6 Questions With Dothraki Creator David J. Peterson

We chatted about creating, deconstructing and learning languages with the conlanger who invented the fictional languages you hear in ‘Game of Thrones.’
Game of Thrones scene

Header Image: HBO

In some ways, the truth about language is stranger than fiction. That is, the process of creating fictional languages can often bring the extreme variability of “real” language into sharp relief. That’s part of what we took away from our conversation with David J. Peterson, fictional language inventor extraordinaire.

Conlangs, or constructed languages that are designed to function like natural languages, are frequently the product of hobbyists who enjoy making up languages for fun. But increasingly, they’re a significant product of our pop culture obsessions (Dothraki and Klingon), and in a few cases, they’re the hopeful projects of visionaries who dream of a language that everyone can understand (Esperanto).

Far from being an exercise in escapism, conlang creation — when done properly — requires a sophisticated understanding of the way natural languages actually work. The more you know about the world’s various languages, says Peterson, the less likely you are to default to specific structures in your native tongue that you often take for granted.

David J. Peterson is the brains behind Dothraki and Valyrian, the complex fictional languages popularized by HBO’s Game of Thrones. He has also created languages for various Syfy and CW shows, Showtime’s Penny Dreadful, NBC’s Emerald City, Marvel’s Thor: The Dark World, and others. He’s the co-founder of the Language Creation Society, a student of more than 20 natural languages, the author of The Art of Language Invention and the producer of the eponymous YouTube channel. He’s also given a TED talk.

In the second installment of our “6 Questions With” series, we spoke with David J. Peterson about his evolution as a language artist, the various linguistic insights he’s gleaned over the years, and what the average language learner can take away from conlangs.

If you’re creating a new language, and there’s no connection to any language in our world, then you don’t just borrow stuff from existing languages. That would be silly. You can certainly be inspired by them, but every single language, if you’re creating a naturalistic language, has its own evolution.

1. How did you get into language creation?

David J. Peterson
David J. Peterson

PETERSON: I created my first language in my first linguistics class. Prior to that, I had taken a number of languages at Berkeley as a freshman, including Esperanto, where I learned that people could create languages. When I started taking my first linguistics course, I was exposed to all of these different ideas about ways that language could work. I came up with the idea of creating my own language for fun; just for my own personal use. Then I kept at it until it ceased being fun, and that hasn’t happened yet. It’s been about 17 years.

I didn’t happen upon any other language creators until probably about half a year in. It was a bit of a rude awakening to learn that there were hundreds, if not thousands, of others who also created languages for fun.

It was another shock to learn that not only was I one of many, but I was also not very good compared to those I was meeting who, of course, had been at it for much longer. Nevertheless, once I learned that, I was able to internalize it, and then start learning other people’s created languages, continuing to study linguistics, continuing to study languages on my own, and continuing to create my own languages. Bit by bit, I got better and better, I guess until I was where I am today. I think that I’m still improving.

2. Why do you think people enjoy creating languages?

PETERSON: Honestly, at the base of it, I think it’s just that people enjoy the activity. It’s like putting together a puzzle, and also building the puzzle pieces that make the puzzle. It’s very difficult. It’s very taxing. It’s the type of thing where, as you can imagine, if you were trying to create all the puzzle pieces by hand, you might find that they don’t fit the way you thought they were going to.

Maybe the image is right, but one of the little tabs is too big for where it’s supposed to insert to. That’s the type of thing that happens with language creation. You have a plan. You sit down. You start doing it. Maybe it works. Maybe it doesn’t. There may be some bit that does work very well, ends up obscuring another bit or making it not work as well. It’s a call for endless tinkering to make sure that the system you have works right.

Then, of course, even if you’ve got the system running pretty well, creating an entire vocabulary is something that can literally last your entire lifetime. No project is ever finished. Unlike with a lot of other art forms, there’s never anything you produce that can’t stand to be improved further. It’s endless, and that’s part of the draw, I think.

We’re gradually getting to a point where every project that could benefit from a created language is thinking to have something in there.

3. Going hand in hand with that, what do you think is behind the growing impetus for movie and television creators to invest in legitimate alternate languages? Why not just make up some convincing-sounding gibberish? Do you think the authenticity of Dothraki accounts at all for its cult-like appreciation?

PETERSON: Part of it was the push for extreme realism, that I think we’re on the tail end of, if I’m being honest. It’s the type of thing that happens every so often in Hollywood, simply due to advancement in technology, and especially visual effects. It’s like, you saw a lot of this in the late ’70s and ’80s. Once they got really good at creating monsters and doing things with robotics, the old monster movies of the ’50s and ’60s, where you just had guys in costumes, looked like garbage compared to “Jaws,” which was terrifying. It blew everybody away.

By the time you got to 2000, visual effects had gotten so good that you could actually return to doing fantasy movies like Lord of the Rings and not have them look campy. Of course, one of the things with Lord of the Rings is that they employed Tolkien’s languages as best they could at the time. They were very important to Tolkien, and they wanted to make sure they were in the movies.

When you have a huge benchmark like that, something that has a big impact, everybody wants to emulate every single bit of it, so that came to the special effects. It also came to the languages. By the time it started to hit TV, with Game of Thrones, they just thought, “Well, we’ve got to have a language for this thing.”

The Dothraki themselves certainly capture the imagination of a lot of people, and partly because of the actors. Jason Momoa is a very attractive man. He’s huge. He delivers things in a really good way, both believable, but that actually make him sound cool, which might not be something that’s ever happened, I think, with a created language before.

To a certain extent, the Klingons, they’re very, very cool. Then there was the Na’vi, of course, kind of goofy, in a way that aliens in an alien show are always going to be a little bit goofy. Whereas with the Dothraki, they’re basically realistic human beings. They’re just very, very buff, so I think that’s part of it. Certainly when it comes to created languages, I think Esperanto is an outlier. Outside of that, the popularity of a language is basically directly proportional to how popular its associated media is. It has very little to do with the language itself.

Frankly, if there was somebody else who had created Dothraki who didn’t know the first thing about created languages, it probably would have been just as successful. We’re just not there in terms of the general public’s knowledge of both language creation and language, in general.

With all of that said, we’re gradually getting to a point where every project that could benefit from a created language is thinking to have something in there.

4. How do you decide which aspects of a language to borrow from and mimic when you set out to create a new language? Do you borrow the syntax of Greek but throw in Hungarian verb tenses and Japanese suffixes, or is it not quite like that?

PETERSON: We don’t borrow anything from a language unless it’s an a posteriori language. For example, there’s a language that I created for The 100. It’s supposed to be a future incarnation of the English language. Consequently, all the words come from English. This is done because it’s appropriate. That’s exactly what the language is supposed to be. It’s not an accident, or it’s not like I just borrowed a different language and am pretending it’s something else.

That would really be inappropriate and honestly, it wouldn’t be creating a language. You know, conlangers often start out like that, but it’s a phase, and not a great one. If you’re creating a new language, and there’s no connection to any language in our world, then you don’t just borrow stuff from existing languages. That would be silly. You can certainly be inspired by them, but every single language, if you’re creating a naturalistic language, has its own evolution.

This is why you’ve got hundreds of languages on the planet that all have a word for “the.” But in no two languages do they work exactly the same. Even similar or very closely related languages like English and German — we both have words for “the,” but they work very differently.

Just in small little things where like, I don’t know if it’s still a type of expression that you can use or if it’s popular with younger people, but at least at one point in time in German, you could say something like, “Wo ist der Gabby?” Where it would literally translate to “Where is the Gabby?” You could say it with people’s names like that, just as a colloquialism to be friendly, I guess.

Whereas you can’t really do that in English. There’s no context where you could say, “Where is the Jason?” You’d have to be making a really stilted, really specific joke in order for anybody to make sense of that at all. It would have to like an in-joke with your community. It’s not something you do with English.

That’s the type of thing I’m talking about. If you just ripped out a structure from one language and put it in your own, the result would be inauthentic. You wouldn’t have earned that result, I guess. You wouldn’t have evolved it. You’d end up with kind of some Frankenstein language. It’s something that can fulfill a lot of the purposes of a language, but if you look at it, it looks totally fake, like a monstrosity.

The more languages you study, the more you know about how languages can vary, the less likely you are to unconsciously assume that a language must always work in some specific way.

5. How many languages would someone have to learn in order to begin inventing new ones? Or is that not really a barometer?

PETERSON: It’s certainly not necessary. You can start creating a language whenever you want, no matter how many languages you’ve studied or spoken, no matter how many linguistics classes you’ve had. The only thing is that you run into cases where you will do things because you cannot imagine a language working any other way. The more languages you study, the more you know about how languages can vary, the less likely you are to unconsciously assume that a language must always work in some specific way.

The more you study, the better. The more linguistics you study, the better. Even just surveying languages, the better you’ll be, and the less likely you will be to replicate some of the unconsciously natural features of your own language. It’s certainly fine to have a structure in your language that’s the same as the language you speak, but ideally, you’d like it to be that way because you decided to make it that way. It was an artistic decision, ideally.

There are things that you never think about. In English, you can say things like, “I want him to do that.” You can’t say “I hate him to do that.” The thing is that’s just a really weird, little quirk of English. It’s an unconscious fact that all English speakers have in their head. You just know that if you say that, you have to insert “for,” for some reason.

6. What have conlangs taught you about the way humans learn languages? Do you feel like you have any takeaways or advice that you could offer to people who are studying a second or third tongue?

PETERSON: If you’re studying a language and really want to be able to use it, you’ve got to have practice just saying stuff.

It’s the type of thing that seems silly, and maybe even a little bit frustrating at the beginning of a language class, where the teacher goes around to everybody and says, “Hola. Me llamo Profesor Profiero. ¿Cómo te llamas?” And then you have to respond, “Me llamo David.” It seems so silly by the time you get to the fifth person, and everybody knows exactly what the professor’s saying and exactly how to respond. But just doing it is helpful, because it’s there for you when you’re ready to respond.

Because often, when you’re speaking a language, part of what you’re saying are things that you’ve said many, many times before. And then part of what you’re saying is brand new, given the context. It’s like the nouns and particular verbs you need to bring up are new, because of what you’re talking about. But you’re used to the little fixed expressions, so you don’t have to focus on that part. You can focus on the parts that are new.

If you don’t get that practice in, everything is going to be new to you and everything is going to be difficult. You’re going to be thinking “what should the intonation be?” “How should my intonation be if I’m setting up this phrase?” “How should I say it exactly if I’m going to follow it with a ‘but’ phrase, and say something else?” If you just have practice with that, you don’t need to do that.

It also helps me to speak to myself a lot. Even quietly, just sounding things out. You do a lot of written exercises usually. It just helps to say them out loud, just to get that feel of it in your head. For me, it helps with speaking.

Otherwise, for just regular study, just being sure that you’re at it, that you do something with it every single day is more important. There’s no real trick to it. Keep at it. It’s better to study one hour a day for five days a week than five hours a day for one day a week, so that you’re constantly hearing it, you’re constantly interacting with it. You’re constantly thinking about it, trying to get that language into your head. That helps me more than anything else.

This interview has been edited for length and clarity. Click here to read the full transcript.

Try these insights for yourself.
Learn A New Language