This was one of my example questions but I now am actually interested in the answer. I am an amateur conlanger; I have one pretty complete language and a couple in the process. But I can't construct a sentence in any of my languages without referencing the lexicon and grammar. How common is it for the creator of a conlang to be fluent in that conlang? (especially for languages that are not meant to be spoken like all of Tolkien's, where there is not a very significant fluent community).

  • 3
    I find the grammar in your last sentence confusing. Are you saying that Tolkien's languages are not meant to be spoken--"languages that are not meant to be spoken (like all of Tolkien's)"--or that Tolkien's languages are meant to be spoken--"languages that are not meant to be (spoken like all of Tolkien's)"? I think it's the former, but the wording could be clarified. – DLosc Feb 19 '18 at 21:18
  • @DLosc The first one – CHEESE Feb 20 '18 at 22:14
  • It might be useful to look for the results of the various Lunatic Surveys conducted over the years on Conlang-L (certainly going back to 1998 or so). As I recall, the experiences of glossopoets in those far off days were little different than Adairain expresses: very few ever learn to speak their own invented languages! Ironically, I think any given invented language is far more likely to encounter someone other than the author who will take on the strange and beautiful burden of learning the language! – elemtilas Mar 2 '18 at 4:04
  • In March 2021, William S. Annis did an online survey about learning one's own conlang. The results can be found here: lingweenie.org/conlang/ConlangLearningSurveyResults.pdf I believe it is worth mentioning that only about two thirds of the respondents said they where actually intending to learn their own language(s). We should keep that in mind when discussing fluency -- for some creators, it's simply a non-goal. – tmh Jun 16 at 21:45

From personal experience (active member in a large server of avid conlangers) very few people ever learn to speak their own conlangs. Language learning is a very time-intensive progress and there is not much gained from learning a language no one else speaks. The occasional exception tends to be found in group projects, in which sometimes at least to a part “creating by using” is used as a technique, which naturally leads to at least rudimentary command of the language. Some languages have also been developed exclusively by using them, such as Viossa, but this is an exception.

You may also be interested in this talk at LCC6, held by Jim Hopkins in his conlang (with an interpreter, of course). Summary of the talk:

In this presentation Jim Hopkins, creator of Itlani, will share his tips and techniques for recording, learning, and living the Language of your Soul. He will discuss the need for finding and ensuring stability in your language’s vocabulary and structure to facilitate easier learning, and how to immerse yourself in your language for mental and emotional engagement, even though you aren’t living full time among native speakers. The presentation will be done entirely in the Itlani language, with interpretation into English by Tony Harris.


It's probably a lot more common for a person to be fluent in his own conlang when dealing with auxiliary languages. Esperanto, Ido, Interlingua, etc. all have large communities of speakers (relative to most conlangs), and I assume their creators also spoke them.

This is likely because the languages were designed for communication, and because the vocabulary is usually designed to be easily recognizable to as many people as possible, which makes learning it easier than learning an a priori vocabulary and grammar.

An example of a relatively recent auxlang is Lingua Franca Nova, created in 1998, which already has a lot of speakers. By contrast, a systematic "engelang" (engineered language) like Ithkuil would be extremely hard for anyone to master, and its creator is not fluent in it.

  • I'd quibble with wikipedia's assertion of 970 users based on Facebook group membership. I'd venture to say no one knows, except the number is probably small with 1 or 2 digits, maybe 3 if fluency was defined loosely enough. – MatthewMartin Mar 1 '18 at 21:12
  • @MatthewMartin Some people in the Wikipedia discussion here 10 years ago seem to agree with your skepticism – b a Mar 1 '18 at 22:02

Ignoring that fluency means a wide range of things (reading, writing, conversing, conversing at a college educated level, etc)--

This is really rare. Outside of Esperanto, which as L1 speakers, the number of people who get to the lower levels of fluency is in the low two digits. Klingon, Na'vi, toki pona have up to tens of people who have various levels of fluency. Getting solid data is also really hard, the numbers you see in wikipedia are often wild ass guesses.

As a thought experiment, you can look at how many languages are even potentially speakable, ie. is the specification complete to say most or all possible sentences? Is there a community? Before the internet, this was usually an insurmountable problem, finding anyone who'd want to learn your language. Are there any "poison pills" in the language put there by the designer specifically to discourage humans from trying to use it? (Pedophilia in the conculture and absurdly difficult pronunciations come to mind)

John Henry did a study of this, targeting more recent personal conlangs.



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.