i have not been working more on Ifadätels due to not needing more details about it yet, but i did do some world-building on a different setting on thursday and then yesterday and today i've been doing different conlanging for that world :3

currently working on the as-yet unnamed proto-language that most of the contemporary languages of the monitor-lizard-people of the western grasslands and forests are descended from. (planned for later: four child languages of it, an unrelated language, and three vulgar-latin-style dialects of one of the child languages in regions where the other child languages and the unrelated language were the main local languages)

the proto-language has a tripartite noun-class system whose classes are i am tentatively calling "warm", "cold", "immaterial". as is common, some nouns make those names seem totally ridiculous. for example, tooth (ekańaš2, pronounced [ekæŋæʃ˧]) is immaterial

@Lady hm. i don’t think i do currently? probably going to start out by putting together one or more languages for names and inscriptions within the Great Labyrinth in my rpg campaign. if you’ve got advice for good formats in which to keep basic notes and a lexicon for easy editing and reference that would be helpful. or honestly just advice in general hehe.

@alyssa
• it can be tempting to make words with similar meanings sound similar, but it’s more useful to make them sound very different. it’s fine that beech and beach sound the same, but not beech and ash, maple, oak, pine, fir…

• when developing the initial corpus, one useful approach is to come up not with words but with a number of semantic “roots” which can be combined to form words. that will give you your basic sound inventory and an idea of how words in the language might be derived from each other.

• come up with rules, but don’t adhere too closely to them; language is messy. choose aesthetically appealing choices over more cumbersome “correct” ones; people have a tendency to change the pronunciation of words they don’t like the pronunciation of

• i find it helpful to define some broad categories of words like “body parts”, “people and occupations”, “plants”, “animals”, “food and drink”, “geography”, “substances and materials”, “emotions”, “sensations”, “actions”, “grammatical words”… to make it easier to browse conceptually-similar words

• the most important thing regarding dictionary/lexicon formats is that it be really really easy to read and edit. i’m a big fan of record-jar: https://datatracker.ietf.org/doc/html/draft-phillips-record-jar-01 (ignore the CRLF line-ending requirement)

i’ve built conlang dictionary viewers/tools before but i’ll save that conversation until you have enough words to where you start needing to use one :P

• for notes, i’ve always just done markdown or a similar lightweight markup language. you could always stick them on the wiki ;)

@alyssa here is a format i used to use and wrote a viewer for, if you’re curious https://github.com/langdev-project/standards/blob/master/DP/0002.txt

@Lady i've looked at that (LREC) before. :3 perhaps i will put my notes on the wiki indeed :)

@alyssa i’ve been meaning to get back into conlanging as well and have some tooling i’m working on for doing it but it’s probably going to be Makefile-based and not really ready yet for actual usage

@Lady is it cool if i stick my lexicon in the wiki repo too? (if i've got my grammar/etc. notes there might as well put my lexicon next to it...)

@alyssa yes totally fine; not sure it will actually get published to the web at this point if it isn’t a .djot but that's something we can work on if needed

@Lady cool. i'm not super concerned about publishing it as part of the wiki since people can go look at the repo if they really want hehe

@alyssa Now I kind of wonder about a language system that uses manual page sections for their word classes.

@aschmitz programming language with word-class agreement requirements