bombobuffoon wrote:tiia wrote:bombobuffoon wrote:Particularly the correct way to change the form of a noun connected to a verb. A teacher might say something like "just take the genitive stem and ..its so easy" well... they ignore the fact that calculating the genitive stem is insanely difficult. There is a lot of randomness in there. But this is common to every language.
Well, I don't think there's a lot of randomness in the stem. If it feels like that, my guess is that the rules were probably not explained well enough. One of the issues is, however, that oftentimes natives don't know the rules so well, so they cannot explain them perfectly. Or they think that a lesd in depth explanation would suffice. So feel free to ask about that here. We may be able to clear up the randomness a little bit.
Forming the genitive stem is complicated, to the point of, it essentially feels random. If there were some clearer shortcut invented by people that could "clear up the randomness" then I think it would be common wisdom. Its not. The best advice I have had so far is to learn by feel because applying rules like the below is too complicated. And to add to that there are also nouns that cannot be calculated analytically. I have to say I agree with this advice of going by feel.
http://users.jyu.fi/~pamakine/kieli/suo ... varen.html
You can't form the genitive stem based on the nominative because it's an arbitrary choice for a dictionary form; other things are not formed from it, and it itself is a derived form. Instead all forms can be said to be derived from a postulated "abstract form" which never surfaces <as itself>. It's a very useful construct in explaining words that are similar on the surface, for example "laki" (mountaintop/hilltop [gen. laen] and law [gen. lain]) which have different underlying or abstract forms; the older word's underlying form would be {lak} whereas the loanword's underlying form would be {laki}. If we then apply regular rules we get:
{lak} + n (genitive)
lak-e-n (epenthesis rule)
lag-e-n (consonant gradation)
"laen" (k-weak grade deletion after middle Finnish)
{laki} + n
laki-n
lagi-n
"lain"
There are simplified schemes, for example for i-nominative words: "stone age words (kivi, tuli, etc.) have genitive in -en, modern words and loanwords have -in".
What I want to illustrate is, that considering the nominative as a basis from which to derive other forms is counterproductive, and you will never know how a word behaves 100% of the time knowing only the nominative. Most dictionaries give nom.sg, gen.sg, part.sg and ill.sg.