We’ve all seen style guides that tell us what to say and what not to say. Has a style guide ever asked you what you wanted to say? Or challenged you to examine your thinking? This one does. It’s the Conscious Style Guide by Karen Yin, and she joins us for this episode.
Timestamps
- Intros: 0:35
- News: 5:52
- Related or Not: 31:13
- Interview with Karen Yin: 48:43
- Words of the Week: 1:30:10
- Comments: 1:44:07
- The Reads: 1:47:45
- Outtake: 1:54:06
Listen to this episode
Videos
Patreon supporters
Thanks to all our patrons!
This time, Daniel classified the alphabet by shape, round ones first. Then partly straight, and then all straight, by number of strokes.
It goes: O C S Q J G U D P B R I L T V X Y Z H N K A F E W M. That’s alphabetical order now.
Here are our patrons at the Supporter level.
- O Tim
- Colleen
- Chris L
- Chris W
- Canny Archer
- sæ̃m
- Sonic Snejhog
- Stan
- Steele
- J0HNTR0Y
- Joanna
- Jack
- James
- gramaryen
- Diego
- PharaohKatt
- Rodger
- Rach
- Rachel
- Rene
- Ignacio
- Iztin
- LordMortis
- Luis
- Linguistic C̷̛̤̰̳͉̺͕̋̚̚͠h̸͈̪̤͇̥͛͂a̶̡̢̛͕̰͈͗͋̐̚o̷̟̹͈̞̔̊͆͑͒̃s̵̍̒̊̈́̚̚ͅ
- Lyssa
- Laura
- Larry
- Tony
- Tadhg
- Helen
- Nigel
- Nikoli
- Nasrin
- Kristofer
- Kathy
- Keith
- Kevin
- Ariaflame
- Aldo
- Alyssa
- Ayesha
- Andy B
- Andy from Logophilius
- Amir
- Amy
- Fiona
- Faux Frenchie
- Felicity
- Elías
- Wolfdog
- Whitney
- Molly Dee
- Mignon
- Margareth
- Manú
- Meredith
And our newest patrons:
- At the Listener level: Justine
- At the Friend level: Greg and Pigeon
- And our newest free members: Taylor, Claire, Kell, Alex, Paul, Hugo, Ben W, Niclas R, Miss Behaving, 達維安, and John R.
Become a Patreon supporter yourself and get access to bonus episodes and more!
Become a Patron!Show notes
How this long-lost Chinese typewriter from the 1940s changed modern computing
https://www.npr.org/2025/07/05/nx-s1-5405452/chinese-typewriter-mingkwai-stanford
Lost and Found: The Unexpected Journey of the MingKwai Typewriter
https://madeinchinajournal.com/2025/05/02/lost-and-found-the-unexpected-journey-of-the-mingkwai-typewriter/
The fascinating evolution of typing Chinese characters
https://www.technologyreview.com/2023/08/23/1078274/fascinating-evolution-typing-chinese-characters/
Medieval Chinese typesetting technique
https://blog.plover.com/IT/Chinese-typesetting.html
ChatGPT Is Changing the Words We Use in Conversation
https://www.scientificamerican.com/article/chatgpt-is-changing-the-words-we-use-in-conversation/
Delving into LLM-assisted writing in biomedical publications through excess vocabulary
https://www.science.org/doi/10.1126/sciadv.adt3813
[PDF] Empirical evidence of Large Language Model’s influence on human spoken communication
https://arxiv.org/pdf/2409.01754
The Conscious Style Guide: A Flexible Approach to Language That Includes, Respects, and Empowers
https://karenyin.com/books/csg-book/
The Conscious Language Newsletter by Karen Yin
https://mail.consciousstyleguide.com
Conscious Language + Design | Facebook
https://www.facebook.com/groups/consciouslanguage/
The Internet has created a new word to deal with Lorde’s pubic hair
https://www.rova.nz/articles/the-internet-has-created-a-new-word-to-deal-with-the-reaction-to-lordes
boomerasking | Dictionary.com
https://www.dictionary.com/e/slang/boomerasking/
What is a micro-retirement? Inside the latest Gen Z trend
https://www.fastcompany.com/91357784/what-is-a-micro-retirement-inside-the-latest-gen-z-trend
If not fren, why friend shaped?
https://www.reddit.com/r/FunnyAnimals/comments/14ubbdi/if_not_fren_why_friend_shaped/
Maitiú’s comment in full:
Hello again!
Just a quick note to say thanks for reading out my e-mail, it was a nice surprise!
I enjoyed your piece on the many words for snow. We have a similar thing here in Ireland, best embodied by the book 32 Words for a Field by Manchán Magan. It’s a huge hit among English-speaking Irish people but most Irish-speakers I know roll their eyes at it.
One more thing – Hedvig asked in the most recent episode if the names Keir and Keiran were “Related or Not”. They are related! Both come from the Irish ‘ciar’ meaning ‘dark’ (as in hair-colour). Ciar was the son of Fearghas (Fergus), the king of Ulster. He was banished to the south west of Ireland where his descendants were knowns as Ciarraighe (the people of Ciar, now spelled Ciarraí) – or ‘Kerry’, another name. Ciar is uncommon in Ireland now, with the diminutive Ciarán much more common. All of those names are Anglicised with a ‘K’, reflecting the hard C pronunciation in Irish. The surname Kirwan is also related – from the patronym Ó Ciardhubháin, descendant of the dark-black one (again, hair). The surnames Mulkerrin and Kerrigan has a similar history – Maolchiaráin and Ciaragáin. An interesting link is the surname Comer, which is common enough in Galway where I live. It comes from a mis-translation of Ciar, meaing dark, and Cíor, meaning a comb (like for your hair).
They are also all related to the Irish word for a beetle – ‘ciaróg’, another diminutive meaning ‘little black thing’.
Really enjoying the podcast!
Le dea-ghuí, (with best wishes)
Maitiú
Transcript
[Transcript provided by SpeechDocs Podcast Transcription]
HEDVIG: I have accidentally reinvented scissors and socks. [LAUGHTER]
DANIEL: That’s two separate things? They’re not like…?
HEDVIG: Yes, yes.
BEN: Yes. This is something terrifying.
DANIEL: Now, I have an idea. I’ll be right back.
[BECAUSE LANGUAGE THEME]
DANIEL: Hello and welcome to Because Language, a show about linguistics, the science of language. My name is Daniel Midgley. Let’s meet the team. Now, Ben Ainslie and Hedvig Skirgård, I want to ask you both a hard hitting and revealing question. And if that’s not okay, if it’s just too close, that’s fine. You don’t have to answer.
BEN: No, I dare you. I double dog dare you. Find my limits, Daniel. My safe word is “more.”
DANIEL: All right. Okay, here we go.
HEDVIG: That’s gross.
DANIEL: I know. Ben plays a dangerous game. I was eating some gumdrops earlier today.
HEDVIG: Uh-huh. I barely know what that is.
DANIEL: They’re these squishy, sugary things with a little bit of sugar on the outside, do you know? And they’re different colours.
HEDVIG: Okay. It’s a lot of candy. [LAUGHS]
DANIEL: I know.
HEDVIG: Anyway, you had some Haribo. Yeah.
DANIEL: Okay, if you will. What’s the best colour and what’s the worst colour? Hedvig first.
HEDVIG: Oh, I don’t like blackcurrant or grape flavors.
BEN: [ASTOUNDED REACTION] Wow!
HEDVIG: And if there’s any colour that correlates with that, so purples.
BEN: Wow, that’s fascinating.
DANIEL: Those are the best?
BEN: No, those are the ones she doesn’t like.
HEDVIG: Those are the worst. I will avoid those.
DANIEL: Those are the worst. Okay, good.
BEN: That is… what an interesting… Because I would have thought that is a safe, positive choice for nearly everyone.
HEDVIG: No, I hate it. The one flavour I really like is artificial pear.
BEN: I’ve never encountered artificial pear.
HEDVIG: It’s popular in some parts of Sweden. Yeah, yeah. Artificial candy, the artificial pear. That’s really nice. And you can taste the esters, it’s not real. You know it’s not real.
BEN: Yeah, yeah. fake banana is so clearly not banana. [LAUGHS]
HEDVIG: Yes. So, if anything is purple, I’ll avoid it. If anything is green, I’ll go for it.
DANIEL: Interesting. Okay, Ben, your answer?
BEN: So, I have… I’m caught in a really tricky spot because my favourite and my most disliked are the same colour. And it’s always… When you get a bag of lollies, whether that’s like snakes or gumdrops or whatever…
HEDVIG: That’s terrible.
BEN: Like Hedvig said, that gelatinous kind of Haribo style of candy or lolli, yellow, it can either be lemon, the greatest of all of the flavors as far as I’m concerned…
HEDVIG: Mm-hmm.
BEN: Or it can be an awful banana flavor. And it just immediately… The dizzying highs to the crushing lows when it ends up being one of the banana ones.
DANIEL: Okay.
HEDVIG: We can all agree that changing expectations is bad. Like, subverting expectations is bad.
DANIEL: No surprises, please.
BEN: But if I don’t want to play the terrible game of Russian roulette that is the yellow gummy lolli, I will go for anything… I really love fake cherry flavor. I adore it. I love it. American sour cherry, anything…
DANIEL: Interesting.
BEN: …and I’m gone.
HEDVIG: Oh, sour.
BEN: Yes. Yes, yes, yes, yes.
DANIEL: Okay. Thank you.
BEN: That dark purple sour cherry. Anyway, sorry. Sorry, Daniel, that was a very long one.
DANIEL: Thank you. The correct answer is orange is the best. Although yellow…
HEDVIG: Orange is good.
DANIEL: …lemon, very solid. Purple is the worst.
BEN: Okay, interesting.
DANIEL: All right.
HEDVIG: agreement.
DANIEL: Thank you both. I hope I didn’t go to some area you weren’t comfortable with, so, [LAUGHS]
BEN: More.
DANIEL: On this episode, we’re talking to Karen Yin, author of The Conscious Style Guide. Conscious style guide.
BEN: Ooh, Conscious in what sense?
DANIEL: So let me take you back to the 1990s. I was working at WordPerfect in Utah. You remember the word processor? I was doing tech support, and I became aware of these corporate language guides that were beginning to go around because people were becoming aware of the need to use language in a way that was more careful or more accurate or didn’t promote or perpetuate old hatreds or, as the people in Utah put it, they were upsetting the natural patriarchal straight order. And why did we all need to be so politically correct all of a sudden?
So, these guides had lists of things not to say and suggestions for other things to say. So, when I found out about The Conscious Style Guide by Karen Yin, I was watching out for that section of the book. When are we getting to that section of the book? Where’s the list? Never came. It’s not that kind of book. No, it doesn’t tell, it asks. It invites you to question yourself relentlessly and quietly and personally and say, “Do you mean to say this? Does this reflect your values? Is there a way of saying this that reflects more accurately what you really mean?” So, it’s not about censuring language terMs It’s about being aware of language choices and maybe becoming aware of some better ones. So, I had a chat with Karen.
HEDVIG: Very nice.
BEN: That sounds really interesting, and I’m really looking forward to digging into that.
DANIEL: You know as I was browsing our Discord server and looking at comments and posting my daily game squares, I was reminded again of how grateful I am for the community we have of our patrons and our listeners. They keep us going, they give us stories and words and ideas. And in a very real sense, as you’re going to see this episode, they power the show. If you’d like to, you can become a part of this community by becoming a patron, even a free one and that’s at patreon.com/becauselangpod. We’d love to have you.
BEN: Come and hang out with all the cool kids.
DANIEL: And support the show. Thank you.
BEN: Before we dig into how we can be consciously stylistic in our writing, Daniel, what’s going on in the world of linguistics?
DANIEL: This story was suggested by Nardog on our Discord, and it’s about a typewriter.
HEDVIG: Oh.
DANIEL: A typewriter has been discovered, but not just any typewriter. This is an old Chinese typewriter from the 1940s. [HEDVIG GASPS] You know this one, Hedvig?
HEDVIG: Have you seen these bad boys. No, but I’ve seen old Chinese typewriters. I’ve seen little documentary videos. They are so cool. Have you ever seen one of these, Ben?
BEN: So, the closest I have to knowing anything about this was I listened to a really fascinating episode of Radiolab that detailed Chinese keyboards and their development in the ‘80s and what I’m assuming the typewriters also struggle with, which is the fundamentally difficult problem to solve of, like, how do you take a language with that many characters and humans with this many fingers and [LAUGHS] make a device that can bridge that gap?
DANIEL: You’re right. Setting type in Chinese is a vexed affair because it’s a logographic language. The characters are not sounds. The characters are words or parts of words. And there’s a lot of them, thousands, just to be literate. There could have been tens of thousands before spelling reform in China.
HEDVIG: And there are parts of Chinese characters that do repeat across characters. So, you don’t have to have an entirely new little type thing for every word. There’s what’s called radicals and little segments, so you can build one up. And there are words that occur more often. You don’t… Even if there’s like 8,000 unique characters, it’s not an unsurmountable challenge, but it’s certainly a lot more complicated than having like 28 letters.
BEN: So, why is this particular typewriter of newsworthiness, as it were?
DANIEL: Well, it turns out that this is a long-lost prototype of an early attempt to get Chinese characters into type. So, this popped up on Facebook, of all places. It was on the group “What’s my typewriter worth?” The photo was posted by one Nelson Felix of New York. This typewriter was about the size of a regular typewriter, and it had Chinese characters on the keys. And it was a long-lost prototype of the typewriter designed by Lin Yutang, who was a linguist and a writer from Southern China. He got the patent for this device in 1946. He called it the MingKwai, which means “bright and clear” in Mandarin. Mr Felix ended up with it because he had a relative who’d worked at Mergenthaler Linotype, and they’re the ones who made the prototype of this typewriter.
BEN: That is fascinating.
HEDVIG: Very cool.
BEN: And I suspect it’s beyond our scope, both as humans with understanding capacity on this show to actually explain how this actually works. Because I listened to a whole podcast episode about this, and at the end of it, I was like, “What is happening?” Daniel, I’m super keen. Give it your best shot. How did these typewriters manage to… I have to assume, like Hedvig was saying, there’s some measure of a single character is not pressed at one time and that it’s built slowly with different components before moving on to another one. Is that sort of what we’re looking at here?
HEDVIG: So, the chords.
DANIEL: So, here’s the NPR article by Emily Fung, and we’ll have a copy of this up in the show notes for this episode. I’m quoting here, “The typist can search for certain combinations of shapes by pressing down on the ergonomic keyboard.” Now, I don’t know if that’s chords or if it’s one at a time, but notice how he’s classifying things by shape, by how they look. So, I guess this is the radicals. Okay, “Then, a small screen above the keyboard, Lin called it his magic eye…”
HEDVIG: Oh.
DANIEL: “…offers the typist up to eight possible characters that might match. In this way, the typewriter boasts the ability to retrieve up to 90,000 characters.”
BEN: Okay. So, they were doing what more contemporary electronic computers did in the 1980s, which is basically a very sort of, for the time, an incredibly advanced autocorrect-suggest window. You’ve pressed these things, here are the things that we think you might have wanted.
HEDVIG: Mm-hmm.
DANIEL: I cannot imagine how this machine in the 1940s would have a screen that would show that.
BEN: That’s absolutely bonkers.
HEDVIG: There is a photograph in the NPR article, and I’m looking at it, and I am also trying to imagine what… I’m wondering if it’s little spools of paper or something, tape, that select and scroll back and forth because there isn’t LED. There isn’t…
DANIEL: Nope. Not at this point. [LAUGHTER]
HEDVIG: There isn’t liquid crystal displays.
BEN: Wow. So, for the listener’s benefit, the best I can offer you here to describe what this looks like is a conventional typewriter that has a protuberance coming out of where you would normally expect to see the sheet and the ream and where the typing happens. And there is a tiny Perspex window that, if anything, looks like it belongs on the side of a boiler or perhaps the side of a submarine. It looks a tiny little porthole that you would only get a very small vision sort of like window into. And then, I’m wondering if they’re doing projection here. They’re back projecting onto this little screen from the inside in some way.
DANIEL: Well, you’re not the only one who’s wondering about how this works, because the machine now lives at Stanford University, and there are people who are sort of looking very carefully into it and how it works.
BEN: Backwards engineer it.
DANIEL: Yeah, exactly.
HEDVIG: That is so cool.
BEN: That is really, really cool.
DANIEL: So, it’s a fascinating look at how people have tried to get written characters into type, but it’s also a reminder of China’s cultural history since all of this was happening during the Maoist takeover. So, check out the articles. It’s on becauselanguage.com. We’ll have some show notes up for you so you can dive down that rabbit hole. And big thanks to Nardog for that story.
HEDVIG: That’s really fun.
BEN: I will say, as a final note, I politely disagree with the characterisation that this is approximately the size of a normal typewriter. Wrong! It large. It big. It’s still a desktop machine. Don’t get me wrong. It’s not like the size of a cabinet or anything.
HEDVIG: Have you seen 1940s typewriters? Like, they are chunky bois.
BEN: I have a 1940s typewriter in my house, and it is a lot smaller than that.
HEDVIG: Okay. Okay, yeah. No, it’s very impressive. Very cool.
BEN: It’s really cool. It’s really cool.
HEDVIG: I also want to note that one of the people at Stanford University who was very happy, he said he studied Chinese computing for two decades and thought this was lost and no one could find it and someone just found it on Facebook.
BEN: I’m trying to think about some sort of special interest area of mine, and someone finds some relic that everyone thought was lost to antiquity and it appears, you would be so pumped, you’d be like “Oh, my god.”
DANIEL: 116 pages. [LAUGHTER] Anyway, let’s go on to our next story. I noticed this one. James noticed it as well and popped it up on our Discord. We keep coming back to ChatGPT and large language models because they really are big news and the impact of these things is socially and technologically explosive. But now, here’s a question. Do you think that large language models will have an impact on the way that we speak? The way that we say words, not just the way that we write?
BEN: Ooh, I thought you were going with a different question there. I thought you were going to say “think” at the end of that sentence.
DANIEL: [LAUGHS] Oh, no. Not that topic again. Well, I mean, it is kind of the same thing, isn’t it?
BEN: I guess.
HEDVIG: [LAUGHS]
DANIEL: So, if they use certain words a lot, will we also…? What? [LAUGHS]
HEDVIG: No, it was just the way we think or the way we speak. Well, it’s kind of the same thing. Let’s move on to the next one. It was just very funny.
DANIEL: That’s not Whorfian by the way. If we think a certain way, then having language follow is totes normal. We would expect that. It’s the other way that we wouldn’t expect.
HEDVIG: Yeah, okay.
DANIEL: Okay.
HEDVIG: Anyway, in what ways could large language models influence people? Well, I’ve been thinking a lot late recently about how sort of sycophantic HR fluff speak large language models are. And sometimes when I use one mainly for programming probleMs yesterday, for example, I spent a number of hours trying to solve cross-platform package installations in different programming languages…
BEN: Oof. Yeah, okay.
DANIEL: Nasty.
HEDVIG: …it is actually a little bit helpful to talk to a large language model about that.
BEN: There we go, Daniel. See, her speech has suffered. The proof is in the pudding.
DANIEL: [LAUGHS]
HEDVIG: And because it suggests things that I tried them out and I’m like, “Nope, that didn’t work. Oh, that did work. Hello.” And it’s nice because you can get feedback about whether it’s actually doing something correct or not, you don’t just believe it. So, that’s one of the ways I do use it.
What I was going to say was I have to tell it to cut down on this kind of fluffy speak. But I’ve been made to understand that for a lot of people who work in a corporate world where they manage people and there’s a lot of interpersonal potential for friction, people use large language models to try and smooth out the rough edges of the way they communicate, to try and be more polite and [BEN LAUGHS] sound less aggressive, etc. So, is that the way people are changing the way they speak? Are they being more American HR in their language use?
DANIEL: Well, there’s the thing. I am really, really skeptical of this kind of notion. I mean, once again, I was sitting with a group of parents and somebody mentioned the Bluey effect, that exposure to Australian English is making American and British kids talk Aussie.
HEDVIG: You wish. We wish. [BEN LAUGHS]
DANIEL: I know, we all wish. But do you find this notion plausible that the way ChatGPT outputs language is capable of, I don’t know, working its way into our patterns of spoken language? How would we know?
HEDVIG: Spoken. Oh, shit, spoken.
DANIEL: Actual spoken language. That’s what we’re talking about.
HEDVIG: That’s a…
BEN: Yeah, that’s a long…
DANIEL: Stretch? [LAUGHS]
BEN: …bow to draw, I think. So, this reminded me of a delightful… that’s what I was just typing to try and find, a delightful… I think it was a Reddit post maybe once upon a time, of either a millennial or a gen Z person who had a corporate job and worked with primarily quite boomer dudes. And she was really struggling with how to engage in corporate lingo. And so, the post is basically a series of things of like, “Okay, I’ve just basically enlisted their help and I’ve said, ‘Guys, I need your help translating this. How do I tell this person that I’ve fucking answered their question like three times already,’ and they’re like, ‘Oh yeah, we got you.’” Dear so and so, as per the previous several emails. [LAUGHS]
DANIEL: That’s a good way of doing it. Yeah.
BEN: But it reminds me of this thing of, like, do we think that person in that post then started speaking that way in their everyday life? I don’t think so.
DANIEL: It’s weird, isn’t it?
BEN: We learn how to code switch if and when the situation demands it.
DANIEL: Like, when I’m talking to my computer to get automatic text from my voice, I don’t then talk like that to other people. I switch.
BEN: I don’t say my punctuation out loud because I’m a tragic old guy who would rather speak his texts into his phone. I don’t then walk around being like, “Excuse me, Ellis comma can you please tell me that you’ve cleaned your room full stop.” [LAUGHS]
DANIEL: Question mark. Yeah.
HEDVIG: This is where we get that divide again where you talk to devices like that and I don’t ever, because it doesn’t work for me. It doesn’t type the things. It’s not a feasible way for me. But let’s step back a bit. Before large language models, there were such things as written genres of writing things, and there were such things as, for example, style guides as we’ll be getting into later. And do we think that those affected the way people spoke? Because I’m not sure they did. They may be affected the way people spoke in some situations in the workplace, but I’m not so sure about everyday spoken conversation. And that was actually intentionally guiding and trying to educate people about how to speak, whereas just large language model effects should be more insidious and subtle.
BEN: Look, here’s how… If people want to see, if some researcher wants to find definitive evidence of large language models changing the way that people speak, what they’ve got to do is figure out how to get large language models to change the way that young, queer, and African American people speak and then everyone else will appropriate that language because it’s cool and edgy and that’s how we’ll get that spread throughout language because that is what we see. All of the changes that I’m detecting are all just people aping AAVE and like queer slang, not even knowing that they’re doing it just because it’s what the cool kids do. And that’s where the change tends to come from.
DANIEL: Interesting attack, Vector Ben. Well, let’s take it step by step. One way to show that this is or is not happening is to first of all assume that ChatGPT and other large language models are using certain words more often than is statistically likely.
HEDVIG: Right. So, this is the delve problem.
BEN: Oh, right. What was the famous giveaway word? Yeah, there we go delve.
DANIEL: This is the delve thing. Now, our friend, Elizabeth, sent me an email, hello@becauselanguage.com. She refers to a paper and she says this paper “delves” into “crucial” changes that have the “potential” to “significantly” affect language. So, all the words that I said are suspected to be ChatGPT words. The paper that we’re starting off with here, link in the show notes, becauselanguage.com, it’s by Dr Dmitry Kobak of the University of Tübingen and a team. This is published in Science Advances. The paper is called “Delving into LLM-assisted writing in biomedical publications through excess vocabulary,” and what they did was they examined the rate of certain words in PubMed abstracts before the introduction of ChatGPT and afterwards to see if certain words just went ticked upward a lot.
According to this paper, there were some that ticked upward sharply: DELVES, UNDERSCORES, PIVOTAL, NOTABLY, also way more verbs than can be explained satisfactorily. Interesting. Okay, so there’s that paper. Thanks, Elizabeth, for that.
BEN: Now, but that’s not saying that they believe therefore wait… Because that could be explained both by people’s language changing, but also just by people using ChatGPT to write a bunch of shit for medical journals.
DANIEL: Yes, exactly.
BEN: [LAUGHS]
DANIEL: At this point, we’re still firmly in the camp of written language.
BEN: Yeah. Gotcha.
DANIEL: But now, James has linked us to a paper. There’s an article in Scientific American, but the paper is on arXiv.org — hasn’t been peer reviewed, but that’s okay, it’s still a good way of getting early work out there and getting comments on it. It’s by Dr Hiromu Yakura and a team from the Max Planck Institute for Human Development.
BEN: Oh, Max Planck.
DANIEL: Yeah, I know there’s a lot of it going around. So, first they confirmed that ChatGPT is in fact favoring certain words. Their list was DELVE, UNDERSCORE, COMPREHEND, BOLSTER and so on.
BEN: BOLSTER.
HEDVIG: BOLSTER.
DANIEL: BOLSTER. I love BOLSTER. I say DELVE all the time. I must be a large language model.
HEDVIG: Can I ask about these delve words? So, we think that large language models are picking up on them, not necessarily because they were common in the input, but because when they output them, people like them. Is that true?
DANIEL: My theory is that people who are paid to do the reinforcement learning, that is to say, they look at a bunch of outputs and they say, “Oh, I like that one the best.” And those people maybe come from communities where words like DELVE and so on are more common. And so then, ChatGPT just picks those up because that’s what the humans who are doing the reinforcement training like.
Now, this team looked at about three quarters of a million hours of human discourse from YouTube academic talks and conversational podcast episodes across multiple disciplines. Now, I saw that and I thought, “Wait a minute. Were the talks and the podcasts written using large language models?” Because that’s going to…
BEN: Yes, that’s scripted, right? If you were doing scripted dialogue, then you’ve got just as much likelihood that it’ll have that.
HEDVIG: He said conversational podcast. We’re a conversational podcast. Do you script what you’re saying, Ben?
BEN: Look, Daniel said conversational podcast.
DANIEL: I do. [LAUGHS] I’m absolutely reading my script.
BEN: I don’t… Yeah, so to Daniel’s point, yes, I think people in conversational podcasts definitely write scripts.
HEDVIG: Oh, god.
DANIEL: Well, here’s what they say. They thought of it. They saw this coming [LAUGHTER] because they’re very smart.
BEN: Yes.
DANIEL: And they said, “For podcasts, only those featuring dialogue were used to examine the influence on spontaneous communication.” They looked at spontaneous dialogue “ensuring that the effect did not primarily stem from reading ChatGPT-generated scripts.”
BEN: Fair enough. Okay.
DANIEL: So, they’re working on controlling that. I still think…
BEN: So, they’re basically… In our show, they would have taken only my dialogue and Hedvig’s dialogue, and they would have clocked you for what you are and they would have been like, “No, not a chance.”
DANIEL: Yes, Ben, that’s correct. Now, let’s delve.
HEDVIG: Ben, what would this show be without the skeleton that is Daniel?
BEN: I’m not saying it’s bad. [LAUGHTER] I am a lamprey attached to the side that is the whale of Daniel Midgley.
HEDVIG: Exactly.
BEN: I make no bones about that.
DANIEL: [LAUGHS]
HEDVIG: Mm-hmm.
DANIEL: Well, let’s see. When they crunch the numbers, they found that… This is the quote. “The discrepancy between, for example, DELVE and the control was maximised when we specified November 30th 2022 as the treatment date.” That’s the introduction of ChatGPT. Before that we see one thing, after that, we see another thing. And if we move that date around, everything just gets worse. The pattern doesn’t show up as much. And they find that the usage of, for example, DELVE in academic YouTube talks coincided with the release of ChatGPT in science and technology podcasts, business podcasts, and education podcasts, but not religion and spirituality podcasts or sports podcasts.
BEN: [LAUGHS]
HEDVIG: That makes sense to me.
BEN: I don’t know why that makes me laugh, but it’s funny. The jocks aren’t using ChatGPT.
DANIEL: [LAUGHS]
HEDVIG: Like, the political podcasts would… Maybe they didn’t have that category.
DANIEL: That wasn’t a category.
HEDVIG: Mm-hmm. Okay.
BEN: Okay, so this could be a thing. We could be seeing some cross pollination from the fact that we, all of us, wittingly or unwittingly, consuming a lot more content generated by large language models. And therefore, we are exposed to words like DELVE at a much higher rate than we might have otherwise been and so they’re starting to just bleed in a little bit.
DANIEL: Input is input, I guess.
BEN: BOLSTER. I mean, you said BOLSTER before, and my brain immediately got tickled. I know that I will be saying BOLSTER at some point in the not-too-distant future to sound like a clever clog.
DANIEL: I still want to know what the mechanism is. I mean, you got to have the mechanism. You can’t just say, “Oh, well, we heard it. So, we did it.” And this team says…
BEN: Can’t we say that? Why can’t we say that?
HEDVIG: Can’t we say that? Sorry, reinforcement learning is the basics of… I mean, can’t we say that? [LAUGHS]
DANIEL: Me personally, I don’t use large language models enough for it to form so much of my input.
BEN: But hang on. Whoa, whooa, whoa, what?
HEDVIG: No. You don’t. But you just read an article from someone online.
DANIEL: Yes, it is.
HEDVIG: I’m not saying this NPR journalist necessarily use a large language model to write their article.
BEN: You did it. You piece of crap.
HEDVIG: Maybe they did. Maybe they drafted part of it. There’re texts that are appearing in front of you, Daniel, that you haven’t been the originator of, and that may have been originating with large language models.
BEN: I think you might be consuming a lot more large language model slop than you might necessarily be thinking you are.
DANIEL: You could be right. I’m also aware that there’s a big moral panic around this area about internet speak somehow infiltrating our… Oh, wait, that is happening actually. [LAUGHS]
BEN: Let’s be really, really clear. Let’s say this is 100% true in exactly the way that it’s been presented in its most extreme form, which is that we’re saying delve and bolster and stuff more than we did before. The outcome of that or like the “so what?” is like not fucking much in my opinion, as in which is not at all to say that the research that has been done here is not interesting. I think it’s really interesting. But what I’m… The moral panic side of things can go fucking jump in a lake, the lake behind me in fact.
DANIEL: Exactly. [LAUGHS]
BEN: If we are saying DELVE and BOLSTER more, that’s not a concern for me or should be for anyone. But that doesn’t mean that there aren’t things that we potentially need to be thinking really thoughtfully and critically about in terms of AI and large language models. And I think we’ll go on to keep talking about that stuff, but it is okay to say, “Mm, and?” [LAUGHS]
HEDVIG: I’m going to go on the record and say that I’m going to say something and if it’s not true, then at least it’s a nice falsifiable hypothesis.
BEN: [LAUGHS] I like that. That’s a great one. That’s a great one…
DANIEL: Okay. Yeah. This is what I want. This is what I want.
HEDVIG: I think that writing like writing, printing press, mass literacy, has a bigger effect on spoken language than large language models. Large language models are sort of just another iteration of that kind of effect on spoken language.
BEN: I guess the only question then becomes, Hedvig, and more so as we go into the future, how distinguishable are those two things as discrete entities?
HEDVIG: Well, for hundreds of years, we had mass literacy, but not large language models.
BEN: Yeah, yeah, yeah. Sure. And I agree with you. For the entire corpus that came before now, we can say that. But moving forward into the future, I think it’s going to be much, much harder to differentiate that as a concept.
HEDVIG: Yeah, absolutely.
DANIEL: Mm. It just reminded me of that thing where they would play records backwards, like rock records backwards…
BEN: Only to find Satanic messages. [LAUGHS]
HEDVIG: The Satanic Panic.
DANIEL: …for satanic messages and you could just find them everywhere and they’re like, “Sweet Jesus, this stuff is everywhere!” And this team is using a method to find patterns in LLM text and it’s like: Oh, my gosh, we’re finding the same patterns in podcasts. Maybe it’s not really a thing both times. It would have to be right twice.
HEDVIG: And also, the things that people notice for language change are often, like we talked before about young alpha people saying, “What the sigma?” Or the influence of queer or other language on like it’s giving and things that.
DANIEL: Mm-hmm.
BEN: Yes.
HEDVIG: That is still like I’m going to guess less than 2% of all spoken language that someone produces. Most sentences are not that. Most things you’re noticing are actually kind of low frequency and will have a little bump and then go away. What’s interesting to me as a linguist is like how the word “to want” becomes the word “to future,” like will. And that takes like years and no one’s paying attention to it. That’s the changes I’m interested in.
BEN: I remember being a teenager, and I remember a bunch of the usually quite a lot older teachers in the mix in my life having a huge problem with my generation’s use of the word LIKE. That was like a massive thing. Do you remember this, Hedvig? Was this true for you as well?
HEDVIG: Oh, yeah, yeah.
BEN: And I haven’t come across that in a really long time. I haven’t come across a person being, [IMITATES AN OLDER PERSON] “People are saying LIKE too much. Blah, blah, blah.” So, there you go. I just did you then. So perhaps, this is just a flash in the pan, a mountain out of a molehill. People are going to say some stuff for a while. And either, it’s not a thing and no one cares and we move on, or it is a thing and no one cares. [LAUGHTER]
DANIEL: Well, let’s not underestimate the potential for humans to talk like each other. It’s something that we really love to do. And algorithmically, this can be a thing. We’re going to find out more of this when we talk pretty soon to Adam Aleksic, who is Etymology Nerd.
BEN: Nerd.
DANIEL: Thanks, James, for this story.
BEN: Cheers, James.
DANIEL: Now, it’s time for Related or Not. Let’s hear our theme. This one’s from Tiago, because I thought we needed some throat singing.
BEN: Okay, here we go.
DANIEL: Here we go.
[LATEST RELATED OR NOT THEME]
BEN: So deep, so resonant.
HEDVIG: Wow. Every time I hear throat singing, it makes me think of how black metal, heavy metal music in general, growling has so much in common with throat singing and why it therefore makes sense that there’s Mongolian throat singing hard rock.
DANIEL: We like doing stuff with our meat pipes. [LAUGHTER] Thanks, Tiago. That’s awesome. The first one comes from John and some help from Vehemently. I’m going to give you three pairs.
BEN: Okay.
HEDVIG: Okay, three pairs.
DANIEL: Only one of them’s related, the other two aren’t.
BEN: I’m going to pick the related one. Okay.
DANIEL: So, you’ve just got to pick the pair that’s related. Here they are.
HEDVIG: Pen and paper.
DANIEL: Curve and curl, like a curve in a road and a curl of hair. That’s one.
HEDVIG: Okay.
DANIEL: Number two. Finger and five. Finger and the number five.
HEDVIG: Mm-hmm.
DANIEL: And the third one from Vehemently, slob and shlub. I’m a bit of a schlub sometimes, and I’m a bit of a slob sometimes. Which one of those is related? Here they are. Curve and curl, finger and five, slob and schlub. You’ve got to pick the pair that are related.
HEDVIG: Can I just ask, I’ve never heard schlub before, but I’m going to guess Yiddish, usually is in English when there’s a S-C-H.
DANIEL: Good guess.
HEDVIG: I don’t get any…
BEN: Finger and five is the related pair. That is my answer.
DANIEL: Okay.
BEN: I’m answering because it seems the most not, and so therefore it is the most.
DANIEL: [LAUGHS] Numbers are weird.
BEN: I think it’s a red herring.
DANIEL: Mm-hmm.
BEN: I also think about the fact that there’s closeness in our other words in English, right?
DANIEL: Yeah.
BEN: So, like one, two, three has some interesting pairings, but five in a lot of those European languages is a bit different.
DANIEL: Okay.
HEDVIG: Mm-hmm.
DANIEL: Hedvig, what do you think?
HEDVIG: I don’t think finger and five are related. [LAUGHTER] I’m pretty sure that the root for five is something like *penkwe-. There’s definitely a P in there at the start. And I think finger, it’s got to be a Germanic invention because, as I recall, it’s not really in Romance. It’s like Manus or something. That doesn’t mean that they couldn’t be related. Germanic could have invented the word. In that case, they would have to invent the word finger from five, I guess.
BEN: Which you know, it’s not impossible.
DANIEL: That would count as related. Remember also, there’s a P to F thing in Germanic languages.
BEN: Oooh. Okay, okay.
HEDVIG: Oh, yeah, yeah. That’s how you get from *penkwe- to five. But you don’t get from *penkwe- to finger.
DANIEL: *penkwe- to finger.
HEDVIG: I don’t know. I just doesn’t… You’re right it could go that direction. You’re not wrong, but it just doesn’t smell right to me.
DANIEL: Okay.
HEDVIG: Slob and schlub, I’m thinking, look so similar and mean such similar things that it’s like…
DANIEL: Yeah, but the Yiddish thing?
HEDVIG: Right. So, I think that’s a red herring because they just like look too similar.
BEN: Too close. Yeah.
HEDVIG: They’re too close, but maybe…
DANIEL: And the Yiddish thing derails it.
HEDVIG: No.
DANIEL: No?
HEDVIG: It doesn’t. Why wouldn’t it?
DANIEL: While If slob came from Old English and then schlub came from Yiddish fairly recently. Yeah, it could still be related.
HEDVIG: Yiddish is Germanic.
DANIEL: Yiddish is Germanic. Yeah, okay.
HEDVIG: They could very well be related, there’s no problem there.
DANIEL: Yeah. Could be a doublet, yeah.
HEDVIG: Also, what Swedish words that are related? I’m doing my cheating.
DANIEL: [LAUGHS] By knowing stuff. Check her out. That’s not right.
BEN: How dare you?
HEDVIG: [LAUGHS] I need to stop doing this because Ben maybe not is paying attention to this, but what happens is I think out loud about Swedish words or other languages, and then I get a little sound clips sent to myself from our lovely transcribers who are, “What did you say? We need to transcribe it.” [LAUGHTER]
DANIEL: It’s actually me, but yeah.
BEN: Oh, there we go.
HEDVIG: Okay. And then, we have curve and curl. I think they’re related.
BEN: Oh, okay.
DANIEL: Okay.
BEN: What did you go with, Daniel?
DANIEL: I’m abstaining. All right, well, first of all, none of us think slob and shlub are related, and they are not related.
BEN: Okay, okay.
DANIEL: Slob, in fact, is Old English and schlub is Yiddish, but not the same source word. The one that was related was finger and five, yep.
HEDVIG: What the hell?
DANIEL: They both go back to *penkwe-. Good job on knowing the proto-Indo European.
BEN: Yes. You had it. You had it.
DANIEL: [LAUGHS] Finger and *penkwe-. Yep. Numbers are interesting because they start out kind of different from each other, and then they can grow together because of perceived similarity or when they entered the language. John points out that six was originally weks, but then it got an S at the beginning because that made it more seven.
BEN: [LAUGHS] I mean, it makes sense, right?
HEDVIG: But wait…
BEN: How many of a thing do you have? I have a hand…
DANIEL: Five.
BEN: …of a thing, right? I have a hand and that means five.
HEDVIG: Yeah, but you’d expect it to go the other way. And also, it’s finger, not hand. So, okay, how does it we get from six to finger? Because six, the root is five, correct?
BEN: I see what you’re trying to do.
HEDVIG: No, no, no. I want to know.
BEN: You’re being a high school debater. You’re trying to take this away from me.
HEDVIG: No, I don’t. I just want to know.
BEN: Adjudicator, point of information.
DANIEL: Etymonline for instance, and OED says the same thing. Perhaps ultimately from Proto-Indo-European, *penkwe-, five. It goes into the Norse languages, Proto-Germanic *fingraz, from *penkwe-, and you can really see the similarity there. And then it gets into Old English, finger. It’s a really super old word. Five, the same kind of thing.
HEDVIG: That’s wild. Okay.
DANIEL: Mm-hmm. it is. It’s very strange. And then, curve and curl. No, totally different words. Curve is from ultimately proto-Indo-European, *sker-, to turn. It’s got the same root as circle and search and even crown. But curl was a totally different word. It just referred to a lock of hair. Completely different words. Thank you to John and thank you to Vehemently. Ben, this one’s yours.
BEN: Yeah, I had one this time.
DANIEL: Go on, then.
HEDVIG: Okay.
BEN: I can’t remember. [LAUGHTER]
DANIEL: Oh, my.
HEDVIG: Amazing.
BEN: I remember texting you this a couple of weeks ago, and that might as well be a century ago from the way my brain went…
DANIEL: It’s a good thing I’m to steer this ship aright. The words were rescue…
HEDVIG: Yep.
DANIEL: …and recuse.
BEN: Ah, that’s right. Rescue and recuse.
DANIEL: Which both share a sense of sort of getting away from something.
BEN: Yeah. Getting out of something or getting… Like, sort of…
DANIEL: Escaping.
BEN: I am putting my flag in the sand. I think I put my flag in the sand with you at the time…
DANIEL: You did.
BEN: …Daniel. And I said yes, didn’t I? I said a very firm they are related.
DANIEL: I said, no, and I’ll wager you an ice cream.
BEN: That’s right. And I took that wager and I say yes, they are related.
DANIEL: Okay. And then, I looked it up. Hedvig, what do you think?
HEDVIG: First of all, I’m very bad at spelling. This is where English gets hard. Are they the same letters tossed around?
DANIEL: They are anagrams which means, obviously, they’re so related every time.
HEDVIG: Then, I have spelled them correctly because I was like that’s suspicious.
DANIEL: Ask and ska are related.
HEDVIG: We’ve had recuse on this segment before for something else.
DANIEL: No, we haven’t. We haven’t. We haven’t done recuse. We’ve done something like it. I think we’ve done escape, but…
HEDVIG: Okay.
DANIEL: Anyway.
HEDVIG: Oh, we’ve done avoid, maybe.
DANIEL: Oh, that’s right. Avoid and evade.
HEDVIG: Yes.
DANIEL: Yes.
HEDVIG: Rescue. Okay, recuse. So, rescue to me is an obvious re, as in like again, or something that.
DANIEL: Oh, I thought the recuse was the obvious re, but okay.
BEN: [LAUGHS] Oh, yeah, I’ve brought a little noodle scratcher.
HEDVIG: Cue to me is more…
DANIEL: And I wouldn’t be surprised to find that the re were the same re. I’m thinking about the other bit.
HEDVIG: And as per Related or Not rules before, Daniel has judged that means they’re not related.
BEN: Yeah, yeah. Totally. Prefixes and suffixes don’t.
HEDVIG: Prefixes don’t.
BEN: Well, I mean, otherwise, ing would make everything related. [LAUGHTER]
HEDVIG: Yes, so just trying the non ING form.
Hedvig What is that cues, by the way, in recuse. I could only think of excuse. I couldn’t think of any other places that root comes up.
HEDVIG: Excuse. What’s the similarity of recuse and excuse? They’re both… You’re moving away. Excuse me from this gathering, I’m going to recuse myself. They’re actually kind of similar.
BEN: Yeah. Basically, it’s the same. Yes, synonymous.
HEDVIG: Rescue. Okay, so scue, some sort of going movement. Okay, maybe they are related. Yeah.
BEN: Ooh. So, there’s two ice creams up for grabs here.
DANIEL: [LAUGHS] Man, I’m going to have a lot of ice cream.
BEN: Oh, damn.
DANIEL: I’m going to be paying a lot for ice cream. Okay. I said that I think the re in recuse is something like to withdraw. I couldn’t think of what the cuse was. I thought, “Mm-hmm. Is that one of those intensifier re’s?” Like, you’re really getting away from something. That’s what I thought.
BEN: Okay.
HEDVIG: Mm-hmm.
DANIEL: Okay, answer, I win, they’re not related. Whoo-hoo.
BEN: Oh, dammit.
DANIEL: Recuse comes from recusare in classical Latin, to make an objection. The re was an intensifier, and the cuse is Latin, causa, meaning cause.
BEN: Okay.
HEDVIG: Cause.
DANIEL: You go away from a cause or you reject a cause. Rescue is a little bit different. This re prefix is also an intensifier. And the scue part, when you rescue someone is from Latin, excutere, to excus, which isn’t a word I knew. The idea is you are liberating someone from legal custody by force.
BEN: Okay.
DANIEL: So, they’re different words entirely.
HEDVIG: Oh, so you’re saving someone from prison.
DANIEL: Yes. You’re busting them out of here.
BEN: There you go.
DANIEL: Let’s go rescue this person.
BEN: Oh, interesting.
HEDVIG: That’s really funny because now if you’re lost at sea and you get rescued… Yeah.
DANIEL: Good old semantic shift. I love it.
HEDVIG: That’s fun.
DANIEL: Our last one is from Finn the Shark, who says, greetings from the Highlands of Scotland. SpeakPipe us you, guys. Come on.
BEN: Yeah, seriously. If ever there was a good reason for SpeakPipe, it is for a Highlands accent. Come on.
DANIEL: “I’m keen on Related or Not. And I’m keen to pose this dilemma. Maybe you’ve done it before. But KEEN, enthusiastic, and KEEN, cry out, for example, in pain.”
BEN: Oh, yes. The keening wail of a…
DANIEL: The keening wail.
BEN: …dying swan or whatever. Interesting.
DANIEL: I’ll go first.
BEN: Okay.
DANIEL: I would say related. I was telling my mom about synesthesia because for some reason she hadn’t heard of it. And that’s where you perceive senses in terms of other senses, seeing the colour three as spicy. Anyway, we’re all a little bit synesthetic. We all talk about cool music or hot jazz or sharp words or a bright student. That’s not weird. So, I think that if KEEN meant something like sharp, and then if somebody made a sound that sounded sharp, they would keen. I’m going to say related, even though it’s only two consonants and a vowel.
HEDVIG: And you’ve got KEEN, sharp.
BEN: KEEN means sharp. Just in case, there is actually a third sense of the word keen, which means…
DANIEL: I have a keen eye. Mm-hmm.
BEN: Or you literally can describe a blade as a keen blade.
DANIEL: Oh, yeah. Or I’m keen to get started. I’m sharp somehow.
HEDVIG: I only knew the keen to get started.
BEN: Like the anticipatory.
DANIEL: Mm-hmm.
HEDVIG: And keen eyes vaguely, but keen knife, I did not know.
BEN: Yeah. So, KEEN, I think its original form would have been sharp. Literally, like you would have described swords and knives that way.
HEDVIG: Sounds possible.
DANIEL: So, I’m saying related.
BEN: I’m going to go. Just because I want to beat Daniel, and I can’t beat him if we both have the same answer. I’m going to say not related, and I’m going to… I’m wondering if KEEN has come to us from Gaelic in some way as a completely… I’m only saying this because they’re quite short, and you could only really get away with this being a double-up with something like really, really short like this. I reckon it’s a complete accidental double-up. And KEEN comes to us from Scots Gaelic or something like that and it’s completely unrelated. And it just so happens that it actually semantically is a little bit similar as well, because a keening call is kind of a sharp call as well.
DANIEL: Okay. Hedvig, what about you?
HEDVIG: So, I also was only mildly aware of KEEN as in the verb to make a sound. So, I’m really clutching at straws here because I’m learning new senses of words [LAUGHTER] and I’m being asked to relate them. But I will note that KEEN, the first one is an adjective or maybe an adverb, sometimes, maybe. And the second one is definitely a verb. And I think that’s sus.
BEN: [LAUGHS]
DANIEL: Oh, interesting! Okay.
HEDVIG: It’s not impossible, but it’s a little bit hard to jump word classes…
DANIEL: In English??
HEDVIG: …especially…
DANIEL: We love transplanting words from one class to another.
HEDVIG: I know, but just like hist…
BEN: Just let her go, Daniel. [LAUGHTER]
HEDVIG: I’m not saying it’s impossible. Definitely not impossible. I’m just thinking it makes it slightly less improbable. I’m going to say it’s no. Also, it’s suspiciously short.
DANIEL: Yes, yes.
HEDVIG: A lot of things become the same when they become short. So, I’m going to say not related.
DANIEL: I said related. You both said not related. You both are correct. It is not related.
BEN: Come on now. Am I double correct? Am I correct in the correct way?
DANIEL: You are double correct in the correct way.
BEN: Yes.
DANIEL: The verb to keen is a borrowing from Irish. The word is keen, stem of caoinim. I wail. It’s just a borrowing.
BEN: Yes.
DANIEL: The adjective keen, meaning eager, comes from an Old English cene, bold, brave, or fearless. Interestingly, the OED, Oxford English Dictionary says it doesn’t have any cognates outside of Germanic, so no Romance reflexes for that one. And English is the only language that developed the sense of sharp.
BEN: So, just to be clear, it meant eager first, and then we then applied it to things like sharp.
DANIEL: Different again. It first meant bold or brave or fearless.
BEN: Yeah. Okay. For me, that’s kind of a semantically same thing. Yeah.
DANIEL: And then to eager. Yeah. Yeah, they’re similar.
BEN: And then, we made it into sharp. I would have thought it was the other way around, for sure.
DANIEL: I would have thought so too, definitely. So, Finn the shark has taken me in. Can’t trust sharks. Tell you what…
BEN: Oooh. I feel hot, it’s so good.
DANIEL: Thanks, everybody. And thanks, Tiago, for that theme song. And if you have words that you’d like us to try out, [LAUGHS] be kind to us. The audience messes with us, they do, they do. Send it to hello@becauselanguage.com and that goes for jingles as well.
HEDVIG: And if you are a Patreon supporter and you hang out in our Discord channel called Related or Not, you’ll also notice that Ste and I have made two submissions that Daniel has ignored.
BEN: Oh! You gotta text him, man!
HEDVIG: So, if you want extra challenges that don’t meet Daniel’s standards…
BEN: Come and have the slop Daniel doesn’t want to do!
HEDVIG: Yes!
DANIEL: [LAUGHS] Yeah, I’ll take that into account.
HEDVIG: They are literally submissions for the segment.
BEN: I mean, not to brag or anything, Hedvig, but that was the first one I ever sent to him, and it just got instantly straight away, weird.
HEDVIG: Okay.
[MUSIC]
[INTERVIEW BEGINS]
DANIEL: I’m here with Karen Yin, author of The Conscious Style Guide. Hey, Karen, thanks for coming on the show with me.
KAREN: Hi. It’s not every day that I get to hang out with a linguist.
DANIEL: Oh, wow. Okay. Aren’t you a linguist? Kinda.
KAREN: Oh, thank you.
DANIEL: I’m very permissive about who gets to be a linguist.
KAREN: Well, I would take that title with gratitude.
DANIEL: All right. I mentioned to Ben and Hedvig at the top of the show that The Conscious Style Guide was not the book that I was expecting. I mean, I’ve seen guides to inclusive language that have long lists of what language to use and what to avoid. And The Conscious Style Guide doesn’t tell you what not to say. It seems to invite you to question yourself, and it asks what do you want to say. So, could you tell me a little bit about the circumstances in which you decided to write this book? What was going on at the time?
KAREN: So, I wrote The Conscious Style Guide in response to this pattern that I had been observing where people were becoming fearful of the language they used, and then some people were also starting to criticize people for the language they used. So, I wanted to, from my perspective as a writer and an editor, kind of put people at ease. With this book, I want to remind people that conscious language is actually not the same as inclusive language because when you talk about inclusive language, a lot of people think about exact terminology, like specific words, like, “Do this, don’t do that.” And as a writer and editor, I know that’s not true because words have no meaning outside of their context. It’s the context that gives them meaning. And the example I sometimes use is, if I draw a line, you don’t know if that’s the number one or a dash or slash until it’s in some context.
So, I wanted to, with this book, separate people’s thinking about inclusive language and conscious language. I wanted to expand the idea of conscious language as something that is contextual and not a list of words to avoid. The offensive example that I like to use is the word “hoe.” Sometimes if you use the word “hoe” or “bitch” or something that in isolation would seem offensive. If you use it with certain groups, like with an in group or if you call your friend that, it can bring you closer together. It can be an expression of affection. It can help you bond, and a lot of marginalized communities do this. Or, even if you look at groups of editors or groups of writers, we have a lot of jokes about ourselves, we have a lot of stereotypes, and we just chuckle over them and it’s because we feel safe.
So, with conscious language, we are always checking in and looking at the anticipated impact instead of saying, “Oh, well, this word has been known to be offensive in some contexts. Therefore, don’t ever use it in any context,” which I have a big problem with as somebody who is creative and flexible with language.
DANIEL: Okay, I can get that when you’re in a situation. I totally get that about context. Context is like the magic ingredient where if you know someone, if you have common ground, yeah, certain things become on the table, and otherwise they might not be.
This is going to be really different in spoken or signed language versus in writing, because in writing, you might not share the context of the person who’s reading you. It might be years from now, and you might not even know who it is. So, it feels like, yeah, if we’re speaking or signing to someone, then context comes into it. But in writing, oh, you’ve got to be even more careful, don’t you?
KAREN: It’s always good to be careful, but I also want people to be able to say, “Well, what is reasonable?” For example, if I’m writing for an LGBTQ publication and I use some ingroup terminology as a member of that community, I don’t want to say, “Well, outside of this community, it’s going to be taken negatively or the wrong way.” I want to be able to say, “Well, I’m talking to my people here. We understand what this word means. I am entitled to that.” I don’t want to say, “Well, will this be able to have the same impact forever and ever across the globe?” I actually kind of resent having to do that, because people outside of the community don’t do that.
Also, you know as well as I do, that’s kind of a failing proposition. There’s no way you can anticipate all the different ways it can be. Like, let’s say the word, isis. How many people knew that isis would now have a tinge to it? Or like the okay sign, how many people knew that? So how are you going to anticipate everything? So, I think we need to consider how reasonable a word is at the moment it’s being used. And then if it becomes skunked, then we learn from that. We go, “Okay, well, see how a completely common word has become skunked or become offensive.”
And I actually think that by learning from all the ways that language evolves, we can become more compassionate toward one another and say, “Well, this is something that we need to deal with, and we need to be able to receive each other’s words in a way that doesn’t make them fearful of using language, period,” which is what I find happening. Because do I want to have people judge me based on a few words that I said? I would like people to look at my whole track record and say, “Well, she has a pattern of doing this for these communities, and she wrote this book, and she has Conscious Style Guide, the website. And then, just because she…” I don’t even know what example to use, “used this questionable word in some podcast, therefore we’re going to disregard everything else she says.” That’s not the kind of world I want to live in.
So, in the book, I talk about how we can extend this compassion to one another and how to call people in and pull them aside and gently say, “Hey, I noticed that you used this word. It might be received this way. What if you use this word instead?” Like, what if we gave that kind of grace to everyone?
DANIEL: Wow. Okay. I like that a lot. In that case, let me ask a question. Have you ever had a time when you disagreed with someone on some point, you both arrived at a different answer, but you were okay with their answer, even though it was different from yours? Has that ever happened?
KAREN: Never. Just kidding.
DANIEL: [LAUGHS]
KAREN: Yeah, it happens all the time. And depending on what is happening, like what the whole point is, I may or may not point it out. I’m not someone who will correct you as we go along. I’m more interested in the feelings and your beliefs behind the language that you use. And I often think about why do I approach people and language this way. And I think it’s because I grew up with parents who… English was not their primary language. So, they often reached for words that were not exact, they were not accurate, they were not appropriate. I didn’t want to criticize their English. Instead, I wanted to be able to put it upon myself to figure out what they really meant by that.
So, I kind of bring this approach to editing as well. I try to figure out what is your intent here and how can I help you execute this better. But then, you’re paying me to actually give you advice. But if were just talking, Daniel, I would not stop you and say, “Hey, I found that really offensive.” [LAUGHS]
DANIEL: Okay.
KAREN: I’m also often in a position where as somebody who is ethnically Chinese and I’m a woman and I’m queer, and now I consider myself an older person, I don’t really… My words are really not received as openly. If people decide that they don’t want to hang out with me or be around me or talk to me or hire me or work with me because I complained about something they said, that really impacts my life. So, that’s another reason why instead of criticizing people and trying to get them to use other types of language, instead, what I did, Daniel, was I wrote a book. And it is your option to read it or not read it.
And the whole book is presented as a philosophy and not “do this, do that.” It’s more like, if you are in this situation, if you are in this context, consider using these words, consider framing your ideas in this manner. Think about these other things. So, I’m trying to help people who want to use conscious language, I’m trying to help them do it better. I’m never trying to convert people. Trying to convert people and trying to correct people all the time just takes so much out of me and I really want to preserve my energy for other things.
DANIEL: Mm-hmm. That really came through in the book. And I sometimes say… like, I talk about skepticism sometimes because I found that useful in my life. And I sometimes say that skepticism is not a bunch of weapons that you take out and use against other people. Skepticism is a set of tools that you use on yourself. You relentlessly question your own self and your own biases and your own thoughts if you’re doing it right. And I feel like The Conscious Style Guide takes part in that. That we’re trying to ask ourselves, what do I really mean to say? Do I really want to have that effect? What would help in this situation? Let’s wake up.
KAREN: Yeah, I really love that way of looking at skepticism, because I do consider myself a skeptic. And I often talk about how I myself, I don’t adopt the latest language shift that comes along. I need to figure out if that resonates with myself at all. Like, sometimes, it’s immediately obvious to me if I like a new terminology, like it makes a lot of sense, it’s logical, it feels good, it feels natural, it feels inclusive. Other times, I look at a new word and I’m like, “You’re kidding me.”
DANIEL: [LAUGHS] I’m glad to hear you say that.
KAREN: Yeah. Because I’m also Gen X, I’m of a certain time. I have a certain way of approaching language already. So, in the book, I do guide people through how to think about language that’s coming at us, how to think about it critically. Like, the steps that you can go through to sus out, “Does this work for me? Can I connect this to something I believe in?” And the reason why it’s important to me to help people understand that the inclusive language that you choose, if you use it without believing in it, what’s going to happen without a doubt is that you’re going to become resentful of this language and your anger is going to be targeting the marginalized communities instead of the systems that created the inequities.
DANIEL: Wow. Okay.
KAREN: So, I recommend that if there’s a phrase that people are saying is offensive and you disagree highly that it’s offensive, great, don’t change it. Move on to something else. Move on. Because sometimes, for your heart and your mind to change, it takes time. It’s not a binary process. It’s not like a switch that you turn off and on. So, it’s really critical for you to figure out who you are, what you believe in, in order to figure out, is this language for me.
DANIEL: Let me ask you then, since you mentioned, what’s an example of a time where you caught a new piece of terminology and you said, “Oh, you’ve got to be kidding me”?
KAREN: [LAUGHS] I like to talk about how when I first became aware that the word “stupid” was disablist, my first thought was, “That’s so stupid.” [LAUGHTER]
DANIEL: For some reason, that is exactly the one I thought of. Because I don’t want to classify someone… That exact example. I don’t want to classify somebody as stupid. And by the way, some people might be stupid.
KAREN: [LAUGHS]
DANIEL: I mean, not everyone is as smart as everybody else. I’m stupider than some people, right? But I don’t want to classify people as stupid. But I don’t mind saying that an idea is stupid or something. But then, I don’t know, maybe I’m still doing it. So, I’m glad that you brought this up. So, where are you now on stupid?
KAREN: So, I like to always think about the impact. And when I say impact, you can divide this into the immediate impact and the long-term impact. You could also think about in the immediate context like with your audience, if you’re talking to your mom or if you’re talking to your friend, or if you’re talking to… If you’re writing an article in a paper, like, what is that context versus its life? We talked about it before, like how it’s going to persist after you let these words out.
I have come to agree that the word “stupid” when describing people is it is disablist, because that’s the impact. The impact is that you’re using this word that is often used to target certain groups. That said, I do also make a distinction between things and people. Like, even with the word “crazy,” like if you say, “Oh, that’s a crazy dance,” or, “Those clouds are crazy,” or, “That line is crazy-”
DANIEL: That’s insane.
KAREN: …I have no problem at all. But if you call a person crazy, depending on the context, I will probably choose other language. Like, I would ask myself, “Instead of saying stupid or dumb or crazy, what do I really mean? What do I really mean by that?” So, as a writer, I love this exercise. I get to know myself better, I get to be more creative with language. And it absolutely does not harm me in any way to challenge myself this way. And it’s come to the point where I no longer even think the word “stupid.”
DANIEL: We do this too on the show sometimes where I ask, “What do you really mean?” For example, let me just try one, a lame excuse. And then, I think, “Oh, you know what? If I weren’t so good at walking, that might hurt my feelings a little bit.” And then, I ask myself, “Okay, what do I really mean? Oh, well, I guess I mean like an unconvincing excuse or an ad-hoc excuse,” or something like that. That feels better to me.
KAREN: Exactly. Yeah, so I think that by always thinking of conscious language as being flexible and creative, really helps people understand that it’s not this rigid thing that you must do because that’s where the resentment builds. That’s where people feel disconnected to language and they have fear about, “Well, who’s going to judge me now? What’s going to happen?” So, I always try to push transparency. Like, if I’m unsure about some language that I used, talk about my uncertainty and say, “Well, I’m not sure if this is the right word to use, but I’m coming at it from this angle. I’m using it because this organization uses it,” like kind of unpack some of that. And by unpacking your approach and your thought process, you’re actually bringing some other people in where they will say, “Oh, I also go through that. So, maybe this conscious language business is not out of reach for me because if Karen Yin also struggles with the best way to approach something, then that’s where I am too.”
So, I always talk about how sometimes I’m a complete failure at choosing the appropriate language, because this is hard. What I don’t like about a lot of language guides is that they do present words to consider, “Do this, don’t do that,” without talking about the struggle behind choosing that list. And in the book, I talk about it openly, like, this is not easy. If it’s easy for you, great. But for a lot of people, it isn’t. So, that’s where we all are. We are all moving forward, or most of us, toward equity in language. And things will happen on our way there and to give each other grace.
DANIEL: That’s cool. I really like that approach. I also feel like that approach has limits. Tell me if you agree with this. There are kinds of languages that a person can use where you say, “Hey, don’t, no. No, we are not doing that.” There’s got to be times when like with slurs or things that would be offensive, sometimes you’ve got to say something.
KAREN: Right. And I think that what’s happening is your anticipated impact is different from their anticipated impact, or they just don’t care. So, it’s all about impact, what makes conscious language conscious. I really thought hard about this, and I came up with five components. Just briefly, it’s content, context, consequence, complexity and compassion. So, we need the context and the consequence components, which is what you’re talking about here, because you’re telling this person, “Hey, you haven’t considered the consequence of the word that you’re about to use. So, let me help you out.”
DANIEL: Yeah.
KAREN: Right? So, that’s what I hear happening here.
DANIEL: Mm-hmm. Okay. And once again, that’s a context where I will have to make a decision as to how to act in that moment. And sometimes… I mean, I haven’t had occasion to call people out, but if I do that, am I somebody from the outside getting mad on somebody else’s behalf? Because in the book, you mentioned that’s not something that we want to do. We don’t want to be somebody who’s coming from the outside saying, “Did you know that those people prefer this label?” I don’t want to be that guy.
KAREN: Yeah. If you have what I call outsider outrage…
DANIEL: Yes.
KAREN: …and that basically means you have not deferred to the group that the language affects the most and affects directly. So, in that case, I would pause and actually consult the group and find out if that word or phrase or description actually is considered harmful. But if you already have checked in and you know, then please go ahead and speak up or rise up on their behalf.
DANIEL: How do you go about the process of gauging when the waters have shifted a little bit? And I’m thinking of Latinx versus Latine. In my own observations, just watching people write about this and talk about this, I feel like Latine is gaining ground over Latinx, but I can’t really be sure. What tools do you use to sort of test the mood?
KAREN: Well, you know, thanks to having to put together a newsletter every month, I actually scour the Internet for articles of interest and I kind of curate the best commentary and thoughts and arguments and debates for the newsletter. And this is my way of helping people out because I do this anyway. Like, if I had no newsletter, I would still be trying to stay on top of how language is evolving in the conscious language world. So, let me help you out. So, I put together this newsletter and this is a very gentle way for people to just stay in touch, stay connected, and if something interests you, please click on the link, read the article, do your own deep dive. I feel like I’m already studying this why not share what I have found. And whatever interests me, that’s what I’m going to put in the newsletter.
And with Latine and Latinx, there is much more support for Latine than Latinx currently for many reasons. And actually, I devote a section of my book to talking about these terms.
DANIEL: Are there any other shifts that you’re noticing that you find really interesting?
KAREN: Actually, there’s nothing that I want to point out that has not already been going on for the last year. But I do want to say that a lot of people are concerned about this list of words that are banned in the government. So, I want to remind people, we’ve always described diversity and equity and inclusion. We may not have always used the words “DEI” or whatever variation you have, like IDEA or the one including the word “belonging.” So, let’s go back to describing those things without the word “DEI.” We don’t need the specific words.
Throughout history, people have always gotten idea across without using specific words. Sometimes, it was because they needed to protect their community. Like, if you were queer, you couldn’t really speak openly about queerness or queer sex or queer relationships. So, we had very coded ways. So, if you’re concerned about using specific words that are supposedly banned, then be creative, be flexible. Like, how else can you get that idea across? What did we say before DEI? So, anytime you ban something, I’m really not concerned because of history and all the examples I see throughout history where people have gotten around that.
DANIEL: I often feel that there’s a kind of asymmetry, that people on the progressive side have to monitor everything they say and get everything right, and people on the conservative side get to do whatever they want. They managed to turn DEI into a slur when most people do think that diversity and equity and inclusion are very good things, I mean, [SIGHS] it just feels like there’s two sets of rules.
KAREN: I really feel your sigh. [LAUGHTER] Yeah, let me tell you. So, I want to put people on the nonconservative side at ease. And the way I do that is in the book, I talk about how to collaborate, how to work across differences. So, if people on the nonconservative side are really concerned about their language and they feel more judged and they feel like they have to be more perfect and on top of everything, if you are able to practice receiving language from your allies, from our allies, from people who want the same equitable future, if you are able to receive their language more openly and understand that there are an infinite number of reasons why they might be using language that does not resonate with you, and I literally have a list of the reasons in the book.
Sometimes, I ask myself, “Well, why did I reach for that word that ultimately I would have chosen something else if I had paused and thought about it?” Well, sometimes it’s because I’m nervous. Sometimes, it’s because I’m really tired. Sometimes, I feel like if I haven’t read the same things that my audience has read, then we both think that different words are actually the most appropriate. Sometimes, I’m using language intentionally outdated to prove a point. There are so many reasons to not be using the most current and the most preferred. There’s also no such thing as the most current and the most preferred and the most ideal because that also brings us back to thinking of language as binary, “Do this, don’t do that.” Whereas I want to constantly be pushing for, “Well, what’s good in this situation?”
And sometimes what I really love talking about to help get the idea across that it’s all about context, is I ask them, think about something that your author has written or that you have written, or you have said, and that was insensitive and see if you can build the context around it so that word can be received better, so that you have built a bridge to that word. So, the word could be offensive or it could be a new inclusive term that a lot of people have never heard of. Instead of just dropping this language in, what if you had a little preface to it? What if you had an explanation to it?
What if, like, let’s say it’s a slur in a novel, what if this slur changed the course of the events so it had some purpose? You showed how it impacted the person that the slur was thrown at. You showed how the user of the slur changed. You showed how the slur rippled through the community. Like, don’t just drop it in there and say, “Aha, I’ve created tension. Great, good for me.” You’ve also offended a lot of people. It has to become meaningful. Can you always develop context to support insensitive language? No. Sometimes, there’s not enough space. Sometimes, it’s just a bad idea. Sometimes, it’s better to think more creatively. And as the writer or as the editor, propose language that has the same emotional resonance of what you’re trying to do. So, it goes back to instead of focusing on the word itself, think about the feeling, like creating the context in other ways.
DANIEL: So, it’s like sometimes I might say in a class, “Hey, everybody, I’m going to use this term, and I’m not sure if I’m doing the right thing here, but my reasons for doing this are such and such and I’ve thought about it.” It’s almost like you can agree with people’s choices as long as you feel like they’re coming from a place of care and they’ve done some thinking about it. That would enable you to approach their language act with more grace than we might normally be wanting to.
KAREN: I think having that kind of pretalk before the talk is really helpful. I also think that if I were a teacher in a classroom, I would be really careful because sometimes, even though you have some kind of warning or you have some kind of preparation, sometimes the student still is offended and they still might complain. So, you probably would know better than I how else to present the information or the material in a way that can preempt that.
I don’t know what that would be for teachers, whether it’s having the student sign a document saying, “I understand that this has potentially offensive content. I understand that what I’m about to view is this. I understand that I can opt out of this.” I don’t know what your options are because you’re the expert in that field, but I would consider all those things because I’m an outsider in this situation. I’m not a teacher. I don’t deal with students regularly.
DANIEL: In the book, you mention about being someone who has the name, Karen.
KAREN: Oh, yeah. [LAUGHS]
DANIEL: I enjoyed that. And this is what I was going to ask you about. How are you feeling watching this term, your name becoming a term over the last five or ten years? What’s it been like for you?
KAREN: This is really interesting because I talk about this pretty often and I always come to the conclusion that if the impact is raising awareness about what black and brown communities have to deal with people… usually, the stereotype is a white, blonde woman calling the cops on them for doing everyday activities. If the impact is social justice, then I’m all for it. I’m pro use of the word, Karen. But it has evolved to Karen being used to describe someone who’s just complaining about something. And sometimes, the complaining is actually for social justice. So, in that case, that person to me is not a Karen at all. So, it has morphed into something that is just so annoying, let me tell you. [LAUGHTER]
Like, because you know how the cocktail effect, if you hear your name or if you read your name, suddenly all your attention goes to it? So, I’m always like… my attention zooms right to that word and I’m like, “Well, how are they using it?” And now more often than not, it’s used to describe somebody who’s complaining, not somebody who is unjustly treating a marginalized group of people or mistreating marginalized group of people. So now, it’s just become a source of annoyance because I read Reddit and it’s everywhere on Reddit where there’s like, “Oh, there was a Karen in line.” And I’m like, “Really? What was she doing?” And she was doing nothing. She was doing something she was completely entitled to do, completely unobjectionable.
DANIEL: And the fact that there’s no parallel name to complain about men who complain. There’s no Kevin, right? There’s no Brad.
KAREN: I’ve heard Kevin used a few places, but you’re right, there is none. But sometimes, I’ll see male Karen. They’ll say male Karen.
DANIEL: Okay. Oh, right.
KAREN: Or they’ll say…
DANIEL: Like a male nurse.
KAREN: Yeah.
DANIEL: Wow. Okay. Yeah.
KAREN: But what I find interesting about Karen is that the whole word “Karen” has become the archetype, whereas in past uses of let’s say like the word, oh, chatty Cathy, or there’s a word for Debbie, is it doubting Debbie? no, I don’t remember.
DANIEL: Debbie Downer.
KAREN: Debbie Downer. So, there’s always been another descriptor attached to the name, but in this case it’s just Karen. It’s not like racist Karen, which I would have preferred. Yeah, anyway, that just makes it super annoying to come across.
DANIEL: Yeah, I’m sorry about that. It may be comforting to know that language will change and move on in 50 years. How about that?
KAREN: [LAUGHS]
DANIEL: Something to look forward to. You mentioned in the book, this is a quote. “A twinge of grief underlies certain shifts in language.” What’s going on there?
KAREN: Yeah, when the language is something that is part of our identity or it’s part of our lives, such as when the term, Chinese New Year, started to shift toward Lunar New Year, I really felt so disconnected to Lunar New Year because it was always Chinese New Year to me. And it wasn’t that I didn’t want to include the other cultures. It was more that I didn’t have that kind of love for the phrase, Lunar New Year. And to let go of Chinese New Year did bring some grief, like letting go of a friend.
But then, I realized being specific is part of conscious language. So, in some situations, it’s absolutely appropriate to say Chinese New Year. It is appropriate to say girl or boy specifically. It is good and conscious to say Asian or black. We need to focus on specific identities and specific groups so we don’t need to let go of the words we are currently using. There’s always an appropriate context for that.
DANIEL: Tell me about a success story where you or someone you know was able to have an influence on outdated, offensive usage. Have you had any?
KAREN: So, the extremely positive person in me wants to say that we all have an effect on language, and you don’t even have to be a parent or a teacher or an editor or writer. Whoever you talk to, you are changing them through the language that you use because that’s how we influence each other.
The more academic response, I suppose, would be the use of the word, Ms That was actually a pretty coordinated thing. It had to be explained, had to be proposed, had to be talked about in media so people would learn, “Oh, there is another option so we don’t have to talk about the marital status when we use feminine honorific instead of Ms or Mrs.” So, all those things. So, I want people to know that you can change the course of language just by the way you use it. And it can be overt or it can be very casual. We all have a say in how language evolves.
Some people have larger platforMs and they can make a bigger impact. Before, I used conscious language to describe respectful, empowering and inclusive language. It was not used that way, and now it’s being used that way. So just by using it, you’re changing the future.
DANIEL: That’s really cool. I’m also aware that while being conscious about the language we use is a very important step, we can’t just change our language and then call the job done. There’s still a lot to do. What do you hope that people will do in addition to using language, in addition to being conscious about the language they use?
KAREN: Yeah. Whenever I talk about this, it’s always like a controversial thing because I actually believe in action more than just changing language. I think language is a great beginning, and I think using appropriate language is a great way to change your thinking and therefore change your actions and change other people’s thoughts and change their beliefs and actions. But it is not as effective as actually changing policies, giving money to marginalized communities, actually going out there and being an activist. We need more concrete things like that to actually get things done.
So, if you told me you didn’t believe in conscious language, but you introduced policies at work that made it more inclusive and more equitable, I don’t care if you believe in conscious language or not, I see the effect of what you’re doing. I see the impact of other amazing things that you’re doing. And I would rather people focus on that than language. I would rather people make peace with words and understand that it’s one part of fighting back, but it cannot be the last thing and the only thing you do.
DANIEL: It feels kind of like something that I’ve been thinking, and that is that I’m not sure that changing language does have an effect on our thoughts. The whole inclusive language thing has felt like, let’s change language and then hope that thoughts and society follow. Whereas what you’re describing is we’re taking stock of ourselves. We’re improving our own thoughtfulness and respect and care. And one outcome of that may be that our language reflects that, but another outcome could be our money reflects that or our political power reflects that, but it starts in the heart. Groove is in the heart.
KAREN: [LAUGHS]
DANIEL: Sorry.
KAREN: Because conscious language, to me, is so many things, it isn’t just the word, it’s also the whole approach. It’s the way you consider the context. It’s the way you choose skillful language. It’s the way you try to describe fairly and accurately. It’s so many things that… It’s also, for example, throughout our conversation, the way you disagree with me is so lovely. I have to say…
DANIEL: [LAUGHS] Have I?
KAREN: …you will first agree and then you’ll say, “But then also.” So, that’s also conscious language because you could have disagreed with me in so many different ways. But the impact is that I would totally hang out with you again. I would talk to you again. And these microunits of relationships impact our entire lives because now we are more open to each other. What we take away is that we will speak kindly of each other instead of me going away and saying, “Oh well, he was such a jerk. Don’t go on his show.” [DANIEL LAUGHS] These little micro units of how people, how we treat each other. When I see my mom, she no longer comments on my weight. I taught her to not do that. And because of that, I see her as often as I can. I don’t have anything bitter or sad in my heart when I see her. It just changes all of our relationships.
And from then on, because she doesn’t comment on my appearance or my weight or anything, I can go away and I can do my work better. I am kinder to other people. I am able to have more energy to do things. All of these relationships just really impact every, every corner of our existence and a lot of it is through language.
DANIEL: True. How can people find your newsletter? Because it seems like that is a really good way to find out what’s going on the latest in the world of conscious language.
KAREN: Yes, you can go to consciousstyleguide.com, which is a website that I launched in 2015, and you can find the sign-up form there. And then, for folks who actually want to be able to talk about language, inclusive and conscious language, with primarily editors and writers, you can join my closed Facebook group. It’s called Conscious Language and Design. As long as you answer the questions and agree to the terMs I will let you in.
DANIEL: Neat.
KAREN: Yeah, you could ask anonymously, so I want to put people at ease. A lot of what I do is putting people at ease and just catching people up. And to me, this is my way of being inclusive. I want you to understand that inclusive language includes you.
DANIEL: The book is The Conscious Style: A Flexible Approach to Language that Includes, Respects, and Empowers. It’s available now from Little, Brown Spark, and I’ve been talking with the author, Karen Yin. Karen, thank you so much for hanging out with me on the show today.
KAREN: Thank you so much. I would do this anytime.
DANIEL: Thank you, Karen.
[INTERVIEW ENDS]
[MUSIC]
Let’s go to our Words of the Week, and this first one was suggested by James. We have a new instance of the -ussy combining form. As you will remember, it was the American Dialect Society Word of the Year in 2022. -ussy meaning some kind of aperture, like a plug, could be a chargussy, something like that.
HEDVIG: Mm-hmm.
DANIEL: Of course, linguists loved it. But we asked the question, “Is this really a thing? Is this really as productive as it seems?” Well, we’ve got a new instance. The musician, Lorde, doesn’t mind putting herself out there. She’s posted photos of herself without makeup as a way of encouraging people to accept their imperfections. Her album cover for Solar Panel was basically her leaping over the camera in a bikini bottom.
HEDVIG: Mm-hmm.
DANIEL: Well, now we have the cover for her album, Virgin, and it has an inner sleeve that’s a photo of Lorde wearing clear plastic pants. And now many people are approvingly commenting on the appearance of the…
HEDVIG: Lordeussy.
DANIEL: Thank you. One TikTok user, “Everyone’s all for free the bush until it comes to the Lordeussy.”
HEDVIG: Number one, not everyone is free the bush.
BEN: Yeah.
DANIEL: No.
HEDVIG: And secondly… yeah, okay, yeah. I don’t know.
BEN: I remain just as profoundly disinterested in this as a productive… It still just stinks, reeks. Positively stenchified of weird American, “Ooh, we can’t talk about genitals ever…
DANIEL: Puritanical. Yes.
BEN: …and so, if we do, it’s very saucy.” And then meanwhile, people from European or in this case, Australian backgrounds are just like…
DANIEL: “I really don’t…”
BEN: …people have vaginas. It’s a thing.
DANIEL: It does.
HEDVIG: Also, artists do things to get attention. I mean, so Lorde, in this case. We also have Sabrina Carpenter’s recent album cover that caught a lot of attention. It seems like big artists are again… I mean, artists are always trying to shock in order to get attention. And we’re just in a wave of album covers of that right now, it seems.
DANIEL: Yep.
BEN: I think the most interesting thing at work here is people are once again talking about album covers. As far as I can tell, to the generations younger than me, albums as a discrete unit of music had basically ceased to exist. So, the fact that people are talking about album covers is mind-blowing to me.
DANIEL: That’s good. The reappearance of album covers and the reappearance of the -USSY combining form. This next one was suggested by listener, Sydney. It’s BOOMERASK. Any guesses?
BEN: [LAUGHS]
HEDVIG: Ooh, is it, can you make this into a PDF? [LAUGHTER]
DANIEL: I have no idea how to do that. I feel like Boomers would be better at making PDFs out of normal documents than young people, than Alphas would, because it’s just all files to them. The idea of a file structure doesn’t exist.
BEN: Yeah, that is true and is a thing that I come up with all the time.
DANIEL: No, it’s not an older person asking you for help.
BEN: Okay. Is it the opposite of that? Is it when you go to an older person and ask them for help?
HEDVIG: Mm.
DANIEL: Oh, that’s a nice idea. I like that one.
HEDVIG: I need putting up my shelves. Yeah.
DANIEL: It’s not that.
HEDVIG: I need to change the oil.
BEN: [LAUGHS] I was just about to say the same one. I’ve got a flat tire, whatever could I possibly do?
DANIEL: I’ll tell you what it is. Sydney points us to Dictionary.com, which says boomerasking is when someone asks you a question, “How was your vacation?” but they don’t really care about your response because they just want to answer it themselves. It’s a conversational boomerang.
BEN: A boomerang.
HEDVIG: Oh, it’s boomerang.
DANIEL: Yes, it’s boomerang, boomerasking.
HEDVIG: Not boomerask.
BEN: So, Daniel, roleplay this one with me. You’re the person asking the question.
DANIEL: Okay. I’m like, “Hey, how’s your car?”
BEN: Oh, thanks for asking, man. Yeah, it’s good. I had to take it to the mechanic recently, but everything got sorted out.
DANIEL: Oh, that’s awesome.
BEN: How’s your car?
DANIEL: I’m glad you asked. And now, I will tell you at some length. That’s a boomerask.
BEN: [LAUGHS] Okay, right! Right right right right. Gotcha.
DANIEL: I think the appearance of BOOMER kind of derails this word and makes it not very intuitive.
HEDVIG: Mm-hmm. Yeah.
BEN: It is a very red herring on this particular one.
HEDVIG: Yeah. I think the difference between BOOMER-ask and boomer-ASK really doesn’t come through in writing.
DANIEL: Okay, well, you know what that means. It is incumbent now upon us to invent a better word for this.
BEN: Uh, I think -ARANG is the more productive and unique thing here.
DANIEL: Nice.
BEN: So the ASKARANG would be your better…
DANIEL: Yep. Not confusing.
BEN: Especially when you’ve got batarangs and other -arang sort of things out there in pop culture.
DANIEL: What other…?
HEDVIG: What about just BOOMERANG QUESTION?
BEN: Daniel, are you about to ask me what a batarang is? Because I am about to reach through this screen.
DANIEL: [BRIEF PAUSE] What’s a batarang?
BEN: The bloody things that Batman throws at goons, Daniel! Come on!
DANIEL: Oh, my gosh. Okay. Yes.
BEN: [SPUTTERING] The bat-shaped weapons that he flings because he never breaks his one rule, Daniel, come on.
DANIEL: What was his one rule?
BEN: He doesn’t kill people. That’s his thing.
HEDVIG: No guns.
DANIEL: That’s his rule?
BEN: Yeah.
DANIEL: You know what? I was thinking about Batman. And I was thinking, you know what would be really cool? I would love it if they investigated how he got to be Batman. Like, if they did a movie or three about…
HEDVIG: He’s super rich. That’s his thing. That’s his superpower.
DANIEL: …his origin story, like how he…
BEN: Daniel, are you being…
DANIEL: …like, his younger years.
BEN: Are you trying to be funny about the fact that these texts already exist in great detail?
DANIEL: I just love it when they explore a little-explored area of their backstory. [BEN LAUGHS] I just think it’s so interesting.
HEDVIG: I thought you were actually meaning not [BEN LAUGHS] what happened to his parents, but how he used his money to make a batarang, because that’s less explored.
DANIEL: And the suit.
HEDVIG: Like, who did he hire to dig out the Batcave?
BEN: [EXPLODES] Julius Fox! None of you watch…? Goddamnit.
DANIEL: That would be interesting. They should totally do a movie about that. [BEN LAUGHS] Next one.
BEN: Guys, The Dark Knight is a good film that you should watch! I don’t think this is particularly out-there take! It’s Christopher Nolan.
HEDVIG: Okay, okay.
DANIEL: Maybe I could write that.
HEDVIG: Oh, my god.
DANIEL: Next one is by Andy from Logophilius. It’s MICRO-RETIREMENT. Fast Company has an article by Kamanzi Constable, link in the show notes for this episode, it’s about this hot new trend of micro-retirement. Have you heard of this?
HEDVIG: Yes, because I am on the Facebook group, Did Silicon Valley Reinvent the Bus? [LAUGHTER] where people post this kind of thing. Because every now and then someone will be like, “What if we all put money into a big pot in a community and then when one of us got sick, we used that money to make that person better?” And it’s like…
DANIEL: You guys. Hear me out. [LAUGHTER]
BEN: You’ve come up with taxation!
HEDVIG: You guys. I don’t know if you’ve heard. [LAUGHTER] So, micro-retirement, in my understanding, was used in one article to describe young people taking some time off.
BEN: To like recharge their social battery.
HEDVIG: And it’s sort of like they’re taking out their retirement early. They’re taking it out in small doses. They’re microdosing their retirement. Whereas a lot of people, especially Europeans, would just be like, “Oh, you mean like vacation? The thing that is proven to make you work longer and be healthier and more productive at your job?” Yeah.
DANIEL: Yeah, but you only get two weeks, so, who’s got time for that?
BEN: Yeah, yeah. So, that’s the one thing that I wanted to sort of acknowledge in the… As much as I would love to dunk on gen Z and be like, “Look at these fucking idiots,” it is worth remembering that this is being executed in an American context. And Americans as a workforce are… Holidays basically don’t exist for them. That is a really important thing to remember. The European model, the Australian model… Americans just think we’re the laziest gits in the world because we’re just like lying on beaches and doing all this stuff. And Americans are like, “What happens to your job?
When I rode across America last year and people were like, “You have what kind of leave?” And I’m like, “Oh, it’s this thing called long service leave where if you’ve worked for a place for a long time, they will like give you money.” And I regularly blew minds. Like, I saw them liquefy inside their brains as they could not understand. So, I’m just a little bit hesitant to dunk too fully because I think the idea of coming up with the idea of micro-retirement is also a little bit because you can’t call it a holiday in America because then your workplace will be like, “Nein, no holiday.”
DANIEL: [LAUGHS] We’re not having that.
HEDVIG: Yeah.
DANIEL: Well, Andy says, who is American, “Hey, y’all. There are certain things that every new generation learns about that they think they actually discovered or invented, like sex or loud angry music or improv. What we don’t usually get though is a generation not realising their great invention has been around forever and then coming up with a name for it. Enter micro-retirement, a less focused term for what could be called a vacation, a sabbatical, a leave of absence, voluntary furlough, or just some time off. As if we needed a new word in this space. Kids these days, am I right? Just wait till these young’uns discover they can achieve orgasm on their own and sexual intracourse or some shit starts showing up in headlines. I hesitate to bring this to you because making it a Word of the Week qualifies it to be a Word of the Week of the Year. And if micro-retirement reaches that status, I may just have to give up on the word business for good. Yours truly, an old fogey, apparently. Andy.”
BEN: Andy. Look, all I would say is I don’t think that this is going to come anywhere near Word of the Year. You’re safe. But having said that, -USSY as a productive fucking suffix won three years ago, and I hated that, so here we are.
DANIEL: Yep.
HEDVIG: But like you just pointed out, Ben, Word of the Year from the American Dialect Society is American, right?
BEN: Mm, very.
HEDVIG: And this says something about American culture, like you just pointed out very clearly.
BEN: That is true.
HEDVIG: So, I think it might make it relatively high.
DANIEL: Might have legs.
BEN: Mm.
DANIEL: Yep. All right. And I will just say, Andy, this could have been a SpeakPipe.
BEN: Yeah. [LAUGHS]
DANIEL: We want to hear your voices, we’ll do it. Last of all, Hedvig, yours. You reminded me of this one.
HEDVIG: Well, I was looking through comments like I do, and I noticed a couple of these that come up again and again. So, there’s one which is, IF NOT X, WHY X-SHAPED?
DANIEL: Why X-shaped?
HEDVIG: So, this is like a cute video of a seal but you’re not supposed touch it because it can actually bite off your fingers. And it’ll say, like, “If not friend, why friend-shaped?”
BEN: Yes.
DANIEL: Yep.
HEDVIG: But it can also be like, “If not fascism, why fascism shaped?
BEN: Yes.
HEDVIG: You can do it for anything. And that one, particularly I like. I had a number of other ones, I don’t know if we’re doing all of them now or we’re just doing.
DANIEL: Go on. Let’s hear it.
HEDVIG: Oh, a classic one for cute animal. If you have a feed with a lot of cute animals, you have, “Not to be dramatic, but I would die for X.”
DANIEL: Okay. Yeah.
HEDVIG: So, there’s a cute Moo Deng or something.
DANIEL: Yep.
HEDVIG: GOD FORBID, AN X HAS A HOBBY. So, it’s like a woman with a lot of collecting something or like having a particular way of setting up their desk or something, and people are like, “Oh, god forbid, an X as a hobby.” That can be a bit more subversive as well. A new one that I don’t know if you’ve seen, Ben, but… um…
[PAUSE]
DANIEL: Doesn’t even ask me.
BEN: [LAUGHS]
HEDVIG: No. No way I am going to ask you.
DANIEL: Not a chance.
HEDVIG: You don’t hang out in Tiktok comment sections!
DANIEL: You’re not on TikTok.
HEDVIG: Yeah, that’s actually really dangerous because the dog could learn and then something like positive comes out of it. Have you seen that one, Ben?
BEN: I haven’t. No, that’s a new one to me.
HEDVIG: Ooh, that one’s really fun.
BEN: So, X could be really dangerous because then, positive outcome.
HEDVIG: Yes.
BEN: Okay.
HEDVIG: And then, there’s a lot of the, “god forbid something, something.” Aengryballs submitted “She’s X on my Y till I Z,” but I got confused by that one.
DANIEL: Well, that’s fun.
BEN: I think I get it.
HEDVIG: Example is like, “She wolf on my clothing till I sheep,” but…
DANIEL: Fascinating.
HEDVIG: I didn’t know that one. And then, there is “X, I hardly know her.” That one is…
DANIEL: Ah, classic.
HEDVIG: Daniel, you might know that one?
DANIEL: Oh, heard that one since…
HEDVIG: Anyways…
DANIEL: …ages ago.
BEN: Because that’s Shakespeare? [LAUGHS]
DANIEL: That’s my vintage.
BEN: Yeah.
HEDVIG: That’s super old. Anyway, I don’t know a good word for these. I mean, they’re clearly like memes but…
BEN: Are they idioms?
HEDVIG: …like comment section memes. Idioms.
DANIEL: I would say that the for example, the “If not X, why X-shaped,” that would be a phrasal template.
BEN: Okay.
HEDVIG: Phrasal template.
DANIEL: Like, the scream I just scrumped. I think that’s what we’re looking for.
HEDVIG: Okay, phrasal template. Yes, the scream I just scrumped exactly, that’s what I’m getting to phrasal template. Thank you. Thank you very much.
DANIEL: I did a little bit of digging on Y shaped or X shaped. The first instance I could find was on Reddit. There was a video of a bear who made friends with some dogs and the title was “If not friend, why friend shaped?” And so, when Elon Musk went full Nazi with his gestures at the RNC, his supporters flooded the comments, “He’s not a Nazi.” And of course, my favourite comments “If not Nazi, why Nazi shaped?”
HEDVIG: Yeah.
BEN: For some reason, this came into my head when Hedvig first started talking about this. In terms of ICE in America, the Immigration Customs Enforcement, if not secret police, why secret police shaped?
DANIEL: Totes.
BEN: Because it is operating in every conceivable way like a secret police force.
DANIEL: If not brownshirt, why brownshirt shaped? So, Lordeussy, boomerask, or perhaps askarang, micro-retirement, and “if not X, why X-shaped”? are our Words of the Week.
Let’s get to a comment by Maitiú: “Hello again. Just a quick note to say thanks for reading out my email. It was a nice surprise. I enjoyed your piece on the many words for snow. We have a similar thing here in Ireland, best embodied by the book, Thirty-Two Words for Field: Lost Words of the Irish Landscape by Manchán Magan. It’s a huge hit among English-speaking Irish people, but most Irish speakers I know roll their eyes at it. One more thing, Hedvig asked in the most recent episode if the names Kier and Ciarán were related or not. They are related. Both come from the Irish, Cir, meaning dark as in hair colour. Ciar was the son of Fergus, the King of Ulster. He was banished to the southwest of Ireland, where his descendants were known as Ciary. The people of Ciar. Ciar is uncommon in Ireland now, with the diminutive Ciarán much more common. All these names are anglicized with a K, reflecting the hard sea pronunciation in Irish.”
There’s a lot here, but I’ll post the comment on the show notes for this episode. “Oh, another little point here. They’re all also related to the Irish word for a beetle, ciaróg, a diminutive meaning little black thing. Really enjoying the podcast. Le dea-ghuí, Maitiú.”
BEN: Aw, thanks, Maitiú.
DANIEL: And that could have been a SpeakPipe. [LAUGHTER]
HEDVIG: It’s such a funny way to bully people.
DANIEL: Here’s a comment from Stevie on SpeakPipe. This was a SpeakPipe.
STEVIE: Hi, my name is Stevie and I’m listening from Seattle. In regards to Hedvig’s “keep your head cool” hat, if you are cheap and can only go to the grocery store, you can make a similar one out of a gallon, Ziploc bag and a lot of Otter Pops or Freezy Pops, because the Freezy Pops come in tubes which are connected in the plastic. And so, if you wear this hat at night while you’re in bed next to your head, as they melt, they will stay evenly distributed and can be refrozen easily. Just my solution for when we had a killer heat dome here in the Northwest. Enjoy the podcast and have a good week. Bye-bye.
DANIEL: Thanks, Stevie.
HEDVIG: Thank you so much.
BEN: Stevie is a person after my heart.
HEDVIG: I’ll have you know that I started… I did this first with Ziploc bags with ice in them and then I upgraded to my actual freeze packs. And then, I was like, “This is so good. Someone must have made like a convenient commercially available product.” And that’s how I solve most of my problems in life, is I’m like, “I can’t be alone with this good idea”
BEN: I’ve basically already invented this, let’s find someone who invented it better.
HEDVIG: Yes, yes.
BEN: [LAUGHS]
DANIEL: That way, I can make my inventions without my brain overheating.
HEDVIG: Yes.
BEN: That’s a pretty good Professor Frink.
DANIEL: Thank you. [LAUGHS]
BEN: I just wanted to say to Stevie, I have a deep abiding affection for the kind of homebrew ad-hoc little solution that you came up with. And by the way, the baffles, which is essentially, I think that’s the right word for what you’ve created, like a series of liquid-filled tubes that are effectively baffles which keep the distribution even, fucking genius. You are an absolute gold-star general innovator. I love it. I love it, Stevie.
DANIEL: And if you get hungry at night, they’re tasty.
BEN: [LAUGHS] Stop it.
DANIEL: Thanks to our guest Karen Yin. Thanks to everybody who gave us stories, words and comments. Thanks to SpeechDocs who transcribes all the words. And thanks to you, great patrons, for being you great patrons.
[MUSIC]
BEN: If you would like to help us here on the old Because Language, you know what to do. You can follow us. We are becauselanguage.com on BlueSky where Daniel spends a lot of his time and becauselangpod just about everywhere else. So many of our ideas come from the wonderful community, the network of listening nerds and you can be part of that network of listening nerds by sending us some ideas. You can do it a bunch of different ways. I don’t know if you’ve heard in this episode, but a particular way that Daniel would very much enjoy is if you left a SpeakPipe for us where you record your voice and drop it in. Now, here’s the thing about SpeakPipe.
DANIEL: That’s a good idea, Ben.
BEN: Thank you, Daniel. Far be it from me to be a grumpy curmudgeon like Daniel Midgley, but I know that every single fucking one of you can record your voice. We live in the halcyon new age world where most of our meetings can be done digitally. You can record your voice. You can all be doing it. So, let’s hear your voice. Stevie just gave us an amazing solution for Otter Pops, and I could hear her quintessentially Pacific Northwest accent, and it was delightful. Please leave us a SpeakPipe with all of your cool and interesting ideas. You can also tell a friend about us and write us a review in all of the places that podcasts live, because that just makes people listen to our show more, which would be really, really good.
HEDVIG: And if you decide to become a patron, you can interact with us on Discord. And technically, I guess you could also record a little audio message and just send it to us on Discord. Is that something having you encourage, Daniel?
DANIEL: Why not?
HEDVIG: Why not? I just invented a new way of harassing and making Daniel’s life more complicated with more platforms.
BEN: Yay.
HEDVIG: Yay.
DANIEL: A new channel for communication.
HEDVIG: If you become a patron, you get stuff. You get to hang out with us on Discord and see discarded content that Daniel doesn’t deem good enough for the show.
BEN: [LAUGHS]
DANIEL: Only from Hedvig.
HEDVIG: You get invited to live shows, of which we may have one coming up relatively soon. We also make bonus episodes. We do mailouts, and we do little shoutouts. And we’re going to get to one of those shoutouts right now. So, because someone, a long time ago, said that it was boring to hear the patron’s name in the same order, Daniel has been inventing new ways of ordering your names.
BEN: They created a monster.
HEDVIG: What has happened now?
DANIEL: I was thinking about the MingKwai keyboard and how Lin Yutang classified the Chinese characters by their shape. So, I did it for the Latin Alphabet. Quite a lot easier, actually.
BEN: Eh? How did you classify the Latin Alphabet’s shapes, Daniel?
DANIEL: Well, I started out with the round letters, and then I went on to the combination round-straight.
HEDVIG: So, P.
DANIEL: And then I went on to the entirely straight ones in order of number of strokes. So, the alphabet kind of went like O, C, S. And then, the straight bits start coming in, Q, J, G. Then, pretty soon we’re down to those just pure straight ones, I, L. And then, we got the W and M at the end. You’ll be able to infer it. You’ll get the hang of it.
HEDVIG: Can you just post this alphabet order?
DANIEL: I will. I will post the alphabet.
BEN: That’s a really fun and interesting…
DANIEL: And I’ll post everybody’s names. Basically, that’s alphabetic order now.
BEN: Okay.
DANIEL: So right back to you, Hedvig.
HEDVIG: O Tim.
Ben and Daniel: O Tim
HEDVIG: O Tim, Colleen, Chris L, Chris W, Canny Archer, sæ̃m, Sonic Snejhog, Stan, Steele, J0HNTR0Y, Joanna, Jack, James, gramaryen, Diego, PharaohKatt, Rodger, Rach, Rachel, Rene, Ignacio, Iztin, LordMortis…
Ben and Daniel: LordMortis.
HEDVIG: Luis, Linguistic C̷̛̤̰̳͉̺͕̋̚̚͠h̸͈̪̤͇̥͛͂a̶̡̢̛͕̰͈͗͋̐̚o̷̟̹͈̞̔̊͆͑͒̃s̵̍̒̊̈́̚̚ͅ, Lyssa, Laura, Larry, Tony, Tadhg, Helen, Nigel, Nikoli, Nasrin, Kristofer, Kathy, Keith, Kevin, Ariaflame, Aldo, Alyssa, Ayesha, Andy B, Andy from Logophilius, Amir, Amy, Fiona, Faux Frenchie, Felicity, Elías, Wolfdog, Whitney, Molly Dee, Mignon, Margareth, Manú, Meredith. Something happened.
DANIEL: What happened?
HEDVIG: This order…
BEN: Broke you.
HEDVIG: …is harder. [LAUGHTER]
BEN: It broke you real bad.
DANIEL: But I just noticed the pleasing movement through the lines, the pleasing shapes.
HEDVIG: No, no. I think that… because each time I looked at one and I was looking at the next, they were so similar that my brain went, “You’ve already done that one. Oh, you’re one ahead.”
DANIEL: So, visually, it threw you off.
HEDVIG: They’re too similar.
BEN: Oh, interesting.
DANIEL: Wow.
HEDVIG: That was weird. That’s never happened before. I don’t like it.
BEN: What’s particularly odd about that is if that was just a normal alphabetical order, you would have the same clumping in a different order, but you would still have the clumping.
DANIEL: Yeah.
BEN: All the M’s would be together, all the O’s would be together, all the Cs would be together, all the Ls would be together.
HEDVIG: Yeah, but…
DANIEL: but it felt wrong.
HEDVIG: But, Ben…
BEN: [LAUGHS] It wasn’t right.
HEDVIG: It was also the second letter. So, like, it’s Andy, Amir, Amy.
BEN: Oh, okay.
DANIEL: All right.
BEN: But hang on…
HEDVIG: I know…
BEN: It’ll still be the same [LAUGHS]
HEDVIG: Yes, but there’s something about the… There’s something. There’s something.
DANIEL: Let’s look at this. Jump on it, science.
BEN: As much as I hate to say it, there’s seems to be something to Daniel’s reordering because we find shit to talk about each and every time.
DANIEL: This is fun.
HEDVIG: Oh, I’m having a bad day.
DANIEL: Next time. Random sort.
HEDVIG: And a special thanks to our latest patrons. At the Listener level, we have Justine joining us, who just upgraded from the Friend level. And at the Friend level, we have Greg and Pigeon. I love pigeons.
DANIEL: Hi, Pigeon.
HEDVIG: Hello also to our newest free patrons, Taylor, Claire, Kell, Alex, Paul, Hugo, Ben W, Niclas R, Miss Behaving, 達維安, and John R. Thanks to all our patrons.
DANIEL: Our theme music was written and performed by Drew Krapljanov, who also performs with Ryan Beno and Didion’s Bible. Thanks for listening. We’ll catch you next time. Because Language.
IN UNISON: Pew, pew.
DANIEL: Thank you all.
HEDVIG: Pew, pew, pew indeed.
BEN: Do you know what Daniel’s “this could have been a SpeakPipe” is reminding me of? And maybe this is a phrasal template that we’ve only just now recognised, because it’s like, “This meeting could have been an email.”
DANIEL: Yes.
BEN: That’s what it reminds me of.
DANIEL: This could have been an X.
HEDVIG: It is.
[Transcript provided by SpeechDocs Podcast Transcription]