[wolfgke]

> I guess that makes sense, particularly if there are not very systematic rules that define those things

There are systematic rules for this, but you don't want to derive the result each time you want to express something. Compare it to applying a formalized multiplication algorithm each time you want to compute something from your multiplication table (up to 100) vs. memorizing the table. You have to understand the rules (just as you have to understand how multiplication works) for conjugating/declinating, but to apply them in practise you have to rote train them as often as you don't have to think about the rules anymore (because you really don't want hesitate for seconds what the correct, say, declination is).

[vram22]

Thanks again for the elaboration. I'm getting what you say, and that was sort of why I referred to learning multiplication tables by rote, above.

It's pretty late for me, so I will read and reply to your and @lorenzhs's (thanks to you too) comments again tomorrow, if I have anything further to say or ask. Good stuff.