It is with great pleasure that I present my interview with David J. Peterson! For those unfamiliar with him, David is the go-to person in Hollywood for bringing fictional languages to life on screen. His work has graced iconic series such as Game of Thrones and the recent movie adaptation of Dune. In this interview, we delve into the details of developing languages for these universes and the challenges that come with it. Join us in unraveling the fascinating details behind the worlds we love!
Here, I extend my gratitude to David for his kindness and availability.
Allison: Thank you for being here! Can’t wait to get started, so I will fire away with some of my burning questions! I know that you didn’t create a Valyrian script before Game of Thrones, but you eventually did for House Of The Dragon. What made you decide on something that was similar to Egyptian hieroglyphs instead of other types of glyphs?
David: It was partly inspired by its description in the books; that’s what started it. Also, it came from the way people who don’t know much about language discuss it. In one of the books, there is a discussion of Arya finding a boat to take over to Braavos and describing it as having “Valyrian glyphs” on the side. Now, technically, any character in any writing system is a glyph, but for those not doing specialized work with writing systems, they often think of it as: a writing system like Chinese has glyphs, but a writing system like Greek has letters. I bore that in mind. That is what George R. R. Martin probably meant when he wrote that — the writing system was different from an alphabet. Then it became clear when Tyrion is walking through a marketplace, and he encounters a Red Priest who is tracing “Valyrian glyphs” in the air. What that told me was that the writing system had logographs: single glyphs that stood for words or concepts as opposed to just sounds.
In terms of its execution, Valyrian was meant to evoke ancient Rome, and so I thought, what if there was a script that actually looked a bit like the Roman script? But instead of functioning like an alphabet, it functioned like a logograph. It also made sense to do something that was closer in structure to Egyptian as opposed to something like Chinese. Chinese, for the most part, is an isolating language which means that it doesn’t have a lot of inflection. High Valyrian, on the other hand, does. So I thought it would work really well if there was a basic glyph that stood for a concept and then there were these extra glyphs to spell stuff out that did the inflections, and so that was where the whole thing came from.
Allison: I am Chinese, so I know what it’s like to learn all the characters from scratch. I am so glad I did that as a kid because I can’t even begin to fathom trying to learn from the very beginning as an adult. I type so much nowadays that I sometimes have difficulty remembering how to write certain characters when I put pen to paper. But I love that you added a little twist to Valyrian in terms of the writing system and made it less difficult in some ways.
David: I never tried to learn Chinese, but I did try to learn Japanese, and I had a similar experience with Kanji as well. I know what certain glyphs look like very well and I can recognize them and read them, but then I try to write, and it’s like “I think something’s up here…I think something’s down here.” So it’s really important to write by hand to achieve that kind of literacy that you have.
Here you can find an entire dictionary of Valyrian glyphs created by David.
Allison: For Dothraki, you were asked to create something that sounded “harsh” and “foreign” to English speakers. I’m wondering if you had similar sound guidelines given to you for Valyrian?
David: No. And I also just ignored the advice for Dothraki. I really just based it on what was in the books. There was enough Dothraki material in there to come up with a really consistent sound system, so I just followed that, and it worked out really well. It worked out mostly well for High Valyrian. There was a lot less material in the books, and the material was less consistent.
For Dothraki, every name was full-on Dothraki, and the Dothraki culture was also insular enough. With Valyrian, it was really difficult to pinpoint names that come from High Valyrian. It could go something like: this is a name that comes from one of the daughter languages of Valyrian; this is a name that was specifically influenced by another culture; this is a Valyrian-esque name that was filtered through the Common Tongue…For example, Daenerys is High Valyrian, for sure. But Baratheon is not High Valyrian, and it’s not clear if it came from a daughter language or if a Valyrian ending was applied to a Common Tongue name, which is something that absolutely could have happened. The Targaryens dominated Westerosi culture for so long, so it’s very likely that there was a Common Tongue name “Barath” who tried to worm their way into higher culture, so they changed it to “Baratheon” to make it sound a bit more Valyrian. That made it very difficult for me to say “this is absolutely High Valyrian” or “this is not.”
Allison: I’m interested in how the verb system works in Dothraki. From looking around on the Internet, it seemed a lot simpler than Valyrian, which is kind of like Latin and that was nightmarish for me in middle school. So how did you decide on things like inflections for Dothraki?
David: When it comes to inflection, there are languages like Chinese where there’s absolutely no agreement between the subject and the verb. One step up from that is English. There is incredibly minimal agreement between the subject and the verb. In the present tense, you know if the subject is third person singular or not, and that’s it, so you still have to list the subject. One step up from that are languages like German and French. There is a lot more agreement but not enough to be able to drop the subject. So there are more verb endings in German and French than there are in English, but you still always have to say the subject. That was really where I stopped for Dothraki because I thought it made sense — especially when I shifted the language from a verb-initial language to a subject-verb-object language. To make it like so, the agreement would be a suffix, and the subject would be following that suffix. It made a lot more sense to move it out front as to avoid redundancy. It doesn’t go as far as High Valyrian, which means that in Dothraki you always have to specify the subject.
The inflection in a language often comes from the shortening of the pronoun being glued onto the end of something. For example, with a language like Finnish, it just wears its inflections on its sleeve: the word for “we” in Finnish is me, and “to speak” is puhum, so “we speak” becomes puhumme — you just throw it on the end. If you were to say puhumme me, that would be redundant. The idea with Dothraki was that the redundancy would end up pushing the subject in front of the verb.
Allison: Is it harder to work on languages like Dothraki and High Valyrian where you have some guidelines based on the source material? Compared to if you were to make something up completely on your own?
David: No, it’s not harder or easier; it’s just different. There’s less that I have to invent, but it also means that I have less control over it. So it’s slightly easier, slightly harder too maybe, but it really is just different.
Allison: Dune was one of your more recent projects. There is a lot of discussion on the languages used during the Sardaukar army scene; could you talk more about that?
David: Yeah. The priest’s chanting was something that I created, but it was just gibberish; it wasn’t a full language. The Harkonnen “battle language” was English but warped by the sound team. I didn’t have any involvement in that. They never asked me about it.
Allison: What about the Fremen language?
David: The theory was that the Fremen spoke Arabic but 10,000 years into the future, and sometimes the words are the same, sometimes they are totally different. It’s like saying 10,000 years in the future, our language doesn’t change at all except when it does change, it changes completely — like the phrase “I love you” now somehow means “a mouse ran across the field.” There is just no possible way that could work.
It’s not as if anything should be traced back to Arabic at all because the time difference is 10,000 years, and the time difference between now and Proto-Indo-European is 6,000 years and there is absolutely nothing recognizable between modern English and Proto-Indo-European. So what I did was I just made it evocative of Arabic. I used the sound system of Arabic but simplified in a way that makes sense for all the words and names you see in the book. The writing system was something I had a lot of fun with. It was meant to evoke the aesthetic of 70s novel covers, which were cool and at the same time bore a surreal resemblance to certain key features of the Arabic script. Then I just tried my best to rescue some of the actual Arabic terms.
Thank you for being on this journey with us! For more of David’s works, check out his blog here.
Disclaimer: This transcript has been modified for clarity and conciseness. While I have made every effort to accurately represent the content and context of the original interview, some edits have been made to improve readability and understanding. These edits do not alter the substance or meaning of the questions or responses. I have made these changes to enhance the transcript’s clarity without distorting the original message.
