Actually, rendaku is (also) one of those little things every language has to help make words easier to understand in noisy environments.
As an English example, if a word ends in a voiced consonant, then the preceding vowel is about three times longer than if it ends in an unvoiced consonant. Just listen to someone saying peas and peace until you're convinced. So if there's some noise, and an English person wonders, "Did he use a s or a z?" they can say, "Well, the vowel was long, so it must be peas."
This is so VERY unconscious that you can confront people on it and they will have no idea it exists. If you get it wrong, you will still be understood, but people will think you have a funny accent, even if they don't know why.
The Japanese are at least aware that rendaku exists. I don't know if you'll still be understood without rendaku, but it helps determine if a word is a compound or just two words next to each other. Compound word? Use Rendaku. UNLESS it's in one of a short list of phonetic environments. (but be aware of the long list of words in those phonetic environments that use rendaku anyways.) Linguists are still studying rendaku.
Fanboy story: In the anime One Piece, there's a character that gets called 青鼻 for the creative reason that he has a blue nose. Half the characters in the show call him あおはな and the other half call him あおぱな. No one comments on it. Ever. This was my introduction to Rendaku.
I *do* have a story of something the Japanese do without realizing involving pitch accent. To make it worse, even the experts don't know what the Japanese are actually doing.
For some background, pitch accent came up shortly before I joined. I wanted to tag to that conversation, but it was already dead.
http://www.japanforum.com/forum/japa...beginners.html
There are actually two pitch accents: The way native Japanese hear it working, and the way they speak it. Japanese tell you that it's a matter of low syllables vs high syllables, and they *do* hear it that way, but if you look at the spectrograms, the truth is a word starts low, rises gradually to the "accented" mora, then drops sharply on the *next* mora, after which it rises gently for the rest of the word. That gentle rise is why pitch accent descriptions tend to show the first mora as low (unless it's the accented mora, in which case the drop appears right after and the first mora seems high.)
That's just trivia. The hair tearing thing is, if the accented mora is voiceless, and the mora after it is also voiceless (and the spectrograms show they're voiceless), a native Japanese person can *still hear* a drop between the two mora. And it's not like they're making something up, they all agree where the "drop" is, even if it's a word none of them have heard before. Last time I read about this was years ago, and the experts didn't really have an explanation why yet. I should probably check up on it.
I was surprised to hear Masaegu say he was taught the accent for every word in school, because they did a survey once, "How do you know the difference between 鼻 and 花?" and a lot of respondents answered, "Because the kanji are different," despite the fact that the survey was 100% oral. Maybe because they're words you learn as a kid instead of having to be taught?
I could go on... frankly I know more *about* Japanese than I know actual Japanese.