5

When creating a world where there are either aliens or fantasy races, you know that they should speak their own language.

But the truth is, that for story purposes, such races will speak their language as little as possible, because you want the reader of a story to understand what is going on.

So, feeling lazy reinventing the wheel, I would like to use already used languages for such cases - Klingon for the aliens and J.R.R. Tolkien's for the fantasy races

What are legal limitations to use these languages?

  • You could always ask the authors of the languages. (Harder for Tolkein, I know, but someone holds the copyright). They might be honored to have you use their language. – Thom Nov 12 '15 at 12:58
  • Constructed Languages Stack Exchange is also useful for questions on constructed languages: just an FYI – FoxElemental May 28 '18 at 17:57
5

In the United States, a language, generally, is described as a "specification"; that is, as a set of facts. For a language, these facts would be a series of statements along the lines of "X is a word having such-and-such a definition". Facts, in and of themselves, cannot be copyrighted (or patented, for that matter). What is protectable is a specific "implementation". Works created to describe the language are themselves copyrightable, such as dictionaries and programming guides, as are works created in the language, like stories and reference material written for use by those literate in the language.

That definition holds, without much question, for "natural languages", developed organically over time by an ethnic or regional group as a part of their identity. Nobody attempting to argue that English itself, or even a modern dialect like Ebonics, is protectable by copyright will be heard in U.S. courts.

However, Klingon and Sindarian are not natural languages. They're invented languages, developed artificially over a short time by a relatively small number of people (as few as one). The argument can therefore be made that these languages aren't subject to the "specification test", because the facts they assert are fictional and exist to serve a copyrighted work of fiction. Just as, for instance, someone cannot publish an unlicensed Star Trek novel for a profit, using characters and worlds from Paramount's protected fictional universe, use of the language without permission for personal gain is copyright infringement.

The closest legal precedent to date is in software copyright law, Oracle v. Google, a case concerning a re-implementation (with some literal copying) of the Java API, owned by Oracle who bought the original creator Sun Microsystems, to create the Dalvik runtime used in Google's Android operating system. The trial court found in favor of Google in 2012 on the majority of the charges, including the one on the API itself, deciding that the API was a declarative specification (similar to a natural language) and thus not subject to copyright.

However, the Federal Appeals Circuit court reversed, holding under an older Third Circuit opinion, Whelan v. Jaslow, that the "structure, sequence and organization" of a piece of software, including an API such as Java's, was in fact copyrightable intellectual property and thus by copying the SSO of the Java API, even if it rewrote every line of code implementing the funcaionality behind it, Google was using copyrighted material belonging to Oracle. It remains an open question of whether a "re-implementation" of a given specification, as Google did by re-implementing the Java API to create Dalvik and the ADK, constitutes "fair use"; the Federal Circuit in Oracle remanded the case back to the District Court to hear those arguments and a decision has not been made yet.

If Google is found in violation of copyright for the use of the Java API itself, that means the structural elements of a software language, essentially the set of facts that various words exist in the language and have a particular meaning, are protectable. That could be used to draw parallels to any non-natural language, including Klingon, Sindarian, Dothraki and other fictional ethnic/racial languages. The same parallels can be drawn if Google is found to be protected by fair use, which is predicated on meeting a four-factor test:

  1. The use of copyrighted work favors the advancement of society over personal gain.
  2. The use primarily copies the facts and ideas of protected work, not their "fixation" or specific implementation.
  3. The amount of the copyrighted property used is a relatively small amount of the full protected property.
  4. The effect on the copyright holder's ability to expolit the value of the protected work is minimally impacted.

The use of an invented language like Klingon within "fandoms" that identify with the original copyrighted work could easily be judged fair use using this test:

  • the usage usually isn't for personal gain but instead to "advance" the fandom by more closely matching the characters of the universe by speaking the language;
  • It's the language itself (the idea), not anything the writers have written in that language, that's used;
  • The amount of the language being used in any given situation is small (though to be fluent, you'd need knowledge of the majority of the language); and
  • The owner of the original property typically isn't economically damaged by the use, such as having sales of his own material using the language reduced.

However, use of the language in a situation like the OP's, where the universe probably isn't identified or identifiable as being the original universe in which the languages were introduced, might tip the scales, because it is primarily for personal gain over furthering the existing fandom. Multi-factor "balancing tests" like fair use can be fairly broadly interpreted, and it's possible, even likely, a judge might rule against the OP even if the other three tenets are met if the OP's intention, as he implies, is to lazily rip off a language and paste it in his universe to avoid having to invent his own.

For the OP to be guaranteed to prevail in a lawsuit, the precedent must be established that there is no meaningful delineation between an organically developed "natural language" and a systematically defined "invented language".

Of course, the OP could very likely claim de minimis non curat lex. Literally, "of the smallest, no care from the law", a doctrine applied in many ways including discretionally by the judge to spare the court having to hear lawsuits over trivial infractions with little potential harm done.

  • Oracle v Google is not really very good precedent, it would correspond to a case where: A one publisher has a reference text book of a language they made that is freely available for uses they approve of B second publisher wants to use the language but the first publisher does not approve the use and refuses the right to use their text books C the second publisher then writes their own text books based on the copyrighted ones they can't use with mostly identical "structure, sequence, and organization" and even some directly copied content. – Ville Niemi Nov 13 '15 at 11:12
  • The language is only relevant as a reason Google did all that. Basically, Google is arguing that if the language and API are not covered by copyright, the violations of copyright they were caught doing will be covered by fair use. They might be right too. So using the language is actually the defense Google uses, not the reason they are in trouble. (IANAL and I am sure Oracle would love to claim copyright on Java itself, just saying the case is not a good precedent for this question) – Ville Niemi Nov 13 '15 at 11:19
  • I agree it's not a great precedent, but the blogosphere has a lot of material drawing very cogent parallels. The hair you split here, where what Google got in trouble for was copying the names of classes and functions in the Java API, would be the equivalent of the existing vocabulary of a language like Klingon; you could argue, if Google could have and should have changed the names of classes and functions in the ADK, that anyone "copying" Klingon could keep the language structure but has to change all the words. As you say in your answer (which I upticked), it wouldn't be Java (or Klingon). – KeithS Nov 13 '15 at 15:51
  • Actually the problem Google had is that copying just the names of classes and functions was not enough, they had to copy the entire class structure to remain compatible, with the result being that they had to to create actual files and directories that were obviously directly derived from the copyrighted ones. Not sure if there is anything analogous, other than the text books I used as example, for languages. But meh, let us not argue when we do not actually disagree. – Ville Niemi Nov 13 '15 at 16:56
2

Serious Edit:

The lowdown, as pointed out by Lauren Ipsum, is that it's probably not a good idea. Since the Tolkien languages - among other constructed languages - are within copyrighted works, and are probably subject to copyright law.

There haven't been any serious legal limitations to using a constructed language, but that's mostly because it hasn't come up in any form of lawsuit. It makes copyright a hazy murk in this part of artistic expression.

There are arguments pro and con for the use of someone else's artificial language, so a strong case for "semiotic democracy" would probably win any potential lawsuits against you, if they would ever show up in court. You can see them in this publication by Harvard Law: http://jolt.law.harvard.edu/articles/pdf/v27/27HarvJLTech543.pdf

If you were to write in such a constructed language, you would most likely have to receive explicit permission to use it in your text. Otherwise, you are opening the gateway to litigation.

  • 1
    yeesh, I really have to disagree about Tolkien. He was a professor of linguistics. He wrote the series to have somewhere to use the languages he created. I think his estate would frown quite severely on using huge chunks of his work without permission. – Lauren-Reinstate-Monica-Ipsum Nov 11 '15 at 17:28
  • @LaurenIpsum Thanks for the insight. I've improved on the answer a bit. I do maintain that there hasn't been a serious issue that's come up in the law books, but you are right in pointing out my presumptions. – Mudly Nov 11 '15 at 19:15
1

I am not a layer, but I doubt the courts would recognize a copyright in an invented language.

I'm assuming you're not actually copying complete sentences and paragraphs, but rather individual words, and applying the rules of an invented grammar.

The publishers of an English dictionary certainly couldn't sue you because you wrote a book consisting entirely of words from their dictionary. Yes, Webster did not invent the English words, while the publisher of the Klingon dictionary presumably did. Is that sufficient difference? I don't think so. The U.S. Copyright office explicitly says that you cannot copyright individual words or phrases. http://www.copyright.gov/title37/202/37cfr202-1.html I think that would rule out copyrighting individual words in a dictionary, even if you used many of them. Likewise you cannot copyright "ideas, plans, or methods", which I think would prevent copyright an invented grammar. (Maybe, possibly, you could patent it. That's a different thing.)

But as I say, I'm not a lawyer. And judges regularly make rulings that strike me as completely contrary to common sense, so "makes sense to me" and "is what a judge would say if it went to court" are not at all the same thing. I couldn't find anything on the Internet about copyrights suits over an invented language, so I don't know if there's any precedent either way.

All that said, even assuming you were assured that you were on solid legal ground, why would you want to? If I was reading a book set in a universe that in no way is derived from Star Trek, and the characters are speaking Klingon, I would find that very distracting and disconcerting. It would be like reading a book set in Canada and all the Canadians speaking Swahili. I'd just constantly be saying, What? Why?

I think it would be much safer to just make up your own language. Unless characteristics of the language itself are an important element of the story, I doubt think there would be any need to actually work out a complete grammar and vocabulary. Just make up some words that have the sound that you want it to have. Beyond making sure that you don't contradict yourself, like in chapter one saying that "nowbarg" means "green" and in chapter two saying that it means "perihelion", what difference does it make? I suppose if I was going to write a story with a made-up language I might invent a few simple grammar rules, like decide on some verb endings or conventional order of words in a sentence. But I wouldn't write a complete grammar and dictionary. Any more than if I set a story in a fictional city that I would find it necessary to draw a complete map of the city with names of the business or resident at every address.

  • Update *

The Oracle v Google decision referenced by @KeithS is fascinating to me as a software developer. I've just spent the morning reading it (instead of doing my job). http://www.cafc.uscourts.gov/sites/default/files/opinions-orders//2013-1021.12-10-12.1.pdf -- a good and scary read for software developers.

This is really getting off-track, but what the case was all about was this: The court ruled that API declarations are copyrightable. Google copied 7,000 lines of Java code from Oracle without permission. Google argued in court that this code is not copyrightable because you cannot copyright an idea. The courts have repeatedly ruled that if a particular set of words is the only possible way to express an idea, or one of a small set of possible ways, than those words are not copyrightable. The court agreed that the Java language of itself is not copyrightable, because it is an idea, but that the specific text of the API declarations is copyrightable.

But here's the kicker: the court said that the principle of "only one way to express an idea" applies at the time the program was written, not at the time it was copied. That is, the court says that Oracle CAN copyright a collection of function names as long as they are not the only possible names that could have been given to those functions AT THE TIME THEY WERE WRITTEN. They give the example that the "max" function, that finds which of two number is larger, could just as well have been called "maximum" or "larger" or many other possible names, and that Google could have used some other name for their implementation of Java. But IMHO, this is absurd. If Google called the function something else, the compiler they wrote wouldn't be Java any more, but a new language. The issue is that, now that Java exists, if you want to make a Java-compatible API, you have to copy the function names. So while the court gives lip service to the principle that a programming language is an idea that cannot be copyrighted, by allowing Oracle to copyright the API they make it practically impossible for anyone to implement an alternative implementation. And, by the way, the court completely contradicts a precedent that they cite: the author of a book on accounting sued other writers for copying his forms, and the courts ruled that, as the layout of the form and the names of columns etc are essential to using his accounting system, and as the system itself is not copyrightable, then the forms and terminology are not copyrightable. Sounds to me like the obvious analogy of their precedent is that an API is not copyrightable. But they somehow came to the opposite conclusion. As a software developer, I find this scary stuff.

But getting back to the case at hand, it's not at all clear that this would be a precedent against using someone else's invented language. Oracle's complaint was that Google used their code to create their (Google's) own version of Java. At no point did Oracle say that anyone who USES the Java language is violating their copyrights. That is, they are not claiming that writing a program in the Java language violates their copyright.

The fair analogy would be this: If you invented your own fake language for a book and wrote a grammar and dictionary for it, and that language used many of the elements of Star Trek's Klingon, and in your grammar/dictionary you copied thousands of lines of text from a Klingon grammar/dictionary published by the Star Trek people, they could use this as a precedent to sue you.

But if you simply USED Klingon in a book that you wrote, nothing in this decision says that Star Trek would have any case against you.

  • 1
    You doubt a language could be ruled copyrightable, but check my answer; Oracle v Google, as decided, applied Whelan v Jaslow and stated that the "structure, sequence and organization" of software, including the Java API which essentially defines practical use of the Java language, is protectable. Parallels can easily be drawn between a computer language and other invented languages. – KeithS Nov 12 '15 at 22:32
  • 1
    I disagree with you on the copyright question. I think that there is no parallel with the English (or any other natural language) dictionary. The dictionary writers did not create the words or decide what they mean. The exact words of a definition would certainly be copyrightable! In contrast invented languages are very much the creation of named individuals or small groups. However I agree with you that, whatever the legal situation, to see Klingon and Tolkien's languages popping up (together!) in a completely different universe would just look ludicrous, unless the story is a comic parody. – Lostinfrance Nov 13 '15 at 9:27
  • @keiths I disagree. See my verbose update. – Jay Nov 13 '15 at 15:44
1

There is a case pending that may answer this in court: Paramount v Axanar.

See http://conlang.org/axanar for more info, including formal legal briefing and a memorandum from Dentons on conlangs & IP law.

(Disclosure: I direct the Language Creation Society's lawyer on this.)

You could also just commission a conlanger to make a language for you, so you have one that's bespoke, fitted to your specific needs / setting / aesthetic / etc, and w/ translations by the person who made the language. (LCS offers that service, FWIW: http://jobs.conlang.org.)

0

You could leave all of this to your agent or editor / publisher.

Here are the easy steps:

  1. write the book
  2. convince an agent to take you and your work on / or convince a publisher to publish your book.
  3. Let them worry about obtaining rights

If your agent or publisher is unable to obtain rights, then, at that point, your book will be done and you can make up your own language. Voila!

Maybe you are self-publishing with no agent or outside publisher? In that case, follow these steps:

  1. Write the book

  2. publish the book

  3. get thousands of people to read it

If you never get thousands to read it, it won't matter if you're not allowed to have the rights anyways.

If you do get thousands of people to read it, then if someone complains, you can remove the references to the borrowed language and replace it with your own.

  • 1
    Retroactively changing a published manuscript, where people downloaded or purchased copies, is no easy feat. More, if you're profiting off of it, simply editing away the infringing content won't be enough; a copyright holder would be entitled to royalties on the work -- or worse, if they don't agree to sharing rights, they could hold grounds for litigation, and you could wind up paying more than any profits you made. – Mudly Nov 13 '15 at 19:39
  • For the 12 copies that you shared freely and the 5 copies that you sold? It would be an interesting court case. The point is, 98% of "writers" worry about these details and never actually finish their manuscript and these details never matter. However, I think you make a good point and glad you commented. – raddevus Nov 15 '15 at 18:50

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.