Woods wrote:But in short, if I learn the general declension types (taivutustyypit), I will be able to make the illative for any nominal (noun, pronoun, adjective or numeral), won't I?
Well, considering the illative form is dependent on the declension type, I would say yes. Although, I have never actually taught this topic in Finnish as a second/foreign language class. The system I presented is actually meant as study material for a Finnish morphology class for native Finnish speakers. It may well be that L2 teachers use another, more simplified system for non-natives that also works fine.
Woods wrote:Or there could still be something that cannot be guessed, like consonant gradation? Maybe they include the illative cause it contains the strong stem, while the genitive contains the weak one, the nominative is the basic form and for the partitive it depends?
This sounds plausible too.
Woods wrote:What do you mean with the ~ sign - that those forms mean roughly but not exactly the same? Kielitoimiston sanakirja lists only ammatteihin with double t.
The tilde ~ means that the forms are equal variants of the same form, so you can use both. So different from the tilde as it is used in mathematics or logic.
Now that I think about it,
ammatti was not the best example here:
ammateihin (weak grade) is indeed rare. But for example both
laatikoihin and
laatikkoihin seem quite normal to me.
Woods wrote:What makes it that loanwords get extra treatment?
I guess because some loanwords haven't adapted fully to the Finnish morphological system, so a speaker is not always sure which native stem type it should equate to, which causes variation. A good example (although not for illative forms) is the word
analyysi, which can take either back vowel or front vowel suffixes:
analyysi-ssa ~ analyysi-ssä. Native Finnish words don't have front and back vowels mixed in a single stem, so that causes confusion and subsequently variation.
Woods wrote:I just googled both in quote marks and got roughly 400 000 results for kauniisiin and 500 for kauniihin. Many of the latter are indeed from grammar explanations where they keep telling that those forms are possible.
I've noticed the same with other options that barely anyone ever uses. Why do they insist so much on them instead of removing them from dictionaries and leaving them in history and in the few books where they (possibly) were used once or twice?
You're right:
kauniihin is much less common than
kauniisiin. As for the question, I can't say for sure. I guess it's still used somewhat so it's kept in the standard. This reminds me: I once read in an older Finnish grammar (like around 100 years old) which stated that words like
matala 'low' can be declined as either
matalissa or
mataloissa. Today, the latter type is gone altogether: it's not mentioned in grammars or dictionaries, I've never heard it used.
Language changes but there's always some delay in when the changes are reflected in descriptions. You probably won't see
kauniihin in a Finnish grammar written in the 22nd century.
Woods wrote:What is a historically contacted diphthong and how to recognise it?
You can find them in plural stems of certain types of words:
- Words that end in -Vs in nom.sg. (+ weak grade) and have -VV- stem in gen.sg. (+ strong grade): rikas : rikkaa-, kaunis : kaunii-, kirves : kirvee-, seiväs : seipää- > (pl.) rikkai-, kaunii-, kirvei-, seipäi-
- Words that end in -e in nom.sg. and have -ee- stem in gen.sg.: sade : satee-, laite : laittee- > (pl.) satei-, laittei-
- Words that end in -VV in nom.sg. and have a similar stem in gen.sg.: vapaa, harmaa, talkoo, hakkuu > (pl.) vapai-, harmai-, talkoi-, hakkui-
Btw, "historically contracted" simply means that there used to be a consonant between the last two vowels in the stem but it has disappeared and left behind a long vowel (in sg.) or a diphthong (in pl.):
- kaunis : kaunii- < kaunihi- < *kaunize- (a loan from Proto-Germanic *skauniz)
- sade : satee- < *saδek : *sateγe- (or *sadek : *satege-)
- harmaa < *harmaγa (or *harmaga) (a loan from a Baltic source; cf. Lithuanian šarma(s))