Slashdot Mirror
Interviews: Ask David Peterson About Inventing Languages
samzenpus writes: David J. Peterson is a language creator and author. He created the Dothraki and Valyrian languages for HBO's Game of Thrones, and more recently has created languages for the CW's The 100 and MTV's The Shannara Chronicles. His new book, The Art of Language Invention, details how to create a new language from scratch, and goes over some of the specific choices he made in creating the languages for Game of Thrones and Syfy's Defiance. David has agreed to give us some of his time to answer any questions you may have. As usual, ask as many as you'd like, but please, one question per post.
Have you considered creating a thread-safe language which avoids buffer overflows? Do you think this would be made easier by making whitespace significant?
What do you feel helps make a language feel natural (as though it was created and evolved by a culture) rather than something created more synthetically for a work of fiction?
Ashdorv sevurish dunutri figgisbu. Codfikow shabto jernyok fludshee?
I guess Americans have an advantage in that there's very little likelihood that they'll unconsciously put bits of an existing language in.
Confucius say, "Find worm in apple - bad. Find half a worm - worse."
qatlh Hol chenmoH, chonayta' chenmoHta' neHqu'.
Silence is a state of mime.
Most people never bother to give a fictional language a second look, they only happen to listen to the way it sounds, in passing. This makes the vocabulary and the distribution of letters / letter combinations the most important part. What, then, is the point of working on the grammar and syntax of a synthetic language, rather than using simplistic ones? Is it for the benefit of the language geeks out there, is it art for art's sake, or does it affect our perception of that language in ways we don't necessarily see?
How do you decide to what extent a language will have less polite/more naughty aspects, and how do you decide what they are?
Almost every language will have some kind of swearing, or double-entendres, or other aspects which aren't purely syntactic.
Obviously a warrior race is going to have much more bawdy aspects to their language than a race of monks.
Surely delivering insults can be as integral to a language as merely conveying an idea.
Lost at C:>. Found at C.
Have you ever worked on a project like a book/app/website/parallel reality game where the reader/player's role would be to decipher one of your synthetic languages, given clues? Does that sound like something you would consider doing?
With thought to thought communication between human beings achieved in 2014 for the very first time, how long do you think it will take before a psudo telepathic language is created between thoughts and have you given any consideration to creating the basis for one of your own when we achieve the widespread technology to utilize such new forms of communication?
Mirab, with sails unfurled.
Temba, his arms open.
Temba at rest.
Isn't it kind of perverse to create languages that people don't understand when the purpose of language is communication?
...but I just have to marvel at the degree of specialization in advanced economies such that "fake language designer" is actually a viable career possibility.
what's the point?
There can be intense debates about the merits and flaws of one computer language versus another. Some languages have tried to be able to do everything and they usually don't catch on. (PL1 might be the first example.)
Natural human languages are not, for the most part, designed, though grammarians may sometimes try to 'fix' them a bit. But they have flaws. The easiest things to point out are the ambiguities and redundancies. (Some redundancy might be a good thing, allowing a listener to guess at meaning when a speaker isn't heard perfectly.)
Do you deliberately put flaws in languages or, on the other hand, try to design 'ideal' languages that are somehow better than the naturally evolved ones?
In theory, theory and practice are the same; in practice they're different. (Yogi Berra & A. Einstein)
Languages are affected by the cultures they are used in. I think this is mostly a matter of vocabulary. In Japanese for instance, you would use a different word for 'brother' if it was your own brother as opposed to someone else's brother. In fact there are different words for older and younger brother. That says something about Japanese culture. Do you incorporate things in your languages that specifically reflect the cultures involved?
In theory, theory and practice are the same; in practice they're different. (Yogi Berra & A. Einstein)
Any thoughts on Esperanto (International), Lojban (Semantic), Solresol (Representative of French), Sindarin (Tolkien), Drow (Dungeons and Dragons), and Klingon (Star Trek)?
I've noticed Esperanto seems to produce propaedeutic effects by either loading quickly (it's *fast* to learn) or directing more attention to the analysis of a language's structure (by nature, it encourages the student to do this). It's a very structured language, in terms of word construction.
Lojban is supposed to be unambiguous; I think Esperanto achieves that exactly as well, due to its grammatical structure, in so much that Lojban is *semantically* unambiguous (we know what in the sentence represents the subject, verb, direct object, adjective, adverb, etc.) but can be *conceptually* ambiguous. Your thoughts?
This leads to things like Solresol, Sindarin, and Klingon. They all seem to have a point: Solresol encodes French to music; Sindarin is supposed to "sound pretty"; and Klingon is supposed to sound harsh. How do people come up with this kind of thing? Is that even a valid concept? Is there any interesting aspect of these sorts of languages which I should consider, or are they just as essentially bland as any other?
Support my political activism on Patreon.
Why do all science fiction shows have to have rubber-faced characters? Hasn't that become horribly derivative? What difference does a language make when it will just be spoken by hokey aliens?
On a Reddit AMA you said that you don't think any constructed International Auxiliary Language has a good shot of becoming a world language "for a million reasons that have nothing to do with language." I would be interested to hear about some of those reasons.
Certain crated languages enjoy at least a limited amount of usage outside of their original contexts, with two obvious examples being Klingon and Tolkien's family of elvish languages. Do your foresee any of the languages you've created thus far being used outside of their respective TV series?
"Can't you see that everyone is buying station wagons?"
Can you design a moo language used by antropomorphic cows?
So THAT's how the Pentacostals come up with all their fake "speaking in tongues" languages ... no wonder they're always looking like they're having fits.
"Transparent" is a shit show that trades on every stereotype going. A man in drag is NOT a transsexual.
It sometimes bothers me that in the movies, people hardly ever make any grammar mistakes. Not even children. And when they do, it usually sounds artificial. Apparently, speaking like an ordinary person does is even harder to imitate than drunkenness. Now our obsession with grammatical correctness is certainly a very recent development in the history of the human species. I doubt very much that ordinary Roman citizens, or ancient Greeks, let alone Egyptians or Babylonians, ever mocked or corrected each other's grammar. I'd rather think that when people understood what you meant, your grammar was considered correct, so to speak. (Actually it wasn't considered at all.) Do the artificial languages you create, when they are spoken in fictional communities more archaic than our own, allow for more realism with respect to how people actually speak in their daily lives?
In your opinion, is any language as ugly as Java?
i'll just leave this here.
I wonder what you think of "the Language" developed by China Mieville in Embassytown. He doesn't create a language in the way you have done (and describe in your book) with the intricacies of grammar, syntax, and vocabulary, but describes it as requiring the speaking of two different words simultaneously. Would it be possible to flesh out a language based on his description, and if so, do you feel it would be worth the effort?
Another purpose of language is identification and differentiation. Language serves a purpose even to Robinson Crusoe. On a cultural level, matters are quite similar: Same language, same culture. Different language, different culture. Moreover, there are words in my own mother tongue whose meaning I don't know. Probably even words that I don't know. What makes the French the French? First and foremost, their language, doesn't it?
Tom Scott, in a Numberphile Video, points out that even compared to the sort of variation you get with how human languages treat numbers, invented languages like Klingon tend to be rather bland - for example, an English-like base-10 number system with different words.
When creating alien languages and fantasy languages which are definitely not supposed to be like English, how do you get the requisite amount of "otherness" to the language? How difficult is it to make a truly "foreign" language, and not simply get the linguistic equivalent of rubber-forehead aliens? ... or is that typically not what you're after?
Can we fix English?
One of the things which distinguishes Tolkien languages from most other fictional languages is that they have a history. Tolkien didn't just construct some languages, he also developed dialects and plausible etymology relationships. Did you try to take this into account in any of your languages, and if so, how successful do you think it was?
sub f{($f)=@_;print"$f(q{$f});";}f(q{sub f{($f)=@_;print"$f(q{$f});";}f});
The spoken language you obviously invent, but what about the alphabet, grammar, tense (past, future), rules, pronunciation, etc?
And if yes, do you use any known language as a template, or try to invent the written language from scratch?
Ever played around with languages that aren't intended for interaction between humans or human-like entities? Programming/deterministic languages being an obvious example, but some other possibilities being human to machine, machine to human, bidirectional human/machine, machine to machine, and vastly more alien actors as well (complex chemical signaling between plants for example)?
Even better, inconsistencies that can be traced to an earlier form of the language. For example, Spanish has some verbs where e becomes ie when accented and others where it doesn't. The verbs where ie appears are verbs that had a long e in Latin, and those where it does not had a short e.
o1D 7EÃ 812#P7E2 ezD1+ w#yN1 `C 8zN`B1FÃZ nÃyN 9r#Ã 1`N 2eV5%Ã weV7NÃ y71Gx%P `C jx#P`Mx#ÃÃ ?
-Dave
Lojban is supposed to be unambiguous; I think Esperanto achieves that exactly as well, due to its grammatical structure
Justin Rye has some choice words about Esperanto, including plenty of ambiguities and latent biases.
This leads to things like Solresol, Sindarin, and Klingon.
One of which is much older than the other two, which is important for the following reason: Tolkien's Elvish languages (Quenya, Sindarin), the Klingon languages of Star Trek (Klingonaase and tlhIngan Hol), and the language of the Drow in Forgotten Realms each exist to serve works associated with one copyright owner. Would a language be considered an uncopyrightable "useful artifact", much like a style of clothing? Or would works written in a language or dictionaries for the language be considered derivative of the work in which the language was presented? To what extent is this affected by WB v. RDR , about The Harry Potter Lexicon?
Sindarin is supposed to "sound pretty"; and Klingon is supposed to sound harsh. How do people come up with this kind of thing?
Quenya and Sindarin sound pretty to the extent that Finnish and Welsh sound pretty respectively, because those are the languages whose phonologies Tolkien nicked. Likewise, Okrand nicked Tlingit's phonology for Klingon because he wanted something that sounded unfamiliar to an English-speaking audience.
and 5% of the profits of this production will be donated to further these efforts
Productions lose money. Studios make money by overcharging productions for distribution and promotion.
On a related note, I've tried coming up with English words for hexadecimal, with words for ten through fifteen starting with A through F, to represent a culture that counts in base sixteen. This way I can represent the convention of translating the viewpoint character's language while still retaining some of the local color. (A base 20 culture would use "score" notation.) These are what I have so far:
Any better suggestions?
Hadjo fdso vda alft dege fd?
I first read "An Introduction to Elvish" (http://tolkiengateway.net/wiki/An_Introduction_to_Elvish) some twenty years ago, and I was astonished to learn the detail behind Tolkein's language work, in particular the history of the language, and the theory behind its evolution. When inventing a language, how much do you consider its development - not as its creator, but as a person observing its changes throughout time?
Oh arse
Now that surprises me. A lot, actually. I see people make mistakes of grammar all the time, and when once I inspected recordings of my own voice, I was shocked to find how many I make, especially if in a situation I wasn't prepared for. It is, for example, a very common mistake to make the numerus of a noun congruent to that of the immediately preceding noun, not the one it is really dependent of, as in "the main cause of errors that go unnoticed remain mysterious".
Have you considered suggesting the use of an extinct or endangered human language in a film or TV show? There are plenty of them, though of course not all are well documented and even those that are would need some new vocabulary words created.
Put these together and you get a fairly direct legal conclusion: information *about* a conlang isn't copyrighted, because it isn't created in form (textbooks, dictionaries) by the creator of the conlang
I just thought of a different legal theory that could be used to claim copyright in a constructed language. A language's lexicon is a set of names of things. In the API of a programming library, the set of functions is also a set of names of things. But in Oracle v. Google, a U.S. court of appeals upheld copyrightability of the "structure, sequence and organization" of the Java standard library's API on May 2014. (It remanded to the district court the question of whether copying said API for purposes of interoperability is a fair use; this remains unresolved.)