Publish a reference grammar or it didn’t happen

There is an internet meme, “pictures or it didn’t happen”, which is also a website for funny photos.

The conlang world is in a similar situation. “Noyatowa” is supposed to be a conlang that is polysynthetic. It has 7 google hits. This language doesn’t exist. It is morally as good as a figment of my imagination. I’m not trying to specifically pick on the author of Noyatowa*, which might be a linguistics grad student or a teenager in Iowa. I’m just tired of trying to research languages that people reference, but don’t exist in any practical sense.

Update: The author of that no-longer published conlang is Jeff Burke. I think the reference grammar is in one of these drawers (well on April 2011, he had a picture of his work area).

The language “Bp@x’àãókxá” is unsearchable with by Google, due to internal punctuation.

CALS is a great swath of language that seem to exist only as a list of features on a CALS page. I suspect the LangMaker archives (since the main site appears to be down), are much the same, but probably more languages that are reference grammar-less because the owners website went down.

So my advice.

If you are writing an article or paper about conlangs, think twice about even mentioning unpublished languages. You aren’t sounding more erudite by listing every name of a conlang you’ve seen on CALS or a forum. You may just be spreading rumors that a language exists, when in fact it is fictional as warp drives, or magic wands or any other setting material that we read in fiction.

If you are naming a language, pick something that is currently getting zero google hits, but isn’t broken up by punctuation, or worse, is only punctuation. “Bieber” is a poor name. Zocovapy, as of right now is a very good name for a conlang.

If you are writing or bragging about your conlang, don’t make many claims about the unpublished part– or if you do, make it clear your talking about an unpublished language. The nitwits reading what you write may repeat it as if your conlang is a published, living language. If you have published a language, release it under a licence that makes it easy for people to grab a copy of it for archiving & republishing.

Update. Someone named Wastrel posted some flame bait questions in my comments which I’ve withheld. I felt real bad that the conversation didn’t go so well– I take online civility seriously, but haven’t worked out what is the right think to do when the feces start flying. I am somewhat reassured that elsewhere Wastrel is getting banned from forums. My advice to bomb throwers and free spirits is that they should put the combative thoughts, their bile in their own blog, and feign civility in comment threads, forums, mailing lists, etc. Know when you are pissed off and refrain from clicking send before giving yourself a chance to cool down.

This entry was posted in conlang community building, conlang design. Bookmark the permalink.

One Response to Publish a reference grammar or it didn’t happen

  1. matthewdeanmartin says:

    I would ask you the question, “Why do you think anyone should realistically know about Noyatowa, those writing conlang blogs or anyone else?”

    Hmm, 10 google hits for Noyatowa and one of them is me. Secret languages are a perfectly legit form of a conlang. My point is a narrow one– if someone is going to write about a secret language, no one can really make claims about it since none of them are verifiable. Pure fiction is also perfectly legit as long as people aren’t trying to foist fantasy off as amateur (or pro) linguistics work.

    I think it is perfectly legit for you to never publish your conlang’s description. But if you or some random person on the net said “So and so” conlang lacks recursion, violates 7 out of 8 of chomsky’s universals, has 800 words for snow, then this is just speculative fantasy until the reference grammar is published.

    re: ZBB as a source
    The ZBB truncates (deletes) the old history, seems to have blocked google indexing, and recently broke all its old links when it upgraded the phpBB. has an incomplete set of the worlds pages and is breathtaking laborious to dig through. AFAIK, only is addressing the problem of disappearing conlangs, by offering long term archiving of conlang docs, which again, means that some sort of reference grammar/sketch/dictionary is required.

    Re: comparisons with unpublished academia work
    I think the practice of referencing unpublished works in academia is just as shady. From my short stint in academia, I found the vast majority of professors take citing sources very seriously and one usually can find the sources of an academic paper given the right academic library.