Back in August, I started running a Pathfinder game in a setting of my designing that I generally refer to as DungeonWorld. It’s a world set up such that a campaign based mostly around dungeon crawls makes a certain amount of sense.
I wanted to limit the amount of world-building work I put into it, and in particular did not want to do have to create a world of ConLangs for it. But places and people still need names, and I’m too much the linguist to throw names haphazardly at them. So I decided to use real world languages, each representing a different in-world language.
For example the local human language (i.e. Common) is played by Portuguese under the name “Lusitano” (with other local human languages played by French, Latin, Greek, etc.), local Dwarven is “Dvergmal” (Icelandic), and Halfling is “Marebi” (Hebrew, giving me a fun chance to play with theophoric names). Gnomish, as a language isolate, is “Kaerugokugo” (Japanese).
After some casting around and debate, I settled on using Māori (“Te Reo”) for the Elven language (elves are a heavily sea-faring elder race, and have a more or less united language throughout the known world). But of course, Elven needs it own native writing system, which Māori doesn’t come with. One of my players lobbied for me to use the Tengwar. While I ruled out using Quenya or Sindarin for the language, using Tengwar seemed an acceptable borrowing.
So, I set out on a quest to create a Tengwar orthography for Māori.
The majority of the Tengwar are part of an ordered table (Phonetic values given for the Classical (Quenya) Mode):
For example Series II are the labials (and labiovelars), or Grade 5 are the nasals; the shapes of the glyphs are likewise featural: Series I glyphs have downward-facing, open bows, Grade 2 glyphs have double bows and descending stems. To use the Tengwar in the proposed context, you’d want a coherent subset of glyphs, or it would have to be interpreted in-world as a borrowed writing system, not one native to the language.
The consonant inventory of Māori is: /p/, /t/, /k/, /f/2, /h/, /m/, /n/, /ŋ/, /ɾ/, /w/. At first glance, this would use:
- Grade 1: Series I-III
- Grade 3: Series II&III (probably, though there are other choice for both /f/ and /h/)
- Grade 5: Series I-III
- Grade 6: possibly Series I and/or IV
- or rómen and úre from the Additional Letters
- The short and long carriers, for vowel-initial syllables.
That approaches a coherent subset. Four letters from the Additional Letters, plus get rid of Series IV, and Grades 2, 4, and 6, use everything else. It leaves only a gap at Grade 3, Series I (súle). You could move /w/ or /ɾ/ to that glyph, from out of the additional letters, but neither makes phonetic sense in that position.
Using Grades 1, 3, 5 is a little weird based on glyph shapes, since Grade 5 glyphs are the double-bow version of Grade 6 glyphs. But if we slip a little influence in from the Beleriand Mode (that is, Sindarin Tengwar), we can just use Grade 6, I-III for the nasals instead of Grade 5 I-III. This eliminates all double-bowed glyphs, simplifying subset.
Additionally, the rómen glyph is based on the óre glyph, so using the derivative without the base would be odd, though subtle and easily ignorable.
So anyway, I got to this point fairly rapidly, and said, “Well, that’s reasonably good, if not perfect.” (And actually, that system was better than I realized at the time, I now know the choices for /f/ and /h/ are better fits then I thought.)
Then I said, “Hmm, I have an intuition that if I looked at the historical linguistics of Māori, I could probably get a better fit. Nah, that’s too much work.”
Then I said, “Wait. Māori is a Polynesian language, it’s not hardly any work at all. The entire sound correspondence for the whole language family probably fits on half a page!”
Researching, I ended up with not half a page, but three full pages (well-space) on the sound correspondences of the whole language family. Polynesian has so few sounds, and such simple syllable structure, there are very few phonetic pressures at work changing the languages.
And sure enough, the sounds of Proto-Polynesian made a pretty good fit to the Tengwar. Notably, we can now fill in the gap in Grade 3, as Proto-Polynesian has the perfect three fricatives: /s/, /f/, /h. (It is true that súle only becomes /s/ in later Quenya, but it is perfectly reasonable for the
dental fricative to have always been /s/.)
Notably, it is also fairly trivial to reconstruct the Proto-Polynesian from of any given Māori word, with the exception of a couple of mergers. The combination of these two facts means that instead of making a Tengwar adaptation for Māori, I can make a Tengwar adaptation for Proto-Polynesian, and easily take any given Māori word and write it in the Tengwar based on its PPn pronunciation.
This gives me not only a slightly prettier fit to the Tengwar, but also endows the orthography with a built in history: the words are now being spelled as they were pronounced in an implied Ancient Elven, at least a thousand years ago (times whatever multipliers you care to add for the counterfactuals in the game world version of the language, notably being spoken by elves instead of humans).
Proto-Polynesian is also easily reconstructed without long vowels: long vowels in Māori, like the diphthongs seem to largely come from consonant deletions between two vowels. So the long carrier is easily repurposed to the PPn /ʔ/. Since PPn /ʔ/ is deleted in Māori, the glyph ends up turning around and becoming effectively a vowel carrier anyway in “modern elven”.
Proto-Polynesian is reconstructed with syllable initial vowels, but since that is one of the cases where reconstructing the PPn form from the Māori is not trivial or even not possible (given a Māori syllable that starts with a vowel, the PPn form could either be vowel initial, start with a /ʔ/, or with a /h/), I have taken the liberty of reconstructing Ancient Elven as requiring a consonant onset in every syllable, and adding /h/s to the Ancient Elven forms of onsetless syllables from Māori.
There is an even easier solution one of the other place Māori -> Proto-Polynesian reconstruction is nontrivial: Māori has merged PPn /l/ and /ɾ/, making it impossible when looking at a Māori word to know which was in the Proto-Polynesian (it can be determined if a Tongan cognate to the word can be found, but it does make it definitely nontrivial). However! That consonant is merged throughout the Nuclear subfamily of Polynesian languages. If we assume the orthographic Ancient Elven corresponds to a Proto-Nuclear-Polynesian instead of Proto-Polynesian, the writing system postdates that merger, and the difficulty can be ignored. As near as I can tell, I don’t even have to make any assumptions about the /l/-/ɾ/ merger happening before other sound changes which are represented in the orthography, the merger is the primary shared, derived phonological characteristic of Nuclear Polynesian languages.
This gives us this chart of PNPn/Ancient Elven Tengwar (usage shown in green):
I could have used rómen or lambe for the merged /l/-/ɾ/ phoneme, as its pronunciation varies throughout the Nuclear languages. I picked lambe through a combination of the visual aesthetic, and adding more fun orthographic/phonetic distinctions.
In the Māori/Modern Elven, orthographic súle is phonemic /h/; orthographic forme is /f/ word initially (Māori orthographic wh), and /h/ internally; orthographic lambe is /r/; and orthographic aha and long carrier (/ʔ/ in Ancient Elven) are silent.
Or, approached from the other direction, taking a known written Māori word and transliterating it to DungeonWorld Tengwar, as in practice I do:
- Māori p, t, k, m, n, ŋ are easily transcribed as the corresponding Tengwar
- Māori w becomes úre (leaving aside a rare-ish complication)
- Māori word initial wh becomes Tengwar forme
- Māori word initial h becomes Tengwar súle
- Māori word internal h can be either forme or súle, I default to súle
- Māori r becomes Tengwar lambe
- Māori long vowels are written as two vowels with a long carrier (Ancient Elven /ʔ/) inserted “between” them (it could also be an aha, but I default to long carrier)
- Māori diphthongs are written as two vowels with an aha “between” them (it could be a long carrier)
And there I have it, Māori to DungeonWorld Tengwar in 8 simple rules.
It was an extremely satisfying little exercise, resulting in an elegant solution with built-in history. What more could you ask for?
1. Since I’m purporting to show the Quneya mode, I should have halla in the additional letters, not hwesta sindarinwa (the glyph without a listed name in the 3rd row of the additional letters). But they are irrelevant to my purposes, as I use aha for PPn /h/.
The glyphs in the second row of the additional letters without names are optional reversed forms or silme and esse.
2. This phoneme is written wh in Māori, and has considerable phonetic variation, but there’s no strong reason not to consider it as basically a phonemic /f/.