Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[FearfulJesuit]

Some time ago on a forum for language geeks, we had a game where one team would create, from scratch, a small (read: 200 word lexicon, short grammar sketch) language, then derive daughters from it, as one might the Romance languages from Vulgar Latin. Then they'd release sketches and vocabulary for the daughter languages and have the community at large try to reconstruct the proto-language they'd made according to something called the comparative method. The way it works is this. Languages change, as a general rule, according to regular and universal sound changes. A good example is what we call æ-tensing in American English. Most of you will probably have noted that the vowel in "cat" is almost but not quite the same in American as in British English. That's because that sound has "tensed" and turned into a diphthong (a double-vowel sequence) in American English, but remained a pure vowel across the pond. Another example, easy to understand but from a bit further afield, is the development of consonants in various Polynesian languages. (I can hear moans in the back of the class, but please, people, this is actually kind of interesting). When you take a wide look at various Polynesian languages- Hawai'ian, Tahitian, Māori, Samoan, Tongan, etc.- you will notice cognates. Some examples:

Code: [Select]

gloss tongan maori samoan hawaiian rapanui
two (r)ua rua lua lua rua
three tolu toru tolu kolu toru
four fā whā fā hā ha
sky laŋi raŋi laŋi lani raŋi
man taŋata taŋata taŋata kanaka taŋata
house fale whare fale hale hare
parent motuʔa matua matua makua matuʔa
canoe vaka waka vaʔa waʔa vaka
sea tahi tai tai kai tai
octopus feke wheke feʔe heʔe heke

Looks complicated; but after a while you can piece together some correspondences. Where Tongan, Māori and Rapanui have /k/, for example, Samoan and Hawai'ian have /ʔ/ (that's a glottal stop, the little catch in the throat in "uh-oh"). Where the others have /t/, Hawai'ian has /k/. Where the others have /ŋ/ (that's "ng" in "singing"), Hawai'ian has /n/. And so on and so forth. Or, another example from closer to home:

Code: [Select]

English Latin Greek Sanskrit
father pater pater pitṛ
tooth dentem dontos danta
three trēs treis trayas
ten decem deka daśa
foot pēs pous padas

Here, again, a pattern emerges: when the English word has an initial fricative like f or th, the Latin, Greek and Sanskrit equivalents will have a regular stop like p or t, and when the English word has a regular voiceless stop like p or t, the Latin, Greek and Sanskrit equivalents will have a voiced one like b or d.

Here's the fun game part! This means that, if all you have are modern-day languages that are related, you can try to reconstruct their parent. It's a puzzle: knowing that sounds change more or less regularly, and given a bunch of related languages, you can start piecing together what their common ancestor probably looked like. That, my dear chaps, is the idea of this forum game. I have, sitting beside me, a notebook with a made-up list of one hundred and twenty words that could be part of a real language we'll call Proto-Mythonesian. You're not going to see it. Instead, I'm going to post fifteen of its descendents in a completely random order and see if the lot of you can piece together what Proto-Mythonesian must have looked like, aided by nearly useless hints from yours truly and the hivemind of whoever pops in to take a look at the wordlist. Here's how this works:

-Proto-Mythonesian split into five branches in prehistory which also aren't recorded. In a sense, then, although this is a family of languages, there are five mini-families, and you're going to have to reconstruct which languages belonged to which. We'll call them Eastern, Northern, Southern, Central and Western Mythonesian. You can determine which languages belonged to which because each family is marked by a set of common sound changes that happened /before/ its own daughters started to drift apart.

-The languages are going to be worked through and then posted in a random order with no regard for which are related to which.

-No grammar will be posted, just a lexicon, and a phonology (the sounds of the language), which you can comb for clues.

-You didn't think I'd make it all that easy, did you? Alas, as in real life, the Mythonesian languages liked to borrow words, mostly from each other. When I am working out the branch languages, each word will have a 1.5% chance that it will be borrowed from another branch protolanguage, but adopted to the borrowing language's sound system. You will need to pick out these false cognates. This goes again for the full daughters, where each word has a 2.5% chance of being borrowed, probably from another within the branch, but maybe from a language from a neighboring branch (just as the Celtic languages borrowed from the Romance and Germanic languages, for example). You'll need to pick out these, too. The borrowings on the branch level will be shared by all the daughters in that branch, unless the dice roll badly and one of them happens to reborrow. That's a possibility, although not a likely one. Rest assured that this is not only for difficulty- it's also for realism!

Who'd be interested?

[GreatWyrmGold]

Potentially intriguing, potentially boring. Presentation is going to be important.

[FearfulJesuit]

Yeah, and I'm not entirely sure what that would look like is the thing...