This idea is to build an open, relational map of every language, human or otherwise.
The basic idea is contained in the schema.
Words are countless. They are a series of unicode glyphs.
Words relate to one another. The relations are defined as followed.
Belongs to (language)
Words can belong to one or more languages.
Belongs to (subject)
Words can be categorized by related subject.
Conjugate / Declension of
This maps one word to another by how it is derived from it. IE, "was" is the past tense of "be". This is also part of a language.
Open Wiki Format
Anyone can contribute, but each contribution is marked by a contributor.
Contributors start off unapproved. Their contributions can be individually accepted or rejected. Anyone can see anyone's contributions and vote to approve or reject it.
Editors are chosen to close voting on a contribution, accepting it or rejecting it.