Valerie Fridland: The Amazing and Um, Complex World of Linguistics

27:44Wait, so you explain what an um signifies. What's an uh? What if I said uh, I don't eat meat. Well, it doesn't tend to have that. That just means you're probably less certain. Really, I tends to what we find. OK, there was really an interesting study done in the 1990s where researchers were asking college students questions that were one word answers. And what they did was they took questions that were easily answered. So most of us would have easy access to that answer. So like, what was the name of your dog growing up? You know, Or what was the name of your pet growing up? Well, most of us can come up with that answer pretty easily.

28:22Or what's the name of your best friend? We have that stored really easy access information. But if I ask you, in what sport is the Stanley Cup awarded? A lot of people won't know that. I mean, I'm not really a sports fan. See, you knew it right away. But for me, I was like, well, I don't know. What sport is that, right? Well, so what you find is people will, uh, if they think they have access to that information pretty readily, but they're not, they don't have it at the tip of their tongue. And they'll um if they need more time to come up with that. So that's using them truly as filled pauses, though, and not as a discourse marker.

28:57Um seems to be preferred as a discourse marker. So while you could say, uh, I don't really like meat, it's much more likely among younger speakers and women to um there. And that seems to be partially because of the sound symbolism of an um, and that's a whole other different topic. But when you say, uh, if you notice my mouth is wide open, my tongue's just hanging out. It's not a very pretty sound, right? It just sort of sounds like someone punched you in the gut. That's not really what we want to walk around saying. So it's not polite. It's not polite. And we do love to be polite. So when I say, um, see how that's much more pretty and pleasant?

29:35It doesn't have that kind of, I'm about to spit at you sort of look. So we think that that actually is playing a role in the preference of, um, as a discourse marker over, uh. So it's not that you couldn't say there, it's just less likely that it would happen. And it might then correlate more to uncertainty. That takes you a minute. It's just sort of then it's marking that you're doing some cognitive searching in your lexicon for that word versus using that may also mean, okay, I know the answer, but I know it's an answer that's dispreferred. And therefore I'm marking it by indicating to you I'm aware it will be a dispreferred answer.

30:09So that's sort of metacommunicative. it's very fascinating isn't it so much more when i answer somebody now well you just can't use those programs right that do it for you if you want to be more more careful about how you do it sorry i'm creating more work for you i'm sorry not sorry actually

30:33so i'm gonna tell you a little true story and you tell me if this is see I'm because I'm thinking about this and I appreciate how cognitively thoughtful you're being this is a true story so I have a friend who's from the UK who has this beautiful British accent and I once asked her to record my voicemail message because frankly I don't think anybody with a British accent can be possibly stupid. I just, I have to admit, whenever I hear a British accent, I think this is a smart, classy person. So I asked her to do my voicemail, and we agreed that we're trying to get the British whore sound, which she achieved.

31:26And so I want your opinion. Is that like, am I being a bad person if I do something like that and ask someone with a British accent to record my voicemail message. Well, as long as she's not a stranger that you accost on a street. No, no, no. As long as you're not stalking her, I think it's absolutely fine. This is actually a really funny story that you have because it really points to how deeply ingrained our social beliefs about language are, right? Because is there really anything more valuable or smarter about a British accent from a purely linguistic perspective? Absolutely not. Right. There was nothing historically about it that's better.

32:06And in fact, it's interesting because in the early days of the colonies and in like the 18th century, when British would come over to America, they would remark on how well-spoken Americans were. Now, it changed later and they remarked a lot after the Civil War about how bad we spoke and how we bastardized British English. But initially, because there was actually a lot of uniformity in the Americas in the way that English had been brought over, although there were actually regional differences. But people often remarked on the uniformity of American English versus British English at the time, that there were a lot less class based differences, a lot less sort of changes.

32:51It just seemed more uniform. So it's funny that that actually has shifted. and after the Civil War, of course, they didn't really think, not the Civil War, sorry, the Revolutionary War, they didn't really think American accents were so hot after that. And a lot of the reason that we think British accents are so wonderful is tied to the status of Britain in the American imagination during the times, the colonial times. So, you know, it was the mother country and many people had a lot of family there. There was, you know, their families were from there, history was from there. They look to Britain as being the cultural capital.

33:29America was kind of rough and tough and you had to fight for survival. So we look towards Britain as being the cultural and economic place to be. And in fact, for many years, people in the South, especially the Virginia colonies, sent their children back to Britain for education, which is one of the reasons that we find some retaining of certain British features that changed since the time of the main mass migration to the colonies because the people that sent their children back, the colonies that tended to send their children back retained or learned new features of British English that came in after most of the colonies had been established.

34:15So, for example, the R dropping in British English. So, you know, ca, pa, ca, instead of heart, park and car. That actually was a 19th century change. It was not considered prestigious in Britain until the 19th century. So the vast number of colonists came before that happened, which is why we have R's in American English. But look in New England or the Virginia colonies. They sent their kids back. So their kids adopted this new London standard and brought it back to those areas, which is what gave rise to some differences in American dialect. So I think what you're reacting to is the historical and cultural significance of Britain in the eyes of most Americans that then got associated with a social belief about the value of their speech.

35:03So we tend to think that British accents are beautiful. They're prestigious. They have high status. They're super smart. You know, they're associated with royalty. They come loaded with these social stereotypes that we have based on the socio-historical development of American speech. So that's a short, that's a long answer to no, don't feel bad. What it says is that you're just valuing where American English initially came from. And I'm with you. I think a good British accent is fabulous. So if Meghan Markle picked up a British accent, maybe the rest of the royal family would accept her? Yeah, I think I think that she's well past that.

35:43Americans might like her better.

35:47I don't think the British are not impressed by British accents, but Americans. Oh, OK. Well, but they're certainly not impressed by American accents. Oh, no, definitely not. No. So the same sociohistorical occurrences that made us like British accents make them dislike American accents. Right. They look at as American English as sort of a bastardization of British English. And as American English has become more global because, you know, the power of this nation has risen far past the power of Britain, which used to be an empire. Right. The value that they put on American English has gone down because it feels like it's invading Britishness.

36:29And so I think, you know, it's just a natural predisposition to social reality of the way that Americans are positioned. So you opened the door to this, which is my, wait, before I go down this path, I want to know if it's okay with you if I crack a tasteless, grotesque, perhaps, tasteless is the best word for joke, depending on how this goes. Are you going to be offended? I don't know. When you preface it like that, it's hard to say yes. Oh, I don't want to. But I have a pretty good sense of humor, so sure. Yeah, okay. So you opened the door to the topic of roticity, which is this. Oh, yes. Right?

37:21Right, big word. I'm impressed, roticity. Hey, that's a big word. I told you, I read your book. Linguistics word, I am very impressed. So I want to tell you that my favorite line in your whole book is, quote, this difference is what linguists call roticity which strangely always makes me hungry for chicken when i read that i just busted up and i don't i don't know how many people got that joke but it is just brilliant i'm glad you like it i do you know my i'd like to joke that if i hadn't become a linguist, I would have been a comedian. Well, at least you get me. At least you get me. Well, I'm glad you like that because it really does make me think of roticity, rotisserie, doesn't it?

38:09It's perfect. Roticity, rotisserie. So we talked about the British accent and now I'm going to flip to the other side, which is I am from Hawaii. I was born and raised there. I left when I was about 18. And so I have the remnants of a pigeon accent. And if anybody from Hawaii heard me, they would immediately pick up that I'm from Hawaii. They can detect this pigeon in me, no matter how conscious I am of trying to not use it. And so honestly, it still bothers me and pains me when I say something like one sentence and somebody can say, you're from Hawaii. What part of Hawaii are you from? And if I were to properly answer that in Pidgin, I would say, yeah, brah, I'm from Hawaii.

38:58I'm from Kali Valley, brah. That's Pidgin, Pidgin to Pidgin. So now, is this wrong? I mean, should you like be proud of your accent and not try to change? And I'm not saying I'm going to go to a British accent, but is there, I think people, if they hear me really use a Pidgin accent, they're going to have a lower opinion of me. so am I like being paranoid that makes me sad that that would be I mean I really why it's a beautiful it's a beautiful accent that tells people about you as a person and isn't that important isn't that connection and and identity isn't that important to us the fact that we would devalue uh accent that shows who someone is where they came from and who they might identify with at least at some point of their life, that makes me sad that people would have stereotypes about that.

39:55What you're talking about is social stereotypes that people hold about Hawaiians and about pidgin, Hawaiian pidgin, that then get transmitted to you despite all your accomplishments, right? That what they value is speech and their beliefs about speech over the reality of speech and the reality of who you are and what you've done. So what I would say to that is that's shame on them, not shame on you, because really it's their loss rather than yours. That said, it's certainly true that we have social stereotypes and we do apply stereotypes about speakers because the thoughts about people being not educated that we assign to a lot of lower class or ethnic dialects or ethnic varieties is based on our beliefs about speakers, not on the actuality of those speakers.

40:44You don't speak pidgin because you're dumb. You speak pidgin because it is a factor of where you grew up, what was the dialect of the area you grew up in, and that it was about connection and solidarity and communication. That's half of the reason we talk to each other. When I wake up in the morning and I see, you know, my spouse or my kids, well, actually, half the time, I don't want to talk to my kids in the morning. So let's take them out of this whole thing. I have teenagers. But if I see my spouse and I say, hey, how's it going? That's not about information. I mean, I saw him the night before.

41:17I can't have changed that much. It's really just about connection. It's about showing solidarity and that we would belong together in some way, that we're friends and we're close and we're intimate. And that's when you learn the dialect that you grew up in. It's an affiliation with region, right? And it's an affiliation with the people that made you and the people that you interact with. And so the value of that is really immense. To devalue it is insulting. So I would absolutely say that you shouldn't be embarrassed about it. In fact, you have an opportunity to educate because you're so successful, right, on your own right.

41:55You do something that's speech-related every day. it gives you the opportunity to reconfigure people's stereotypes about a Hawaiian pigeon speaker. Oh my God.

42:10Maybe we should end this podcast now. You're giving me so much homework to do. Well, although you reduced my, you reduced the amount of work I have to do about taking out filler words. So we'll add pigeon into it. Exactly. And how lucky you are. I would love I'd love to have been born in Hawaii. I love Hawaii. It's Honolulu and Waikiki are two of my favorite spots. And actually, I love Hawaiian pigeon. I think it's amazing. And, you know, I don't think the other thing is I actually don't think that most people that aren't familiar with Hawaiian pigeon, which is many, many people, would know that that's the accent you have.

42:45I think people might notice that there is an accent, but I don't think the majority of American English speakers would be able to place it. You know where it's from, and maybe more people on the West Coast, California, because they travel to Hawaii more often. But the vast majority would not think that. They would think, oh, there's an accent, but I don't think they know it, like where it's from. I wish I had learned this 30 years ago, but anyway. It's never too late to start. Okay. Next topic. Let's discuss what I consider a most fascinating case study in linguistics. Not that I know much about linguistics, but the evolution of the meaning and connotation of the word woke.

43:30Because to me, being woke is a positive thing. I'm not asleep. I'm not stupid. I'm not irresponsible. I'm not a lot of things. Now, all of a sudden, it's bad to be woke. So what happened here? How did this happen? Well, it underwent a process that we call semantic pejoration in linguistics. And you can have it either go amelioration or pejoration. Words can go either direction. So some words start off bad and become good over time in meaning. Other words start off fine and become bad in meaning over time. So, for example, another word that that's happened to over time is the word mistress. So it used to be that mistress simply meant the female of the house, like the master and the mistress of the house, of the manor.

44:20So it was good to be a mistress. But today, not many people want to say, yes, I'm proud of being a mistress because it has undergone this pejoration of meaning over time. And the reason it happens is because of sort of social cultural context that takes that word and shifts its meaning in a way that either is just organic over time, that it gets used in various ways. It can either happen because of cultural beliefs. So women being masters of anything are often sexualized. And that's when how mistress became sort of pejorative over time is its association with women and sexuality and sort of women who were in control of things would have been, have sexual control and also our tendency to think of women in a sexualized fashion.

45:12So that's really where a sociocultural belief impacted the word and made the meaning get worse over time. But with woke, which actually has been around a lot longer than, you know, it's not a new word. It's a new word to us, but it's not a new word in Black culture. And in fact, I don't know the etymological history in terms of the exact date, but my understanding of woke is it's something that has been around for over 50 years. in Black culture to mean to another African-American, be aware, be aware, you know, be out there, be woke, be safe, be safe. It sort of means be safe because with the reference to the kind of cultural consequences of being Black in America, particularly in like the 1940s and 1950s, when I think with the initial use of that word happened, where it was like, just be safe out there.

46:00You know what happens to a person that's Black in America. So keep your eyes out. And white people didn't use that word, right? It was really in Black culture. But with, you know, things like Black Twitter and TikTok, where there's been sort of a really valuation of African-American culture and has given more speakers access to words in Black culture. And a lot of those words, especially with, you know, the last two decades with sort of the reconsideration of Black culture's role in America and, you know, a lot of sort of prevalence of hip hop and music becoming really something that all races enjoy and embrace.

46:45I think a lot of words that come from Black culture have been embraced as hip, cool, edgy, rebellious, which again is a lot to do with the stereotypes we put on Black culture rather than anything truly emergent from that culture itself, right? This idea of danger and rebelliousness. And so woke is one of those words that got taken out of context and repurposed in a way that took on already a new meaning. It subtly shifted in meaning when it was removed from the initial context of one Black person to another Black person saying, be safe out there because we both recognize this overarching cultural threat to our safety, to being a sort of word used more widespread among all sorts of races to mean being aware of, you know, having a social consciousness about race relationships and race issues in America, to then being something that was applied by usually people, you know, that lean right in politics to refer to anything they saw as overreach in terms of social consciousness.

47:48and so they would use woke as a way to destroy that word's credibility in those other contexts where it was previously used. So if you start using a word in a different way, and we see this a lot of times with words that are previously bad. So the N word is a good example of a word that's been reclaimed and repurposed and has shifted in meaning. But again, it has a very specific cultural context and I as a white person shouldn't be throwing it around, right? So it has a certain meaning. And by reclaiming that meaning, you've shifted it to be a sort of look, a statement of brotherhood and of being proud of that word if you're in that culture.

48:27Well, I think woke has been under a similar kind of shifting from being a word that's used by a certain in-group to elicit solidarity and sort of suggest safety to a word that's used in a larger social conscious way to a word that then is repurposed by those that are opposed to those kinds of ideas and ideologies to expand the use to refer to things outside of the original reference of that word. So, you know, woke was often termed a term just to say, I'm socially aware. But now it's a term used by the far right, typically, to say anything that's very left leaning, anything that seems sort of overreach in terms of social consciousness is now woke.

49:13And that by doing that, they've sort of turned the word woke on itself. So I think it's not that to the original speakers, it still it means anything bad. It's that now it's expanded to be used by speakers to use it to mean something bad. And therefore, you cannot be sure when talking to someone which meaning of that word they're going to take from it, which has affected the use of the word overall. Does that make sense? It makes total sense. And it just it just pisses me off how the right has changed the meaning of that word or at least the, you know, the. This happens in language all the time.

49:50It's certainly far from the first time. But here's what's beautiful about language. And hopefully this makes you feel better is for every word that's appropriated by somebody else for different meaning, which has happened throughout the course of English. I mean, English word stock is like 80 % different than what it started as. And many of those original words don't mean anything like they mean anymore. So, for example, the word girl means a young female today. But originally, girl actually meant just young person. But it started to get used more often to refer to only girls. And that's how it changed in meaning over time.

50:24So, and knight, the word knight meant servant or young boy in Old English. English, but often those young boys would be knights, take on roles in the king's court, and they became known as knights, young men doing that job. So that's how that word shifted. So words are supposed to shift as we use them in different ways. So I think the lesson there is someone else can take the word and change its meaning. So you come up with a new one, and we always do. You just develop new language to talk about the same thing, and it just takes a little bit of time for that language to maybe effuse or, you know, be more diffused through the population.

51:02So, I don't think it really means anything bad that language changes. In this one moment, it might feel bad that that word has been appropriated in a way that's negative and sort of colored or tainted its use previously. But in the grand scheme of language, this is a very typical process and we always find ways to surmount it. Okay. There's your pep talk. Does that help? Yeah, that's bad. Well, you know, well, I guess is, is the interpretation of that. If you're talking to black people, you'd say, don't bother trying to get it back, find a different, better one. No, I think actually, here's a situation where I think context is key.

51:45So I think you would, I think most African-Americans would say, you know, I can still use woke like that because it's obvious when I'm using it, what I'm meaning. So I think what happens is that sometimes though, people move away from a word that has been taken over and appropriated and shifted meaning because they no longer feel that word represents them. So, you know, I think it would just depend on the individual speaker. I think, you know, one great example of that is the word riz. Have you heard of the new word riz that became really popular on TikTok, I think maybe in January or so. Nope, I'm too old.

52:23Well, I don't really tend to, but I have teenagers, so I learn a lot of new things through them, which is what inspired me to write the book in some ways. So Riz is, you know, if you have maximum Riz, that means you're really basically chill with the ladies. You know, you have a lot of allure. It's sort of a new - No wonder I never heard. it means you have a lot of attractiveness to to uh it was really originally used to sort of talk about male attractiveness to the ladies um and there were all these tiktok videos that were actually pretty entertaining you should look them up where people would say yeah i got riz and it would be a young you know adolescent boy saying that to the camera and then turning it to look at a bunch of girls running to him on a track team so they'd be at a track meet or a cross-country meet.

53:10And then all of a sudden these girls would come running to them like, oh, my God, I love you. But actually, they're just running the track meet or the cross country. You know, the guy's just superimposing himself on there. So it was kind of used in jest and in humor. It actually started with I can't remember the name of the originator, but an African-American social influencer who was talking about his own skill with the ladies in sort of a tongue in cheek way. And he was sort of talking about how you start talking to a lady and it's not going well, but then by being so suave and cool, you win them over.

53:44And that means you have serious riz because you're able to get the ladies to be attracted to you. Or the best type of riz, according to this guy, is if you have unspoken riz, which means you just give them a look like, you know, that Zoolander blue steel look. Do you remember that movie? And it sends the ladies flocking. So, you know, he was doing it to be funny, but he said now that it's gotten so appropriated by, you know, anybody under the age of 30, he doesn't even use it anymore. So it's not that it doesn't mean similar things to what he used it for, but once everybody else has it, the whole point of novelty and words like that is that they are remarkable and they stand out and they set you apart in some way.

54:28So once everybody starts using them, even if they're using them in the same way, who wants to use it that way anymore, which is why then language changes over time. So the bigger question is if someone would want to use woke still or not, because it has expanded beyond its original confines of use. But I think that probably happened in the African-American communities once woke was appropriated outside that community. I think there were a lot of African-American initial users of that word that didn't like its appropriation, even by white, liberal, left-leaning people, because it was an appropriation.

55:03So I think, you know, appropriation with language is always tricky to navigate in terms of who's going to be happy and who's not. Up next on Remarkable People. Lower pitch voices are overwhelmingly preferred for traits like dominance, authority, competence, and even intelligence and even trustworthiness. We find that lower pitch voices get higher ratings on those credibility, things like that.

55:33Listeners of the Remarkable People podcast will learn from some of the most successful people in the world. They provide practical tips and inspiring stories that will help you be more remarkable. If you live in the U.S. or Canada, text 831-609-0628 to get notified of each new episode. You're listening to Remarkable People with Guy Kawasaki. Next word. We've got a lot of words to cover. Next word. Next word. My kids in a five-minute conversation might say the word like 40 or 50 times. And there have been times where I sat down and I tallied how many times they said it. And at the end of this two or three minutes, I said, you just said like 25 times in two minutes.

56:32and then I read your book. So now I am not sure as a parent who wants my kids to be successful and happy and considered intelligent and articulate, et cetera, et cetera. Is it just that I'm 68 years old and just give it up guy? They can say like 20 times in two minutes and it's not a problem? Well, first let me ask you how old your child is. um that i have i have kids who are the ones that do it a lot are under 20 like okay so that's an important distinction to make because research on like you suggests that it peaks in adolescence and then levels out as kids age up so that's something in linguistics we call age grading and what that means it's a feature that has a really youth identity associated with it And it often has a purpose in terms of modeling that form for younger kids in the long view of how language change occurs over time.

57:36There are a number of different important factors that help language either, you know, shift from one generation to the other or change over time. And one of those seems to be what we call an adolescent peak, which is adolescents tend to use an incoming feature at a much higher rate than the speakers immediately before them in age groups. So the slightly older people, because they're learning that new feature, they're playing with it. And then it actually serves as a model to younger speakers to highlight the use of that form. And then as that age group ages up, they tend to use it less. And we refer to that as age grading in linguistics.

58:13Now, like is an age-graded feature in that way. So it does seem like adolescents use it to a higher degree than older speakers. But it's not fully age-graded in that as those speakers age up, almost all speakers under the age of 40 today use like quite a bit more than speakers over 40. And as you get younger, it's even more extreme. At some point, that levels out and that change has come to completion. So it's not going to progress any further, meaning we're not going to use it at a higher rate. So two things in reference to your kid that uses it. They're young and therefore they're in the adolescent peak.

58:50So naturally as they age up, it will change. Typically that happens. Everybody points to some extreme case of some person that's 30 and uses like to a point that's crazy. But that's rare. Honestly, if you walk around and talk to a bunch of 30-year-olds, they don't do that. They use like. And in fact, several of the episodes I listened to on your podcast, I was listening to two psychologists, evolutionary psychologists, I believe. And I actually noted several like usages in their interview. But they're older speakers, right? So they're using it to a lesser degree, but younger speakers just use it time exponentially more because it's an incoming feature.

59:30And by the time they age up and become a 60-year-old, use of like will be so normal that people won't notice it to the degree you're looking at it now. So that's short for saying, no, you shouldn't worry. Second thing is that I'm going to give you some homework. I'm giving you some homework. In addition, because I do this to my students, I give my students this like homework. So I'm going to give it to you. And anybody listening can try this out too. You've read my book. So you know the different uses of like. So it's not purposeless and random, which is what we tend to believe. The problem is there are about three major different forms of like used non-traditionally.

1:00:07So traditional uses, of course, are things like using it as a verb. I like ice cream. Using it as a noun, I have a dislike or I have a like for something. Using it as a preposition, he has eyes like the sky, that kind of thing. Using it as a conjunction, which is still actually grammatically in prescriptivist terms, not well loved, but that's saying things like it seems like, or I feel like, or I pretend like. Those kinds of usage are more traditional in terms of grammatical use. The newer ones are using it as an adverbial approximator. And that's basically where you use like in place of about to approximate or estimate something.

1:00:51So he's about 10. He's like 10. It's exactly one-to-one. There's nothing useless about it. It's serving exactly the same function as an approximating adverbial as about. People over 40 use about at a much higher rate than people under 40 who tend to prefer like. So it's just a shift that's happening over time. Second is as a sentential adverbial, which I don't know if you've seen that movie, I think it's the Banshees of In Sharon. Have you seen that with Colin Farrell, I think is his name? No. It's really funny. It's Irish and they use like a lot at the end of their sentences. So they'll say something like, it was smooth like, I was uneasy like, or just any sentence with a like at the end.

1:01:33That's a sentential adverbial. And there's a hypothesis and it's British. It's British and Irish. Actually, we find it in court documents from the 1700s in that area. And you still find it in Ireland, especially more rural areas. So it's an old feature. And we think that's probably where the modern like comes from. It's not a Valley Girl feature. That's actually a myth. It comes from this early like. But what does that do? It's providing estimation over the whole Senate. So when I say it's, he's like 10, that's providing estimation over the noun phrase or the adjective or whatever you're following it with right after.

1:02:10It's like blue. He's like 10, right? It's the clause or the phrase that follows it immediately after. But if you use it at the beginning or the end of the sentence, it's providing approximation or estimation over the whole sentence. Like he was a lawyer or he was like a lawyer. Do you see how you're taking this from approximation or estimation in the sentence and moving it to the front. So that's, again, serving for approximation. And the final knee use is quotative like, which is where you use it instead of the verb to say. Again, a one-to-one substitution. It's just a new form coming in. Younger speakers use like in that context a lot more.

1:02:46But older speakers actually say things like, it's like, I don't know, five or six miles from here. That is actually a quotative or a discourse marker use of like that adults, Older adults tend to have adopted pretty early. So the long story short is your homework, now that you know the three different types of like, is to not just count how many likes they use, but do an analysis of the function of those likes in their speech. and I am going to tell you that you will find that they're really much more useful than you think and the reason they're they feel like they're so populous is because there's three major purposes that are being served completely different purposes by their likes it just happens to be the same word that's being used and so it sounds like they're using like a lot so that's your homework if you choose oh my goodness but does it make you feel a little better about them like you's makes me feel worse about myself but okay so you you you you said something there that i i i have to take this opportunity it just triggered a thought in my brain so you said that you listened to two or three episodes of my podcast right right so as a linguist first Did you have this reaction like, oh my god, this podcast is overedited.

1:04:09There are no filler words. This is bullshit. This is not true to life. And it's over done. He's overed it too much. No, no, because we go with whatever we hear in these contexts. So people don't stop and think, oh, my God, this is weird. It has no ums. They just sort of listen, especially in contexts that are expected to be more formal or more, you know, kind of broadcast speech. So when we go see a movie, how many times do you sit in a movie theater? I'm like, I'm out of here. They're not saying like, there's no ums. They never say, I mean, I don't want to listen to them. It's bullshit. it. But we don't, we don't, we're used to using sort of suspended belief in those contexts to, you know, listen to this model speech.

1:04:57But what, here's what the problem with it is. It sets us up to believe that real conversation works that way, which it doesn't. It doesn't work that way. In real conversations every day that we have with people, we don't edit ourselves. But the thing is, we actually don't notice those ums and uhs. So what happens when you edit them out is you're doing what we already do. We filter it out unless it's overly populous for our own personal likes and dislikes. So for example, you hear your daughter or your son's likes more because you dislike them. And when we dislike something, it's called a frequency illusion.

1:05:33Once you hear it, you hear it everywhere, right? Arnold Zwicky is a linguist that coined that term, the frequency illusion, because once someone draws your attention to something, and my students laugh because we do this. We'll talk about a feature like, um, or like in class. And then everybody laughs for the rest of the class period. Every time someone uses those features, because it's so apparent to you because we were just talking about it. And that's, what's happening with like when you're reacting to your daughter is because you dislike that feature. Every time she says it, it's like she's stabbing you some, you know, with a knife, but you don't typically do that in general speech.

1:06:09So when I hear someone talk in a conversation, you don't walk away from a conversation with your neighbor thinking, oh my God, they used um 50 times. In fact, if I asked you to estimate how many ums they use, you wouldn't be able to tell me. If you were interested in what they were saying, you probably didn't hear the ums at all or very little. So, so I don't think it's that different from what we perceive in day-to-day speech when you filter them out of your podcast. So, you know, I think we have an expectation in cases like television or podcasting or news broadcasts, that those are contexts where people do more speech planning.

1:06:47And in cases of speech planning, we would predict fewer uses of um and uh, which is probably why we notice them more in those contexts. So when you assume someone has rehearsed what they have said or someone knows what they're talking about very deeply, then we anticipate people to have to do less cognitive processing. And therefore, when they signal a lot of cognitive processing by using um and uh, we think, wait a minute, don't they know what they're talking about? So I think that's probably one of the reasons why it's okay to filter them out in those contexts if you feel strongly about doing it.

1:07:20Oh, well, that's 180 degree shift. Okay. Oh, I'm trying to make you feel better about yourself. I personally wouldn't do it, but I don't think there's anything wrong. You're doing a good job. Okay. I have two topics more. Okay. If you still have time. Okay. Sure. So topic number one is let's say that, and this has to do with vocal fry. Let's say that Elizabeth Holmes hired you at the start of Terranos. What would you have advised her vis-a-vis lowering her voice or not? Okay, so many things I would advise her on beyond just her voice. But I would probably... You mean lying? Let's have a technology that's not deceptive.

1:08:10But beyond that, you know, that's an interesting question because, you know, she claims she's not lowering her voice, but it is extremely low pitched. And, you know, the people have said that they knew her earlier. She didn't lower her voice as much. My answer would be don't lower your voice because it's inauthentic sounding. I think a lot of people saw that as sort of an unusual choice for her to make. But especially when people started being suspicious about her product, then it sounded incredibly inauthentic and put on. And that is something that does not hit people very well. It's like me putting on a British accent.

1:08:52That's not my authentic self. So much as I might love using a British accent, it's not my natural speech. And people understand that and they sense that. And it makes me seem inauthentic and disingenuous. And I think that's actually how her voice shift spins. So that's, but that's interesting because here's the other part of that. The reason she did that, I think, is because there's a huge amount of research in psychology on our preferences for voice pitch. And overwhelmingly, if you look at how preferences on voice pitch play out, lower pitch voices are overwhelmingly preferred for traits like dominance, authority, competence, and even intelligence and even trustworthiness.

1:09:39We find that lower pitch voices get higher ratings on those, credibility, things like that. In a workplace context, we also find that people that adopt a lower voice tend to correlate with those that take on a more hierarchical high role in a discussion. So there was a really interesting experiment, and I think it was published in the Journal of Nonverbal Behavior, oddly enough. But no, it was in the Journal of Evolutionary Psychology, I think. Anyway, I'll look that up. But regardless, it was a really interesting experiment where they had people do tasks as a group in a business setting. And they said, all right, here's something that needs to happen.

1:10:19Here's your job. You guys figure out how to do it. You all figure out how to do it. What they discovered, and they had done an analysis of everybody who was participating in the experiment speech and established their baseline pitch. So in your normal talking voice, what's your pitch? We all have sort of an average pitch. That's why we recognize speakers when they call us and they just say, hello, we're actually analyzing their average pitch. And we know our friend's average pitch. We know our husband's average pitch. We know our wife's average pitch. And that allows us to estimate who they are.

1:10:47But anyway, so they did an average pitch analysis going into the experiment. Then they measured their pitch while they were performing the experiment. And what they found was an interesting correlation between those who took on a higher, more organizationally managerial role in performing the task and those that took a lower ranking, lower status role as sort of a worker bee rather than an organizer. And they discovered that the organizers lowered their pitch over the course of the task and the worker bees essentially raised their pitch over the course of the task. And this replicates studies that find that those in higher roles organizationally or those who take expert roles tend to have lower pitch voices or even lower their voice pitch during that expertise showing or display than those that are not in those roles.

1:11:37And why wouldn't you then as a woman who comes in with a higher pitch voice, just generally speaking, on average, women have higher pitch voices. Why would you not want to reap the benefits in an environment where you're already fighting against odds to be taken seriously and be viewed with authority. So what an Australian study done in the 1990s found was over the last 50 years, women have generally lowered their pitch by, I think it was almost 40 hertz. And think of Margaret Thatcher, who took voice lessons to learn to lower her pitch. Because the message we get as women is, if you want to be taken seriously, if you want to have authority, you need a lower pitch.

1:12:17You need to do something to make yourself seem more authoritative, to seem to be more competent and seem to be more in control. And dominance is strongly associated with lower pitch. Now, is there anything inherent in lower pitch that makes you more dominant? No. Now, an evolutionary psychologist would say that this association with lower pitch and dominance has to do with larger bodied animals having more power and being sort of, you know, A lion has a deep roar. A mouse has a squeak. That has to do with the size of that animal. And so animals will try to growl or bark or whatever at a lower pitch to come across when they're upset versus their sort of happy noises.

1:13:01I don't know if you ever heard coyotes when they're kind of just howling to each other. They're yippee. They're like yippee. But when they're in the kill mode, they have a deeper pitch to their voice because it correlates with dominance and fear. And so the correlation from the evolutionary psychology standpoint is that we still associate that socially with dominance, even though it really has no correlation to duality anymore. So anyway, that's sort of what I would tell Elizabeth Holmes that, you know, probably she's not doing herself any favor because it's not natural. But I do understand why she did it.

1:13:36And I think that probably is what drove her decision. That was a long answer, but I hope it was an interesting one because I don't think people are that aware of how much voice matters in the world. CEOs generally of larger companies who make more money, they've been correlated with having lower pitch voices. Presidential elections in many countries are correlated with the candidate that had lower voices. Now, clearly, that's not the only thing that's going on, but it does seem to play a role. You know, as a woman, that's awful if you're trying to get a job, right? You're having to compete against these people that naturally get associated with competence, intelligence, and trustworthiness over you.

1:14:15I know Phil Zimbardo. So maybe he and you can get together and redo the prison experiment and see if the guards got lower pitch and the prisoners got higher pitch as a result of this role playing. Oh, that would be interesting. with i'm not finished with elizabeth holmes yet okay okay now fast forward she's arrested she's in trial she's even in sentencing so now you know basically everybody thinks she's guilty and her legal team during the trial or during sentencing hires you and and says okay so what's your advice as a linguist should she ditch the low-pitched voice should she go back natural or should she stick with it you know what's your advice for her now that she's in deep shit yeah well uh if only i could be that helpful to someone who's in that deep of shit but that's a really interesting question because my advice to her if there wasn't if i had heard her talking in her deeper voice in a context where it wasn't so widely publicized.

1:15:28My advice to her on a witness stand would be probably for her not to even testify at all, but it would be to be natural with your voice and to have a higher pitch voice because a higher pitch voice is considered more vulnerable. And I think jurors might react more favorably to it. Just if you look at how those higher pitch voices, they're also considered for women in particular, or a higher pitch voice. I mean, it's definitely not true for a man, so I wouldn't give him this advice. But women with higher pitch voices are generally perceived as more attractive and more vulnerable, more submissive.

1:16:02And so I think her argument was in that trial, I think her argument was that she was kind of being pushed into doing it by her boyfriend at the time, that he was really the mastermind behind it all. And so if you're going with that play, then coming across as submissive and vulnerable is probably to your benefit. So from a purely mercenary position, if I'm hired by the defense team, that would be the advice I give them. And it's interesting because actually linguists are often hired by defense teams, by prosecutors in trial. So I have actually testified as an expert witness in court cases before.

1:16:39I've done some forensic linguistics as well. I feel a little dirty doing it sometimes, you know, because you know what the people want and, you know, you have to go with what's true as a linguist. And it just feels like you're, you know, people might think you're for sale, which I think is not true of almost every forensic linguist I know. They're very dedicated to their craft and their skill as a linguist and they wouldn't lie on the stand. But that is a field, actually, because people do want analysis of speech and they know there is research that suggests how people come off and people do do find a distinct difference in the way that they're viewed.

1:17:19So my advice to her would be definitely to go with her natural higher pitch voice during that trial. OK. I promise you, if I get arrested in Hawaii, I'm going to revert to pigeon. I think that very smart move very smart move however if you don't get arrested in Hawaii that's probably not the way to go yeah okay my last question it has to do with chat GPT and AI which in in text written form can produce absolutely grammatically perfect text and we're not far away from it producing perfect vocalism too. So what's going to happen to language in a world with AI? Well, you know, I think that's, of course, the hot button question these days.

1:18:13And again, it's a very different thing to have a computer that's doing it in text versus spoken assistance. And when you call, you know, those lines and you get the endless buttons, press three. Right now, we get frustrated because those virtual assistants stuck, right? They're not very good at understanding you. They're not very good at responding you to in a natural way. And I think we're on the cusp of that being completely revolutionized with chat GPT type technology based on large language models driving this. My concern with these types of of computational sort of language models, these virtual assistants is that, you know, as a professor, obviously it concerns me for plagiarism, you know, for basically taking from students, the ability, the forced ability to think through things and do their own writing.

1:19:05But I think there are ways around that from a interactive standpoint. I don't think it threatens us as people to have conversations with, with each other one-to-one or in business context on zoom, talking to each other, because those are cases where virtual assistants won't be doing the work we're doing. And we get a lot of value from interactions. We get a lot of social value. And if you look, you know, at why babies don't learn language from TV or radio, it's not because they're not hearing language. It's because they're not having interaction with language. And how would that be different than playing a virtual assistant with them?

1:19:40It won't because what they learn from is the interactive nature of eye gaze, of touch, of emphasis. And those are things that you have to be a conscious being to do. So I'm not worried about it replacing us in those ways. What I'm worried about is that we're going to lose the empathetic humanness of interactions in service encounter. So, you know, when I call, for example, American Airlines and my mother is dying and I need to get to her in 24 hours. What I'd get now is after, you know, an hour on the phone with virtual assistants, I might get through to a real human and they can have empathy.

1:20:22They can understand the emotional content of what I'm saying and respond back to me in that alike emotional sense that shows empathy, sympathy, and maybe even work with me doing something that traditionally would be disallowed because they recognize the value of doing that as a human. Well, virtual assistants are not going to do that. So I think we're going to enter a phase of human computer interaction where we're going to very much miss that human connection. And it's going to change the way that we engage with people that are not in our close inner circle. So our neighbors, our friends, our family, we're still going to have interactivity.

1:21:01But think about when you go to the store. And I have a grocery store near my house that I go to a lot because I'm lazy and I don't like to drive anywhere, you know, to go to the grocery. If I need milk, I don't really care if it costs more at that place. It's convenient and it's small. It's easy. I run in. And over the years, they've had the same staff and I've gotten to know them and I have these lovely conversations with them. You know, when I know about their lives and, you know, a couple of them, when I was pregnant with my child, they were pregnant with their child. And so we have this human connection and it makes going to that grocery store more pleasant for me because who wants to go grocery shopping?

1:21:32No one. But if you have this interaction and this human emotion, you have connection and that is something we're going to lose when you go through a grocery store and i don't know whole foods here now is already mostly digital and you know there's one person that you can flag down if something happens but who enjoys that interaction none of us so you know i i mourn the loss of that kind of connection because for me that's what language is about am i worried that it's going to reshape the way that we humans talk to each other. I don't worry about that. No, I don't. You know, I think what we may value might change, but I'm not too worried about the impact of that on just sort of the shape of language to come in our inner circles.

1:22:15How about you? I'm curious to think what you, because you have, you work in that area. What's your thought? I use ChatGBT as a research assistant, not as a writer. So I'm looking for, you know, if I'm writing a topic about what are the benefits of being envious chat gpt will give you five reasons maybe i thought of three and i'll grab the next two not not grab the text but i'll grab the idea and rewrite it i use it like that all the time i also use it to replace online thesaurus because it can find synonyms just like thesaurus.com can but let's say that i want to have i i always use a phrase in my talk about innovation jump to the next curve right so go ice harvesting to ice factory to refrigerator so i would go to chat gpt and say give me another phrase for jump to the next curve you can't do that with a thesaurus you can do that with chat gpt but now let me ask you this really my last question i promise you let's say the ceo of american airlines calls you up and says you know help us improve our virtual assistant so would you say to him well let's say somebody calls up and says you know uh can you upgrade me to first class and you train the virtual assistant to say um well let me give that a look the i mean by saying um with the pause and you know let me get Let me give that a shot.

1:23:53Do you think you could make virtual assistants, quote, better? More human? I think they actually are. There are programs that do add that to the virtual assistant. I've heard, I think it's Google Assistant, in fact, does use ums and uhs. And they've tested these models where they've had ones that don't do that versus ones that use things like discourse markers and build pauses. And actually, those are considered better communicators. because we find that when we test people on speech without any of those things versus speech with those things, that people tend to find the speech with those things more sociable.

1:24:28So we miss them if they're not there. And I think that that is exactly what models will do. And interestingly enough, most of my graduate students over the past few years don't work as professors in linguistics because those jobs are like, you know, one in a million these days. Most of them work for companies developing voice recognition and virtual assistants. I have students at Amazon. I have students at Google. That's really a prevalent employment for linguists who act as expert trainers to help improve the models based on what we know from linguistic research that humans do and how those things are construed.

1:25:09In fact, people do call people like me up and do that. And some friends of mine who are also very, you know, sort of senior linguists, I was going to say old people, but we're all senior linguists. We have been talking about forming a company to help these large language models do better work with their training data. Because I think, as you probably know, there are lots of issues with the training data that are used for these large language models. One question, of course, is undersourced languages and dialects are not adequately represented. And that is something that people like me, sociolinguists in particular, are very good at doing.

1:25:52And we have data just standing in our research labs with undersourced languages and dialects. So one idea is if we could connect researchers that use those data as sources for their research with companies that would need training data in those areas that we could actually make a good connection there and help support some research at the same time. We can also build better databases to be more adequately representative of those things. But the other thing is to actually also help eradicate bias that those training data have, because we know the research that looks at co-allocation, for example, which is where two words come together and they're more frequent.

1:26:35So, you know, co-allocates are often biased in the direction. So, if you have doctor, the word doctor, more often in training data, just from a historical perspective of the type of training data they use, more often in that training data, a pronoun like he would co-occur with a word like doctor. And for a hairdresser, it would be more often a pronoun like she will co-allocate. And that's just a natural bias in the data because of social historical context. But what if you could equalize that training data so that you had an equal number of those different situations? Then when ChatGPT is doing this predication of what the next word would be, which is my understanding of how it operates, right?

1:27:17It's doing a predictive modeling. It would equally predict a he or she in that context and get rid of the training bias. So there's a lot, I think, that a linguist actually can do to help make those models better. Yeah. And then chat GPT will be banned in Florida because it's woke. That's probably true. Well, I don't know. Is equal the same thing as woke? Well, then they'll have different versions. They can have the Florida version of chat GPT. That's right. The world is flat. Donald Trump is president. If you had, so here's the interesting thing, because I'm curious if you've had this experience and maybe I know they just released the fourth version.

1:28:00So maybe it's better, but I was playing around with it and I was asking it the same questions I would ask my students because I'm assuming that my students next semester, I'm not teaching this semester, but when I go back into the classroom next semester, I'm 100 % sure that chat GPT will be part of what's helping my students. And I think it, as you said, can be a research aid, but I wanted to see, well, if I ask it the questions that I send home my students with, you know, to answer in their journals, their daily journals that they do, what are the answers I will get and how correct are they?

1:28:33Because I'm assuming half of my students will go home and ask Chad GPT these questions and write down what they say. I found that actually it was only correct some of the time. Like it did not actually give me the correct answers that it should have based on, you know, the type of data it had access to. Some of them more tricky questions like, you know, what are the phonetic correlates of, what did I ask it, of pharyngealization or something like that. And what it did is give me, I said, what are the causes of this happening in speech? And it instead gave me the outcomes of what happened. So what are the examples of this?

1:29:13And I wrote it back and I said, no, those are the outcomes. What are the causes and then it said oh yes you're correct sorry here are the causes now i know it because i knew the answer but my students won't know that so in your research how do you cross check it how do you know that it's right how about on the first day of class in the first assignment you say okay these are the questions and right in the class you type it into chat gpt and you say okay so this is what chat gpt is going to say if you took this home and did it so and then you'll understand holy shit she already knows what's going to say i gotta i gotta figure out a better way right well hopefully i mean i'm less worried about that about my students using it because a lot of my a lot of stuff i have my class do it's things they have to do with with their own active imagination and chat gpt is not going to help them there but my worry is that they won't know when they're using it as a research aid, what's correct and what's not.

1:30:12Because, you know, when you go look up in a book, you can probably have a pretty good idea in that textbook or that original research article that it's giving you the correct information in terms of what you're supposed to tell me. But when you just type in chat GPT for it to do that work for you, you don't know that as a student, whether it's correct or not. So, you know, that's the bigger question. I was wondering, you know, in your own personal experience, have you found it to be incorrect and the things that you've asked it to research? Well, I did one test about 60 days ago and I asked chat GPT, is Steve Jobs a college graduate?

1:30:50And it said yes. And I know he's not. And then I just asked this last week and it said, no, he attended Reed College, but he did not graduate. So somehow it got fixed. But, you know, in this book I'm writing, I use a lot of quotes and a lot of examples of, you know, this is what Justin Timberlake said in an apology. And I got a lot of those. I got a lot. He had a lot of practice or Bill Clinton. I got a lot of those quotes. And then Madison checked each one to be sure that chat GPT is not just making shit up. And I'm attributing a quote to Justin Timberlake that he never said. And some of it we had to throw out actually.

1:31:34Wow. Okay, Madison, I need you. I need to borrow you. See, not all of us have a Madison.

1:31:42This is the problem we need. What we really need is not chat GPT. We need a Madison. That's truly what we need. We need Madison GPT. Yes. Yes. So listen, we've been going an hour and 34. You may now be the longest episode. Yeah. Oh my gosh. Hopefully you can cut out that useless part. Yeah, except you didn't say very many useless parts, if any. So I just, I want to close with this one story. I am friends with a guy named Steve Wolfram. Maybe you've heard of him. He is a physicist and he invented a software application called Mathematica. he was the youngest MacArthur fellow winner at 18 or 19 or something like that and he got his PhD at 18 or 19 anyway when I talk to Steve Wolfram I need to rest about four years because his conversation is just so intense and so deep and all that I'm telling you this whole story you you google or chat GPT Steve Wolfram after we're over okay and you're gonna see you're gonna see how high this praise is.

1:32:50I am telling you that you are the Steve Wolfram of linguistics. Thank you for joining us on this episode of Remarkable People, where we had the pleasure of speaking with Professor Valerie Friedland for a very long time. Her extensive expertise in linguistics, coupled with her passion for language, has made her a terrific resource for professionals, academicians, students, and podcasters. She opened up a whole can of worms regarding whether I should remove or keep filler words. If you have an opinion about this, please send me an email at guykawasaki at gmail.com. I want to thank the Remarkable People team.

1:33:41Peg Fitzpatrick, Jeff C., Shannon Hernandez, Louise Magana Alexis Nishimura and Madison Nismer the drop-in queen of Santa Cruz I'm Guy Kawasaki this is Remarkable People until next time mahalo and aloha one more thing I may slip into pigeon in this podcast from now on I hope that's okay with you brah no hoo hoo okay This is Remarkable People.