I am working on three built worlds or conworlds.
Two, Adpihi and Reptigan, are related science-fiction “worlds”.
Adpihi is a humans-only one-planet world accidentally settled from Earth. The people don’t know exactly where they are and don’t know how to get back in touch with Earth. They have their own language.
Reptigan is a thousand-light-year-diameter multi-species Union or Empire or Federation one of whose capitol planets is Adpihi and one of whose leading species is Humans.

The other is a multi-“racial” magical fantasy world called Ataivsh.
It has four intelligent species.
Three are hominoid:

“Humans” or “Men”, extremely similar to various species that have been called “human” IRL *here*, who are native to plains and grasslands and savannahs and prairies and steppes and pampases and other flatlands without big masses of tall trees;

“Dwarves”, similar to African great apes, especially bonobos and chimpanzees and gorillas (not for size, though!), and Homo florensis “hobbits”, who dwell in mountains and caves and intramontane valleys and do the mining and smithing and a lot of the engineering;

“Elves”, similar to orangutans and gibbons and other lesser apes and non-African great apes, who dwell in woodlands and forests and jungles.

The fourth are MerCentaurs, who are native to freshwater lakes and rivers and other navigable streams and bodies of fresh water. They are a three-way chimera of centaurs and Merfolk and Hippocampi.

The three hominoid species couldn’t have arisen without magic, but they don’t now (or “now”) need magic to live and reproduce.
The MerCentaurs can live without magic now, but they still can’t reproduce without magic.

All four species occur on every continent and every ocean and every climate.
Each has their own global auxlang, though I’m only developing Arpien, which is the Human global auxlang and is used as the inter-racial interlang or Lingua Franca.
I believe I’ll have to at least develop the other three at least to the point of having naming-languages for all four species.
I may also have each species have a continental and oceanic language for each continent and ocean. I haven’t even considered developing those yet, though. I’ll probably just say they exist but never say anything else about them.
I have even considered having regional or areal dialects of these continental and oceanic languages for a few really big areas within the big continents and oceans; but not separate languages.

I’m not planning for these people to ever get to space flight in any of my stories about them.
But they probably will get to the Age of Steam. Maybe also to internal combustion engines.
They may start before the Neolithic; but not before bows and arrows and atlatls and clothes and ropes and bodypaint and tattoos and language and so on. I expect they’ll be just about ready to domesticate lots of stuff at the earliest.

....

Anyone else have anything?

@CZM:
That’s very interesting! I want to follow.

@Riven:
I’ll keep an eye out for your progress updates!

@Tek S.: Congratulations!

@Jet P.: What Tek said.

@ravenspirit: sounds like an accomplishment!

Is the fortified chicken-coop idea one you could use in a sword-and-sorcery fantasy story or game?
Who’d defend it —— the chickens?

It’s nice to have a dog who takes her job seriously! And can be affectionate toward her nonhuman noncanine homies!

No kidding! Tell us a little detail about one of them?

I am working on a way to construct a “semantic map” of a language’s lexicon; and a (related) way of connecting two such maps for different languages.
I have no idea how to write about it or where to discuss it.

@Riven:

Not exactly.
For maps I want something like distances and something like directions.
Basically comparative semantic distances and analogies presented as “ratios”.

Distances:
For any three lexemes A B and C which pair are closest and which pair are furthest apart out of AB and AC and BC
And for any four lexemes A B C and D which pair are closest and which pair furthest apart out of AB and AC and AD.

“Ratios”:
For any three lexemes A B C which is the best, and which the worst, “mean proportional” between the other two;
B:A::A:C or A:B::B:C or A:C::C:B

For any four lexemes A B C D which pair of B C D is A the best or worst mean proportional between;
B:A::A:C or B:A::A:D or C:A::A:D

For any four lexemes A B C D which of the following is the best analogy and which the worst;
A:C::C:B A:D::D:B A:C::D:B

For any four lexemes A B C D which is best and which is worst from among;
A:C::D:B A:B::D:C A:B::C:D

For any five lexemes A B C D E which of the following three analogies is best and which worst;
A:B::B:C A:B::B:D A:B::B:E

And for any five lexemes A B C D E which is best and which is worst out of the following three analogies;
A:B::C:A A:B::C:D A:B::C:E

….

The alternative analogies give an idea similar to parallel directions.

……

To connect two languages we’d need something like this;

For any lexeme A in either language and any three lexemes X Y Z in the other which pair are closest and which furthest apart out of
AX AY AZ?

For any lexemes A and B in either language and any lexemes W X Y Z in the other which analogy is best and which worst out of
A:W::B:X A:W::B:Y A:W::B:Z ?

….

I know I still have some design work left to do about the analogies questions regarding two languages.

….

A “lexeme” can be a word or a morpheme or a phrase, provided it would (or could or might) get its own entry in a dictionary.
But I don’t think it could be a very long phrase, because I’m not sure I could handle more than about 50,000 of them.
Maybe a two-word phrase of five-phoneme words? Or a three-word phrase of three-phoneme words?

I might have to limit my attention to lexemes at most seven or eight or nine phonemes long, depending on how big the languages’ phoneme inventories were and how permissive their phonotactics were. Maybe ten phonemes long at the outside if there weren’t too many of those. Or maybe just the most commonly used 3000 or 5000 or 30000 or 50000 lexemes.

….

The whole thing is meant to be crowd-sourced. Volunteers who are fluent in the language (or in both languages, depending whether we’re mapping one language or connecting two languages) would interact with a program that would present them with sets of three options and ask them to rank them from best to worst. The program could make choices about what to ask them next based both on their own previous answers and the answers of other volunteers.

…..

This is for languages with many fluent speakers; a hundred, or a few hundred, or a thousand, or a few thousand, at minimum. So not many conlangs.
I can’t imagine needing fewer than twenty participants, nor more than a hundred thousand.

…..

I’ve also been thinking about a programmatic way to build a corpus of syllables and morphemes and words and phrases for a conlang, by taking data on what you’ve already created and proposing potential new items.
That could work with just a few, or even just one, creator. It might be most beneficial for collaborations among small groups; maybe just two or three people.
Right now it’s all in my head or my mind’s eye.

Maybe I can make that ratio business simpler.

I want to solve the following “ratios” or “proportions” or analogies, for the best X, given any A and any B and maybe any C.

A:B::B:X
A:X::X:B
A:B::C:X

I am glad I re-read this! I’m still interested in the projects people talked about!