society of Paris banned debate on the origins of language, of how language got started, you’ve heard this, because there were lots of hypotheses about how it happened, and there was no basis of evidence for any of them. But we haven’t stopped asking about how language started. And usually, the theories that people advanced at the time were things like the heave-ho theory, that people started language because they needed to coordinate their cries as they were moving a thing. And then, there was the… you know… what else?
EMILY: Bowwow.
DANIEL: They were trying to imitate sounds. And when I thought about non-lexical vocalisations, I thought about these different hypotheses for how language got started. Do you have any opinion, or does this research have any clue for us? What do you think about the origins of language?
EMILY: Oh, boy. Um… I mean, I’m fascinated by this question as well, and… I mean… ohhhh, because the scientist in my brain is saying: [DARK VOICE] ~You can’t comment on this. You can’t really say.~ And because I don’t have recordings of protohumans using them.
DANIEL: Ah, come ahhhhhn. But this is entertainment, so you get to do whatever you want!
EMILY: Woohoo! Um…
DANIEL: [CHUCKLES]
EMILY: So definitely, people have also…
DANIEL: Woohoo, hey, there’s another lexicalised one.
EMILY: It’s true.
DANIEL: Sorry, I just… sidetrack here. I feel like before Homer Simpson, people said WAHOO. But after Homer Simpson, they said WOOHOO.
EMILY: Oh.
DANIEL: I don’t know.
EMILY: Oh, gosh.
DANIEL: Okay. Corpus study, we need to do the research. More research needs to be done.
EMILY: For sure.
DANIEL: Anyway, where were you going with that?
EMILY: I was thinking about Swedish, which is CHOHOO instead of WOOHOO. Um, I mean… There’s lots of theories about whether or not language started with gesture or with sound, specifically fronting of some kind or whatnot. And comparing that to great ape vocalizations and things. And, um… for starters, I think it’s always a bit of a false thing to compare to, like the alarm calls that apes have been found to have, because they’re not… I mean, I wouldn’t call them non-lexical. They’re a word, or at least there’s some evidence that they have some degree of conventionalization potentially in some species, or at least that they have some means of specifying different things. And so, they actually have words, is one way of interpreting that, rather than that they have these sorts of vague grunts.
DANIEL: Rrrrr… I am uncomfortable, but I cannot disagree. [LAUGHS]
EMILY: Oh, that they would have words, you mean?
DANIEL: …Yes.
EMILY: Yeah.
DANIEL: The language communicator in me is… unc… sees a headline coming on, but okay, that’s all right.
EMILY: Oh, no. Yeah, no.
DANIEL: That’s all right. Sentences! Ape sentences.
EMILY: Oh, dear. Yeah, not quite.
DANIEL: Okay, but this is not meaningful. Please. Please continue.
EMILY: Oh, no, but I mean anyway. Obviously the reasonable thing is that language was a multimodal thing at any point in history. They used whatever was available to them, be it material resources, their bodies or their voices and whatnot. So, there’s some way in which vocalizations were part of that and how they came to be specified as words is an interesting question. I think what non-lexicals today show us is that there has always been, or I think there was always, and there always will be a place for things that are not 100% conventionalized, that are not words in some kind of classic way, because we need sounds that do something that is less committed to a specific meaning. We need sounds that are kind of in the middle. And the word that I really love for this is “defeasible,” which I think Jack Sidnell uses a lot. He’s one of Nick Enfield’s colleagues, thinking of Nick Enfield. And the defeasible meaning that you can always backtrack from an interpretation. You can only half go on the record with any kind of commitment to a meaning. And so, you can always be like, “Well, that wasn’t what I said,” or something with one of those things.
DANIEL: Plausible deniability.
EMILY: Yeah. And non-lexicals do this so well because they get across something that’s a bit more of a gut feeling or a timing aspect or coordination, without necessarily, for instance, making an assessment all the time of what you’re doing. You’re not saying it’s good or bad or anything in particular. And so, I think those were always part of what we did. And whenever we exclude those from our theories of language, we’re going to have a partial view because we won’t be including a lot of the really necessary resources people had and would have needed to have in order to get things done. So, I think always… the heave-ho hypothesis might not be literally how words originated, but it definitely had to have happened. And we still use HEAVE-HO now for a reason, and it’s not going to disappear. Even if we conventionalise these sounds, new ones will show up that will take their place in doing the work that is not super specific or that is not focused on a meaning-based sound, but something that’s a little bit broader and a little bit more based, more focused on doing coordination work or emotion or something like that.
DANIEL: So, they’re going to be with us for a long time.
EMILY: Yeah, exactly.
DANIEL: Tell me something that everyone gets wrong about your work that you keep bumping up against and you have to keep correcting it.
EMILY: Oh, gosh. [CHUCKLES] I mean, prescriptivism is one of the big ones, for sure. People always assume that UM and UH and things are something that should be gotten rid of, or are bad about language, or indeed that there is anything bad about language, but of course there isn’t. And maybe that’s just sort of a general linguist thing. People are really uncomfortable when they find out that you’re studying interaction because the immediate thought is, “Is that what you’re doing right now?”
DANIEL: Oh, no. [CHUCKLES]
EMILY: Or, “Are you manipulating our conversation?” I wish I… I mean… I don’t know that I do wish that I could do it that fast, but I can’t do it that fast.
DANIEL: Right. And yet, you are manipulating people’s attentional state.
EMILY: Yes.
DANIEL: People’s internal knowledge and beliefs by just simply engaging in conversation at all, whether lexical or non-lexical.
EMILY: Yeah, exactly. And it’s something that we do automatically. It’s really hard to express the idea that what we’re doing is somehow unconscious or subconscious without implying that there’s some kind of secondary brain functioning. And it’s more just that it’s a really practiced skill. I mean, something that people get wrong a lot or that I come up against a lot is that there’s something about timing where people are like, “Oh, if there’s a few seconds pause, then that’s a problem.” No, if there’s a very, very tiny pause of less than a few tenths of a second, we already know there’s a problem. So those are some of the big things, I think, that people struggle with. So yeah.
DANIEL: So, I noticed this sentence. I think we’re ready to wrap it up with this sentence that I found in your paper. I’m just going to read it. “Vocalizations constitute a specific means of accomplishing sociality that has so far flown below the radar of studies and communication, having been treated as perhaps too subtle and evasive or too chaotic.”
EMILY: Mm.
DANIEL: I mean, they are subtle and they are a little evasive and they are a little chaotic. So, I guess I understand why we have been excluding them from some levels of analysis. But, man, what I’m taking away from our conversation is that they are so important.
EMILY: Yeah, for sure. And, I mean, it can be really difficult to analyze something that is in any way evasive or inconsistent because then that makes it hard to know what you’re working with or whether you’ve got another thing that is the same or whether you have something new. But I think it’s important to include looking at things that are so messy because there’s a reason for them to be so. And it gives us a resource as people to do stuff with each other and to be able to be noncommittal if we want, or to be really expressive in a creative way if we want. So, this is very much in the vein of language is infinite. That means that these things are part of that.
DANIEL: We’re talking to Dr Emily Hofstetter of Linköping University. Emily, how could people find your work and what you’re up to?
EMILY: Um, I have become one of those academics that never updates their website. Oh, god. [CHUCKLES]
DANIEL: Oh, no.
EMILY: It’s very out of date. I will fix that.
DANIEL: [LAUGHS]
EMILY: I have a Twitter account, which is @EMdoesCA, which is a pun on my name and methodology does conversation analysis, which is my area. And I technically have a YouTube channel that I update maybe once a year with a new video, but not super often, with just sort of different ways to think about pieces of media and how they relate to interaction, like movies and stuff like that. And then, we have a website for our project, which is nonlexicalvocalizations.com.
DANIEL: nonlexicalvocalizations.com. Vocalizations with a Z, not an S, -zations.
EMILY: Yes. Sorry, British folk.
DANIEL: And Australians.
EMILY: Oh. [LAUGHS]
DANIEL: Oh, what’s also cool is a lot of your work is open source.
EMILY: Oh, yeah.
DANIEL: So, that means people can read it, which is good.
EMILY: That’s the goal. I mean, sure contact if there’s anything that’s missing, we can always send personal copies. That’s one’s always an option, but yeah, that’s definitely a big goal to make sure it’s available.
DANIEL: Emily, thanks so much for hanging out with me today and talking. This has been super illuminating. I think I’ve added a few different nodes to my bulletin board of red string.
EMILY: Okay, thank you so much for having me. It’s been so much fun to talk about it.
[TRANSITIONAL MUSIC]
DANIEL: And now, it’s time for Words of the Week. Our first one, suggested by Diego. Thanks, Diego. The word is a pronoun. And the pronoun is IT. Um… We’ve talked a lot about pronouns on the show, and for many of us, if you call somebody an “it.” That just feels really super duper wrong.
HEDVIG: Unless it’s a child, maybe.
DANIEL: Yes. Because before a child is born and you don’t have any indication as to gender… which people just obsess over. I don’t know why, but you say, like, “What color are you painting its room?” And then people often say, “Oh, sorry, I said ‘it’. That’s kind of odd, isn’t it? I don’t know what to say.”
HEDVIG: Yeah. Or like, “It doesn’t have a heartbeat yet.” Like, “The fetus doesn’t have a heartbeat yet.” I remember having… that there were discussions maybe 10-15 years ago. I heard discussions about using IT as a pronoun for people, like a non-binary pronoun. But then I didn’t hear about it for a long time. But then people were talking about XE, like ze, zie, and then THEY and then other ones, and I was like, “Okay, well, I guess we’re leaving behind this IT thing.” But then, apparently it’s popped up again.
DANIEL: It’s popped up again. Many people do feel like being referred to as “it” is kind of dehumanising, but some people are really liking IT and choosing to be referred to by the pronoun IT. This is on yahoo.com, link on our website, one user says, “IT instinctually gives me the most gender euphoria. When someone uses that to refer to me, I get so happy and excited.”
HEDVIG: Okay.
DANIEL: And a lot of people feel that way. Not many. I checked out some data from the Trevor Project. They surveyed forty… well, they’re always doing surveys, but in this one survey, they studied 40,000 young people, one in four of them, so about 10,000, said they were nonbinary. Among them, the most popular pronoun was THEY. One third of them chose THEY. IT was very small in the 2020 survey, only 50 people said: My pronoun is IT. But in 2021, 150 people said they like to be referred to as IT. That’s three times as many. So, small but growing, can we say?
HEDVIG: Yeah, but to be fair, if we talk about the recommendations, I don’t know, but I’m going to guess that people who prefer IT would not maybe be as bothered by THEY as they would with either HE or SHE. So. if you don’t know, a good bet is probably THEY, right?
DANIEL: Yeah. I imagine that I think somebody who likes IT wouldn’t mind THEY. But I think if somebody likes THEY they would probably really mind IT.
HEDVIG: Yeah. And someone who likes IT would probably prefer THEY over HE or SHE.
DANIEL: HE or SHE, yeah.
HEDVIG: Right. So, if in doubt, probably… but maybe this survey asked this question, something like, “What’s your second most preferred?” or something.
DANIEL: Oh, that’s interesting.
HEDVIG: You and I are theorising now about what these people prefer and we don’t know, but I’m going to guess also just by the numbers you just said that the people who identified as nonbinary, the most common one, a third, you said, was THEY/THEM.
DANIEL: That’s right.
HEDVIG: Right. So probably, if you don’t know, general recommendation is probably that you’re safest with THEY/THEM then.
DANIEL: You got to be careful with IT but if people like it, you use IT. So, there you go.
HEDVIG: Yeah.
DANIEL: Now, let’s see. This one, Hedvig, I’m going to need your help with, because it’s NYEAH.
HEDVIG: NYEAH.
DANIEL: Nyeah.
HEDVIG: NYEAH.
DANIEL: I remember bratty kids on the playground singing the taunting song. “Nyah nyah nyaaahh.”
HEDVIG: Yeah, I feel like that’s a different one. This is actually no and yes.
DANIEL: It’s no and yeah?
HEDVIG: It’s no and yeah. It’s no and yeah. Yeabh. It came up in my feed from these really funny people on TikTok. Angelo Marasigan.
DANIEL: It’s @_angelo.marasigan. Yes.
HEDVIG: There’s these funny siblings who make fun of the Kardashians and they just make up fun things and they talk with, like a super creaky voice and they do, like, all the things. And they say, “You took my charger. That’s not ethical or professional.” They always say that, and I love it.
DANIEL: I checked out a few of their videos and they always do that. They say, “Nyeah,” to each other. Nyeah.
HEDVIG: Nyeah. And it sort of means if someone’s like, “I didn’t take your charger, I only borrowed it,” you could say, “Nyeah, that’s not right.”
DANIEL: “Nyeah, you kind of did.”
HEDVIG: “You kind of did. You kind of did.”
DANIEL: Nyeah. So almost like, “I don’t believe you,” kinda?
HEDVIG: I’m still trying to learn what it means because in Swedish we have NYOH, which is a combination of NO and YES. And that is like if you say something like… if you ask me if my home is in Germany, then in English I might say, “Yes and no. My home where I reside right now is in Germany, but I’m from Sweden. I’ve lived longer time in Sweden. My home country is Sweden.” Right?
DANIEL: Oh, yeah. Is your home country Australia? And I’d be like, “I mean, it is now and it’s complicated. Nyeah.”
HEDVIG: Often in English, I hear people say, “Yes and no,” and in Swedish you could say NYOH.
DANIEL: Do you get a feeling of whether this use of NYEAH is particular to that duo, that sibling duo, or do you think it’s spreading a little bit? Do you think anybody else is using it?
HEDVIG: I think other people are using it. I think the Kardashians are using it. I think that’s why they’re doing it. Right? Anyway…
DANIEL: I have no idea. I’m out of that world.
HEDVIG: As listeners of this show know, I’m desperately trying to get Gen Z to like me.
[CHUCKLES]
HEDVIG: And I’m trying to keep up with their trends.
DANIEL: They love you, Hedvig.
HEDVIG: And I’ve heard this word and I want to know what it means so I can use it, so I can be cool.
DANIEL: Well, you’re going to have to infer it.
HEDVIG: So, if you’re a Gen Z and you know what NYEAH means, does it mean yes and no or does it mean something else?
DANIEL: I feel like from the videos that I’ve watched from @_angelo.marasigan, it seems like it’s kind of confrontational a little bit. When they say it to each other, they’re like, “Mm. Mm. I’m not sure I believe you,” or, “I’m kind of acting in opposition to you at this moment. Mm, nyeah.”
HEDVIG: Yeah, it’s definitely not agreement.
DANIEL: No, it’s not. Nyeah. Let’s do our last one. This one was suggested by PharaohKatt and Ariaflame: GLASS CLIFF.
HEDVIG: Okay. I have not clicked a thing. I’ve seen it in our run sheet, but it makes me think of GLASS CEILING.
DANIEL: Well, how would you describe the glass ceiling?
HEDVIG: The glass ceiling is this idea that there is something limiting your career advancement upwards or maybe real estate buying or something. There’s something where you’re trying to advance upwards to get more free time or earn more money or get a higher rank or more authority or something, and that there is some sort of invisible level at which you cannot advance beyond, and that that is structured by something that is beyond your control. It’s not that you’re not good enough, it’s not that there are many other good candidates. It’s something about your gender, your background, your ethnicity. Something like that.
DANIEL: Yep yep yep, that is the glass ceiling. First reference in the Oxford English Dictionaries from 1984 in Adweek. Family Circle Editor Gay Bryant used it. However, also the Wikipedia article on GLASS CEILING backdates it, coined by Marilyn Loden during a speech in 1978. So that’s where we start seeing it. But since then, we’ve seen a lot of different glass metaphors. Like a GLASS FLOOR where if you are white and male, there’s a limit to how far down you can go if you fail. There’s the GLASS DOOR, the initial hiring process that keeps women and people of colour out. The GLASS WALL, a stratification of careers based on gender or ethnicity. And the GLASS ESCALATOR where men get promoted faster than women. Okay, so lots of glass metaphors.
HEDVIG: Yeah.
DANIEL: Now, the GLASS CLIFF.
HEDVIG: Cliff!
DANIEL: If you had to take a guess, do you have any feeling?
HEDVIG: If we take the example with men and women, but I think it applies to other dichotomies of oppression as well. But if we take the example with men and women, is glass cliff when women get overworked and burnt out and get thrown over the cliff?
DANIEL: You are right there. So, let me describe it here. “We are happy to promote women and people of colour when things aren’t going well in the organisation.”
HEDVIG: Oh, I’ve seen this. Yeah. Yeah yeah yeah. Oh, no, I know what you mean. That’s more specific than what I said.
DANIEL: It is. When the risk of failure is highest, we just let them get burnt out and stressed out, and then when everything falls down, we blame them for the failure. So, we’ll promote them during risky times and push them to the edge of that glass cliff.
HEDVIG: And then say, “You know, we tried a woman CEO two years ago and then went into bankruptcy.” And it’s like, “Yeah, but you promoted her to CEO like three months after your auditor said things were bad.”
DANIEL: [LAUGHS] That’s right. “Oh, we tried but this just fits into our narrative that there aren’t any qualified Black candidates or no qualified women candidates,” blah, blah, blah. So, it fits into that narrative.
HEDVIG: That’s a little bit different because I was gonna say that… I was thinking more of the thing of that there’s a little bit of a tendency, for example, women in workplace to take on more pastoral duties and to pick up other people’s slack and to get the coffee and to be emotional supportive so that they end up not spending time in their work on things that they should be spending time on in order to get promoted.
DANIEL: That happens.
HEDVIG: That happens too. But that’s not this glass cliff.
DANIEL: Not necessarily the same thing.
HEDVIG: Maybe not even a glass thing at all. What is it?
DANIEL: What would it be? What kind of glass thing would it be? Something that comforts someone. Like a glass pillow.
HEDVIG: What about a glass weight?
DANIEL: What’s something that you use… a glass dummy? A glass binkie? What’s something that cares for people, a glass teddy bear? Something… I don’t know. What am I thinking of here?
HEDVIG: No, but I think the focus should be on the weight on the person, not the support to others.
DANIEL: Oh, so a glass anvil.
HEDVIG: Right. Glass collar. Like something that’s on the other person… Anyway.
DANIEL: A glass yoke?
HEDVIG: I don’t know.
DANIEL: I don’t know.
HEDVIG: But a glass yoke. Oh, that’s not bad.
DANIEL: That’s what you put on oxen, not the egg part. So IT, NYEAH, and GLASS CLIFF: our Words of the Week. Let’s get to some comments. This one’s from Pontus on Facebook. He’s affectionately known as Moon-Moon around these parts. Pontus says, “Just heard a fascinating phrase for Word of the Week. Acoustic toothbrush.”
HEDVIG: [GASPS]
DANIEL: Acoustic!
HEDVIG: Yeaaaaaah.
DANIEL: You’ve heard this, right? I think we might even mention this before.
HEDVIG: I love it.
DANIEL: Acoustic as opposed to electric. People are talking about: it might better to brush your teeth acoustically. Don’t you love it?
HEDVIG: I love it.
DANIEL: So at first we had guitars. Then, we got electric guitars. So now what do you call the normal guitars that you don’t plug in? Well, that’s an acoustic guitar, because acoustic means using sound energy in its operations. So now, we’ve got acoustic as opposed to electric. Then, we started seeing lots of references to acoustic bikes. Meaning the old kind of bikes that aren’t electric.
HEDVIG: Oh, my god. That’s so funny. I have an acoustic bike.
DANIEL: People are being funny, but it’s exciting for me to see the meaning of a word get diverted and take on a completely new meaning. And by the way, this is called a retronym. Like, we used to have phones and then we got smartphones. So now, what are we going to call the old phones? So, we call them dumbphones and that’s a retronym.
HEDVIG: That’s cute.
DANIEL: I also see ANALOGUE in this context. An ANALOGUE bike instead of a digital bike. So thanks, Pontus, for that one.
HEDVIG: That’s really fun. Yeah, I really like it. I was just thinking in my head, if like acoustic,… it’s so funny that it’s acoustic because the toothbrush doesn’t make any sound but anyway.
DANIEL: I guess they all make a sound. I was going to say also manual toothbrush. But then, manual implies hands and you still use your hands for an electric toothbrush. So, I don’t know. What do you think, listeners? This one’s from Franco via email. In our last episode, we talked to Jack Grieve about fake news. Franco asks, “What similarities and differences do you see between fake news and propaganda? As I remember, you defined fake news something like the sharing of intentionally false information with the intention of making listeners believe it as true. Please help me see the differences between that and propaganda?” What do you think, Hedvig?
HEDVIG: I’m not sure that they are that different. Depending on how we think of propaganda. Propaganda could be a number of different things. One of them could be only the positive kind, only saying like, “Our leader is the best, he’s the best, our country is the best, everything’s great.” That could be fake news, or it could be real.
DANIEL: It could be.
HEDVIG: There’s also the negative. “Other people are bad. They’re sabotaging us. They’re really cruel. They’re doing this and this.” Both of those things could be fake news and propaganda. Propaganda doesn’t necessarily have to be all lies. It could just be lies by omission as well.
DANIEL: Propaganda could be true, couldn’t it? Whereas fake news isn’t.
HEDVIG: Well, fake news, I think, also could be… I think they’re really similar. I’m wondering if fake news is just a subcategory of propaganda because…
DANIEL: Could be.
HEDVIG: …when we say fake news, we called it sharing of intentionally false information with the intention of making listeners believe it as true. I think that sometimes things go in there that are lies by omission…
DANIEL: Or shading it a certain way to promote an agenda.
HEDVIG: Which is not technically false. Right? Like, the interpretation and the context and the implication of it is false but you could share a small amount of… do you know what I mean?
DANIEL: Truth. Yeah, I think that too. I think propaganda can be used to promote an organisation or a group or an ideology or a cause or a person. Fake news doesn’t have to do that. It can promote information about something other than itself just to sow confusion and discord. Whereas propaganda seems to me to lean more towards promoting a thing and fake news doesn’t have to.
HEDVIG: But I do agree with Franco that they are quite similar and maybe we should even consider them. Because, for example, at least in Swedish news, they often talk about that one of Russia’s common intelligence tactics the last 10 years have been to sow chaos, to just, like, not promote any specific candidate in other people’s election but just make people doubt the election system in general.
DANIEL: Yeah yeah yeah. Which isn’t propaganda. That’s fake news.
HEDVIG: Which isn’t really propaganda, but it is probably fake news and it is maybe propaganda in a very roundabout way because then you can go back to your own domestic public and be like, “Hey, the US has a shitty election system but ours is great.”
DANIEL: Yeah, that’s true. They really do feed into each other. I got one more thing and that is that you ignore propaganda because it’s an unreliable source. And with fake news, you often don’t know what the source is.
HEDVIG: Yeah, propaganda is a little bit more associated with Himmler said blah, blah, blah, you know where it’s coming from.
DANIEL: And I’m going to ignore it because he would say that. That’s what I got. But again, there’s a lot of overlap, but the propaganda does lean a little bit that way, and fake news leans a little bit the other way.
HEDVIG: Yeah.
DANIEL: Thanks, Franco. Last one from Bill on Twitter. The background here is that in the last episode, Ben and I talked about how President Zelenskyy of Ukraine had signed legislation to ban Russian placenames and make knowledge of Ukrainian language and history a requirement for citizenship. On the show, we are strongly against Russian aggression. We realise that Ukraine is literally fighting for its existence, so strong measures are called for. But we are also aware that measures like these are often used to punish linguistic minorities. And many people in Ukraine speak Russian as a first language. So, Bill said something I hadn’t realised. Bill says, “Hey, @becauselangpod, apologies if someone else has mentioned it, but President Zelenskyy’s first language is Russian. That doesn’t mean I agree with the language requirement for citizenship.” Okay.
HEDVIG: It is. Yeah. And he’s made several speeches in Russian to Russians. So recently, for example, there was a hack of a lot of emergency systems in Russia where a specific speech by Zelenskyy was broadcast, and it was in Russian, directed at Russians saying, “Don’t send your soldiers to die in Ukraine because it’s only for your corrupt leaders, and we are really going to fight.”
DANIEL: That’s right.
HEDVIG: And it’s obvious to me whenever I hear things like this that it really can go bad ways. I mean, in this case, I believe that President Zelenskyy knows that there are people who speak Russian in Ukraine who don’t support the Russian regime because he’s one of them. So, he’ll know.
DANIEL: And he wouldn’t want to obliterate the Russian language out of Ukraine because he would presumably share a connection with Russian speakers. So, that’s an important bit of context, I thought.
HEDVIG: It is, but it is still… Yeah. I wonder if this law is going to stick.
DANIEL: The most significant thing to me, and I mentioned this on the show, was that the presence of Russian history and placenames and language is being taken by Russia as justification for their occupation, and that’s not acceptable.
HEDVIG: Right. And placenames change all the time. And street names and placenames are some of the primary ways that nation states signal ongoing politics. We recently had a talk here at our department by Melody Ross about placenames in East Timor. So East Timor has been occupied by Indonesia. It’s been independent. It’s been under UN and everything like that. And all the time, they keep changing the street names at every time to signal what’s going on. You know? So, when they became independent, they removed any Indonesian place names to be like, “We’re definitely not Indonesian.”
And people get hung up on placenames and street names, but they change. I think if you look up the street you live on, just check how long it’s been called that or how long your neighbour, because it’s not always that long. Especially in Soviet times, a lot of things change. Just south of where I live, I live in Leipzig in East Germany, and south of here, there’s a place called Chemnitz that used to be called Marx-Stadt for several decades and now it’s back to Chemnitz. It changes a lot. So place names… mmm, yeah.
DANIEL: They’re a real political weathervane. Bill, thanks for adding that context. Thanks also to Dr Russell Gray, Dr Eleonora Beier, and Dr Emily Hofstetter for the chat about their work. Thanks to everyone who gave us ideas for the show. Thanks to Dustin of Sandman Stories for recommending us to all and sundry. By the way, his podcast, Sandman Stories, very worth listening to. Thanks to the team from SpeechDocs who transcribes all the words. Thanks to our patrons who support the show and keep us going. And, Hedvig, thanks to you for coming on and having a chat. I always enjoy making a show with you.
HEDVIG: Aww. God, Americans, hey?
DANIEL: So wholesome.
HEDVIG: Like, I just want to sympathise with all of our listeners. He does this all the time, and it makes me and Ben incredibly awkward because we don’t know what to say back. You’re like, [IN DANIEL’S VOICE] “Oh, I really like making a show with you. You do such a great job.” And we’re like, “Uhhh, same? We don’t know how to be earnest about emotions! Go away!”
DANIEL: I can’t compliment either one of you. It’s so hard. How am I supposed to be effusive? Ugh, I don’t know.
HEDVIG: Anyway.
DANIEL: We are incredibly grateful for the many ways that people help us in making this show. And if you would like to be someone who helps us out, there are loads of ways you can do that and I’m going to tell you about some of them right now. You can send us ideas and feedback. We’re becauselangpod on all the socials. Leave us a message with SpeakPipe on our website becauselanguage.com. Send us an email, hello@becauselanguage.com, I’ll do my best to get through all of them. People are sending us stuff. And the comments that we get, the mailbag questions, keep sending them, folks, you’re doing great. Another thing you can do is tell a friend about us or leave a review. We really depend on word of mouth and good recommendations. So, thanks to everybody who’s doing that. If you do leave a review, I’ll find out about it and I’ll read them on the air. Uh, if it’s good. The next thing you can do is become a patron. And as I’ve mentioned, you’ll be getting bonus episodes, you can come to our live episodes like the one we’re doing next. Don’t sleep on it, it’s going to happen soon. You can hang out with us on Discord, which is super-duper fun. I post all my Wordle scores, and I post memes. I’ve been on the meme beat. It’s been great. [HEDVIG LAUGHS] And also, by becoming a patron, you’ll be giving us money so we can afford to pay the bills, webhosting, transcripts by SpeechDocs, they’re doing a great job. They help us to keep our shows readable and searchable. So, we’re really grateful to all of our patrons.
HEDVIG: Speaking of all of our patrons, we have a lot of them, and we like to give a special shoutout to the top patrons. So, those are the people who give us above… what’s the threshold again?
DANIEL: These are our patrons who are at the Supporter Level.
HEDVIG: Ah, okay. So here they are. Iztin, Termy, Elías, Matt, Whitney, Helen, Jack, PharaohKatt, LordMortis, gramaryen, Larry, Kristofer, Andy, James, Nigel, Meredith, Kate, Nasrin, Joanna, Ayesha, Steele, Margareth, Manú, Rodger, Rhian, Colleen, Ignacio, Sonic Snejhog, Kevin, Jeff, Andy from Logophilius, Stan, Kathy, Rach, Cheyenne, Felicity, Amir, Canny Archer, O Tim, Alyssa, Chris, Laurie, and Aengry Balls. And our newest patrons at the Listener Level are Benedict and Terry. And we have a new one at the Friend Level, and they are called Boo88 or Bob, I’m unsure. Then, we also have FauxFox, who has bumped up their pledge to the Listener Level and Tadhg doubled their pledge. Thanks to all of our wonderful patrons. Our theme music has been written and performed by Drew Krapljanov, who’s a member of Ryan Beno and of Didion’s Bible. Thanks for listening. We’ll catch you next time. Because Language.
DANIEL: Woo hoo! Thank you very much.
HEDVIG: [SIGHS] Okay. All right.
DANIEL: Big show.
HEDVIG: Big show. I think I need to go eat something.
[BOOP]
HEDVIG: Do you know what my cat just did?
DANIEL: No.
HEDVIG: I bought a potted plant. Like, it’s a sort of tree, small tree with daisies, and she loves to just sit inside the pot.
DANIEL: Oh, yeah. Dirt Cat.
HEDVIG: So, I’ve had to put cardboard so she doesn’t kick up all the dirt and kill the plant.
DANIEL: Ah.
HEDVIG: And now she’s decided to just have a nap there.
DANIEL: Ah, like on the cardboard?
HEDVIG: Yeah. It’s very cute. Slightly, probably not great for the plant.
DANIEL: I am Cat. My name is Dirt.
HEDVIG: She’s very cute. I’ll send you…
DANIEL: Okay. We always kill the thing we love.
HEDVIG: [LAUGHS] Okay!
DANIEL: Like democracy or the fandom.
HEDVIG: I’ve put pictures in our Discord channel, Cats and Animals.
DANIEL: Thank you. That’s great.
[Transcript provided by SpeechDocs Podcast Transcription]