Derivation

[Dedalvs]

Is there a way we can make it so that anyone can edit the first post in this thread? It'd be nice if we can just have a list anyone can edit so we get a collection of all the derivational processes floating around.

[Miatato]

I am not sure how to allow people to edit other people's posts without giving them moderator privileges, but I will see what I can do. (I might make a few more people moderators anyway, since most people thus far seem like they are pretty level-headed and not likely to go around maliciously editing/deleting posts.)

[kadani]

About go-... The way I thought of it was that it means something like 'symbol of' or 'representing'... and probably that it is no longer productive.

[PeteBleackley]

Some of my early examples had a process
From X derive "product of X" by inserting /i/ after second mora. eg zavusa "bee" => zavuisa "honey"

Where this would create a three-vowel sequence there would have to be allomorphs

[Dedalvs]

[Matthew Turnbull]

that seems like a good idea, it'll limit the number of roots using it and let the person who decided it should exists decide how far the root has diverged from it's former meaning.

[kadani]

[Dedalvs]

[alavda-isere]

I just created a suffix in the word creation thread; -no attaches to nouns to create adjectives.

[alexfink]

If we cut out Pete's case system as cases [or even if not], we could keep them around for more derivational operators. For example, we might try to use the genitive in ways which come out close to "adjectiviser" (which is a dangerous word to use -- have we decided "adjective" is even a word class?):
podji 'gold' (substance) > piodji 'made of gold'? (and then I guess audja 'gold-coloured'?)
txifom 'hair' > txixifom (using kadani's hiatus resolution) 'hair-covered'?
kami 'moon' > kiami 'nocturnal'? 'monthly'?
etc.

... Ooh, and here's an idea. It would be quite cool if gosana 'crown' came first, and the old language had a form giosana 'crowned' which got substantivised to 'king', reflected as josana in the modern language by a very natural palatalisation process!
Could xami be 'month'?

EDIT: or txami. In the history that's fleshing itself out in my mind faster'n I can keep up with, initial dj was lenited to j without a corresponding tx > x -- we seem to have tx-initial words but no dj-initial right now. Not surprising in the least, there're several natlang examples of voiced affricates leniting but not voiceless. And either this old palatalisation affected only [kg]iV and not all [kg]i, or else there was another source of ki, since there's lots of ki- initial words. (Old Kenakoliku had a /k_w/ or a /q/ or some such?)