The philosophy behind NaNoWriMo ( is to jump
right in with no advance planning and plow full steam ahead on writing
a novel. The idea that occurred to me applies a similar philosophy to
the creation of a conlang from scratch, and with no advance

The idea is to set a time limit (say for example, one month), then
choose a particular existing short literary work, fiction or
non-fiction, say 3,000 to 5,000 words in length. Then, beginning with
the first sentence of the work start translating it into a new,
previously non-existent conlang which you create, sentence by
sentence, as you go along.

Only AFTER the translation is complete may you go back and describe
the phonology, morphology, syntax, grammar, etc. that emerged during
the process of spontaneous translation. The idea is to "use the
language into existence" rather than planning it in advance of using
it. (You could, of course, go back and revise the translations of
earlier sentences when called for by the evolution of the conlang as
the translation proceeds.)

Ideally, to make the project more interesting, it the original conlang
created by this process should not be related to or descended from any
of your own previous conlangs. The goal is to conlang straight through
with wild abandon! (And with no idea of where you might be headed
until after you arrive.) This would, of course, require one to attack
the problem of creating a conlang in a considerably more casual manner
than usual, but wouldn't it be fun to just do something completely
crazy for once?

(And how cool would it be to have the complete short story like, for
example, Poe's "Pit and the Pendulum", or Mark Twain's "The Jumping
Frog" translated into a new conlang by this time next month?)