I wrote a post somewhere about personal conlangs, i.e. conlangs that are intended to be used and used by only one person. Thanks to a comment (albeit from a very angry elf), I started thinking about something not so different, the secret conlang or crypolang.
A cryptolang would be designed mainly to keep something secret, either from everyone (making it very much like a personal conlang), or from some one in particular, kind of like how Navajo was a cryptolang during World War II.
A cryptolang would probably want to have the reference grammar and dictionary secret.
Some conlangs are cryptolangs by accident. For many conlangs, the reference grammar hasn’t been published because it’s been lost, destroyed in fire, isn’t ready yet or what ever. So in practice, it’s achieving the goals of a cryptolang, without explicitly calling itself a cryptolang.
Conlanger like to make claims about their own and other conlangs, which often lack published descriptions, a reference grammar or what not. What should we do with these claims? For the real scholar, or the amateur linguist or even a half-ethical journalist, reporting on fabrications is problematic. For someone completely in a fantasy world–sure. The esteemed linguist George Lucas had some interesting things to say about the language of the Jawa’s in the novelization of “New Hope” and sure I can enjoy nonsense with everyone else. But for the amateur scholar, cryptolangs, (either on purpose or accidental) are problematic. You really can’t say much about them, except that someone has made some unverified claims about them.
A cryptolang author probably could still gain some glory by publishing the methodology. That wouldn’t compromise the author’s cryptolang, but would give some substance to the cryptolang’s claims. Knowing the techniques for creating a top notch crypolang is something worthy of study, even if the language itself isn’t available.