Create a language, go to jail

« previous post | next post »

I've received several messages with links to this NYT piece since its appearance online on Sunday. The piece is on Dothraki, a constructed language used in the HBO series "Game of Thrones" and invented by David J. Peterson, founder and President of the Language Creation Society and (as it happens) a former PhD student here in the Extreme Southwest Wing of Language Log Plaza. The piece also talks about constructed languages ("conlangs") and language constructors ("conlangers") a bit more generally, and most specifically with respect to their use in Hollywood. (That 'their' is purposely ambiguous.)

One of the messages I received was a Google+ post from a non-linguist friend with the humorous imperative, "+Eric Bakovic This seems lucrative. Hop to it." Seriously, I've been tempted. But I doubt that I have what it takes to win an open call to create a language for a fictional universe, as Peterson did. And besides, I have as close to no experience with constructing full-fledged languages as you can possibly have; the real conlangers would spot me as an interloper from miles away. And besides besides, I have job security. And besides besides besides … well, you get the idea.

One of the other messages I received was an email message from a linguist colleague, with a more sarcastic version of the same sentiment — something along the lines of, "here's another career opportunity that we can talk to linguistics undergrads about." [Exact content withheld to protect the innocent.] Why the sarcasm? Well, there's an attitude among some linguists — and also plenty of non-linguists, as is evident from many of the comments on the NYT piece — that engaging in conlang activity is a waste of time, perhaps even detrimental to the real subject matter of linguistics. I don't share this attitude, in large part because one could say much the same about any human undertaking, including large swaths of what we might call "the real subject matter of linguistics". If I had a conlanger in one of my classes, I certainly wouldn't discourage them from doing what they're doing — at least, not any more than I would discourage any of my undergrads from following any other specific path in linguistics, especially given that the career opportunities in this field in general aren't exactly thick on the ground (with the possible exception of speech-language pathology and computational linguistics, as pointed out by Stuart Robinson in the comments below).

I also happen to think that asking the type of question that I believe conlangers must (= should) ask themselves — what does it take to construct a language? — is probably a great way to get people to think critically, creatively, and scientifically about language; in other words, to comprehend the real subject matter of linguistics. I know that Peterson thinks deeply about all this, which is part of what makes him a successful conlanger and which would have made him a successful "mainstream" linguist had he chosen to pursue that route with the same vigor with which he has chosen to pursue the conlang route.

A large proportion of the 119 comments (and counting) on the NYT piece can be lumped into two large groups. (Some of the commenters really got wound up about these issues, though apparently not sufficiently to look for fellow travelers among the previous comments — the number of times each type of comment is repeated is pretty astounding, but that's probably true of long comment threads in general.)

  1. Klingon (Star Trek) and Na'vi (Avatar) and their inventors (Marc Okrand, Paul Frommer) are discussed, but why is there no mention of J.R.R. Tolkien's invented languages from The Lord of the Rings? [Internal LLog plug: Ben Zimmer wrote about Na'vi in his NYT 'On Language' column back when Avatar was released.]
  2. Why are these conlangers sitting around inventing languages, when there are so many endangered languages in need of "saving", and/or why is Hollywood paying to have new languages invented, when they could be using and promoting existing (and preferably endangered) languages?

The first of these is a bit obvious, perhaps, but the point of the piece wasn't to review the history of conlangs in fiction — it's a fluff piece in the Television section, for crying out loud. The second is simply curious. Do we really want Hollywood to appropriate existing languages and cultures and attach them to fictional, stereotypifiable individuals/cultures more than they already do? And why are the commenters sitting around commenting, when there are so many better things they could be doing with their lives?

[Comments are open to allow commenters to sit around and comment instead of doing better things with their lives.]


  1. Aaron Boyden said,

    December 15, 2011 @ 2:46 pm

    I was going to go save a language, but comment threads are so rarely open on Language Log that I just couldn't pass up the opportunity to comment.

    [(myl) You either can't read or can't count — of the last 20 LL posts, comments have been open on 16. Or could you be making a joke that's going over my head. Whatever the source of the confusion, you're not exactly making a strong argument for the value of keeping comments open.]

  2. Dan T. said,

    December 15, 2011 @ 2:50 pm

    I'm used to snarky endnotes about why comments are closed, but this is the first time I recall seeing one about why comments are open.

    [<EB> That's just one of many, many differences between me and Geoff Pullum.]

    The headline of this article is rather unnecessarily inflammatory, isn't it? There's nothing in the content that even vaguely hints at anybody being sent to jail for creating a language.

    [<EB> It's an allusion — perhaps too obscure — to the "play an accordion, go to jail" bumper sticker (and variants thereof). The sentiment is that the accordion is a waste of time compared to "real instruments", kind of like the sentiment I talk about in the post. And now you've made me go and explain it, which takes all the fun out of making an obscure allusion in the title of a post. Thanks a lot, Dan T.]

  3. Sam said,

    December 15, 2011 @ 2:51 pm

    I wonder if the Architecture profession has a similar divide? A valid career path for many architects is designing virtual landscapes and structures for entertainment – mainly movies and video games. These aren't "real" buildings, but when they are designed by a "real" architect, they vastly improve the quality of the entertainment, by improving the credulity of the world, and deepening the immersion.

    Some games I've played recently have simply staggering architecture and set design involved, like Assassin's Creed and Skyrim. It was worth the effort put into it. This seems like something that is well understood – movies can win awards for set and costume design, for instance. In the movie Inception, architects are brought in to design the virtual dream worlds that they work in.

    Why is the linguistics connection any different? A good invented language can add a lot to a work of fiction – why is that not a valid pursuit for a linguist?

  4. Flink said,

    December 15, 2011 @ 3:02 pm

    At least nobody's mentioning Dark Skies.

    [<EB> Except the NYT piece did.]

  5. Cy said,

    December 15, 2011 @ 3:06 pm

    I've met a few of these people in my undergrad classes. I guess I'm against it, sort of on the "waste of time" side, but just slightly – and only because these constructed languages always make so much sense – they're always very formal, and I think the "danger" is that these students create these great things with all the rules they've absorbed, but the thrill of learning linguistics (I think) is discovering all the exceptions and all the weirdness, and then trying to discover the principles in action behind them.The rules are sort of post hoc, they just model and describe the real situation, which is always amazingly complex. The playground is too safe when you construct your own fake language. They seem to be great exercises in markedness and segment inventorying, though – it seems like a rough sketch would be great fun for an advanced intro phonology course.

  6. Stuart Robinson said,

    December 15, 2011 @ 3:14 pm

    I take issue with this:

    "given that the career opportunities in this field in general aren't exactly thick on the ground (with the possible exception of speech pathology)"

    You're overlooking computational linguistics, for which career prospects are quite good (especially if you lump in natural language processing and machine learning–arguably part of comp sci).

    [<EB> Point taken. Corrected in the post. Thanks.]

  7. Barry said,

    December 15, 2011 @ 3:29 pm

    I think what should be asked about the "waste of time" people:

    How is it any more a waste of time than any other hobby? Should gardeners be expected go out and do reforestation/native plant rehabilitation rather than growing gardens? What about wine and beer aficionados? Surely they're wasting their time on their hobby when they could be going out and helping in food kitchens?

    To call this hobby a waste of time is a rather hollow opinion and says to me that there's a rather big chip on some folks' shoulders about other people's business.

  8. John Lawler said,

    December 15, 2011 @ 3:35 pm

    I'm interested in the Dothraki vowel system. It seems like [u] is an allophone of /o/ that occurs after postvelar consonants, which is exactly the opposite of the behavior we'd normally expect.

    Clearly, Na'vi is out, Dothraki is in. Now I've got to brush up my NorWesCon lecture on Fictional Linguistics (handout from last year's talk here). 
    The show must go on!

  9. David J. Peterson said,

    December 15, 2011 @ 3:38 pm

    @Cy: If I may summarize your point (please feel free to correct this if it's not a fair summary), your contention is that conlangs are frivolous because they're too regular (and natural languages, of course, are not). This is true of conlangs that are regular. However, if one is creating a language that's supposed to look like a natural language, then it should not be regular—and if it looks too regular, the conlanger has failed.

    And even though there are many such attempts that do ultimately fail, the art is young yet, and conlangers are learning and improving all the time. If you come across an early attempt that looks a bit too cookie-cutter, that's a perfectly good teachable moment (e.g. "This language looks fake. Want to know what tipped me off…?").

    Nevertheless, you do raise a good point about the nature of conlang descriptions which, I think, conlangers (if they're reading) should consider. The way I see it (and this is probably an unfair analogy, but I'm using it for purely pedagogical purposes, so please don't take offense!), languages are like a hunk of cheese, and a linguistic analysis is like taking that hunk of cheese and grating it (the result being grated cheese). Starting out with an analysis and building a language from it is rather like putting the block of cheese back together from the gratings, and is probably not the best way to go about creating a language.

    @EB: Hello! Hope all's going well in San Diego. I wanted to let you know I really did (and still do) appreciate the support you gave me while I was at UCSD. I feel like I got more than I gave, but as a student, I suppose that's usually the case.

    In a blog post, I added some commentary addressing your second point specifically (but perhaps from a different angle):

    In short: I think such comments greatly undervalue and underestimate the incredible work that field linguists have done and continue to do by suggesting that any old person could, on a whim, decide to record and describe an endangered language.

  10. Greg said,

    December 15, 2011 @ 3:40 pm

    "I also happen to think that asking the type of question that I believe conlangers must (= should) ask themselves — what does it take to construct a language? — is probably a great way to get people to think critically, creatively, and scientifically about language"

    This is exactly what got me interested in studying linguistics to begin with. I've been creating conlangs since I read Lord of the Rings when I was a kid and thought that Elvish was kind of a neat idea, and at some point I decided that I needed to understand how language actually *does* work in order to do a good job of creating an imaginary one. So because of conlanging I'm actually more likely to be the sort of person who has both the interest and training to do fieldwork with speakers of a threatened language.

  11. Jake said,

    December 15, 2011 @ 3:43 pm

    What on earth is going on here- the dude is talking backwards! Fluently! I didnt know where else to put this but on this language blog and hope youd have something smart to say about it. I was blown away.!

  12. Josh Birch said,

    December 15, 2011 @ 3:44 pm

    In response to Cy's comment, about conlangs "always [being] very formal," I want to add the note that this is entirely a matter of the goals, personal artistic aesthetics, and maybe level of experience of the conlanger. Personally, I am much more drawn to create languages approximating the full gamut of oddity I've observed in my study of natural languages, including etymologies from a proto-language and huge amounts of irregularity. I see the same amount of variation between the creations of conlangers as I do between those of, say, visual artists. I've never been able to understand why people deem conlanging less worthwhile than any other artistic pursuit.

  13. Aaron Toivo said,

    December 15, 2011 @ 3:46 pm

    Correction: while Mr. Peterson is the president of the LCS, he did not found it. The founder was Sai Emrys (or whatever he's calling himself on the internet these days), who was the first president and remains on the board of directors. Not that I'm an LCS member, but as a participant on another conlanging forum I'm familiar with the basics of their history.

    [<EB> Thanks, corrected.]

    @ Cy:

    There are a great many conlangers who do as you describe – create logical or regular languages that make too much sense and avoid any irregularity, etc. But others among us revel in the exceptions and the weirdness and seek not to improve on nature but to model and emulate it, as a form of art. And this does require some knowledge of linguistics, let me tell you. You have to keep learning and learning, it never ends. In the pursuit of conlanging I have read through numerous linguistics texts and especially, grammars documented by field linguists, hunting both for ideas and for understanding, and various of my fellows do the same. Naturalistic conlanging is the main trend at the forum I go to, and among its members I know of more than a dozen people who have gone on to major in linguistics or pursue higher degrees in it, all of whom were first exposed to the field through conlanging.

  14. Aaron Toivo said,

    December 15, 2011 @ 3:53 pm

    [Ugh, I must remember to refresh before I post and see what's happened since I started typing. Which was before Mr. Peterson posted.]

  15. KWillets said,

    December 15, 2011 @ 4:07 pm

    There was another category of comments complaining about the words-for-snow allusion in the number of Dothraki words for horse, (which may or may not be more than English).

    The question is whether a conlang should add imaginary linguistic rules, like n-words-for-X or no-word-for-Y which are lamented regularly on this blog. Is it part of the fun to add these features, or do conlangers strive for linguistic accuracy?

    Also, linguists construct proto-languages out of existing ones, so the line isn't as strict as some think.

  16. Leland Paul said,

    December 15, 2011 @ 4:08 pm

    It's great to see conlanging get a bit more press, even in TV section fluff pieces!

    As a prospective Serious Linguist ™ with a focus on (wait for it) endangered language documentation and revitalization who also happens to be a conlanger, the argument that the two are at odds has never made much sense to me. My conlanging, if anything, makes me more concerned about the issue and more eager to do something about it – the more linguistic diversity there is in the world, the more for me to be inspired by! (Which, of course, is an awfully selfish reason to be mucking in other people's linguistic lives, and not actually the main reason I'm interested in language documentation.)

    I took an undergraduate class in linguistic typology that was taught via conlanging, and I thought it was pedagogically inspired. It's a great way to engage with the patterns (and exceptions!) of natural language. I definitely recommend it as a teaching tool. (Personally, I wish someone had taught a class like this when I was in high school – I think it'd be a great way to get more young people thinking critically about language!)

  17. Theodore said,

    December 15, 2011 @ 4:31 pm

    I'll echo what Barry said above: Back when I was playing with conlangs, questions of form #2 would come up in usenet discussions, etc. I always thought it was like telling a model railroad club they're wasting their time playing with toys when there's so much real freight to move.

  18. Jonathan Badger said,

    December 15, 2011 @ 4:39 pm

    In biology, the current hot topic is synthetic biology — the quest for the creation of an artificial living cell. Much like with linguistics, there are those who think that biologists should focus on saving endangered species rather than creating new ones. However, as Richard Feynman wrote, "What I cannot create, I do not understand" — building anything requires a deeper understanding of a subject than merely describing existing examples.

  19. Arnold Zwicky said,

    December 15, 2011 @ 5:33 pm

    A couple more bits of recent bibliography, on my blog here. (The Okrent is mentioned in the NYT piece.)

  20. David J. Peterson said,

    December 15, 2011 @ 5:45 pm

    @KWillets: There are 14 words for "horse" in English, too: horse, mount, steed, charger, filly, colt, mustang, stallion, thoroughbred, mare… Probably over 100, in fact. They don't all just mean "horse", and they're not all basic terms (basic in the sense of not the product of derivational processes), and the same is true of Dothraki. But the whole "Language x has y words for z!" thing looks fun on a press release (lots of people like to hang their hat on such facts, whether they're indicative of anything linguistically or culturally interesting or not). Plus, in this case, it gives an idea of the culture of George R. R. Martin's Dothraki people (in which case the statement is basically synonymous with "the Dothraki people value horses and use them a lot").

    Personally, I think there's much more mischief to be had with the "x words for y" trope. For example, in Southern Californian English, there's 23 different words for "surfboard", but only one word for "business formal".

  21. J. W. Brewer said,

    December 15, 2011 @ 6:01 pm

    No doubt there are many different rooms in the mansion of the conlanging subculture. Speaking for myself, I suppose what mostly rubs me the wrong way is what you might call the Esperantist/Lojbanist side of the activity, where the whole point seems to be create something "better" that lacks the messiness and irregularity of a natural language in order to promote Global Harmony or something. But obviously that's not the only sensibility out there, as witness many of the comments above.

    I pretty much stopped reading science fiction and fantasy decades ago in part because I found that the various invented worlds, however painstakingly constructed, lacked the richness and depth of even a small portion of our actual world as created by the complex interaction of large numbers of actual people. I think it's more or less impossible for any single author or small group of authors to generate that sort of richness, because it's an emergent property of large-scale social interaction over time with no single conscious designer. I expect I would have the same take on the more naturalistic style of conlanging, but that's perhaps a matter of taste.

  22. Michael Cargal said,

    December 15, 2011 @ 6:09 pm

    One way to look at the X words for Y issue would be to look at different languages' words for male and female naughty bits. If Eskimos did have more words for snow because they live among snow, we would expect every culture to have the same number of words for naughty bits, since every culture has naughty bits,

    If they do not, then there would be no relationship between the number of words in a language and the ubiquity of the referent, so there would be no reason to expect people to live in snow to have more words for snow.

  23. Stephen D. Rogers said,

    December 15, 2011 @ 6:24 pm

    There are as many reasons to construct a language — and as many types of languages constructed — as there are people constructing them. I researched over a hundred conlangs over the past couple of years, and the stories behind the stories are more entertaining than many Hollywood productions.

  24. Stephen Nicholson said,

    December 15, 2011 @ 6:25 pm

    I don't know about other people, but for me, created languages are less onerous in film and tv than in print. Film and tv usually provide a translation via subtitles, which aren't a problem for me. In print, however, they usually just end up doing things like calling a common item a "smeerp". It's really not a constructed language, but changing out a few words for flavor. I find it annoying.

  25. Mia said,

    December 15, 2011 @ 6:28 pm

    I think that the first part of question #2 consists of two issues that it doesn't really make sense to conflate.

    I used to be somewhat interested in conlanging when I was younger (and obsessed with Lord of the Rings), but I last tried my hand at it about 6 years ago, and that was before I started studying linguistics. I still think it would be cool to try seriously, especially now that I know something about language typology and non-Indo-European languages. In particular, since starting grad school this past semester I've been keeping a mental list of interesting things that would be fun to play with in creating an "alien" language: I'd give it some blatantly non-conservative determiners, a vowel harmony system that seems logical but that we don't find in natural languages, etc.

    I wouldn't want to learn an artificial language created by someone else, though. I'd always prefer to study a natural language than, say, Klingon or Na'vi or even Sindarin; in that sense I can understand saying that it's a "waste of time". I don't really understand why people want to expend the time and effort to learn to speak Na'vi (if such a thing is possible; apparently there are people who try).

    That doesn't mean that I think that the creation of fictional languages is a stupid activity that people should be ashamed of engaging in; on the contrary. But I think the "you should be shipping actual freight and not playing with model trains" analogy only really applies to the creation of artificial languages, and less so to their…consumption(?).

    And, yeah, if someone wants to try to become fluent in Quenya, it's nobody's business but their own, so I don't actually care; on the other hand, I really don't get why they would want to. What it actually reminds me of quite a lot is the notion that people who play Guitar Hero are wasting their time and should be playing real instruments. I think the answer to that is that Guitar Hero isn't some kind of poor substitute for making music; rather, it's adding interactivity to create a new and fresh way to *listen* to music (as has been argued by, for instance, indie game designer Jeff Vogel: For real vs. fake languages, though, there isn't as much of a difference in what you're actually doing as there is for real vs. fake guitars. You're still learning a language; the difference is whether the speech community it primarily corresponds to is real or fictional.

  26. Wm Annis said,

    December 15, 2011 @ 6:32 pm

    In the same vein as Leland Paul…

    Once upon a time it was usual for classics students to compose a good deal of Latin and Greek prose and verse. The point of this was not to produce Greek stylists of course, but to improve the students' understanding. I still think everyone specializing in, say, Homer or Greek drama, should have to produce the major meters of the genre for a semester or two.

    In the same way, advanced classes in typology or syntax could use invented language projects as an excellent pedagogical tool. I once spent an entire month reading papers about relative clauses before I could finally decide how to do those in a language I was working on.

    How many non-conlanging linguistics grad students could do the following on a one-day take-home exam?

    (1) Using the phonology, phonotactics, nominal and pronominal system above, complete the verbal system of this language. Note from the pronoun table that this is a split-ergative language with the first and second person pronouns nom-acc.

    (2) Briefly describe the syntax of coordinate and subordinate structures with special attention on correferentiality. Don't copy Dyirbal's.

    Don't forget to include negative examples, with reasoning.

  27. Leonardo Boiko said,

    December 15, 2011 @ 6:50 pm

    I’ve met some conlang hostility by people who have a commitment that the real subject of true linguistics are “natural” languages exclusively, defined as those learned by a child in the critical period and that arised without (much) conscious individual intervention etc. Therefore, the study of artificial languages could lead to theories that do not properly model proper natural languages, and therefore artificial languages are dangerous toys and linguists should steer clear from them.

    Personally I’d think this hypothesis in itself would be worthy of investigation—how much and in what ways are conlangs similar to or different than natural languages? Contrary to Cy’s observation, I don’t feel like most artlangs I’ve heard are more logical or regular than the natural languages I know (auxlangs and other non-artlangs with utilitarian purposes, such as Lojban or (bleh) Esperanto, being a very different case). What happens when you apply your favorite framework to Klingon or Quenya or Toki Pona? Does it break (as a model) in interesting ways? or does it reveal something surprising about the conlang that even the author didn’t see?

    Nunberg once proposed that the system of written language, as delimited structurally by punctuation etc., is an “application” of natural language—“is what follows, roughly, from setting language down.” I think conlangs could be described in similar terms, and would love to have more linguists exploring such questions.

  28. George Corley said,

    December 15, 2011 @ 7:03 pm

    I host a podcast on constructed languages (Conlangery: and most of our discussion topics are not so much about conlangs as about various features from natural languages that one can incorporate into a conlang. For me personally, I feel that conlanging has deepened my interest in and knowledge of linguistics, and ultimately led me down a path where I am now applying for graduate programs in linguistics.

    I would most definitely not identify myself as a linguist (at least, not until I get the credentials to back that up), and really the skills to be a good conlanger and the skills to be a linguist are not entirely the same. I'm perfectly content to pursue my hobby while wildly hoping for the dream job of conlanging for movies, while at the same time seeking to study the more practical areas of linguistics and possibly working to make contributions there (including, yes, documenting endangered languages).

    Those people who pooh-pooh conlanging should consider several things:

    1) Not all people creating languages would necessarily be interested in or capable of documenting endangered languages.
    2) There are not unlimited funds and opportunities available to study endangered languages, nor for any "real" linguistics pursuit.
    3) The vast majority of conlangers conlang in their own time, as a hobby. Those few lucky enough to get money for it are an extreme minority, and often have contributed to "real" linguistics.

    Even if they still think of conlanging as a rather odd hobby (which it is), I would hope that most people would agree, as it seems most of the commenters on this post agree, that people are free to pursue whatever non-dangerous hobbies they like in their own time.

  29. Nick Lamb said,

    December 15, 2011 @ 7:21 pm

    Should we interpret the "Extreme Southwest Wing" as being a splinter group who believe the mainstream Language Log aren't taking their linguistic blogging seriously enough and have resorted to measures that ordinary Language Loggers frown on ?

    Maybe liveblogging as they rudely interrupt people who are making common but fallacious claims about language in a private telephone conversation. Or uploading Youtube videos of themselves protecting a sign reading "5 items or less" from vandalism by people who wrongly think the sign ought to read "or fewer".

  30. peterv said,

    December 15, 2011 @ 7:29 pm

    Does constructing artificial languages for machine-to-machine communications (languages such as HTTP, TCP and FIPA ACL) count as conlang activity? If so, the activity has already proved its value to society.

  31. maidhc said,

    December 15, 2011 @ 7:35 pm

    I recall an anecdote about Tolkien meeting a private in the trenches of WWI who was constructing his own language, but wasn't inclined to discuss it with anyone else. So it's not a new hobby.

  32. The Ridger said,

    December 15, 2011 @ 7:44 pm

    @Mia: You say "For real vs. fake languages, though, there isn't as much of a difference in what you're actually doing as there is for real vs. fake guitars. You're still learning a language; the difference is whether the speech community it primarily corresponds to is real or fictional."

    I would argue that the difference is whether the community is vibrant and composed of like-minded people, or on the other side of the planet/composed of people you don't actually want to have anything to do with.

    Sure you could all decide to learn Zulu instead of Quenya, but the community is still not going to be native Zulu speakers.

  33. Christopher said,

    December 15, 2011 @ 8:50 pm

    "You're overlooking computational linguistics, for which career prospects are quite good (especially if you lump in natural language processing and machine learning–arguably part of comp sci)."

    From what I've read (e.g. here) that those jobs are just engineering jobs and you can't really go far with a BA in Linguistics.

    Of course, that's just my impressions, and if I'm wrong let me know.

  34. Trey Jones said,

    December 15, 2011 @ 9:05 pm

    Conlanging, like any hobby, is okay as long as you don't take it too far. David has turned his personal obsession into public success, so good on him. But for the conlanger whose avocation takes away from, rather than contributes to, their overall happiness, there is help: Conlangers Anonymous

    Though maybe not everyone can be helped: "There is hope! (At least for those who don’t write to us in their conlang. Those poor bastards may be just too far gone, alas.)"

  35. Matt G said,

    December 15, 2011 @ 9:24 pm

    One of the Bouvier sisters (I forget which one) speaks Dothraki, according to a recent episode of The Simpsons

  36. Sai said,

    December 16, 2011 @ 1:05 am

    I've shared my comments on G+:

    As a side note: yes, I founded the LCS back when it was just a UC Berkeley student group. David however was one of the founding Board members when we became a nonprofit, and was my main co-executive for four years before he took over as President. He's done a lot for it, and I'm glad to see him take the lead.

    And for the record, my name is just Sai. It's on my passport and everything. :-P

  37. Cy said,

    December 16, 2011 @ 1:20 am

    @Josh Birch
    I didn't mean to imply that conlangers NEVER do it, or that they don't try, or that they're always too formal – sorry. If there were a war and I absolutely had to pick sides, I'd go with "waste," but I think it sounds like a super fun hobby. If I didn't like real languages so much, I'd totally do it. Because the problem I had was, I started reading stuff, and realized I would have to start making dialects, and registers, and then contact languages for word borrowings – at a certain point I realized it would be a better use of time to study real languages. But that's just my own personal problem.
    @Theodore So much freight to move! So apt.

  38. Joyce Melton said,

    December 16, 2011 @ 2:31 am

    I've played around with conlanging for about half a century, mostly as a way of creating "real" sounding place names for fantasy stories. I also kept a journal for a while that was in a semi-construted, semi-encoded artificial language in an artificially created script. Waste of time? Not compared to commuting through L.A. traffic….

    mulon laram ika tireska ava dazoli. ("I wanted to build a language and it is done" in "tireska cusayot")

  39. J. Goard said,

    December 16, 2011 @ 2:51 am

    David Peterson…long time, dude! Still remember some great conversations…

    Got a lot of ideas bonking around in my head, but I'm flying out tomorrow (Incheon to San Francisco) and I haven't started packing…

    LL is one blog that actually picks up around the holidays!

  40. LDavidH said,

    December 16, 2011 @ 3:43 am

    I have always invented languages, since I was a kid, and the main reason was simply "because it's fun" – I rarely invented any worlds to go with them. As I got older, it was more of "I wonder what a language with no consonants would look like" or "I'm going to create a language where all verbs are inherently negative" etc. Most of them got thrown away after a while, but I think I kept one in a folder somewhere – one where the accusative case was marked by changing a hard consonant in the middle of the noun to a soft one. I once met two other guys who also invented languages for fun; otherwise I thought it was just me and Tolkien…

  41. David J. Peterson said,

    December 16, 2011 @ 5:06 am

    @J. Goard: NO WAY! Man, it has been a long time! I found you on Twitter. Hit me up!

  42. Heidi said,

    December 16, 2011 @ 7:14 am

    I think I remember reading somewhere that comments have to be approved before they appear. If I am correct, then the comment judges have successfully and systematically established the rule in my mind that linguists are an amazingly bright, thoroughly interesting, clearly interested, and often hilarious group of individuals. I’m a layperson (art teacher by trade) but love languages and linguistics and the Language Log is one of the first sites I check each day. Thanks for giving me so much to think about! In particular, in response to today’s question of whether conlangers are “wasting” their time (as model train builders), I have been contemplating new perspectives on scientists’ creation of seedless watermelon when there’s so much cancer to cure, the federal government’s crackdown of family farmers selling raw milk when there’s so much… other federal government stuff to do, and so on. Thank you again, linguists! So glad I found your site.

  43. P. Orbis Proszynski said,

    December 16, 2011 @ 8:27 am

    Today's comment section is so heartwarming, my joy can only be expressed adequatly in Toki Pona, which, I hear, has 123 words for YAY!

  44. Martin J Ball said,

    December 16, 2011 @ 8:46 am

    Please note, the profession in the US is "Speech-language Pathology" and in the UK "Speech-Language Therapy" – it's been a long time since the term 'speech pathology' has been used. SLPs deal with both speech and language disorders – an ideal professional path for linguists.

    [<EB> Just my own ignorance, primarily because I use "speech" in both the narrower and the broader senses. Anyway, it's fixed. Thanks.]

  45. Leonardo Boiko said,

    December 16, 2011 @ 8:57 am

    @Proszynski: toki sina li musi tawa mi. taso mi pilin e ni: toki pona li jo e wan taso.

    ona li "pona" :)

    (quasi–word-for-word rendition: « Your talk was funny to me. But I think this: Toki Pona has one only. It is “pona”. »)

  46. Ron said,

    December 16, 2011 @ 10:41 am

    "[W]hy is Hollywood paying to have new languages invented, when they could be using and promoting existing (and preferably endangered) languages?"

    Wouldn't it be cool to see Avatar in Yiddish?

  47. JanSilipu said,

    December 16, 2011 @ 10:53 am

    @LeonardoBoiko pona! taso toki pona li jo e nimi "a" kin. Incurable since responding in a constructed language and on a totally trivial point ( and possibly incorrectly at that).

  48. R said,

    December 16, 2011 @ 12:10 pm

    This reminds me of an old humorous interview between Ali G (Sacha Baron Cohen) and Chomsky. Chomsky lays the smackdown on conlangers, in typical linguist fashion.

    If you've never seen it, it's pretty amusing…only 3 1/2 mins :

    Here's a quick script of when they start talking about it though.

    Ali G: Why don't you create a new language where instead of the word bread, you have something like methynol, or methla, or no, lafina.

    Chomsky: You could do it

    Ali G: [more nonsensical rambling]

    Chomsky: You could do it if you like, and no one will pay the slightest attention to you because it would just be a waste of time.

    Ali G: You could earn a lot of money…

    Chomsky: No you wouldn't because no one would pay the slightest attention to you because it's a waste of time.

  49. ENKI-][ said,

    December 16, 2011 @ 1:45 pm

    There's a distinction to be made between making a fantasy conlang for a book or film and making a philosophical language. That distinction is primarily in intent: when a fantasy conlang is created, it is intended to entertain (or sometimes to sell books), while when a philosophical language is created the intent is to test something about language. The actual attributes tend to merge: Klingon is a good example of a conlang intended to sell dictionaries to trekkies but that is legitimately interesting insomuch as it's designed to be speakable and understandable by humans but *just barely*.

    Nevertheless, there are lots of fluent speakers of Klingon and at least one native speaker of Klingon — and whether that says something about Klingon or something about people, it's nevertheless striking that there's more Klingon fluency than Lojban fluency. So, in a sense, both of these conlangs have provided data for linguists to study. This is before we even get into the very strange conlangs like Ithkuil and Ilakash, which nobody can speak. While I'm not sure Ithkuil and Ilakash provide a lot of information about linguistics as such, studying people's attempts to learn or speak them might be useful in the domain of cognitive science.

    Then, on the extreme practical end, you have the various international auxiliary languages. Esparanto is a conlang, but is spoken by many people fluently and some people natively.

    While the New York Times may laugh off conlangs as the domain of the linguistic equivalent of hack writers, I would hope that actual linguists would generally recognize the usefulness of having full 'laboratory' languages the way physicists have standardized experimental apparatus.

  50. Janice Byer said,

    December 16, 2011 @ 2:27 pm

    There ought to be a name – any logicians in the house? – for the category of erroneous reasoning epitomized by conflating the work of generating a language with what's required to keep one alive. For me, it's analogous to equating the building of a humanoid robot with successfully reversing multiple organ failure of a gravely ill human, both equally beyond me but otherwise not even apples and oranges.

    A general logic term for errors of thought that can be algebraically seen independent of context is "formal fallacy". Let L be all living languages; G all languages formerly gleams in conlangers's eyes; and D, the dearly departed:

    L + G = L – (-D)

  51. Adam said,

    December 16, 2011 @ 4:44 pm

    Dan T.: I'm used to snarky endnotes about why comments are closed, but this is the first time I recall seeing one about why comments are open.

    Yes, it's a great improvement!

  52. Adam said,

    December 16, 2011 @ 4:46 pm

    Barry: To call this hobby a waste of time is a rather hollow opinion and says to me that there's a rather big chip on some folks' shoulders about other people's business.

    Most (99%) people's hobbies are less harmful to others than some people's businesses!

  53. Adam said,

    December 16, 2011 @ 4:48 pm

    If I had a conlanger in one of my classes, I certainly wouldn't discourage them from doing what they're doing

    But you'd encourage them to construct a language with a good general animate singular pronoun, right? ;-)

    [<EB> No thanks, already have one.]

  54. Army1987 said,

    December 17, 2011 @ 1:55 pm

    @Stephen Nicholson:

  55. George Corley said,

    December 17, 2011 @ 5:43 pm

    @Army1987 Note that in the alt text he mentions Tolkien is an exception. I myself would restrict myself in how much of a conlang I would actually present in a book. It is much easier to reserve conlang terms for important cultural concepts and then put dialogue in English with taglines mentioning what language the characters are speaking.

  56. Ed said,

    December 17, 2011 @ 7:25 pm

    i just finished teaching my fourth semester of my freshman seminar "How To Build A Language" at Cornell. i've now been indirectly responsible for the creation of some 70-odd new conlangs. the whole goal of the course is linguistics by doing rather than by rote, and i find that it generally succeeds (although these courses are required, and i always have a couple students stuck with my course as their last choice).

    the fact that it's a writing course sets up an interesting juxtaposition, too. striking down all of the horrid grammatical misconceptions that my students have been subjected to in high school while simultaneously having them build up rules for their own projects really makes them puzzle.

    of course, despite my best efforts, i always get a certain amount of dry, uninspired auxlangs, Lojban clones, and Esperanto reduxes. but i have had some really fascinating projects. some fall into the fantasy/s.f. sphere (including a really convincing implementation of Parseltongue, and a couple for fictional universes that my students had already been developing as hobbies), but others definitely don't (e.g. a secret language for a sports team to use instead of signals, a language that assigns syntax and semantics to beatboxing).

    this semester, i had something happen that never did before: three of my students said they have plans to keep working on their projects past the end of the course. i am not concerned for their health or well-being. but maybe conlanging isn't just for nerds anymore.

  57. Larry said,

    December 18, 2011 @ 12:04 am

    I think I've noticed a pattern in these comments.

    A says: "Hey, leave activity / group X alone! They're not hurting anybody. Except, of course, sub-activity / sub-group Y, which is useless / pointless / dangerous / stupid / wrong."

    B says: "@A: Hey, leave sub-activity / sub-group Y alone! They're not hurting anybody. Except, of course, sub-sub-activity / sub-sub-group Z, which is useless / pointless / dangerous / stupid / wrong."

    Through this process, I think we've established so far that we should all stop hating on the conlangers. Except for those idiots who like to create logical or engineered languages. THEIR creative aspirations really ARE illegitimate and they really DO deserve derision, because what they produce don't look like natural languages, which is the stated goal of all conlanging (says so right there in the rules).

    Have I got that right? ;-)

  58. Finlay said,

    December 18, 2011 @ 1:01 am

    I had a good giggle at the comments line. As for the particular subject matter, I like to invent languages myself, but I don't quite know why I do it, and it's purely for my own amusement, at the end of the day. It's completely not like Peterson or Okrand inventing a language for something, it's just me amusing myself. I've learnt not to advertise it particularly, though, which is perhaps a shame, but I had an interview one time for an exchange program, which I didn't get a place on, and I think this was partly because I had mentioned conlanging in my application and didn't really know what to say when they challenged it. Like, "why would you do that?" "but nobody speaks it", etc. As for advertising it to my friends, I don't bother, and again that's perhaps a shame, because I'm sure at least some of them wouldn't mind, but I don't like the potential embarrassment. It does mean that when people ask what my hobbies are I tend to sheepishly deny having any (although even if I acknowledge conlanging as a hobby I 'should' probably have more than one hobby! and to that end I will claim reading and film-watching as my hobbies…..)

    The comments about endangered languages do seem to be symptomatic of the public's view of linguistics, though: I use a language, therefore I have a valid and unfalsifiable opinion about language. :(

  59. Larry said,

    December 18, 2011 @ 10:12 am


    About a year ago I had the ideal "explaining conlanging" experience. I was at an art show hosted by several artists. One of them, a close friend, was explaining to another that I invent languages. The other was not dismissive, but was perplexed. He turned to me and said, "But why do you do it?" I pointed over his shoulder to his own painting and said, "Why do you do that?"

    He got it. :-)

  60. Psi Wavefunction said,

    December 19, 2011 @ 2:42 am

    In defense of conlanging and its possible effects on one's understanding of natural linguistics (in absence of the wonderful irregularities that natural languages have), I think it depends on one's approach. I've spent a significant portion of my childhood constructing languages (and worlds) for fun, and this has been a powerful encouragement to learn about natural languages and how they relate to each other, change, function — ultimately leading to a strong interest in linguistics and taking a few courses in it on the side. Perhaps being a biologist by day led to a greater fascinating with variance than regularity, and many of my languages (which I never 'finished', leaving them a constant work in progress, a sandbox) were related in a particular way, leading to fun with phylogenies and sound change.

    In terms of introducing irregularities, it prompted consideration of *how* those irregularities could arise, whether other parallel irregularities should also be present in the language by sharing mechanisms, etc. This whole process seemed to get along nicely with learning formal historical linguistics material, kind of like a place to apply the theories, if you will. It's also fun to think of the languages in the context of their social and cultural environment, and its changes.

    Perhaps what makes historical and evolutionary linguistics more interesting to me, personally, than, say, structural linguistics, is much of the same stuff that keeps me in biology — this awesome interplay of structure and variation within a system, and the inherent stochasticity within. Thus, some of us design languages not to create something logical or 'better', or even to have a complete language, but to fool around and run simulations in a sort of a thought experiment.

    In other words, coglanging need not interfere with one's interest in and learning about natural languages, and the process can actually be quite enhanced by attempting to emulate the erratic or otherwise quirky habits and features of natural languages.

    (Now I have a sudden urge to read something linguistics-y — alas, there's no time or energy in a research life to pursue two distinct fields =( And yes, my languages lie neglected, save for the occasional night of creative inebriation…)

  61. GAC said,

    December 19, 2011 @ 9:55 pm

    @Larry: I see your point there. I generally like to figure in the purpose of a conlang when I try to evaluate it. Most conlangs created today aim to be naturalistic. But there is room for those conlangs that are designed to be as dense as possible (look at Ithkuil for a good example), and for the international auxiliary languages (think Esperanto). I, personally, think that the first is an intellectual curiosity that typically proves unlearnable, and the second really don't have a realistic chance of reaching their more ambitious usage goals, but there's no conlang I really hate, so long as the creator doesn't make it into something it's not (such as Paolini claiming he had created a language, when he had really just relexed English with badly pronounced Old Norse).

  62. Richard Littauer said,

    December 20, 2011 @ 11:11 am

    As someone mentioned in the NYT article, I feel obliged to say something here about it not being a waste of time. My work founding, being a member (and on the board) of the Language Creation Society, and more than that, developing the dictionary that's used by the Na'vi fan community (over 40 pages, translated into over a dozen languages), has not been a waste of time any more than any other hobby. Most of my conlanging time was spent in off-hours while pursuing an undergraduate degree at Edinburgh. Sure, I could have worked harder at my school work – but then I wouldn't have made a ton of friends, been to California, and developed a love for languages beyond the classes in my degree.

    I'm currently pursuing a masters in Computational Linguistics; this decision was very much influenced by the fact that I felt I could work on the skills I developed working on conlangs to help out endangered languages. And even now, I'm working with a few linguists, developers, and language communities making resources for endangered languages. That wouldn't have happened if I didn't conlang.

    Hobby? Sure. Avocation? probably a better word. Helpful to endangered languages? I'm trying to make it so!

    This comment may be too far down the list for anyone to read it, but I hope I got my point across for those of you who read so far, anyway.

  63. Bill Chapman said,

    December 20, 2011 @ 11:41 am

    It’s a shame that all the time and energy invested in these fictional languages is not dedicated to Esperanto which has demonstrated its usefulness over many years.

    Esperanto is celebrating its 125th anniversary next year. I’m sure we’ll hear more about that shortly.

  64. Simon said,

    December 20, 2011 @ 4:28 pm

    One thing that we in linguistics are quick to overlook when it comes to conlangs is that they are quite simply fun to invent, study, figure out, and learn. Fun counts for something! Right? Recently Wizards of the Coast released some tidbits of a Magic conlang, and a community of Magic fans obviously had a blast trying to figure out what they could (a surprising amount) with just a couple of Rosetta Stones and a very small corpus to work with (see

    Despite the fun, there is nonetheless a didactic component to endeavors like this as well: the skills applied in figuring out the details of a conlang, particularly as a part of a group of people working together, are exactly the skills that linguists must bring to deciphering real languages now known (e.g., Etruscan, Linear B, etc.) or not yet discovered, to say nothing of cryptology, and myriad other fields. A conlang provides a venue for practice and for students having their minds opened to new possibilities.

    Real-world languages that are dying off would benefit from similar attention, of course, but these cannot usually be studied by amateurs from their desks at home with access to a fully digital corpus online. Dying languages would benefit from some kind of wiki that made sound files and any existing printed materials freely and easily accessible.

    The use of languages in literature has sometimes run into intellectual property right problems, though not particularly in the United States (see, in particular Article 31 of the UN DRIP, as well as the section on Lego copyrighting of Maori words), to say nothing of charges of racism, colonialism, etc. Issues like this tend to make conlangs more appealing.

    Perhaps the most important thing, however, is this: conlangs are an art form. Their purpose and use transcend formal linguistics and are integral to many literary and artistic creative processes. In that way, it's difficult to begrudge anyone the enjoyment of conlangs, either in creating them or studying them.

  65. Aaron Boyden said,

    December 22, 2011 @ 11:43 am

    Hmmm. I suppose this shows how misleading one's perceptions can be. Perhaps it is the snarky comments when comment threads are closed that make the closed comment posts come to mind so that they seem more frequent, or perhaps for some reason there's a pattern of the posts I feel like I might want to comment on being frequently closed, and if I'm not interested in commenting I usually don't click through from my reader and don't see if there is a comment thread or not. Or perhaps they used to be closed more frequently, and I haven't caught up with the new pattern; I suppose that explanation would be testable.

  66. Terra said,

    December 24, 2011 @ 3:45 pm

    I'm inventing a language myself for my micronation. And I'm only 12, already running a nation and inventing a language.

    Man, what kind of 12-year-old boy am I?

RSS feed for comments on this post