Episode 5 [Unedited]

This is episode is the full, unedited interview with David Peterson, language creator. If you haven’t listened to the fully produced episode yet, we strongly encourage you to do so before listening to this one. They’re shorter in length and much more refined.


*David Peterson Photograph by Gage Skidmore


Guest Starring David J. Peterson, Language Creator

Produced & Hosted by Adam Greenfield

Executive Produced by Patrick Yurick, Instructional Designer – MIT OGE

Executive Produced by Heather Konar, Communication Director – MIT OGE

Special thanks to the following editors who provided us invaluable feedback that aided in the development of this show:

Christopher O’Keeffe, Co-Founder of Podcation

Kristy Bennet, Manager – MIT Women’s League

Jennifer Cherone, Phd Candidate – MIT Burge Laboratory

Erik Tillman, Phd, Formerly of the Kim Lab & Currently A Fellow at Vida Ventures, LLC

The Great Communicators Podcast is a part of Gradcommx. Gradcommx, targeted at enhancing research communication, is the first offering of Gradx – a professional development project created for the graduate student population at the Massachusetts Institute of Technology by the Office For Graduate Education.


“Divider” by Chris Zabriskie is licensed under Attribution 4.0 International License (http://freemusicarchive.org)


[MUSIC PLAYING] ADAM GREENFIELD: Hello. Adam Greenfield, here, host of the Great Communicators podcast series. And what you’re about to hear is the full, unedited interview with one of the guests we spoke with. If you haven’t listened to the fully-produced episode yet, I definitely encourage you to do so before listening to this one. It is shorter in length, and much more refined. You can find them all at gradx.mit.edu/podcasts.

The idea behind these longer, unedited conversations is to give you an opportunity to hear the entire talk, [INAUDIBLE] and all. This is not only a fun way to hear the full flow of the conversation, but it also emphasizes the importance of the points made in the shorter, produced episodes– which, again, can be found at gradx.mit.edu/podcasts. Thanks for listening, and enjoy the conversation.

DAVID PETERSON: My name is David Peterson, and I’m a professional language creator and author.

ADAM GREENFIELD: Author. And what have you written?

DAVID PETERSON: Well, initially, Living Language– Dothraki, which was a teach-yourself guide for the Dothraki language. But most recently, The Art of Language Invention, which is an instructional book about how to create a language.

ADAM GREENFIELD: Actually, that’s a question I’ll get to in a second, because I want to go through the thought process of creating a language. But to go back a little bit, when did you realize that there was this love of language and creating languages?

DAVID PETERSON: My initial interest in language happened rather spontaneously. I grew up in a bilingual household, but due to an early divorce, I was disconnected from the Spanish-speaking side of my family. And so I didn’t gain full bilingual fluency, I became what’s called a heritage speaker. And so for many years, I was annoyed by the fact that my relatives could speak Spanish so well and I couldn’t, and so I didn’t like the idea of second languages at all.

ADAM GREENFIELD: Is that your background?

DAVID PETERSON: Just a second.



ADAM GREENFIELD: Is that your background– the Spanish heritage?

DAVID PETERSON: Yeah, Mexico, specifically.


DAVID PETERSON: And so I just ignored language and rolled my eyes at trying to learn it. I took Spanish in high school because I thought it would be easy, and it was. But when I was 17, I woke up one morning after a dream and was just struck by the fact that millions of people spoke French and that I wasn’t one of them. And so it immediately became my goal to learn every language on the planet.

It happened literally that quickly. And I started right then. I just grabbed any kind of language book I could and tried to start learning them, trying to do the exercise and things like that. And then the next year, in addition to AP Spanish, I took German 1.

I wanted to take French 2. It was the only one available. The instructor wouldn’t let me. I thought that I could catch up, and I’m sure that I absolutely could have caught up, but I wasn’t allowed in. And then when I went to Berkeley, I continued to take languages.

I took Arabic, because I was really interested in it. I took two semesters of Arabic, a semester of Russian, a semester of Esperanto– which was the first created language I’d ever heard of– semester of French, a semester of Middle Egyptian. And unfortunately, that was it. There were many more offerings. I should have taken many more language classes. It’s so easy when you’re in college, not in the real world.

But it was during my sophomore year that I learned about language creation, or I hit upon the idea. So I heard about Esperanto. I took a course in it. And so I knew about languages like Esperanto that had been created for international communication. But while I was taking linguistics, I hit upon the idea of creating a language just for my own personal use– just for fun. And so I started creating it basically as soon as I thought of it. And it’s been 16 years now, and I have yet to lose interest.


DAVID PETERSON: Let’s pause, for just one moment.


DAVID PETERSON: Roman, Roman, Roman! Roman, this behavior is not acceptable. Now, Roman, this is your last chance, you understand? I’m going to put you in the bedroom. I guess it’s just not going to happen. Come here, my boy.

ADAM GREENFIELD: You know this.

DAVID PETERSON: Come here, my boy. Yes. I see it, too. Don’t worry about it. All right. He must have smelled the cat on you. He got real friendly real quick.

ADAM GREENFIELD: Yeah, he’s– yeah.

DAVID PETERSON: My cat, she’s 11. She’s not much of a shedder, but she is very much into rubbing and all that stuff, so it wouldn’t surprise me.

ADAM GREENFIELD: So at this moment, how many languages do you speak?

DAVID PETERSON: I don’t know. Depends on how you define “speak.” I really need to learn what those ratings are for language ability and use those. But I’ve studied more than 20. Right now my project is Finnish, which I hope to learn well enough to be able to do an interview next summer.

ADAM GREENFIELD: So when you say it depends on how you define “speak,” what do you mean by that?

DAVID PETERSON: Well, for example, I think that I know the grammar of Hawaiian and the ins and outs of it a lot better than I do German. But if you talk about dropping me in a country and having to get by, I will take German over Hawaiian any day. Just because when you’re a linguist, you study a lot of grammars abstractly and you know the ins and outs of them, but that doesn’t mean that you’re very good at speaking the language or that you’re going to recall vocabulary.

But with German, I had an entire year of interaction. I got used to saying phrases and stuff. In fact, I just went back to my high school and spoke to a couple German classes there, because my teacher’s still there. And he addresses the class in German, and I discovered that 16 years later, having not done anything with German since then, I still understood everything.

ADAM GREENFIELD: What do you attribute that to? Is that just a brain function that allows you to do that, or? I took a year of Latin in college, and I remember some of it. Is it attributed to just going back and being in that atmosphere?

DAVID PETERSON: It’s just memory. I have a really good memory for everything. It’s the type of thing where it’s like– for example, when you talk about Latin, when I think of it, I think of the four, five months I spent studying Latin on my own from a book that was published by Barnes & Noble in the ’50s, so it wasn’t Wheelock’s Latin. But when I think of the declensions, I remember how it looked on the page. I can still see them. The first one was “rosa,” “rosi,” and so on. Just all there.

ADAM GREENFIELD: So it’s just a memory thing. Do you you have a photographic memory, then, I take it?

DAVID PETERSON: I don’t know. I think so, or it’s just very visual. This is how I did it– I was an English major. I’m a very slow reader, so I could never keep up with the pace that you’re supposed to keep up with in English classes. So I would just get the reading list ahead of time and be reading all of them, say, over the summer or over the previous semester. And then during class, I would just remember it. And then the way I would remember quotes– because I didn’t like marking up my books or anything– I just remember where it was on the page and what that page looked like, and then that would be enough for me to go and find it if I needed to get the wording exact.

ADAM GREENFIELD: OK, OK. So you were then in, I would assume, your early 20s when you first created your first language, or still in teens?

DAVID PETERSON: I mean, I guess I would have been 19 still, yeah. Yeah, because my birthday’s in January, yeah.

ADAM GREENFIELD: And you said that was just for fun? Were your friends involved in any of that, or?

DAVID PETERSON: No, up at Berkeley. I’m from Orange County, so all my friends were down here. So four years that we were just totally disconnected. So yeah, I started creating a language then. I did it totally on my own.

I thought initially that perhaps my girlfriend and I would speak it– my girlfriend at the time– not understanding that this was– I mean, this was really– it was really just for me. This is not something that you give as a present that somebody is going to appreciate. Because you know, learning a language is a pain in the butt.

ADAM GREENFIELD: English is hard enough sometimes.

DAVID PETERSON: Yeah. And I don’t think I actually told people about– just general people– about how I created languages until at least a couple years later. Eventually, I did find the community online– language creation community online– so I had lots of people to talk to there. But yeah, I don’t think I actually mentioned it or told people about it in person until maybe my second-to-last year of college.

ADAM GREENFIELD: So can you go into a little bit of the– not just the thought process, but the process itself of creating a language?

DAVID PETERSON: Sure. So the language itself– the most important thing that you have to do at the beginning is to determine what type of language you’re creating, because that’s actually going to make a lot of your decisions for you. There are a lot of things that languages have in common, but depending on your project, they can be very, very different enterprises.

So for example, if you decide that you want to create a language as if Latin had never been expelled from the British Isles but continued on there into the present, then that’s going to be a very different enterprise than if you’re creating a fantastical language for aliens that have no eyes and ears but 49 tentacles. Totally different process, but they’re still probably– both are going to have words and such.

It’s just that with one, you’re going to be working with the actual Latin language and evolving it up using sound changes that happened in– I always get this wrong. I think it’s P-Celtic is in the west and Q-Celtic was in the east. Whereas for the 49 tentacles, obviously you’re going to have to be working– OK, if they can’t use sounds, if they can’t use their mouths, how are they going to convey language? It’s going to be with their tentacles, right? Well, how is it going to work? And then you go on from there.

So it could be that different when it comes to creating a language if you don’t specify at the outset exactly why you’re doing it. But let’s say that you do languages that I do most of the time– or that I’m called on to do most of the time– which is a more-or-less realistic language that takes place in some sort of a fictional setting. In that case, you have to figure out as much as you can about who’s speaking it, where they’re from, and why.

For example, creating a language in the universe of George R. R. Martin means that you can really start from scratch, whereas if you’re creating a language that takes place– let’s say it’s just a fictional race of people that exist in our world, then, just like our languages, they’ll have borrowings from real-world languages. Their word for “television” is probably going to sound a lot like the English word “television,” so on and so forth.

So that’s another top-level decision. If you’re creating a fantasy language like I do, then what you do is you start from a very early stage. So you create a proto-language. This is something like– in English and Russian and German and Greek, those are all descended from one language that we call Proto-Indo-European.

So as a language creator, I try to create that earliest stage of language. At some very, very early date you have to choose a random cut-off point, because we don’t know how language got started. We can only guess. But from that point, you start with a sound system.

This is what sounds are in the language and how they’re used, how they’re deployed in syllables and words. And then you have a grammar– both the inflection of nouns, the inflections of verbs– to the extent that the language has any– and also how words are put together into phrases, and phrases into sentences. And then the lexicon, which is the entire host of words that the language has. As well as derivational strategies– how you form new words from old words.

Then you evolve it forward. So you start 2,000 years in the past, and then you change the sounds of the language as they would have changed over time. You change the meanings of words as they would have changed over time, and you evolve the grammar incrementally as it would have evolved over time. And the end result is a very believable, very authentic modern language.

ADAM GREENFIELD: Is there some– I don’t want to say is it easier, but is there– is it easier, for lack of better words, to create a fantasy language where you’re even creating the prototype, or is it easier to already have that prototype in place?

DAVID PETERSON: I wouldn’t say that either is easier than the other. They each present unique challenges– unique and different challenges. So on the one hand, it is easier if you’re starting with a complete blank slate because you get to invent everything, and you don’t have to take anything into account except for what you’re doing. But there’s also a lot more work.

If you’re doing something different, like what I described with Latin, then you’re starting with a lot of material– and that’s great. You don’t have to create it. But you’re also beholden to that material. So you need to make sure everything that you’re doing is accurate and makes sense based on the existing material, and also the existing timelines so that you know everything is actually going as it should be.

When you’re creating your own language, you have to take into account the history of the people that you have there, but you may get to invent it as you’re going along. If you’re evolving Latin forward, you really need to be up on the exact timeline of events of what happened over in the British Isles, starting from the Roman invasion.

ADAM GREENFIELD: You mentioned the language creation community. What is that community like? Will they call you out if you’re not following– I guess the rules of the prototype language?

DAVID PETERSON: Oh, yeah, yeah. The language creation community is a very– it’s, at this point, a loose-knit community, and it comprises several different listservs, bulletin boards, and then random people on Twitter and Tumblr and Facebook. And so there was only one community at first, and that was the listserv. And then as time wore on, it expanded to many, many other communities.

All of them essentially have the same goal in mind, which is to produce the best languages possible– advance the state of the art. And indeed, in general, the community has got better at it as time has gone on. So when popular projects come out, I mean, they of course look at them and tear them apart. And usually, tearing them apart is in order, because honestly, before Game of Thrones, there wasn’t a case where there was an actual language creator working to create a language for any famous project. It was just somebody who’d never even considered the prospect before. And so tearing those languages to shreds was really just child’s play.

At this point, I was already in the community for about 10 years by the time I started working on Game of Thrones, and it already established a reputation there. So I think most of the people, more or less, trusted me to do a good job. Even so, I keep that community in mind whenever I’m working on anything now or anything new, because I want to be sure myself that I am holding myself up to the standards that the community has set. And if I ever slip, I expect them to call me out on it, because if they don’t, no one will.

ADAM GREENFIELD: So you’re definitely keeping your audience in mind as far as when you create new languages. You’re keeping not just the general audience, but the people in your field in mind when you’re doing that.

DAVID PETERSON: Yeah, and I think that the best way to describe that is anytime that you’re producing anything– whether it’s a language or a television show or what have you– there’s going to be different levels of viewership, or different levels of audience. Concerning just how the languages are used on Game of Thrones, there is a large percentage of the audience that just watches the TV show, hears the language, and doesn’t give it a second thought. It’s just like, all right, well, whatever. They’re speaking something, and I don’t care about it. There is going to be that level of audience participation.

And there’s a next level where they will pay attention to the fact that, for example, if one person is speaking Valyrian here and then somebody else is speaking Valyrian here, it should sound about the same. Without knowing any idea what the grammar is or what the words are or paying attention to anything like that, they should develop an ear for it so that they can actually– especially after many episodes– pick out, that person isn’t speaking it very well, or that doesn’t really sound like Valyrian. So that they would notice if, for whatever reason, somebody just started writing down gibberish for them to speak.

And then there are further levels beyond that. There are people that will actually write down all of the words. There are people– there’s a fan of Valyrian in particular– he can spot grammatical errors on the first airing, where he’ll just hear it and say, that doesn’t sound right. In fact, that sounds like what they said was this. And if that’s what they said, that was a grammatical error– that I probably made. I try my best to say that. Oh, no, the actor just made that error. They just screwed it up. It wasn’t me. But sometimes it’s me.

So I think with that in mind, you have to keep all of the audiences in mind. It’s got to be accessible to every single level so that if somebody wants to dig into it, there’s something to dig into, but it’s not a total barrier to somebody who doesn’t want to do that. For example, if subtitles were not included for any of the lines on Game of Thrones, I think that would be totally unacceptable. That would just be too high a barrier, and it would be off-putting to a large portion of the audience. And you want everybody to be watching it. That goes the same for anything that you produce, I think.

ADAM GREENFIELD: As a side note, I was listening to an interview with– oh, her name is totally escaping me– the actress that plays Daenerys.

DAVID PETERSON: Oh, Emilia Clarke.

ADAM GREENFIELD: Yeah. I was listening to an interview with her, and she mentioned that she’s gotten to a point now where she can spot her own mistakes when she’s speaking, and it’s interesting. Did you have any involvement in teaching the actors the languages?

DAVID PETERSON: Not on Game of Thrones, no. It’s depends. I’ve been involved with a lot of the actors on The 100 and a lot of the actors on Defiance, and then some actors for other productions. It depends. It’s an entirely– it depends on the show writers or director and what they want, and the actors and what they want.

In the case of Game of Thrones, though, it’s been running for six seasons now, with a seventh season coming up. And certain of the actors have just had so many lines that they’re developing their own– I mean, I wouldn’t say fluency in it, because they don’t know what they’re saying, but they’re developing their own ear for it. And so absolutely, I believe if she can just tell that something doesn’t sound right– even if she can’t put her finger on exactly what’s wrong, she might have a better ear for it than I do at this point.

ADAM GREENFIELD: Wow, interesting. OK, so as I mentioned, this project– it’s essentially about the importance of effective communication of technical or highly-specified subjects. And that includes, again, both audiences in a general sense and within the field. So when you’re asked to create a language, do you ever go into the specifics of how you’re creating that language in order to get feedback and ensure that you’re giving them what they want, or can too much background muddy that process?

DAVID PETERSON: Sometimes going into detail is warranted, and sometimes– most of the time it’s not. For example, when settling on the sound of a language– and if I know I want to employ some sort of phonological process, whether it’s dissimulation of stops before other stops or word [INAUDIBLE] voicing, I don’t need to tell them that so long as I can say, here’s what I’m thinking sounds like. Does that sound good to you? And they can just say up or down, yes or no.

And I always tell them to give whatever feedback they want, and I know how to interpret it. So if they say, oh, it needs to sound softer or harsher or there’s too many stringent consonants– they’ll come up with any type of adjective. It’s totally meaningless, but at this point in time, I get what they’re after, and I can look at what I’m doing and interpret what they’re saying.

So usually we’re working at that level, especially when you talk about production. But sometimes, if I need to really make it point-specific, then I’ll have to get into details, in which case it’s best to start off with what they know and then build from there. Just essentially teach a mini lesson so you can get to the point where you can say, here is the issue that I am having. I want to do this, but right now, what’s in the script calls for this, which is the exact opposite. Could we perhaps alter this so that we can do what I want to do here?

ADAM GREENFIELD: Have you found them to be a little flexible in that?

DAVID PETERSON: Yeah, actually. Usually– I mean, and we both have to be flexible there. But usually, if I ask for some sort of change and it’s not important to anything else– important to the script– they’ll just go ahead and do it without even asking why. But also sometimes– for example, this happened a couple seasons ago.

We translated this line about– it was originally, you sit before Daenerys Storm-Born, blah, blah, blah. It was this and a whole bunch of titles. And since Valyrian is a case language that inflects with suffixes, her name changed, as it would, because it was in a different case. And I wrote up, I could say, we really wanted her name to be recognizable, can it not change?

And so then the response to that is not, OK, let’s rip out the case system, but I say, OK, can I recast the sentence, then? Can I just, for example, make it a passive sentence? So elevate her to subject position so it sounds the way it should. And that’s when they said– as I explained it, they said yes, that’s fine. Just so long as we still hear her name and it means approximately the same thing, then it’s fine.

So that’s the type of discourse that usually happens when it comes up. And this type of thing comes up maybe once or twice a season. And I found usually they’re pretty amenable to the changes that I want to make.

ADAM GREENFIELD: Do you ever find yourself under a time restraint of how long something is being said? Because in television, you have to move it along– move the process along– and not be so extensive in what you’re trying to say. So is that ever a limitation when creating language?

DAVID PETERSON: Yeah. I’ll give you three different answers, here. So on something like Game of Thrones, they’ve actually been very, very, very, very tolerant of what I do with it. In other words, they’ll just take whatever I translate, and that’s what it becomes. And however long it takes, that’s how long it takes. And that’s great, but they, of course, also don’t really need to be worried about how long the show is. They can say to HBO, hey, we want another 10 minutes for this episode, and HBO says fine. They don’t have to worry about commercials or anything like that.

I have discovered, in other shows, that there is a loose constraint– which is if the director, on the day of the shoot, is hearing something and decides it’s too long, he may just decide to indiscriminately lop off a word. And sometimes it absolutely destroys the grammar of the sentence. And so in that case, I keep in mind– if something is getting very long, I’ll try to recast it on my own so that it’s close to the same length of what the English would be.

The nice thing is that acted English often takes up much more time than just casual conversation English. I think it just goes back to theatre. That theatrical strain is still in there a lot– where when they’re acting the English, it doesn’t sound like two people speaking. It sounds a little bit more like they’re performing.

The third thing, though– one of the things that I think I’ve done in my work that’s the most interesting and that I also hate the absolute most is translating songs, and I’ve done it a lot. I counted up– I think I did more than 20 songs on Defiance. And some of those were original lyrics, but most of those was translating either popular songs that they got the rights to or original lyrics written by the show writer. And that was just a nightmare– an absolute nightmare– because English is so compact, so amazingly compact. You can say so much with so few syllables, and it’s the syllables that are really important, rather than the number of words.

ADAM GREENFIELD: And there’s context, too.

DAVID PETERSON: Yeah. So for example, I had to translate the song “Doll Parts” once, which is a song by Hole. And lyrics of this song is like, “I am doll legs, doll eyes, doll mouth.” Four lines, each one two syllables. And I was translating this into a language which is, first of all, SOV, so the verb comes at the end. So you can start with something like, I am. And second, the shortest I could get just the title “Doll Parts” was five syllables.

ADAM GREENFIELD: Which language?

DAVID PETERSON: This was Castithan for a show called Defiance. And so there you have to be very, very clever. With that one, I ended up looking at just the entire first two stanzas as a whole and saying, how can I convey this content in roughly the same way? And I just broke it up differently so that by the end of the first stanza, I had said basically, “I doll’s legs, eyes,” and kept going so the verb was at the end of the second stanza. Had to cut out a word. I think I ended up cutting out “legs,” because there was no way to get it super short. And so after two stanzas, it fit the syllable count and mostly meant the same thing. Very, very difficult.

ADAM GREENFIELD: So when you explained that to them, were they adamant about just making sure that it worked, or were they?

DAVID PETERSON: Yeah. I mean, there were two different levels there. So the one who was writing the lyrics was not the songwriter, right? Or the one who requested. I know, obviously, the lyrics– it would be the show runner. So he just says, let’s do a version of this and translate it. So he just cares that the meaning is the same.

Then I had a separate working relationship with Brendan McCreary, who was the one who actually ended up doing the songs. He was great, because he could pronounce my languages very, very well, so that was nice. But also, he was good with working with what I gave him.

So I would give him my translation, and then I would say, here it is. If you have to start cutting syllables, these are the ones to go first. If you have to do even more, these are the ones you can do if absolutely necessary, but it’s going to sound a little funny. And he would be able to work with it and get something that was still recognizably the language.

He wasn’t cutting too much, but also managed to fit and sound natural. He was a genius, and I think I was very lucky to have him on the team. So it was very much a collaboration.

ADAM GREENFIELD: Sometimes your audience isn’t necessarily part of the world you’re in when it comes to language creation, or even linguistics as a whole. So what are some of the requirements of convincing that audience– whether they’re clients or public or even students– that you really are someone of authority on the topic, and not just some guy who’s throwing sounds and things together?

DAVID PETERSON: There’s two different ways to do this. I mean, first you have to demonstrate mastery of the subject matter, which means that you know all the terminology and you can use it. But for people that don’t know the terminology, it’s just going to sound like gobbledygook. And to a certain extent, they’ll hopefully trust that you know what you’re talking about. But it helps to be able to demonstrate exactly what you’re talking about in a way that somebody who can understand.

Language, I think, is a very– it was a very easy field to do that in, because whatever you’re saying about language data, there’s always the data. You can always just write down some words, pronounce them, and say, this is what’s happening. This one’s changing like this– using very, very simple terminology. And it’s like, well, why is it changing? Well, it’s changing to make it a little easier to pronounce, just like we change our T’s and D’s in the middle of words to something that sounds like a D.

So we don’t say something like– I always have trouble coming up with an example off the top of my head, and “matador’s” a bad example, because that comes from Spanish. Like “little,” for example. We don’t go around saying “lit-tle” with a very clear and obvious T. It just gets reduced a little bit to something that’s D-like to make it easier to pronounce. “Little.”

That’s kind of what’s happening with this example that I’m showing you. It’s getting easier to pronounce, and sometimes sounds change like that. So you can work with it at that level to make sure that they just understand what’s happening. They don’t necessarily need to understand, in a very complicated way, why it’s happening or why, for example, one is easier than the other, or what other options there were, or why it doesn’t happen or when it doesn’t happen– like why the T changes in “little” bit it doesn’t change in “photography.”

They don’t need to understand that, but they need to understand that there’s something going on there. And I think, especially with language, it’s easy to do, because the data is always going to be– here’s a word, or here’s a sentence, or here’s a sound. And since we’re both human beings, we can pronounce them and figure out what’s going on with it.

ADAM GREENFIELD: And the understanding of what you’re trying to say is still coming across.


ADAM GREENFIELD: We interviewed a linguistics professor at MIT, and we talked a little–


ADAM GREENFIELD: What was his name? Ted Gibson. And we got into context. I was an English major, and there are certain things that, like I guess a lot of people in society, bug me. Like Oxford comma– when that’s not being used. And he told me the story about– or he was telling us a story about a guy who was like, apostrophe’s aren’t necessary because context is still there. People are going to understand it either way. It might take them a little longer, but they’re going to get it.

So when you were talking about that, that reminded me of that conversation– that context is still going to be there. But when it comes to this technical stuff, do you think that things can get lost in context or translation if you’re not being specific about not only just your writing, but your speaking as well?

DAVID PETERSON: Yeah, so I’m trying to think about how to respond to this. You think you can re-ask this question? I’m trying to think. So give me a scenario.

ADAM GREENFIELD: Sure. So if I’m contracting “that is” to “that’s–”

DAVID PETERSON: Oh, yeah, no, no, no, I know that.

ADAM GREENFIELD: Yeah, but if I don’t use that apostrophe, people are still going to understand. In memes you see things grammatically incorrect, but you still get it.

DAVID PETERSON: Yeah, absolutely.

ADAM GREENFIELD: Do you think the context can ever get lost if you’re not speaking effect– [INAUDIBLE] communicating?

DAVID PETERSON: Yeah, one of the reasons that we have these things like apostrophes which you don’t hear in speech is we use them for disambiguation. So there’s always going to be things that are ambiguous. And so since writing is a static medium and you have all the time in the world to do what you want there, we employ these simply for the purpose of disambiguation, because things can be ambiguous when you speak, since it’s not there. And so I think there is definitely that potential there.

It’s hard to– especially on-the-fly– construct the perfect example of ambiguity, but it absolutely can happen where you could have two people coming away with two totally opposite interpretations of what you’re saying. I think it’s helpful, especially, to know where those problem situations can emerge so that you can be aware of them beforehand and make it absolutely clear what it is you’re talking about.

It is also, of course, always helpful when people listening ask questions, because that’s the best indicator that something you said wasn’t clear. But yeah, there are certain cases, especially– so responding to things that you earlier said, where if we got rid of apostrophes, well, context will always help you to determine the difference. If you add an infinite amount of time in there, I will agree.

However, if all you’ve got is an hour, and you said something that is slightly ambiguous or vague, and then the hour ends and there’s no possible way to disambiguate what was said, then it remains forever ambiguous– until maybe another hour comes along later and you can clear it up. So it’s something that I think that one should be mindful of– as mindful of as possible ahead of time– so that it’s maximally clear.

ADAM GREENFIELD: OK, yeah, because sometimes, at the end of that hour, they won’t have– just as you said– an opportunity to clear that up, so. So when you create a language, how much of the utility– or, I guess, usefulness– of that language comes into play as you’re parsing everything out? So speaking of the ambiguity part, are you allowing for some of that when you create a language?

DAVID PETERSON: When I create a naturalistic language, it should be indistinguishable from a natural language that occurs on Earth. And that means that it will function fully– so it should be perfectly usable for translation. But it also means that it has all of the irregularities, all of the vagueness, and all of the potential for ambiguity that exists in any natural language. If you’ve done anything else, then you haven’t created a naturalistic language. You’ve created something that’s rather fake.

ADAM GREENFIELD: OK, yeah. What about slang? Do you ever throw that in?

DAVID PETERSON: That’s more of an issue for what we are able to represent on screen. So if you have one person and this one person is speaking the language and they only speak it in one type of context, then you really only get one type of language. If you have a bunch of different types of speakers who come from different places or different backgrounds, then that’s when you have the opportunity to do different registers.

So one of the things that I had a lot of fun with was, in Defiance, we had an age group of people that come from a different planet and were born there and raised their whole lives there, and came to Earth and learned English as a second language. And then you had their offspring– were raised bilingually at best, sometimes maybe as heritage speakers with English as their dominant language. And so they spoke their home languages differently from their parents.

And in fact, with traits that were definitely influenced by English– which, of course never would have happened with those languages back on the home planet. And so when I was doing a speech for the younger generation, I could demonstrate that. And that was really cool, and it was a lot of fun.

And also, on something like Game of Thrones, we’ve had some fun with people who are non-native speakers of a language who don’t always get everything right, and so that’s another little registry you can do. So when I’m working for television shows and movies, I of course am always thinking about it, but I’m able to do it to the extent that the script calls for it. And then in that case, it’s just totally on me to represent it.

ADAM GREENFIELD: OK. So to flip the script some, when you’re part of the audience, what are some things that a speaker or writer can do when it comes to the language they’re using that can distract you from the message or topic?

DAVID PETERSON: The more tokens you have for a language– and I guess this is more for audio than visual, like looking at a book. The more tokens you have– a token being just a sentence– the more likely it is that the casual viewer will pick up on inconsistencies. And not inconsistencies like, wait a minute, you were supposed to use the dative with that post-position, and in this one, you used the allative– no, nothing like that; nothing like that.

But just the fact that if you hear five sentences of Russian spoken throughout the course of a movie, you can tell that it’s Russian even if you’ve never studied the language. And it’d be really weird if the fifth sentence, some new actor comes out and is saying–


And it’s like, that’s not Russian. I don’t speak Russian. I don’t know a word of Russian, but that’s not Russian. So you can tell. So the more sentences you have, the more opportunities to hear the language you have in a movie or in an episode of a show, the more likely it is going to be that you can pick up on the inconsistency of sound. And so that’s what, as a listener, I hear– the inconsistency from sentence to sentence and speaker to speaker, and also just the fluidity.

And that’s something that– it’s easier to be fluid with a language that works like a language than with just a bunch of random words, because languages will have repetitions of the same types of consonant clusters to begin a word and consonant clusters or consonants to end a word. And even with something like English, where you think about, well, English can have anything to end a word. This is absolutely true.

But in a random sentence, certain codas– that is, certain consonants that end a syllable and end a word– are going to pop up more than others, because they’re more common. And so you hear that and internalize it, and it helps to build a rhythm for the language; as well as, of course, the intonation for the language in general, which should be consistent from sentence after sentence after sentence.

And this is something that, typically, those who are constructing gibberish and saying it’s a language– they don’t think about this. They’re just like, whatever. Just put that in there. And then the actors treat it very inauthentically so that they’re speaking their English lines with their acting English voice and doing very well, and then suddenly they come up to this created line, and they’re saying–


I don’t even know what that was, but that sounded awful. It sounded fake. And that’s what I think that I am looking for, just even as a casual viewer. It just needs to sound right. It can’t sound obviously fake. It has to sound authentic.

ADAM GREENFIELD: Do you think things like– I heard you a couple years back on the Nerdist episode, and you were talking with Jonah Ray about the Pidgin English that they speak there.

DAVID PETERSON: Hawaiian Creole English, yeah.

ADAM GREENFIELD: Yeah. Does that distract you, knowing that– or do you consider it, even– I don’t want to say lazy, because I think that’s a bad word for it. But is that a distraction for you when you know that it’s a bastardized version of the native language, I guess?

DAVID PETERSON: I’m going to try to answer this question in less than an hour.

ADAM GREENFIELD: Did I open up a can of worms, there?

DAVID PETERSON: Little bit. First of all, the state of Hawaiian Creole English at present is basically just a dialect of English, which means that it has the exact same status of any dialect of English. It’s not better than any others It’s not worse than any others. It’s simply English. The earliest forms of Hawaiian Pidgin English were very much more a Pidgin in that they did include a lot of Hawaiian words, and there was some influence of Hawaiian grammar.

That influence has been watered down a lot over the centuries– a couple centuries– to the point where I would say that HC is pretty much just a dialect of English at this point, but very much a proper and appropriate dialect of English for those that speak it natively. So there’s no connection, or there’s no real direct connection any longer between HC and the Hawaiian language proper– just a lot of vocabulary.

Pidgin languages themselves are called Pidgins because they lack the consistency of a language, and also just the range of expressivity and the vocabulary. So that means that if it’s at the stage of a Pidgin, one person may be speaking it one way. Another person may be speaking it another way, and it hasn’t really solidified yet. The stage of a Pidgin happens at the very earliest stage, when– and it happened a lot, historically– when people who spoke many different languages were brought to a location and they didn’t share a language in common.

And at the same time, there was often a power imbalance. So there were people that spoke one language who were in charge, and then people who weren’t in charge who spoke many different languages. And the result of that was a Pidgin, which was the only language they shared in common was the overseer’s language. But they were never taught it– in fact, specifically were not taught it. And so they just picked up whatever they could as best they could and formed a makeshift language. But that’s a Pidgin.

A Creole is something different. A Creole is when a Pidgin essentially solidifies. It solidifies and becomes a full language that has a very consistent grammar. In other words, a Pidgin you can’t really speak wrong as long as you’re using the same words. A Creole you absolutely can speak wrong, so that somebody can tell you, no, that’s not how you say that.

Probably the most famous and most robust Creole spoken in the world today is Tok Pisin, which is one of the official languages– and probably the main language for most people– in Papua New Guinea. And Tok Pisin is something that evolved in precisely this way over on those islands. English was used as this lexifier language, but its grammar is entirely distinct from English.

And so even though you can look at a sentence of Tok Pisin and recognize– well, that word came from this English word. That word came from this English word. They don’t mean the same thing. The grammar is entirely different. The grammar is consistent, and it’s basically– yeah, children learn it, children use it, and it has the full range of expression.

There’s an entire literature in Tok Pisin, in addition to radio programs, news programs, other TV programs, and all that. So at this stage, Tok Pisin as a language and the fact that its vocabulary was derived from English is basically a historical footnote. It would be wrong to call it a bastardized form of English. It just wouldn’t make any sense.

Like I said, at this stage, it’s just historically related to English– in a similar but less significant way than something like Spanish is related to Latin. Spanish clearly evolved from vulgar Latin, but there were a lot of influences from other places. And if you look at the Spanish future tense, for example– the composite future tense– and look at how it’s formed and see what its history was, a Latin speaker– a pure Latin speaker from the days of the Roman Empire– will look at that and say, well, that’s just bad Latin.

I mean, we have a future tense. Why aren’t you using that future tense instead of this made-up gobbledygook you just did? But that doesn’t mean anything. It certainly doesn’t mean anything to modern Spanish. They’re just two different languages at this point that have a historical connection.

So does it bother me? Far from it. I mean, Pidgin and Creole languages are one of my favorite areas of linguistics. And I was very fortunate. I was able to take a course taught by John McWhorter at UC Berkeley while I was there, and it was amazing. And not only that, Creole languages have been very informative when it comes to demonstrating how languages evolve, because their time depth is such that we can actually go back and study them at the very beginning.

And not only that, we know the sources that they were drawn from, as opposed to if you start going back with Latin, you get to Proto-Indo-European, which we can just guess at, and there are absolutely no written records of. And going further back than that, it’s even greater guesswork. So anyway, that’s about my response to that.

ADAM GREENFIELD: That was good for now. I’m sure you could have talked for another hour on that, so. Just a side comment you brought up– Spanish. I just finally caught up with Orange is the New Black–

DAVID PETERSON: I’m not fully caught up.

ADAM GREENFIELD: Oh, OK. Well, I won’t give anything away, but–


ADAM GREENFIELD: –sometimes when they speak, there’s a mix, and within a sentence, they’ll go from English to Spanish. And I really enjoyed that. I had never heard, really, many people doing that.

DAVID PETERSON: Yeah. . Hear a couple of cool examples of code switching happening in Game of Thrones, but it’s fun. It’s fun when you can do it, and I did all kinds of stuff like that in Defiance. That was great. Loved it.

ADAM GREENFIELD: So when you do that, then, do you ever end up repeat– because you said that there’s a– for “Doll Parts,” or legs, it was a five-syllable word. So I mean, do you ever end up– when you’re code switching like that, do you ever end up repeating a word, or are you able to get through one half of one language and the other half of the other language without mixing?

DAVID PETERSON: Oh, yeah, no, code switching only really works if you said what you said. So I know exactly what you’re thinking of. But for example, let’s say that you started out in English right, and then you say the verb and then you switch to Castithan, it’s not going to make sense to repeat the verb again. That wouldn’t work.

So usually what happens is you can switch, and you just entirely switch to the new grammatical form of the language that you’re switching to without repeating any of the old elements. And so usually it happens when it’s more convenient so that you don’t have a bizarre thing where you’ve said the verb in one language and say it again in a different language. There have actually been a lot of studies on code switching in natural languages that demonstrate that there’s a lot of internal consistency to it across languages. And so it’s a very interesting literature if you have time for it.

ADAM GREENFIELD: All right. Finally, do you have anything– I guess advice for grad students when they’re trying to communicate something to an audience? Are there effective rules, or rules that are very more effective than others?

DAVID PETERSON: Audiences– and I should say, I give talks on language creation all the time– all the time. And most of the time, the audience has absolutely no background in linguistics. And actually, I just knew this already from being an audience member. Audiences know when you’re dumbing things down, and they don’t appreciate it.

So the goal is to not dumb things down, but to recognize the areas where things are going to be complex– so that means both when new terminology has come up and when the concept itself is complicated. And in those circumstances, with the terminology, of course, repetition helps. But also, the will and testament method– so in other words– the reason we say will and testament is one has a background in English, one has a background in French. And back during the days of Norman-French, you needed to use both so that everybody would understand what you were talking about, even though they meant the same thing. So that’s why we say will and testament.

So I do that, too. I start with the formal term, and I usually have it up so people can see it so they know how it’s spelled. But then immediately say what it is in a simpler way and keep referring to it until you can gauge that everybody knows what you’re talking about. But then the best thing is to follow up immediately with a concrete example.

And again, this is going to differ depending on your field and language. It is so wonderful to be able to just have language data that people can work with and look at. And it also helps to start with a language that everybody speaks.

So when I’m around America, I do English. But when I went to Spain and Mexico, I used Spanish examples. And then when I was in India, I used Hindi examples. Just something that– or, in northern India. I guess I’ll have to switch to Tamil when I go to southern India. But it helps so that people say, all right, I know this. And then you say, all right, you know this. Now let me demonstrate something that you don’t know about this. And that’s going to illustrate my point.

So with sound change, this is something that I think a lot of people have no idea about– that sounds change over time. One thing they do know is that English has funky spelling. So it’s wonderful to take an example like knight and night– knight spelled with a K, night spelled without a K, but both of them spelled with a funny GH– which, who knows why? And demonstrate that these words were actually spelled this way because they were pronounced differently at some point in time in the past, and show them how gradually, the sounds changed so that we got from “kuh-nee-ght” to “nite,” but the spelling didn’t. And that is a really easy way to demonstrate both why we have irregular spellings and how sounds change over time.

Same thing with grammatical evolution. That is a crazy complex idea– the idea that grammar subtly and slowly evolves over time. But we have a wonderful example in English with our go future, where everybody knows that the future tense is something like, “I will eat,” or, “I shall eat.”

But we also have a different one with go. And you can show how we would have started off with some sort of prolix expression like, “I go to London to eat,” where you’re saying that actual definition using the verb on its own– the way that it should be used– and then expressing what you’re going to do when you get there. And then you show how you can just, well, take away the destination.

Now you’re saying you’re going for some intention. And then things slowly change, so you say– instead of “I go to eat,” you say, “I am going to eat,” and then “I am going to eat,” then “I am going to eat,” and then “I am gonna eat,” “I’m gonna eat,” “I’m gonna eat,” “I’mma eat.” And you get all the way to the present.

And so you can demonstrate that and people get that, because they know it, without having to go straight into the technical terminology and without showing them am example that might potentially be more interesting. I think it’s tremendously interesting to demonstrate how the cases evolved and merged– the accusative and genitive cases– evolved and merged in Finnish. But I think it’s a little opaque for English-speaking audiences.

So you start with the easy example, then you can move on to great examples, and just basically bring them along with you. And I’ve had a lot of success doing it in precisely that way. So that most of the time, especially what people say at the end– wow, there was a lot of terminology I didn’t understand, but I was able to follow what you were saying, and that’s really cool. And now I get what you’re doing. And I think that’s the best part, because then if they want to know about the details, they can go investigate themselves.

ADAM GREENFIELD: Do you think language is ever going to get to a point where there’s so little words, but people will still be able to understand? Like the, “I am going to eat” part– it’s all of a sudden down to two, three syllables? Where do you see this– where’s this endpoint going to be?

DAVID PETERSON: A lot of people have noticed that. This was George Orwell’s whole thing with his fool novel, 1984. Anyway–


DAVID PETERSON: Nah. But they often don’t recognize the other end of it, which is this– when things start to get too small, that’s when we add more words. So you notice, especially with– you talk about this future tense– the most natural thing to do at this point is not to say “I’mma eat,” but, “I’mma go eat.”

So in other words, it got too short. We just threw in another go. Why not? But yeah, at a certain point, things get too short, and so people feel that they’re not being explicit enough, and so they just add more verbiage. And this is something that, actually, writers complain about every single century.

Where it’s like, why say this? It’s too prolix. Instead you could just say this simply. Well, you spotted something. And there was something else I wanted to say. Oh, yeah, yeah– and the most productive– usually the most productive– language creators, shall we say, in this macro sense– are teenagers, actually. And I don’t think that teenagers are ever going to stop being that way.

They want to distinguish themselves as the new generation, so they say things a little differently. That means shortening certain things and expanding other things and changing the meaning of things. Some of it sticks, some of it doesn’t. But that is absolutely never going to change.

ADAM GREENFIELD: Yeah, I listen to a podcast called A Way with Words. Are you familiar with that one?

DAVID PETERSON: I’m not, actually.

ADAM GREENFIELD: It’s basically these two highly intelligent wordsmiths, I guess. I can’t think of another, better way to explain them. One of them, his name is Grant Barrett, and he’s written dictionaries and works on dictionaries and things like that. And they talk about the etymology of words and sayings and things like that. And recently, there was one episode where they were reciting graffiti from back in medieval times and even further back than that, and it’s that kind of stuff that doesn’t change. How it’s said changes, but it’s always going to be there. I thought that was interesting.

DAVID PETERSON: That’s really cool.

ADAM GREENFIELD: Yeah, that’s all I have. I really appreciate it.

DAVID PETERSON: Yeah, no problem.


This podcast was written and produced by Adam Greenfield. The Executive Producer of this podcast is Patrick Urich. The Great Communicators podcast–

The Great Communicators podcast, Grad Comics live Grad Comics The Game, and the Technically Speaking comic book series are part of a professional development initiative called GradX. GradX is–

GradX is made possible by the Office of Graduate Education at the Massachusetts Institute of Technology. To find out more about–

To find out more about GradX, as well as get access to other episodes of the Great Communicators podcast, go to gradx.mit.edu. For more information–

For more information and links on the music used in this episode, please see the show notes.