So. I'm going to be running a game for the first time in fifteen years. I first said, "hey, I can do that." having brainstormed on a setting with my sister and having come up with a campaign model that seemed fairly simple and contained.
However, once i got my hands on it, it didn't remain simple. Well, the campaign model maybe still is. I have many very useful tools for containing the game: I am putting the players in contained environments, they belong to an organization (thus having a reason to stay together as a team), and they have superiors (so i can tell them where to go and what to do).
But it's also a campaign model (and world) that is very past-focused. The player-characters' jobs will be exploring old ruins from many different ancient civilizations. So this drives me to come up with lots of history.
Also, being me, I decided that random fantasy names were not acceptable, I needed a naming language. But I didn't feel like making one, so I said "umm..... Portuguese! Portuguese will stand in for Common." All well and good and simple.
But then I needed a name other than Portuguese for it, well how about Lusitano? And then I needed other languages, and before I knew it I was creating Hebrew theophoric names for halflings and a system for writing Māori in Tengwar.
Māori Tengwar
After some debate, I finally settled on using Māori for the Elven language. I considered writing Māori in Arabic Script and calling it Elvish, but then decided, well maybe just Tengwar.
So now I needed a Tengwar orthography for Māori. Tengwar is a featural script with fairly orderly 4 by 6 grid which hold most of its consonants. This is also reflected in the glyph forms for the consonants. So I can't just pick out a handful of scattered consonants to use, they would form an incoherent system.
But I couldn't find a coherent subset I liked, each approach had one problem or another. I suspected that if I researched the philology of Māori I could make some of those problems disappear in historical forms of the language, but that seemed like too much work.
Then I realized, it's Māori. It's a Polynesian language, they don't have any sounds to start with and they never change, the entire history of the language family fits on a page!
So I pulled up Proto-Polynesian. And PPn fit very nicely into the Tengwar grid! Just drop the fourth column and the second, fourth, and sixth rows of the grid, and you have a 3x3 grid of the nine normal-y consonants of PPn., excepting only the liquids and the glottal stop (which can be easily assigned Tegnwar glyphs that also are not part of the grid.)
But then, for some reason, the 5th row is the double-bow glyphs and the 6th row is the single-bow, so we'll use the glyphs of the 6th row to be the sounds of the 5th, and leave us with only single-bow glyphs.
And we'll say that the Tengwar wasn't invented until after the Eastern Polynesian split, so */l/ and */r/ have already merged, so I don't have to try to alternate between two letters for the /r/ in Māori.
That leaves only one real problem, which is that both PPn. */h/ and */?/ have disappeared in Māori. So there are going to be some silent letters that I can't reconstruct with perfect accuracy. But PPn. is recontructed with a strict CV syllable structure, so I can just add in and wherever there is a consonant missing.
For the sake of simplicity, I'll use where there is hiatus, and where there is a long vowel. The other way around might make more sense from a purely phonetic point of view, but the long vowel carrier glyph from Tengwar makes an excellent glyph for /?/.
Other than those two/three mergers, the PPn. form is ridiculously easy to reconstruct from Māori, so I can just look at a word, and know the old form I should be spelling.
So Elven is now known as:
or, Te Reo "The Voice"
And I now have not just a coherent system for writing Māori in Tengwar, but I have a system as if it has been place for a thousand years!